- Email: [email protected]
Membrane dopamine β-hydroxylase: a precursor for the soluble enzyme in the bovine adrenal medulla
fnf. J. Biochem. Vol. 16. No. 6, pp. 64-650, Printed in Great Britain. All rights reserved
1984
0020-711X/84 $3.00 + 0.00 Copyright 0 1984 Pergamon Press Ltd
MEMBRANE DOPAMINE P-HYDROXYLASE: A PRECURSOR FOR THE SOLUBLE ENZYME THE BOVINE ADRENAL MEDULLA
IN
K. B. HELLE, R. K. REED, K. E. PIHL and G. SERCK-HANSSEN Department
of Physiology,
PKI,
University of Bergen, Arstadvei [Tel. (05) 29- 17001
19, 5000 Bergen,
Norway
(Received 2 September 1983) Abstract-l. Searching for endogenous proteolytic activities converting the membrane form of dopamine /j’-hydroxylase (dopamine p-monooxygenase, DBH) into the soluble and releasable form, DBH was monitored enzymatically and immunologically in aqueous and detergent-solubilized extracts of the adrenomedullary fractions. 2. Degradation of the soluble DBH and acidic chromogranins by activation of endogenous proteases occurred during lysis in H,O. 3. Shifts in the hydrophobicity of the membrane DBH were also apparent. Loss in enzyme protein or activity was, on the other hand, not observed for buffer-dialysed CG (pH S-6). 4. Limited proteolysis within the membrane phase was, however, indicated by the shift towards dominance of the intermediate hydrophobic DBH in the buffer-dialysed CC. 5. By two-dimensional, crossed immunoelectrophoresis with cationic detergent the microsomal DBH was immunologically identical to the granule-bound enzyme but differed from the latter in molecular heterogeneity and in susceptibility to proteolytic solubilization by endogenous protease activities. 6. DBH in the membranes of the chromaffin granules was proteolytically solubilized at pH 68 and the soluble DBH further degraded at pH 5. 7. The results indicate that a post-translational conversion of the amphiphilic DBH into the soluble form, initiated at the level of the microsomes, may continue within the light and the heavy granule fractions which contain several DBH-converting and degrading proteolytic activities with acid optima.
INTRODUCTION Dopamine fl-hydroxylase (DBH) occurs in two distinct states in the bovine chromaffin granules, one releasable and another confined to the surrounding membrane (Ksnig et al., 1976; Aunis et al., 1977; Blakeborough et a/., 1981). The two states have large parts of the primary structure in common and are immunochemically identical but can be distinguished by differences in substrate affinities and pH stabilities (Miras Portugal et al., 1976) and in hydrophobicity
during
charge-shift crossed electrophoresis (Bjerrum 1979; Skotland and Flatmark, 1979). The hydrophilic and releasable DBH retains the immunoreactivity and enzyme activity during limited tryptic proteolysis (Helle et a/., 1977). The two membrane forms, the amphiphilic and the intermediate hydrophobic DBH, both integral membrane proteins, can be proteolytically solubilized by thermolysin or chymotrypsin in vitro and by activation of endogenous proteolytic activities in the large granule fraction (Bjerrum et al., 1979; Helle et al., 1982a). These findings suggest that the membrane and soluble DBH may arise from a common, hydrophobic precursor in viva and that solubilization may occur by posttranslational conversions, analogous to the processing of other secretory proteins (Steiner et al., 1974; Blobel and Dobberstein, 1975; Docherty and Steiner, 1982). A search for endogenous conversion systems within the adrenomedullary fractions should therefore be undertaken in order to pin-point the stage(s) in granule biogenesis where solubilization of DBH takes place. et ul.,
641
Much of our knowledge of the molecular organization of the chromaffin granules, their formation and role in secretion rests on studies with the soluble and membrane DBH as markers for the released core and retained membranes (Gagnon et al., 1976; Winkler 1977; Ledbetter et al., 1978) without taking into account that the soluble DBH may in part arise from proteolytic conversion of the membrane enzyme during the isolation procedures. Hence, it is first of all essential to re-examine how the in vitro conditions during lysis affect the soluble and membrane bound DBH. The present study, of which a brief account has already been presented (Helle et ul., 1982b), describes to what extent solubilization and recovery of the active, solubilized enzyme in the chromaffin granules depend on the lysis conditions. Furthermore, activation of the endogenous proteases at pH 5-8 has been used to elucidate possible conversion effects on the membrane enzyme in the microsomal, light and heavy granule fractions. MATERIALS AND METHODS Bovine adrenals Bovine adrenals were obtained from the local slaughter house and processed as previously described (Bolstad et a/., 1980). The large granule fraction (LG) in 0.3 M sucrose was layered on 5 volumes of 1.6 M sucrose and centrifuged to give two membrane-rich fractions after 6 x lo6 g,, mitt; the first (LG-I) at the interphase between 0.3 and 1.6 M sucrose and the second (LG-II) as the pellet in the high density layer. These fractions served as sources of light, immature gran-
K. B. HELLE er al.
642
ules and membrane ghosts (LG-I) and heavy, mature chromaffin granules (LG-II = CC) respectively. Microsomal (I’,) and post-microsomal (SJ fractions were obtained from the supernatant (S,) above the LG pellet after further centrifugation (6 x 106ga, min). LJJS~Sqf CG All procedures were carried out at +4”C. Two different procedures were employed: (1) dilurionin 10 or 20 vol of the appropriate buffer (< 1 mg protein/ml in final dilution), or (2) dialysisof a concentrated granule suspension (> 10 mg protein/ml) against 3 shifts of 500 vol for 48 hr in tubings of restricted porosity ( 100yO retention of mol. wt > 15,000). The lysed suspensions were centrifuged at 6 x 10hg,, min and the high speed supematant (first lysate, SN I) was taken as the soluble fraction. The pellets were resuspended in detergent-containing buffer at PH 8.7 (2% Triton %-100, v/v in 100 mM glvcine. 38 mM Tris) and left overnight at +4”C. The Triton&luble (TX-SO] aft& high speed ~~trifugatio~ (6 x 106gaV min), and the residual protein, resuspended in the detergent-containing buffer (TX-r), represent the total membrane ,fraction. Activation
of endogenous proteases
a~ pH 5-8
Undialysed preparations of P,, LG-I and LG-II membranes (CG,) were diluted in 10 volumes of buffer (f4”C). One aliquot of each preparation was kept as control and the protease inhibitors were present in the buffers during dilu&on. The other aliquot was diluted without inhibitors and incubated for 2 hr at 37°C before addition of inhibitors and transfer to ice. All further procedures were carried out at +4”C and separation of the soluble and membrane fractions was made as described above for the lysed CG. Aprotinin KIU (100 Trasylol “?/ml), phenylme~anesulfonyi fluoride (1 mM) and p-mercuro~nzoate (1 mM) were added as inhibitors of trypsin-, chymotrypsinand thiol-subtilisin-like protease activities respectively. Folin protein Folin (Bolstad
protein was determined et al., 1980).
Dopamine P-hydroxylase EC 1.14.17.1, DBH)
as previously
(dopamine
described
geneity while more than one peak or a shoulder in the continuous precipitate were taken as indications of molecular heterogeneity in the antigen. The relative mobilities of the peaks were determined for each run. using the active soluble DBH (hydrophilic enzyme, pH 5) as the internal reference for peak 1 in each pattern. The amphiphilic (peak 3) and intermediate (peak 2) hydrophobic DBH were numbered according to Bjerrum et al. (1979).
The probability (P) for significance of difference groups was obtained by Wilcoxon’s non-parametric appropriate tables.
between test and
RESULTS
DBH in the lysed CG ~oIub~iization. Chromaffin granules are commonly lysed in hypotonic solutions by dilution or dialysis. Dilution leads to solubilization of _ 50% of the total enzyme activity concomitant with _ 77% of the total protein after several cycles of freezing and thawing of the water-insoluble fraction (Winkler, 1976). As shown in Fig. 1, solubilization of the immunorea~tive enzyme protein into the first supernatant was slightly dependent on pH (slope = 5.6”/, per unit pH, P < 0.001) in buffers ranging from pH 5-8. At pH 6-7 the solubilization of DBH was 42 & 5% (n = 7) of the total enzyme protein. Dialysis of the lysate is essential for recovery of a high specific DBH activity (Serck-Hanssen et al., 1980) and leads to the same ratio of free DBH/total activity as three cycles of freezing and thawing (Aunis ef al., 1977). As shown in Table 1, the soluble enzyme was highly active after lysis by dialysis in acid buffers (pH 5-6) but largely inactivated in Tris-HCI (pH 8), in accordance with the acid optima for the substrate affinities and V,,,,, (Miras Portugal et al., 1976). The
f%monooxygenase,
%
DBH was assayed enzymati~ally by the spectrophotometric method, with some modifications (Helle ef al., 1977). Immunoreactive DBH (DBHI) was quantitated with a specific rabbit anti-bovine DBH (Helle et al., 1978, 1982a) by two different techniques: (I) by nephelometry with the Beckman Immune-Analyzer and (2) by crossed immunoelectrophoresis (Bjerrum et al., 1979). Antibody was used in excess, at dilutions I:32 and I:800 by methods 1 and 2 respectively. Samples containing 0.1-5-,g DBHI were assaved by method 1 in 0.17 M NaCl(O.6 ml) and the turbidity oc&rr&g from immunopr~pitat~on after 2 hr at room temperature was quantitated. Samples with a high background turbidity (TX-r) could not be quantitated by method 1 and DBHI was therefore estimated by method 2. Purified soluble DBH Purified soluble DBH was used as the reference immunotitrations (a gift from Dr T. Skotland).
in the
70
r-
lI
.
I
I
I
I
5
6
7
8
PH Moieculrtr heterogeneity Hvdro~hilic and hydrophobic DBH were differentiated by &osskd immunoeiectrbphoresis with the cationic decetvltrimethvlammonium bromide (CTAB, tergent. 0.0125% v/v) and Triton X-100 (0.2x, v/v) in the agarose gel during the first dimension electrophoresis (for further details, see Bjerrum ei al.. 1979). According to the de~nitions given in this study, a single, symmetric peak of immunoprecipitate was taken as evidence for molecular homo-
Fig. 1. Effect of pH on solubilization of DBHI. Chromaffin granules were lysed by dilution in 20 vol of buffer (I = 0. I. final protein concentration O.O60mg/ml). The first lysate and the Triton-solubilized pellets were quantitated by crossed immunoelectrophoresis for sDBHi and mDBHI respectively. DBHI, = sDBH1 + mDBH1. Buffer ions: Kacetate(O), Na-succinate (Cl), K-phosphate (a), Tris-HCl (A) and lysine -HCl (A.). H20-diluted granules, final pH 6.4 (arrow). Calculated regression line (---), P c: 0.001.
