and functions as a transmembrane electron transporter (Apps et al., 1980). It is thought to supply reducing equivalents to the intravesicular mono-oxygenase ...
The EMBO Journal vol.7 no.9 pp.2697-2703, 1988
The structure of cytochrome b56 1, specific electron transport protein
Mark S.Perin', Victor A.Fried3, Clive
A.Slaughter' and Thomas C.Suidhof' 2 'Howard Hughes Medical Institute and 2Department of Molecular Genetics, University of Texas, Southwestern Medical Center, Dallas, TX 75235 and 3St Jude Children's Research Hospital, Department of Biochemistry, Memphis, TN 38101, USA Communicated by P.De Camilli
Cytochrome b561 is a transmembrane electron transport protein that is specific to a subset of secretory vesicles containing catecholamines and amidated peptides. This protein is thought to supply reducing equivalents to the intravesicular enzymes dopamine-3-hydroxylase and apeptide amidase. We have purified cytochrome b561 from bovine adrenal chromaffin granules by reverse phase chromatography and have determined internal amino acid sequences from peptides. Complementary oligonucleotides were used to isolate two cDNA clones from a bovine brain library. The structure predicted by the sequences of these cDNAs suggests a highly hydrophobic protein of 273 amino acids which spans the membrane six times with little extramembranous sequence. Cytochrome b561 is not homologous to any other cytochrome and thus represents a new class of electron carriers. RNA blotting experiments indicate that cytochrome b561 is expressed in the adrenal medulla and all brain regions of the cow, but not in visceral organs. This result agrees well with the putative function of this unique cytochrome and with the notion that this protein is localized to large dense-core synaptic vesicles. Key words: chromaffin granule/cytochrome b561/electron transport/secretion/synaptic vesicle
a
secretory vesicle-
reduction to ascorbic acid it functions as cofactor for dopamine-,B-hydroxylase, the enzyme that hydroxylates dopamine to norepinephrine (Njus et al., 1983; Srivastava et al., 1984; Wakefield et al., 1986a,b). This mechanism explains the presence of high intravesicular ascorbic acid concentrations in chromaffin cells. In vivo, electron flow is unidirectional from the cytoplasm into the vesicles in order to sustain mono-oxygenase activity, while in vitro, reconstituted purified cytochrome b561 shows no directionality in the electron transport across the membrane (Kent and Fleming, 1987). Therefore, cytochrome b561 appears to be a comparatively simple, functionally symmetric electron transport protein. The central role played by cytochrome b561 in dopamine,(-hydroxylase activity, suggests that this protein is present in the secretory vesicles of all neurons containing adrenaline or noradrenaline as neurotransmitters. In addition, cytochrome b561 should be present in cells that secrete neuropeptides that are a-amidated by an intravesicular monooxygenase called peptide a-amidase which a-amidates peptides such as vasopressin, oxytocin and a-MSH (Bradbury et al., 1982; Murthy et al., 1986). This enzyme has recently been shown to require ascorbic acid for reduction, and evidence has been presented that vesicular cytochrome b561 may also be necessary for co-amidase activity (Russell et al., 1985; Kent and Fleming, 1987). Cytochrome b561 and ascorbic acid, accordingly, play a central role in the biosynthesis of several catecholamine and peptide neurotransmitters, which belong to the functionally slow, 'neuromodulatory' class of neurotransmitters. We have set out to study cytochrome b561 as a model transmembrane electron transport system and as a marker for a subset of large dense-core synaptic vesicles. Here, we report the purification and molecular cloning of cytochrome b561, and present a model for its transmembrane structure.
Introduction Cytochrome b561 was first described as an integral
Results
membrane protein in chromaffin granules, which are the secretory vesicles of chromaffm cells (Flatmark and Terland, 1971; Hortnagl et al., 1973; Terland et al., 1974). The protein comprises 15 % of the granule membrane proteins and functions as a transmembrane electron transporter (Apps et al., 1980). It is thought to supply reducing equivalents to the intravesicular