Membrane dopamine P-hydroxylase Table Dialysis
I. Elkt
of lysis on recovery
medium
Hz0 Na-succinate. Na-succinate, K-phosphate, Tris~HCl, 50
of active enzyme DBH activity (units/mg SP)
PH 5.3” * 0.1 5.0 6.0 6.0 8.0
50 mM 50 mM 50 mM mM
0.04 0.54 0.50 0.52 0.03
(n)
k 0.06 * 0.04 F 0.23 f 0.08 +_0.01
(4) (6) (15) (4) (3)
Values are means *SD of the activity in the lysate. I unit = I ymol,‘min. SP = soluble protein. Number of preparations (n). “pH determined in dlalysed suspension.
percentages of solubilized enzyme protein, 41 and 45% at pH 5 and 8 (Table 2B), were within the range obtained by dilution (Fig. 1). However, the pH difference in solubilization by dialysis was not statistically significant. Lysis of CG by dialysis in H,O (final pH 5.3) resulted, on the other hand, in low recoveries of immunoreactive enzyme and total protein (Table 2A), particularly of the soluble fractions (Table 2B). As shown in Table 3, the loss in total protein could be prevented by dialysis in presence of inhibitors of
Table
2. Distribution membrane
the serine and thiol proteases during dialysis. In the absence of inhibitors about half of the soluble protein was lost, without significant decline in membrane protein. Inactivation of DBH by H,O dialysis (Table 1) was on the other hand not prevented by the inhibitors of trypsin and chymotrypsin-like proteolysis (Table 3). In the buffer-dialysed preparations the total and soluble immunoreactive enzyme accounted for 5-6 and 34% by weight of the total and soluble protein respectively, without significant differences between at pH 5 and 8 (Table 2B). This is in accordance with previous estimates by different techniques (Winkler, 1976; Helle et al., 1978). The buffer-dialysed CG revealed on the other hand a marked pH difference in extraction of total protein (Table 2B). Hence, these results revealed that extraction of the enzyme could be dissociated from that of the total protein by a change in pH, leading to large differences (6 and 16%) in the concentration of enzyme protein in the membrane phase without a change in ratio of soluble and membrane-bound DBH. All but 5% of the CG protein could be solubilized in aqueous solutions in the presence of a neutral detergent (Table 2B). Immunoreactive enzyme was
of immunoreactive DBH (DBHI) and total protein in soluble fractions of chromaffin granules lysed by extensive dialysis Dialysis
~~~.._
Constituents (A) In 7; ofrecowr.v af pH 8.0: Total protein Total DBHI (DBHI,) (B)
In
643
and
conditions
in H,O (a)
in Na-sac&ate (PH 5.0) (b)
80 +_ 5 65 f 3”
113 f 15 93 f I
in Tris-HCI (PH 8.0) (c) 100 100
mg/IOOmg: 4.8 8 I .4 6.5 30
DBHIJTP sDBHI/DBHI, sDBHI/SP mDBH/MP SP/TP RPjTP rDBHI/RP
f + * k f
0.2” 2“ 0.5’ 0.5” 3” 5 2
5.2 * I.9 41+3 4.2 i 0.7 6.5 f 1.4” 51 f5” 5 2
6.1 f 45* 3.4 f 16.0 k go+
0.2 1 0.2 I.0 1 4 2
SP = soluble protein; MP = membrane protein. RP = residual protein and rDBH1 = mDBH1 in RP. i.e. remaining in the high speed pellet after extraction of the membrane fraction with 2:, Triton X-100 (v/v) in Tris-glycine, pH 8.7 (see “Methods”). Values are means k SD of 3 preparations. Each preparation was divided in three aliquots and the concentrated suspensions were lysed by dialysis: (a) in absence of bulk ions or(b) and (c) at I = 0.10-O. 15 for 48 hr at f4’C with three changes of 500 vol of medium. “I’ < 0.01 for differences from the values obtained for granules dialysed at pH 8.0.
Table 3. Effect of protease inhibitors on recoveries of protein and DBH activity after H,O dialysis of the chromaffin granules DBH activity (pmoliw) Granule
fractions
Soluble lysate Membranes Total Lvsate/total
6)
(-) 0.04 k 0.06 0.29kO.15 0.11 f0.10 (4)
(+) 0.07 + 0.07 0.17f0.14 0.12+0.10 (4)
Recoveries in % (protein) (-) 55 f 7” 91*11 13 * 7” 37 f 5” (4)
(+) 100 100 100 52 + 2 (4)
Heavy granules (LG-II, CC?) were suspended in H,O in the absence ( - ) or presence ( + ) of I mM p-mercuribenzoate, 1 mM phenylmethylsulfonylfluoride and 100 KIU Trasylol. The suspensions were dialysed against H,O + inhibitors (0.01 mM p-mercuribenzoate, 0.01 mM phenylsulfonyltluoride, 1 KIU Trasylol) in the cold (3 changes of 200 vol). After dialysis the lysate pH was 5.3 i 0.1. Number of preparations (n). “P < 0.01.
644
K. B. HELLE etal.
present in this residual fraction and accounted for about 2% of the residual protein after dialysis against H,O and in buffers at pH 5 and 8. Molecular heterogeneity. The effects of the dialysis procedure on the relative distribution of hydrophilic and hydrophobic forms of the enzyme were examined by two-dimensional crossed electrophoresis. In accordance with the method (Bjerrum et al., 1979) the cationic detergent, CTAB, was present in addition to the neutral detergent (Triton X-100) in order to reveal minor differences in hydrophobicity. Only one, continuous immunoprecipitate was obtained for each fraction (Figs 2-5) consistent with previous reports on the specificity of the antibody and the immunological identity between the hydrophilic and hydrophobic forms of the enzyme (Helle et al., 1978, 1982a; Skotland and Flatmark, 1979; Slater et al., 1981; Bjerrum et al., 1979). The buffer-dialysed preparations revealed the hydrophilic enzyme as symmetric precipitate peaks (Fig. 3, lysate), with slightly faster relative mobility for the inactive enzyme (pH 8) than for the enzymatically active lysate (pH 5). The membrane fractions were extracted (TX-sol) and the residual membrane protein (TX-r) resuspended under identical conditions with Triton X-100 in the Tris-glycine buffer at pH 8.7. Any difference in the immunoelectrophoresis patterns between the TX-sol and Tx-r fractions should thus refer to earlier preparatory events. As shown in Figs 2 and 3 a peak corresponding to the hydrophilic enzyme was conspicuous in the patterns of the HzO- and buffer-dialysed membrane fractions. Marked differences were on the other hand apparent in the distribution of DBH protein in peaks 1, 2 and 3 in the H,O- and buffer dialysed membranes. The amphiphihc DBH (peak 3, Fig. 2A) was inconspicuous in the membranes of CG dialysed at pH 5 and 8 and apparent only as a shoulder at the catodic side of peak 2 (Fig. 3, TX-sol and TX-r). Membrane DBH corresponding to the intermediate hydrophobic DBH (peak 2, Fig. 2A) was on the other hand abundant, notably after dialysis at pH 8. Thus, these immunoelectrophoretic patterns indicated that the molecular heterogeneity of DBH in the CG membranes was modified by the dialysis conditions. After dialysis in pH 8 the intermediate hydrophobic peak was the dominating membrane DBH. A shift towards the hydrophilic peak was on the other hand characteristic of the membrane DBH after dialysis of CG in Hz0 in absence of protease inhibitors.
Comparison of DBH in the microsomal, heavy granule fractions
light and
Solubilization. The preparations were lysed by dilution in pH 5 and 8 buffers in the presence of the three protease inhibitors. As shown in Table 4 the microsomes, the light granules and the membranes of CG were dominated by the buffer-insoluble forms of DBH (78-88x of the total enzyme). The percentages of buffer-insoluble total protein were lower than the enzyme protein in the microsomes and CG membranes. These two fractions contained similar percentages of residual protein. The light granule frac-
tion was on the other hand lower in buffer-soluble and residual protein than the two other fractions. Molecular heterogeneity. The patterns obtained under identical electrophoretic conditions were photographically superimposed to reveal differences in relative mobilities between the fractions. Figure 4 compares the patterns of the total microsomes and the total CG. In accordance with the quantitations in Table 4 the microsomal fraction (Fig. 4, upper pattern) was abundant in hydrophobic DBH, in contrast to CG (Fig. 4, lower pattern) being dominated by the hydrophilic enzyme. As shown in Fig. 4 the relative mobilities of the microsomal peaks were lower than those of CG. The membrane fractions of the microsomes (Fig. 5A, lower pattern) and the light granules (Fig. 5B, lower pattern) were compared with the partially proteolysed, H,O-dialysed CG membranes (Figs 5A and 5B, upper patterns). After dilution in presence of protease inhibitors the microsomal membrane fraction was virtually freed of the hydrophilic DBH (Fig. 5A) and was characterized by the intermediate enzyme peak. Amphiphilic DBH was indicated in the microsomal membranes from the asymmetry on the catodic side of peak 2. Dilution of the light granules in presence of protease inhibitors resulted in membranes with hydrophilic (Fig. 5B) as well as the intermediate DBH. Judging from the apparent symmetry in peak 2 the light granule membranes were devoid of the amphiphilic enzyme. Effects of protease activation
on DBH
Activation of endogenous protease activity was achieved by incubation for 2 hr at 37°C before addition of inhibitors. Preparations of microsomes, light granules and CG membranes were compared with respect to solubilization of the enzyme and total protein after activation at pH 5-8. Sofubilization. The microsomal enzyme protein was not significantly changed by incubation at pH 558 (Table 5) although a substantial loss in total and membrane protein was apparent after 2 hr at pH 5. The membranes of CG was on the other hand significantly lower in total enzyme protein than in total protein at pH 68. At pH 5 an apparently parallel degradation of the total and membrane protein had taken place. The light granule enzyme protein was resistant to proteolytic degradation at pH 6-8 (Table 5). At pH 5 the degradation of the total enzyme protein was proportional to the protein concentration during incubation. The declines in the membrane enzyme were larger than in the total enzyme. As shown in Table 6, incubation of the light granules a pH 5 caused significant increases in soluble DBH and Folin protein. However, a loss of solubihzed enzyme occurred in the more concentrated preparations. Molecular heterogeneity. The immunoelectrophoretic pattern of the light granule membrane fraction (batch III, Tables 5 and 6) is given in Fig. 5C (lower pattern). The intermediate DBH, conspicuous in the control (Fig. 5B, lower pattern), was no longer apparent after the 2 hr incubation at pH 5 (Fig. 5C, lower pattern). The remaining enzyme corresponded in mobility to the hydrophilic enzyme (peak 1). The post-microsomal fraction (Fig. SC, upper pattern),
Fig. 2. Crossed immunoelectrophoresis patterns of membrane DBH in H,O-dialysed chromathn granules. (A) The Triton X-IOO-soluble (TX-sol) and (B) the -insoluble, resuspended fraction of residual protein (TX-r). The patterns of (A) (11 pg) and (B) (9 JL~Fohn protein) were obtained in parallel runs in the same electrophoresis experiment for comparison of relative mobilities of the peaks, the bars representing the mobility of the soluble DBH in a parallel run. The peaks are numbered according to previous definitions for DBH in the chromaffin granules run in the presence of cationic detergent in charge-shift immunoelectrophoresis (see Bjerrum et al., 1979).
Fig. 3. Crossed immunoelectrophoresis patterns of soluble and membrane DBH in buffer-dialysed chromaffin granules. The granules were lysed by dialysis against buffer, 50mM Na-succinate, pH 5 or 50 mM Tri-HC1, pH 8. The lysates, the Triton-soluble (TX-sol) and residual membrane protein (TX-r) were run in presence of cationic detergent and Triton X-100 (horizontal direction) for separation of hydrophilic and hydrophobic components (see Bjerrum et al., 1979). Bars indicate relative mobility of the soluble DBH (peak 1, lysate pH 5) run in parallel in the same electrophoresis experiment. The protein concentrations of the samples were: lysates, 14~g; TX-sol, 11 pg and Tx-r 9pg respectively.
Fig. 4. Comparison of total DBH in microsomal and chromaffin granules by crossed immunoeiectrophoresis. The immunopr~pitate patterns were obtained under identical electrophoresis conditions and photographically superimposed for comparison of relative mobilities. The total, Triton X-100 solubilized microsomes (upper pattern, 42 pep Folin protein) and chromaffin granules (lower pattern, 13pg Folin protein) were run in the presence of cationic detergent in the Triton X-100 containing gel in the first dimension (ho~zontal) el~trophoresis. Bar indicate relative mobility of the soluble DBH (peak I) of the lysate, pH 5 in a parallel run in the same electrophoresis experiment.
Fig. 5. Comparison of membrane DBH in microsomal, light and heavy (chromaffin) granule fractions. The immunoprecipitate patterns were obtained under identical electrophoresis conditions and photo~aphical~y superimposed for comparison of relative mobihties of the peaks. The samples were (A) and (B) the membranes of the H,O-dialysed chromaffin granules (upper patterns, 12 fig Folin protein); in (A) the microsomal membrane DBH (lower pattern, 16 pugprotein); in (B) and (C) the membranes of the light granule fraction (lower patterns; (B) 11 pg Folin protein before and (C) 5 fig Folin protein after 2 hr activation of proteolysis (see “Methods”). The soluble DBH in the post-mi~rosomal fraction (S,} is given in (C) (upper pattern, 94 pg Folin protein).
646
647
Membrane dopamine /I-hydroxylase Table
4. Membrane
DBHI
and protein
in adrenomedullary inhibitors
Buffer-insoluble Soluble mDBH1
Fraction
fractions
components
in 2% Triton
in the presence
of protease
in “/, of total in fraction
X-100 Folin protein
Residual fraction” Folin protein
Microsomes (P,)
88 * 5
64+
IO
Light granules (LG-I)
78 +8
87 k 5
2+0
Heavy granule membranes (LG-II,; CC,)
82 f 8
53 * I4
II *2
1212
Four preparations of each fraction were diluted IO times in ice cold buffer containing protease inhibitors (see “Methods”). The high speed supernatants and buffer-insoluble pellets, solubilized in 2% Triton X-100, were assayed for DBHI by nephelometry and Folin protein. “The residual fraction, resuspended in 2% Triton was not free of light scattering particles and could therefore only be assayed for protein.
assumed to represent soluble DBH from fragmented organelles (Winkler, 1977; Ledbetter et al., 1978) revealed a symmetric peak of hydrophilic DBH, with slightly faster mobility than peak 1 in the proteolysed light granule membranes. DISCUSSION
The present
experiments
have made it evident
Table 5 Effect of activation
that
proteolytic activities are present in the purified preparations of CG and contribute to DBH solubihzation and degradation when activated during lysis. This conclusion imposes a number of questions related to previous quantitations of the soluble and membrane components and to the current concepts of the soluble and membrane DBH as stable and well defined proteins.
of endogenous proteases on the total and membrane in membrane fractions of the bovine adrenal medulla Recoveries
after 2 hr at 37°C in “/, of unincubated
in total fraction Fraction Microsomes (P,) pH 6.5-8.0” pH 5.0 Heavy granules, membranes (LG - 11, = CG,) pH 6.1lX.O pH 5.0 Light granules, (LG-I) pH 6.1-8.0 pH 5.0
DBHI,
(1) (I)
91 _+6 93
(II) (II)
79 + 9 71
(III) (III)
102 * 2 102 85 +9 69 f 5
(IV)
W)
bound DBHI and protein control
in but&-insoluble
Total protein (TP)
mDBH1
86 * 2 56
fraction Membrane protein (MP)
94 f 4 84
85 k 2 49
(3) (1)
IO1 +2 61
72k 67
85+ 49
(3) (1)
90+ I3 52 Ill is III *4
955 IO 52 65 + 1 541 I
I2
I2
83 k 3 43 78 k 2 67* I
(3) (1) (3) (3)
Aliquots from each hatch incubated for 2 hr at 37°C before addition of protease inhibitors (see “Methods”). Control aliquots were suspended in ice-cold buffer containing the protease inhibitors. “Final pH during incubation. Number of incubation experiments (n) from each batch (1-V). Final protein concentration during incubation (mg/ml): I, 1.3; II, 1.5; III, 0.5; IV, 6.9; V, 8.1. Each value is the mea” of duplicates. Means f SD of values for the incubations (n > 2). Table 6. Proteolysis
dependent
increase in soluble eranules
DBHI Recovery
Component
Batch No.
mg/TP/ml”
pH5
and protein
in the light
in ;< of control pH 6-8
sDBH1
III IV V
0.5 6.9 8.1
500(l) 176 k 30 (3) 149 _+40 (3)
ll7kl1(3)
SP
III IV V
0.5 6.9 8.1
109(l) I98 F 2(3) 201 f I8 (3)
177 + 105(3)
Three different batches of LG-I were incubated for 2 hr at 37°C before addition of protease inhibitors (see “Methods”). Control aliquots were suspended in ice-cold buffer containing the protease inhibitors. “Final protein concentration during incubation. Number of incubation experiments from each batch (n). Means k SD for the incubations. Each value based on assay in duplo.
648
K. B. HELLE el 01.
(1) Does the proteolytic activation affect the solubilization and the activity of the soluble DBH during lysis? Inactivation of the H,O-solubilized enzyme was not prevented by dialysis in presence of the protease inhibitors. Nor was there any indications of proteolysis in the enzymatically inactive pH 8 lysate. The soluble DBH withstands trypsin treatment at pH 7.5 without decline in tyramine affinity and increases further in specific activity upon digestion of the trypsin-sensitive chromogranins (Helle rt cd., 1977). Hence, it is concluded that proteolysis per se is not a likely reason for the low specific activity of the soluble DBH after H,O dialysis. Our experiments have shown that proteolytic activities degrade the soluble DBH as well as the chromogranins during H,O-dialysis of CC but leaves the membrane DBH relatively intact. Dialysis in buffers, leading to the same degree of DBH solubilization (4145%) into the first high speed supernatant as dilution (-42x), does not activate proteolysis. This shows that proteolytic loss of the soluble DBH during lysis by H,O dialysis is an exception which does not apply to buffer-dialysis. (2) Are there any eflects of mdogenous proteases on solubilization of DBH ,from any’ of the other grunule fractionsV Most of the microsomal and light granule DBH was insoluble in buffer and less sensitive to proteolysis than the total protein. Incubation for 2 hr near neutral pH (6-8) had insignificant effects on the solubilization of microsomal or light granule DBH. Proteolysis-related changes in solubilization of DBH in the light and heavy granule fractions were on the other hand apparent at pH 5, being -5 times more potent in the CG membranes than in the light granule fraction. These results indicate that DBH-converting and degrading proteases with acid optima were insignificant in the microsomal fraction while seemingly concentrated in the membranes of CG. (3) Is there a specific proteolytic system ,for the conversion of hydrophobic to hydrophilic DBH in any of the adrenomedullary ,fractions? A slow conversion of the hydrophobic DBH into the hydrophilic enzyme was first observed after incubation of the large granule fraction for 24 hr at pH 7.5 (Bjerrum et al., 1979; Helle et al., 1982a). The present experiments have shown that a more rapid conversion may occur in the light granule fraction, leading to considerable increases (149-500%) in soluin excess of the concentrationbilized enzyme, dependent DBH degradation at pH 5 and without a parallel degradation of the total protein. The subcellular fractionation scheme employed is closely similar to that described by Bolstad et cd. (1980) and by Ledbetter et al. (1978). By these protocols the newly formed granules and membrane ghosts of ruptured CG, cosediment with mitochondria and lvsosomes into the 0.3-1.6 M sucrose boundary. This light granule fraction is therefore assumed to be rich in the lysosomal proteases, notably in cathepsin D (Smith and Winkler, 1966). Inhibitors of cathepsin D could not be obtained for
use in the present study. A selective conversion effect by cathepsin D was on the other hand indicated from in vitro experiments with the bovine spleen enzyme. After cathepsin D treatment of CC for 30 min at pH 6 there was an increase in soluble DBH (132% of control) without concomitant degradation of the chromogranins. This effect was contrasted by that obtained with chymotrypsin which concomitant with a similar increase in the soluble enzyme also degraded the chromogranins (unpublished experiments). Hence, cathepsin D appears as a plausible candidate for the DBH converting acid protease in the lysosome-rich light granules. Thiol proteases resembling lysosomal cathepsins have been implicated in the conversion of proinsulin to insulin on the basis that they are localized in the secretion granule/mitochondrial fraction and that they can convert pro-insulin to insulin in vitro (Docherty and Steiner, 1982). Some contamination of lysosomes in the CG pellet cannot be avoided (Smith and Winkler, 1966) consistent with recoveries of I l”/, of lysosomal acid protease (Lindberg et cd., 1982) and 217; of [j-glucuronidase (Bolstad et al., 1980) in the CG preparations. More recently there have, however, been reports on specific “trypsin-like”, acid proteases in CG, cleaving pro-enkephalins within the soluble core (Lindberg et al., 1982; Evangelista et al., 1982). As limited tryptic proteolysis of the chromogranins did not affect the enzyme activity or the quantity of immunoreactive DBH (Helle et al., 1977), a “trypsinlike” degradation of the soluble DBH in CG seems rather unlikely. A conversion of the amphiphilic to the hydrophilic DBH in CG was not apparent after the trypsin treatment but was readily obtained with chymotrypsin or thermolysin (Bjerrum et al., 1979). As higher concentrations of chymotrypsin also partially degraded the solubilized enzyme at pH 6 (Helle et nl., 1982b), a “chymotrypsin”-like activity might be implicated. This assumption lends support from the observed, parallel degradation of soluble DBH and protein during H,O dialysis in the cold. (4) Is the umphiphilic or the intermediate DBH serving us the immediate precursor ,for the proteolytically solubilized enzyme? The hydrophobic and hydrophilic DBH differ in a short sequence at the C-terminus (Slater et al., 1981) and small portions of this hydrophobic “tail” which spans the membrane, may be cleaved off within the membrane phase and without loss of enzyme activity (Zaremba and Hogue-Angelletti, 1981). Complete absence of the hydrophobic sequence, as in the hydrophilic DBH, prevents incorporation into lipid bilayers (Albanesi et al., 1980). The intermediate and amphiphilic DBH peaks have both been shown to be integral membrane components, differing however in their binding of the cationic detergent (Bjerrum et al., 1979). These two forms were present in equal proportions in the total extract of CG while a shift towards the intermediate peak was apparent for the membranes The present study has revealed further shifts towards a dominance of the intermediate peak, in the pH 8 dialysed CG membranes and in the microsomes. The amphiphilic peak, seen as a shoulder at the catodic side of most membrane patterns, was virtually absent from the light granule fraction. Lysis by
Membrane dopamine /I-hydroxylase
dialysis in acid buffer or in H,O caused additional shifts in the membrane enzyme and quite unexpectedly, peak 1 was conspicuous in every membrane fraction but the microsomal and was also present in the Tx-r fractions of the dialysed CG. Marked differences in composition between the Golgi and CC membranes have been reported (Trifaro et al., 1976). The apparent difference in solubility of peak 1 from the microsomal, the light and heavy granule membranes may thus reflect such differences in the hydrophobic environment. Aiternatively, the dominant peak 1 in the H,O-dialysed CC membranes accounts for a proteolysis-related conversion of the intermediate DBH to a less hydrophobic type within the insoluble fractions during solubilization in the Triton buffer. Our results on the light granule fraction shows that the enzyme protein is shifted from the insoluble to the soluble form concomitant with loss of the intermediate peak as a result of acid proteolysis which is arrested in the presence of inhibitors to serine and thiol proteases. Thus, taken together, the present findings
provide
strong
indications
of a sequential
conversion system for DBH within the membrane compartments, the amphiphilic DBH being the source and the intermediate hydrophobic form being the immediate precursor of the hydrophilic DBH. (5) IS it possible
to pinpoint where the conversion begins
the stages in biogenesis
in vivo? Assuming that the amphiphilic peak represents the undigested precursor of the soluble DBH and taking the dominance of the intermediate peak in the membranes as evidence of partial conversion, this posttranslational proteolytic modification must be initiated at an early stage in the granule biogenesis, analogus to the processing of prohormones (Docherty and Steiner, 1982). In the rat adrenal medulla (- 10% soluble DBH; Grzanna and Coyle, 1976) the steady state level of DBH is regulated in vivo by glucocorticoid-modulated proteolysis, presumably by trypsin-like activities (Wong and Ciaranello, 1982). Whether this also holds true for the bovine adrenal DBH remains to be established. The close similarity in DBH patterns between the microsomes, the Golgi (Helle et al., 1982b) and the pH 8 dialysed CG suggests that a conversion process within the membranes, as presently postulated, may be regulated and possibly reach a state of equilibrium with the soluble fraction which increases in ratio over the membranes in the course of CG maturation. In conclusion, the present study has shown that the membrane and soluble DBH in the bovine adrenal medulla are susceptible to attack by several proteolytic activities inherent in the adrenomedullary fractions and this calls strongly for control of proteolysis in studies of the bovine adrenal DBH. Furthermore, the present work provides substantial support for the “conversion hypothesis” and may serve as a basis for further studies of proteolytic activities involved in a sequential conversion of the membrane DBH in granule biogenesis and during secretion and subsequent retrieval of the emptied membranes. ~ckno~vle~~em~~~s-Thanks are due to the Norwegian Research Council for Science and the Humanities for financial
649
support (C.13.11.10-012) and to Dr T. Skotland, Department of Biochemistry, University of Bergen, Norway, for the gift of highly purified soluble DBH of bovine adrenomedullary origin. REFERENCES
Albanesi J. P., Robinson E. C. and Kirshner N. (1980) ~ri~cation and analysis of membrane-bound dopamine /I-hydroxylase. Fedn kroc. 39, 1920. Aunis D.. Bouclier M.. Pescheloche M. and Mandel P. (I 977) Properties of membrane-bound dopamine P-hydroxylase in chromaffin granules from bovine adrenal medulla. J. Neurochem. 29, 439447. Bjerrum 0. J., Helle K. B. and Bock E. (1979) Immunochemically identical hydrophilic and amphiphilic forms of the bovine adreno-meduliary dopamine j-hydroxylase. Biochem. J. 181, 231-237. Blakeborough P., Louis C. F. and Turner A. J. (1981) The structure and organization of dopamine p-hydroxylase in the chromaffin granule membrane. Biochim. biophys. Acta 669, 33-38.
Blobel G. and Dobberstein B. (1975) Transfer of protein across membranes. I. Presence of proteolytically processed and unprocessed nascent immunoglobulin light chains on membrane-bound ribosomes of murine myeloma. J. Cefi. Biol. 67, 835-851. Bolstad G., Helle K. B. and Serck-Hanssen G. (1980) Heterogeneity in the adrenomedullary storage of catecholamines, ATP, calcium and releasable dopamine /I-hydroxylase activity. J. Autonom. Nero. Syst. 2, 337-354.
Docherty K. and Steiner D. F. (1982) Post-translational proteolysis in polypeptide hormone biosynthesis. A. Rev. Physiol. 44, 625-638.
Evangelista R., Ray P. and Lewis R. V. (1982) A “Trypsinlike” enzyme in adrenal chromaffin granules: A proenkephalin processing enzyme. Biochem. biophys. Res. Commu~. 106, 895-902. Gagnon C., Schwartz R., Otten U. and Thoenen H. (1976) Synthesis, s&cellular distribution and turnover of dopamine /I-hydroxylase in organ culture of sympathetic ganglia and adrenal medulla. J. Neurochem. 27, 1083-1089.
Grzanna R. and Coyle J. T. (1976) Rat adrenal dopamine /I-hvdroxvlase: purification and immunological characieristics. j. Neuiochem. 27, 1091-1096. Helle K. B.. Serck-Hanssen G. and Reichelt K. L. (1977) Resistance of dopamine /?-hydroxylase to proteolysis during tryptic digestion of the adrenomedullary chromogranins. ht. J. Biochem. 8, 693-704. Helle K. B., Serck-Hanssen G. and Bock E. (1978) Complexes of chromogranin and dopamine P-hydroxylase among the chromogranins of the bovine adrenal medulla. Biochim. biophys. Acta 533, 396-407.
Helle K. B., Bjerrum 0. J. and Bock E. (1982a) Immunochemical characte~~ation of dopamine ~-hydroxylase. &and. J. Irnrn~~. 15, 97-123. Helle K. B., Pihl K. E. and Serch-Hanssen G. (1982b) Hydrophililic and amphiphilic forms of dopamine fl-hydroxylase in early and mature stages of the bovine adrenomedullary granules. In Synthesis, Storage and Secretion of Adrenal Catcholamines (Edited by Izumi F. et al.). Ad;. Biosci. 36, 217-224.
_
Kiinig P., Hsrtnagl H., Kostron H., Sapinski H. and Winkler H. (1976) The arrangement of dopamine P-hydroxylase (EC 1.14.2.1) and chromomembrin B in the membrane of chromaffin granules. J. Neurochem. 27, 15391541. Ledbetter F. H., Kilpatrick D., Sage H. L. and Kirshner N. (1978) Synthesis of chromo~anins and dopamine ~-hydroxylase by perfused bovine adrenal glands. Am. J. Physioi. 235, E475-E486.
650
K. B. HELLE ef al.
Lindberg I., Yang H.-Y. T. and Costa E. (1982) An enkephalin-generating enzyme in bovine adrenal biophys. Res. Commun. 106, medulla. Biochem. 186193. Miras-Portugal M. T., Jorda A. and Santos Ruiz A. (1976) Propietes cinetiques de la dopamine j?-hydroxylase membranaire. FEBS Let!. 72, 267-270. Serck-Hanssen G., Helle K. B., Sommersten B. and Pihl K. E. (1980) Quantitative aspects of the acetylcholineinduced release of dopamine b-hydroxylase and catecholamines from the bovine adrenal medulla. Gen. Pharmat. 11, 243-249. Skotland T. and Flatmark T. (1979) On the amohiohilic . _ _ and hydrophilic forms of dopamine b-monooxygenase in bovine adrenal medulla. .I. Neurochem. 32, 1861-1863. Slater E. P., Zaremba S. and Houge-Angeletti R. (1981) Purification of membrane-bound dopamine fl-monooxgenase from chromaffin granules. Relation to soluble dopamine /?-monooxygenase. Archs Biochem. Biophys. 21!, ?88-296. Smith A. D. and Winkler H. (1966) The localization of
lysosomal enzymes in chromaffin tissue. J. Physiol. 183, 1799188. Steiner D. F., Kemler W., Tager H. S. and Peterson J. D. (1974) Proteolytic processing in the biosynthesis of insulin and other proteins. Fedn Proc. 33, 2105-2115. Trifaro J. M., Duerr A. C. and Pinto J. E. B. (1976) Membranes of the adrenal medulla: a comparison between the membranes of the Golgi apparatus and chromaffin granules. Molec. Pharmac. 12, 536-545. Winkler H. (1976) The composition of adrenal chromaffin granules: an assessment of controversial results. Neuroscience 1, 65-80. Winkler H. (1977) The biogenesis of adrenal chromaffin granules. Neuroscience 2, 657-683. Wong D. L. and Ciaranello R. D. (1982) Regulation of synthesis and degradation of catecholamine synthetizing enzymes in the adrenal medulla. In Synthesis, S/oruge und Secretion of Adrenal Catecholamines (Edited by Izumi F. et al.). Ado. Biosci. 36, 233-241. Zaremba S. and Hogue-Angeletti R. (1981) Transmembrane nature of chromaffin granule dopamine /J-monooxygenase. J. biol. Chem. 256, 1231&-12315.
1984
0020-711X/84 $3.00 + 0.00 Copyright 0 1984 Pergamon Press Ltd
MEMBRANE DOPAMINE P-HYDROXYLASE: A PRECURSOR FOR THE SOLUBLE ENZYME THE BOVINE ADRENAL MEDULLA
IN
K. B. HELLE, R. K. REED, K. E. PIHL and G. SERCK-HANSSEN Department
of Physiology,
PKI,
University of Bergen, Arstadvei [Tel. (05) 29- 17001
19, 5000 Bergen,
Norway
(Received 2 September 1983) Abstract-l. Searching for endogenous proteolytic activities converting the membrane form of dopamine /j’-hydroxylase (dopamine p-monooxygenase, DBH) into the soluble and releasable form, DBH was monitored enzymatically and immunologically in aqueous and detergent-solubilized extracts of the adrenomedullary fractions. 2. Degradation of the soluble DBH and acidic chromogranins by activation of endogenous proteases occurred during lysis in H,O. 3. Shifts in the hydrophobicity of the membrane DBH were also apparent. Loss in enzyme protein or activity was, on the other hand, not observed for buffer-dialysed CG (pH S-6). 4. Limited proteolysis within the membrane phase was, however, indicated by the shift towards dominance of the intermediate hydrophobic DBH in the buffer-dialysed CC. 5. By two-dimensional, crossed immunoelectrophoresis with cationic detergent the microsomal DBH was immunologically identical to the granule-bound enzyme but differed from the latter in molecular heterogeneity and in susceptibility to proteolytic solubilization by endogenous protease activities. 6. DBH in the membranes of the chromaffin granules was proteolytically solubilized at pH 68 and the soluble DBH further degraded at pH 5. 7. The results indicate that a post-translational conversion of the amphiphilic DBH into the soluble form, initiated at the level of the microsomes, may continue within the light and the heavy granule fractions which contain several DBH-converting and degrading proteolytic activities with acid optima.
INTRODUCTION Dopamine fl-hydroxylase (DBH) occurs in two distinct states in the bovine chromaffin granules, one releasable and another confined to the surrounding membrane (Ksnig et al., 1976; Aunis et al., 1977; Blakeborough et a/., 1981). The two states have large parts of the primary structure in common and are immunochemically identical but can be distinguished by differences in substrate affinities and pH stabilities (Miras Portugal et al., 1976) and in hydrophobicity
during
charge-shift crossed electrophoresis (Bjerrum 1979; Skotland and Flatmark, 1979). The hydrophilic and releasable DBH retains the immunoreactivity and enzyme activity during limited tryptic proteolysis (Helle et a/., 1977). The two membrane forms, the amphiphilic and the intermediate hydrophobic DBH, both integral membrane proteins, can be proteolytically solubilized by thermolysin or chymotrypsin in vitro and by activation of endogenous proteolytic activities in the large granule fraction (Bjerrum et al., 1979; Helle et al., 1982a). These findings suggest that the membrane and soluble DBH may arise from a common, hydrophobic precursor in viva and that solubilization may occur by posttranslational conversions, analogous to the processing of other secretory proteins (Steiner et al., 1974; Blobel and Dobberstein, 1975; Docherty and Steiner, 1982). A search for endogenous conversion systems within the adrenomedullary fractions should therefore be undertaken in order to pin-point the stage(s) in granule biogenesis where solubilization of DBH takes place. et ul.,
641
Much of our knowledge of the molecular organization of the chromaffin granules, their formation and role in secretion rests on studies with the soluble and membrane DBH as markers for the released core and retained membranes (Gagnon et al., 1976; Winkler 1977; Ledbetter et al., 1978) without taking into account that the soluble DBH may in part arise from proteolytic conversion of the membrane enzyme during the isolation procedures. Hence, it is first of all essential to re-examine how the in vitro conditions during lysis affect the soluble and membrane bound DBH. The present study, of which a brief account has already been presented (Helle et ul., 1982b), describes to what extent solubilization and recovery of the active, solubilized enzyme in the chromaffin granules depend on the lysis conditions. Furthermore, activation of the endogenous proteases at pH 5-8 has been used to elucidate possible conversion effects on the membrane enzyme in the microsomal, light and heavy granule fractions. MATERIALS AND METHODS Bovine adrenals Bovine adrenals were obtained from the local slaughter house and processed as previously described (Bolstad et a/., 1980). The large granule fraction (LG) in 0.3 M sucrose was layered on 5 volumes of 1.6 M sucrose and centrifuged to give two membrane-rich fractions after 6 x lo6 g,, mitt; the first (LG-I) at the interphase between 0.3 and 1.6 M sucrose and the second (LG-II) as the pellet in the high density layer. These fractions served as sources of light, immature gran-
K. B. HELLE er al.
642
ules and membrane ghosts (LG-I) and heavy, mature chromaffin granules (LG-II = CC) respectively. Microsomal (I’,) and post-microsomal (SJ fractions were obtained from the supernatant (S,) above the LG pellet after further centrifugation (6 x 106ga, min). LJJS~Sqf CG All procedures were carried out at +4”C. Two different procedures were employed: (1) dilurionin 10 or 20 vol of the appropriate buffer (< 1 mg protein/ml in final dilution), or (2) dialysisof a concentrated granule suspension (> 10 mg protein/ml) against 3 shifts of 500 vol for 48 hr in tubings of restricted porosity ( 100yO retention of mol. wt > 15,000). The lysed suspensions were centrifuged at 6 x 10hg,, min and the high speed supematant (first lysate, SN I) was taken as the soluble fraction. The pellets were resuspended in detergent-containing buffer at PH 8.7 (2% Triton %-100, v/v in 100 mM glvcine. 38 mM Tris) and left overnight at +4”C. The Triton&luble (TX-SO] aft& high speed ~~trifugatio~ (6 x 106gaV min), and the residual protein, resuspended in the detergent-containing buffer (TX-r), represent the total membrane ,fraction. Activation
of endogenous proteases
a~ pH 5-8
Undialysed preparations of P,, LG-I and LG-II membranes (CG,) were diluted in 10 volumes of buffer (f4”C). One aliquot of each preparation was kept as control and the protease inhibitors were present in the buffers during dilu&on. The other aliquot was diluted without inhibitors and incubated for 2 hr at 37°C before addition of inhibitors and transfer to ice. All further procedures were carried out at +4”C and separation of the soluble and membrane fractions was made as described above for the lysed CG. Aprotinin KIU (100 Trasylol “?/ml), phenylme~anesulfonyi fluoride (1 mM) and p-mercuro~nzoate (1 mM) were added as inhibitors of trypsin-, chymotrypsinand thiol-subtilisin-like protease activities respectively. Folin protein Folin (Bolstad
protein was determined et al., 1980).
Dopamine P-hydroxylase EC 1.14.17.1, DBH)
as previously
(dopamine
described
geneity while more than one peak or a shoulder in the continuous precipitate were taken as indications of molecular heterogeneity in the antigen. The relative mobilities of the peaks were determined for each run. using the active soluble DBH (hydrophilic enzyme, pH 5) as the internal reference for peak 1 in each pattern. The amphiphilic (peak 3) and intermediate (peak 2) hydrophobic DBH were numbered according to Bjerrum et al. (1979).
The probability (P) for significance of difference groups was obtained by Wilcoxon’s non-parametric appropriate tables.
between test and
RESULTS
DBH in the lysed CG ~oIub~iization. Chromaffin granules are commonly lysed in hypotonic solutions by dilution or dialysis. Dilution leads to solubilization of _ 50% of the total enzyme activity concomitant with _ 77% of the total protein after several cycles of freezing and thawing of the water-insoluble fraction (Winkler, 1976). As shown in Fig. 1, solubilization of the immunorea~tive enzyme protein into the first supernatant was slightly dependent on pH (slope = 5.6”/, per unit pH, P < 0.001) in buffers ranging from pH 5-8. At pH 6-7 the solubilization of DBH was 42 & 5% (n = 7) of the total enzyme protein. Dialysis of the lysate is essential for recovery of a high specific DBH activity (Serck-Hanssen et al., 1980) and leads to the same ratio of free DBH/total activity as three cycles of freezing and thawing (Aunis ef al., 1977). As shown in Table 1, the soluble enzyme was highly active after lysis by dialysis in acid buffers (pH 5-6) but largely inactivated in Tris-HCI (pH 8), in accordance with the acid optima for the substrate affinities and V,,,,, (Miras Portugal et al., 1976). The
f%monooxygenase,
%
DBH was assayed enzymati~ally by the spectrophotometric method, with some modifications (Helle ef al., 1977). Immunoreactive DBH (DBHI) was quantitated with a specific rabbit anti-bovine DBH (Helle et al., 1978, 1982a) by two different techniques: (I) by nephelometry with the Beckman Immune-Analyzer and (2) by crossed immunoelectrophoresis (Bjerrum et al., 1979). Antibody was used in excess, at dilutions I:32 and I:800 by methods 1 and 2 respectively. Samples containing 0.1-5-,g DBHI were assaved by method 1 in 0.17 M NaCl(O.6 ml) and the turbidity oc&rr&g from immunopr~pitat~on after 2 hr at room temperature was quantitated. Samples with a high background turbidity (TX-r) could not be quantitated by method 1 and DBHI was therefore estimated by method 2. Purified soluble DBH Purified soluble DBH was used as the reference immunotitrations (a gift from Dr T. Skotland).
in the
70
r-
lI
.
I
I
I
I
5
6
7
8
PH Moieculrtr heterogeneity Hvdro~hilic and hydrophobic DBH were differentiated by &osskd immunoeiectrbphoresis with the cationic decetvltrimethvlammonium bromide (CTAB, tergent. 0.0125% v/v) and Triton X-100 (0.2x, v/v) in the agarose gel during the first dimension electrophoresis (for further details, see Bjerrum ei al.. 1979). According to the de~nitions given in this study, a single, symmetric peak of immunoprecipitate was taken as evidence for molecular homo-
Fig. 1. Effect of pH on solubilization of DBHI. Chromaffin granules were lysed by dilution in 20 vol of buffer (I = 0. I. final protein concentration O.O60mg/ml). The first lysate and the Triton-solubilized pellets were quantitated by crossed immunoelectrophoresis for sDBHi and mDBHI respectively. DBHI, = sDBH1 + mDBH1. Buffer ions: Kacetate(O), Na-succinate (Cl), K-phosphate (a), Tris-HCl (A) and lysine -HCl (A.). H20-diluted granules, final pH 6.4 (arrow). Calculated regression line (---), P c: 0.001.
Membrane dopamine P-hydroxylase Table Dialysis
I. Elkt
of lysis on recovery
medium
Hz0 Na-succinate. Na-succinate, K-phosphate, Tris~HCl, 50
of active enzyme DBH activity (units/mg SP)
PH 5.3” * 0.1 5.0 6.0 6.0 8.0
50 mM 50 mM 50 mM mM
0.04 0.54 0.50 0.52 0.03
(n)
k 0.06 * 0.04 F 0.23 f 0.08 +_0.01
(4) (6) (15) (4) (3)
Values are means *SD of the activity in the lysate. I unit = I ymol,‘min. SP = soluble protein. Number of preparations (n). “pH determined in dlalysed suspension.
percentages of solubilized enzyme protein, 41 and 45% at pH 5 and 8 (Table 2B), were within the range obtained by dilution (Fig. 1). However, the pH difference in solubilization by dialysis was not statistically significant. Lysis of CG by dialysis in H,O (final pH 5.3) resulted, on the other hand, in low recoveries of immunoreactive enzyme and total protein (Table 2A), particularly of the soluble fractions (Table 2B). As shown in Table 3, the loss in total protein could be prevented by dialysis in presence of inhibitors of
Table
2. Distribution membrane
the serine and thiol proteases during dialysis. In the absence of inhibitors about half of the soluble protein was lost, without significant decline in membrane protein. Inactivation of DBH by H,O dialysis (Table 1) was on the other hand not prevented by the inhibitors of trypsin and chymotrypsin-like proteolysis (Table 3). In the buffer-dialysed preparations the total and soluble immunoreactive enzyme accounted for 5-6 and 34% by weight of the total and soluble protein respectively, without significant differences between at pH 5 and 8 (Table 2B). This is in accordance with previous estimates by different techniques (Winkler, 1976; Helle et al., 1978). The buffer-dialysed CG revealed on the other hand a marked pH difference in extraction of total protein (Table 2B). Hence, these results revealed that extraction of the enzyme could be dissociated from that of the total protein by a change in pH, leading to large differences (6 and 16%) in the concentration of enzyme protein in the membrane phase without a change in ratio of soluble and membrane-bound DBH. All but 5% of the CG protein could be solubilized in aqueous solutions in the presence of a neutral detergent (Table 2B). Immunoreactive enzyme was
of immunoreactive DBH (DBHI) and total protein in soluble fractions of chromaffin granules lysed by extensive dialysis Dialysis
~~~.._
Constituents (A) In 7; ofrecowr.v af pH 8.0: Total protein Total DBHI (DBHI,) (B)
In
643
and
conditions
in H,O (a)
in Na-sac&ate (PH 5.0) (b)
80 +_ 5 65 f 3”
113 f 15 93 f I
in Tris-HCI (PH 8.0) (c) 100 100
mg/IOOmg: 4.8 8 I .4 6.5 30
DBHIJTP sDBHI/DBHI, sDBHI/SP mDBH/MP SP/TP RPjTP rDBHI/RP
f + * k f
0.2” 2“ 0.5’ 0.5” 3” 5 2
5.2 * I.9 41+3 4.2 i 0.7 6.5 f 1.4” 51 f5” 5 2
6.1 f 45* 3.4 f 16.0 k go+
0.2 1 0.2 I.0 1 4 2
SP = soluble protein; MP = membrane protein. RP = residual protein and rDBH1 = mDBH1 in RP. i.e. remaining in the high speed pellet after extraction of the membrane fraction with 2:, Triton X-100 (v/v) in Tris-glycine, pH 8.7 (see “Methods”). Values are means k SD of 3 preparations. Each preparation was divided in three aliquots and the concentrated suspensions were lysed by dialysis: (a) in absence of bulk ions or(b) and (c) at I = 0.10-O. 15 for 48 hr at f4’C with three changes of 500 vol of medium. “I’ < 0.01 for differences from the values obtained for granules dialysed at pH 8.0.
Table 3. Effect of protease inhibitors on recoveries of protein and DBH activity after H,O dialysis of the chromaffin granules DBH activity (pmoliw) Granule
fractions
Soluble lysate Membranes Total Lvsate/total
6)
(-) 0.04 k 0.06 0.29kO.15 0.11 f0.10 (4)
(+) 0.07 + 0.07 0.17f0.14 0.12+0.10 (4)
Recoveries in % (protein) (-) 55 f 7” 91*11 13 * 7” 37 f 5” (4)
(+) 100 100 100 52 + 2 (4)
Heavy granules (LG-II, CC?) were suspended in H,O in the absence ( - ) or presence ( + ) of I mM p-mercuribenzoate, 1 mM phenylmethylsulfonylfluoride and 100 KIU Trasylol. The suspensions were dialysed against H,O + inhibitors (0.01 mM p-mercuribenzoate, 0.01 mM phenylsulfonyltluoride, 1 KIU Trasylol) in the cold (3 changes of 200 vol). After dialysis the lysate pH was 5.3 i 0.1. Number of preparations (n). “P < 0.01.
644
K. B. HELLE etal.
present in this residual fraction and accounted for about 2% of the residual protein after dialysis against H,O and in buffers at pH 5 and 8. Molecular heterogeneity. The effects of the dialysis procedure on the relative distribution of hydrophilic and hydrophobic forms of the enzyme were examined by two-dimensional crossed electrophoresis. In accordance with the method (Bjerrum et al., 1979) the cationic detergent, CTAB, was present in addition to the neutral detergent (Triton X-100) in order to reveal minor differences in hydrophobicity. Only one, continuous immunoprecipitate was obtained for each fraction (Figs 2-5) consistent with previous reports on the specificity of the antibody and the immunological identity between the hydrophilic and hydrophobic forms of the enzyme (Helle et al., 1978, 1982a; Skotland and Flatmark, 1979; Slater et al., 1981; Bjerrum et al., 1979). The buffer-dialysed preparations revealed the hydrophilic enzyme as symmetric precipitate peaks (Fig. 3, lysate), with slightly faster relative mobility for the inactive enzyme (pH 8) than for the enzymatically active lysate (pH 5). The membrane fractions were extracted (TX-sol) and the residual membrane protein (TX-r) resuspended under identical conditions with Triton X-100 in the Tris-glycine buffer at pH 8.7. Any difference in the immunoelectrophoresis patterns between the TX-sol and Tx-r fractions should thus refer to earlier preparatory events. As shown in Figs 2 and 3 a peak corresponding to the hydrophilic enzyme was conspicuous in the patterns of the HzO- and buffer-dialysed membrane fractions. Marked differences were on the other hand apparent in the distribution of DBH protein in peaks 1, 2 and 3 in the H,O- and buffer dialysed membranes. The amphiphihc DBH (peak 3, Fig. 2A) was inconspicuous in the membranes of CG dialysed at pH 5 and 8 and apparent only as a shoulder at the catodic side of peak 2 (Fig. 3, TX-sol and TX-r). Membrane DBH corresponding to the intermediate hydrophobic DBH (peak 2, Fig. 2A) was on the other hand abundant, notably after dialysis at pH 8. Thus, these immunoelectrophoretic patterns indicated that the molecular heterogeneity of DBH in the CG membranes was modified by the dialysis conditions. After dialysis in pH 8 the intermediate hydrophobic peak was the dominating membrane DBH. A shift towards the hydrophilic peak was on the other hand characteristic of the membrane DBH after dialysis of CG in Hz0 in absence of protease inhibitors.
Comparison of DBH in the microsomal, heavy granule fractions
light and
Solubilization. The preparations were lysed by dilution in pH 5 and 8 buffers in the presence of the three protease inhibitors. As shown in Table 4 the microsomes, the light granules and the membranes of CG were dominated by the buffer-insoluble forms of DBH (78-88x of the total enzyme). The percentages of buffer-insoluble total protein were lower than the enzyme protein in the microsomes and CG membranes. These two fractions contained similar percentages of residual protein. The light granule frac-
tion was on the other hand lower in buffer-soluble and residual protein than the two other fractions. Molecular heterogeneity. The patterns obtained under identical electrophoretic conditions were photographically superimposed to reveal differences in relative mobilities between the fractions. Figure 4 compares the patterns of the total microsomes and the total CG. In accordance with the quantitations in Table 4 the microsomal fraction (Fig. 4, upper pattern) was abundant in hydrophobic DBH, in contrast to CG (Fig. 4, lower pattern) being dominated by the hydrophilic enzyme. As shown in Fig. 4 the relative mobilities of the microsomal peaks were lower than those of CG. The membrane fractions of the microsomes (Fig. 5A, lower pattern) and the light granules (Fig. 5B, lower pattern) were compared with the partially proteolysed, H,O-dialysed CG membranes (Figs 5A and 5B, upper patterns). After dilution in presence of protease inhibitors the microsomal membrane fraction was virtually freed of the hydrophilic DBH (Fig. 5A) and was characterized by the intermediate enzyme peak. Amphiphilic DBH was indicated in the microsomal membranes from the asymmetry on the catodic side of peak 2. Dilution of the light granules in presence of protease inhibitors resulted in membranes with hydrophilic (Fig. 5B) as well as the intermediate DBH. Judging from the apparent symmetry in peak 2 the light granule membranes were devoid of the amphiphilic enzyme. Effects of protease activation
on DBH
Activation of endogenous protease activity was achieved by incubation for 2 hr at 37°C before addition of inhibitors. Preparations of microsomes, light granules and CG membranes were compared with respect to solubilization of the enzyme and total protein after activation at pH 5-8. Sofubilization. The microsomal enzyme protein was not significantly changed by incubation at pH 558 (Table 5) although a substantial loss in total and membrane protein was apparent after 2 hr at pH 5. The membranes of CG was on the other hand significantly lower in total enzyme protein than in total protein at pH 68. At pH 5 an apparently parallel degradation of the total and membrane protein had taken place. The light granule enzyme protein was resistant to proteolytic degradation at pH 6-8 (Table 5). At pH 5 the degradation of the total enzyme protein was proportional to the protein concentration during incubation. The declines in the membrane enzyme were larger than in the total enzyme. As shown in Table 6, incubation of the light granules a pH 5 caused significant increases in soluble DBH and Folin protein. However, a loss of solubihzed enzyme occurred in the more concentrated preparations. Molecular heterogeneity. The immunoelectrophoretic pattern of the light granule membrane fraction (batch III, Tables 5 and 6) is given in Fig. 5C (lower pattern). The intermediate DBH, conspicuous in the control (Fig. 5B, lower pattern), was no longer apparent after the 2 hr incubation at pH 5 (Fig. 5C, lower pattern). The remaining enzyme corresponded in mobility to the hydrophilic enzyme (peak 1). The post-microsomal fraction (Fig. SC, upper pattern),
Fig. 2. Crossed immunoelectrophoresis patterns of membrane DBH in H,O-dialysed chromathn granules. (A) The Triton X-IOO-soluble (TX-sol) and (B) the -insoluble, resuspended fraction of residual protein (TX-r). The patterns of (A) (11 pg) and (B) (9 JL~Fohn protein) were obtained in parallel runs in the same electrophoresis experiment for comparison of relative mobilities of the peaks, the bars representing the mobility of the soluble DBH in a parallel run. The peaks are numbered according to previous definitions for DBH in the chromaffin granules run in the presence of cationic detergent in charge-shift immunoelectrophoresis (see Bjerrum et al., 1979).
Fig. 3. Crossed immunoelectrophoresis patterns of soluble and membrane DBH in buffer-dialysed chromaffin granules. The granules were lysed by dialysis against buffer, 50mM Na-succinate, pH 5 or 50 mM Tri-HC1, pH 8. The lysates, the Triton-soluble (TX-sol) and residual membrane protein (TX-r) were run in presence of cationic detergent and Triton X-100 (horizontal direction) for separation of hydrophilic and hydrophobic components (see Bjerrum et al., 1979). Bars indicate relative mobility of the soluble DBH (peak 1, lysate pH 5) run in parallel in the same electrophoresis experiment. The protein concentrations of the samples were: lysates, 14~g; TX-sol, 11 pg and Tx-r 9pg respectively.
Fig. 4. Comparison of total DBH in microsomal and chromaffin granules by crossed immunoeiectrophoresis. The immunopr~pitate patterns were obtained under identical electrophoresis conditions and photographically superimposed for comparison of relative mobilities. The total, Triton X-100 solubilized microsomes (upper pattern, 42 pep Folin protein) and chromaffin granules (lower pattern, 13pg Folin protein) were run in the presence of cationic detergent in the Triton X-100 containing gel in the first dimension (ho~zontal) el~trophoresis. Bar indicate relative mobility of the soluble DBH (peak I) of the lysate, pH 5 in a parallel run in the same electrophoresis experiment.
Fig. 5. Comparison of membrane DBH in microsomal, light and heavy (chromaffin) granule fractions. The immunoprecipitate patterns were obtained under identical electrophoresis conditions and photo~aphical~y superimposed for comparison of relative mobihties of the peaks. The samples were (A) and (B) the membranes of the H,O-dialysed chromaffin granules (upper patterns, 12 fig Folin protein); in (A) the microsomal membrane DBH (lower pattern, 16 pugprotein); in (B) and (C) the membranes of the light granule fraction (lower patterns; (B) 11 pg Folin protein before and (C) 5 fig Folin protein after 2 hr activation of proteolysis (see “Methods”). The soluble DBH in the post-mi~rosomal fraction (S,} is given in (C) (upper pattern, 94 pg Folin protein).
646
647
Membrane dopamine /I-hydroxylase Table
4. Membrane
DBHI
and protein
in adrenomedullary inhibitors
Buffer-insoluble Soluble mDBH1
Fraction
fractions
components
in 2% Triton
in the presence
of protease
in “/, of total in fraction
X-100 Folin protein
Residual fraction” Folin protein
Microsomes (P,)
88 * 5
64+
IO
Light granules (LG-I)
78 +8
87 k 5
2+0
Heavy granule membranes (LG-II,; CC,)
82 f 8
53 * I4
II *2
1212
Four preparations of each fraction were diluted IO times in ice cold buffer containing protease inhibitors (see “Methods”). The high speed supernatants and buffer-insoluble pellets, solubilized in 2% Triton X-100, were assayed for DBHI by nephelometry and Folin protein. “The residual fraction, resuspended in 2% Triton was not free of light scattering particles and could therefore only be assayed for protein.
assumed to represent soluble DBH from fragmented organelles (Winkler, 1977; Ledbetter et al., 1978) revealed a symmetric peak of hydrophilic DBH, with slightly faster mobility than peak 1 in the proteolysed light granule membranes. DISCUSSION
The present
experiments
have made it evident
Table 5 Effect of activation
that
proteolytic activities are present in the purified preparations of CG and contribute to DBH solubihzation and degradation when activated during lysis. This conclusion imposes a number of questions related to previous quantitations of the soluble and membrane components and to the current concepts of the soluble and membrane DBH as stable and well defined proteins.
of endogenous proteases on the total and membrane in membrane fractions of the bovine adrenal medulla Recoveries
after 2 hr at 37°C in “/, of unincubated
in total fraction Fraction Microsomes (P,) pH 6.5-8.0” pH 5.0 Heavy granules, membranes (LG - 11, = CG,) pH 6.1lX.O pH 5.0 Light granules, (LG-I) pH 6.1-8.0 pH 5.0
DBHI,
(1) (I)
91 _+6 93
(II) (II)
79 + 9 71
(III) (III)
102 * 2 102 85 +9 69 f 5
(IV)
W)
bound DBHI and protein control
in but&-insoluble
Total protein (TP)
mDBH1
86 * 2 56
fraction Membrane protein (MP)
94 f 4 84
85 k 2 49
(3) (1)
IO1 +2 61
72k 67
85+ 49
(3) (1)
90+ I3 52 Ill is III *4
955 IO 52 65 + 1 541 I
I2
I2
83 k 3 43 78 k 2 67* I
(3) (1) (3) (3)
Aliquots from each hatch incubated for 2 hr at 37°C before addition of protease inhibitors (see “Methods”). Control aliquots were suspended in ice-cold buffer containing the protease inhibitors. “Final pH during incubation. Number of incubation experiments (n) from each batch (1-V). Final protein concentration during incubation (mg/ml): I, 1.3; II, 1.5; III, 0.5; IV, 6.9; V, 8.1. Each value is the mea” of duplicates. Means f SD of values for the incubations (n > 2). Table 6. Proteolysis
dependent
increase in soluble eranules
DBHI Recovery
Component
Batch No.
mg/TP/ml”
pH5
and protein
in the light
in ;< of control pH 6-8
sDBH1
III IV V
0.5 6.9 8.1
500(l) 176 k 30 (3) 149 _+40 (3)
ll7kl1(3)
SP
III IV V
0.5 6.9 8.1
109(l) I98 F 2(3) 201 f I8 (3)
177 + 105(3)
Three different batches of LG-I were incubated for 2 hr at 37°C before addition of protease inhibitors (see “Methods”). Control aliquots were suspended in ice-cold buffer containing the protease inhibitors. “Final protein concentration during incubation. Number of incubation experiments from each batch (n). Means k SD for the incubations. Each value based on assay in duplo.
648
K. B. HELLE el 01.
(1) Does the proteolytic activation affect the solubilization and the activity of the soluble DBH during lysis? Inactivation of the H,O-solubilized enzyme was not prevented by dialysis in presence of the protease inhibitors. Nor was there any indications of proteolysis in the enzymatically inactive pH 8 lysate. The soluble DBH withstands trypsin treatment at pH 7.5 without decline in tyramine affinity and increases further in specific activity upon digestion of the trypsin-sensitive chromogranins (Helle rt cd., 1977). Hence, it is concluded that proteolysis per se is not a likely reason for the low specific activity of the soluble DBH after H,O dialysis. Our experiments have shown that proteolytic activities degrade the soluble DBH as well as the chromogranins during H,O-dialysis of CC but leaves the membrane DBH relatively intact. Dialysis in buffers, leading to the same degree of DBH solubilization (4145%) into the first high speed supernatant as dilution (-42x), does not activate proteolysis. This shows that proteolytic loss of the soluble DBH during lysis by H,O dialysis is an exception which does not apply to buffer-dialysis. (2) Are there any eflects of mdogenous proteases on solubilization of DBH ,from any’ of the other grunule fractionsV Most of the microsomal and light granule DBH was insoluble in buffer and less sensitive to proteolysis than the total protein. Incubation for 2 hr near neutral pH (6-8) had insignificant effects on the solubilization of microsomal or light granule DBH. Proteolysis-related changes in solubilization of DBH in the light and heavy granule fractions were on the other hand apparent at pH 5, being -5 times more potent in the CG membranes than in the light granule fraction. These results indicate that DBH-converting and degrading proteases with acid optima were insignificant in the microsomal fraction while seemingly concentrated in the membranes of CG. (3) Is there a specific proteolytic system ,for the conversion of hydrophobic to hydrophilic DBH in any of the adrenomedullary ,fractions? A slow conversion of the hydrophobic DBH into the hydrophilic enzyme was first observed after incubation of the large granule fraction for 24 hr at pH 7.5 (Bjerrum et al., 1979; Helle et al., 1982a). The present experiments have shown that a more rapid conversion may occur in the light granule fraction, leading to considerable increases (149-500%) in soluin excess of the concentrationbilized enzyme, dependent DBH degradation at pH 5 and without a parallel degradation of the total protein. The subcellular fractionation scheme employed is closely similar to that described by Bolstad et cd. (1980) and by Ledbetter et al. (1978). By these protocols the newly formed granules and membrane ghosts of ruptured CG, cosediment with mitochondria and lvsosomes into the 0.3-1.6 M sucrose boundary. This light granule fraction is therefore assumed to be rich in the lysosomal proteases, notably in cathepsin D (Smith and Winkler, 1966). Inhibitors of cathepsin D could not be obtained for
use in the present study. A selective conversion effect by cathepsin D was on the other hand indicated from in vitro experiments with the bovine spleen enzyme. After cathepsin D treatment of CC for 30 min at pH 6 there was an increase in soluble DBH (132% of control) without concomitant degradation of the chromogranins. This effect was contrasted by that obtained with chymotrypsin which concomitant with a similar increase in the soluble enzyme also degraded the chromogranins (unpublished experiments). Hence, cathepsin D appears as a plausible candidate for the DBH converting acid protease in the lysosome-rich light granules. Thiol proteases resembling lysosomal cathepsins have been implicated in the conversion of proinsulin to insulin on the basis that they are localized in the secretion granule/mitochondrial fraction and that they can convert pro-insulin to insulin in vitro (Docherty and Steiner, 1982). Some contamination of lysosomes in the CG pellet cannot be avoided (Smith and Winkler, 1966) consistent with recoveries of I l”/, of lysosomal acid protease (Lindberg et cd., 1982) and 217; of [j-glucuronidase (Bolstad et al., 1980) in the CG preparations. More recently there have, however, been reports on specific “trypsin-like”, acid proteases in CG, cleaving pro-enkephalins within the soluble core (Lindberg et al., 1982; Evangelista et al., 1982). As limited tryptic proteolysis of the chromogranins did not affect the enzyme activity or the quantity of immunoreactive DBH (Helle et al., 1977), a “trypsinlike” degradation of the soluble DBH in CG seems rather unlikely. A conversion of the amphiphilic to the hydrophilic DBH in CG was not apparent after the trypsin treatment but was readily obtained with chymotrypsin or thermolysin (Bjerrum et al., 1979). As higher concentrations of chymotrypsin also partially degraded the solubilized enzyme at pH 6 (Helle et nl., 1982b), a “chymotrypsin”-like activity might be implicated. This assumption lends support from the observed, parallel degradation of soluble DBH and protein during H,O dialysis in the cold. (4) Is the umphiphilic or the intermediate DBH serving us the immediate precursor ,for the proteolytically solubilized enzyme? The hydrophobic and hydrophilic DBH differ in a short sequence at the C-terminus (Slater et al., 1981) and small portions of this hydrophobic “tail” which spans the membrane, may be cleaved off within the membrane phase and without loss of enzyme activity (Zaremba and Hogue-Angelletti, 1981). Complete absence of the hydrophobic sequence, as in the hydrophilic DBH, prevents incorporation into lipid bilayers (Albanesi et al., 1980). The intermediate and amphiphilic DBH peaks have both been shown to be integral membrane components, differing however in their binding of the cationic detergent (Bjerrum et al., 1979). These two forms were present in equal proportions in the total extract of CG while a shift towards the intermediate peak was apparent for the membranes The present study has revealed further shifts towards a dominance of the intermediate peak, in the pH 8 dialysed CG membranes and in the microsomes. The amphiphilic peak, seen as a shoulder at the catodic side of most membrane patterns, was virtually absent from the light granule fraction. Lysis by
Membrane dopamine /I-hydroxylase
dialysis in acid buffer or in H,O caused additional shifts in the membrane enzyme and quite unexpectedly, peak 1 was conspicuous in every membrane fraction but the microsomal and was also present in the Tx-r fractions of the dialysed CG. Marked differences in composition between the Golgi and CC membranes have been reported (Trifaro et al., 1976). The apparent difference in solubility of peak 1 from the microsomal, the light and heavy granule membranes may thus reflect such differences in the hydrophobic environment. Aiternatively, the dominant peak 1 in the H,O-dialysed CC membranes accounts for a proteolysis-related conversion of the intermediate DBH to a less hydrophobic type within the insoluble fractions during solubilization in the Triton buffer. Our results on the light granule fraction shows that the enzyme protein is shifted from the insoluble to the soluble form concomitant with loss of the intermediate peak as a result of acid proteolysis which is arrested in the presence of inhibitors to serine and thiol proteases. Thus, taken together, the present findings
provide
strong
indications
of a sequential
conversion system for DBH within the membrane compartments, the amphiphilic DBH being the source and the intermediate hydrophobic form being the immediate precursor of the hydrophilic DBH. (5) IS it possible
to pinpoint where the conversion begins
the stages in biogenesis
in vivo? Assuming that the amphiphilic peak represents the undigested precursor of the soluble DBH and taking the dominance of the intermediate peak in the membranes as evidence of partial conversion, this posttranslational proteolytic modification must be initiated at an early stage in the granule biogenesis, analogus to the processing of prohormones (Docherty and Steiner, 1982). In the rat adrenal medulla (- 10% soluble DBH; Grzanna and Coyle, 1976) the steady state level of DBH is regulated in vivo by glucocorticoid-modulated proteolysis, presumably by trypsin-like activities (Wong and Ciaranello, 1982). Whether this also holds true for the bovine adrenal DBH remains to be established. The close similarity in DBH patterns between the microsomes, the Golgi (Helle et al., 1982b) and the pH 8 dialysed CG suggests that a conversion process within the membranes, as presently postulated, may be regulated and possibly reach a state of equilibrium with the soluble fraction which increases in ratio over the membranes in the course of CG maturation. In conclusion, the present study has shown that the membrane and soluble DBH in the bovine adrenal medulla are susceptible to attack by several proteolytic activities inherent in the adrenomedullary fractions and this calls strongly for control of proteolysis in studies of the bovine adrenal DBH. Furthermore, the present work provides substantial support for the “conversion hypothesis” and may serve as a basis for further studies of proteolytic activities involved in a sequential conversion of the membrane DBH in granule biogenesis and during secretion and subsequent retrieval of the emptied membranes. ~ckno~vle~~em~~~s-Thanks are due to the Norwegian Research Council for Science and the Humanities for financial
649
support (C.13.11.10-012) and to Dr T. Skotland, Department of Biochemistry, University of Bergen, Norway, for the gift of highly purified soluble DBH of bovine adrenomedullary origin. REFERENCES
Albanesi J. P., Robinson E. C. and Kirshner N. (1980) ~ri~cation and analysis of membrane-bound dopamine /I-hydroxylase. Fedn kroc. 39, 1920. Aunis D.. Bouclier M.. Pescheloche M. and Mandel P. (I 977) Properties of membrane-bound dopamine P-hydroxylase in chromaffin granules from bovine adrenal medulla. J. Neurochem. 29, 439447. Bjerrum 0. J., Helle K. B. and Bock E. (1979) Immunochemically identical hydrophilic and amphiphilic forms of the bovine adreno-meduliary dopamine j-hydroxylase. Biochem. J. 181, 231-237. Blakeborough P., Louis C. F. and Turner A. J. (1981) The structure and organization of dopamine p-hydroxylase in the chromaffin granule membrane. Biochim. biophys. Acta 669, 33-38.
Blobel G. and Dobberstein B. (1975) Transfer of protein across membranes. I. Presence of proteolytically processed and unprocessed nascent immunoglobulin light chains on membrane-bound ribosomes of murine myeloma. J. Cefi. Biol. 67, 835-851. Bolstad G., Helle K. B. and Serck-Hanssen G. (1980) Heterogeneity in the adrenomedullary storage of catecholamines, ATP, calcium and releasable dopamine /I-hydroxylase activity. J. Autonom. Nero. Syst. 2, 337-354.
Docherty K. and Steiner D. F. (1982) Post-translational proteolysis in polypeptide hormone biosynthesis. A. Rev. Physiol. 44, 625-638.
Evangelista R., Ray P. and Lewis R. V. (1982) A “Trypsinlike” enzyme in adrenal chromaffin granules: A proenkephalin processing enzyme. Biochem. biophys. Res. Commu~. 106, 895-902. Gagnon C., Schwartz R., Otten U. and Thoenen H. (1976) Synthesis, s&cellular distribution and turnover of dopamine /I-hydroxylase in organ culture of sympathetic ganglia and adrenal medulla. J. Neurochem. 27, 1083-1089.
Grzanna R. and Coyle J. T. (1976) Rat adrenal dopamine /I-hvdroxvlase: purification and immunological characieristics. j. Neuiochem. 27, 1091-1096. Helle K. B.. Serck-Hanssen G. and Reichelt K. L. (1977) Resistance of dopamine /?-hydroxylase to proteolysis during tryptic digestion of the adrenomedullary chromogranins. ht. J. Biochem. 8, 693-704. Helle K. B., Serck-Hanssen G. and Bock E. (1978) Complexes of chromogranin and dopamine P-hydroxylase among the chromogranins of the bovine adrenal medulla. Biochim. biophys. Acta 533, 396-407.
Helle K. B., Bjerrum 0. J. and Bock E. (1982a) Immunochemical characte~~ation of dopamine ~-hydroxylase. &and. J. Irnrn~~. 15, 97-123. Helle K. B., Pihl K. E. and Serch-Hanssen G. (1982b) Hydrophililic and amphiphilic forms of dopamine fl-hydroxylase in early and mature stages of the bovine adrenomedullary granules. In Synthesis, Storage and Secretion of Adrenal Catcholamines (Edited by Izumi F. et al.). Ad;. Biosci. 36, 217-224.
_
Kiinig P., Hsrtnagl H., Kostron H., Sapinski H. and Winkler H. (1976) The arrangement of dopamine P-hydroxylase (EC 1.14.2.1) and chromomembrin B in the membrane of chromaffin granules. J. Neurochem. 27, 15391541. Ledbetter F. H., Kilpatrick D., Sage H. L. and Kirshner N. (1978) Synthesis of chromo~anins and dopamine ~-hydroxylase by perfused bovine adrenal glands. Am. J. Physioi. 235, E475-E486.
650
K. B. HELLE ef al.
Lindberg I., Yang H.-Y. T. and Costa E. (1982) An enkephalin-generating enzyme in bovine adrenal biophys. Res. Commun. 106, medulla. Biochem. 186193. Miras-Portugal M. T., Jorda A. and Santos Ruiz A. (1976) Propietes cinetiques de la dopamine j?-hydroxylase membranaire. FEBS Let!. 72, 267-270. Serck-Hanssen G., Helle K. B., Sommersten B. and Pihl K. E. (1980) Quantitative aspects of the acetylcholineinduced release of dopamine b-hydroxylase and catecholamines from the bovine adrenal medulla. Gen. Pharmat. 11, 243-249. Skotland T. and Flatmark T. (1979) On the amohiohilic . _ _ and hydrophilic forms of dopamine b-monooxygenase in bovine adrenal medulla. .I. Neurochem. 32, 1861-1863. Slater E. P., Zaremba S. and Houge-Angeletti R. (1981) Purification of membrane-bound dopamine fl-monooxgenase from chromaffin granules. Relation to soluble dopamine /?-monooxygenase. Archs Biochem. Biophys. 21!, ?88-296. Smith A. D. and Winkler H. (1966) The localization of
lysosomal enzymes in chromaffin tissue. J. Physiol. 183, 1799188. Steiner D. F., Kemler W., Tager H. S. and Peterson J. D. (1974) Proteolytic processing in the biosynthesis of insulin and other proteins. Fedn Proc. 33, 2105-2115. Trifaro J. M., Duerr A. C. and Pinto J. E. B. (1976) Membranes of the adrenal medulla: a comparison between the membranes of the Golgi apparatus and chromaffin granules. Molec. Pharmac. 12, 536-545. Winkler H. (1976) The composition of adrenal chromaffin granules: an assessment of controversial results. Neuroscience 1, 65-80. Winkler H. (1977) The biogenesis of adrenal chromaffin granules. Neuroscience 2, 657-683. Wong D. L. and Ciaranello R. D. (1982) Regulation of synthesis and degradation of catecholamine synthetizing enzymes in the adrenal medulla. In Synthesis, S/oruge und Secretion of Adrenal Catecholamines (Edited by Izumi F. et al.). Ado. Biosci. 36, 233-241. Zaremba S. and Hogue-Angeletti R. (1981) Transmembrane nature of chromaffin granule dopamine /J-monooxygenase. J. biol. Chem. 256, 1231&-12315.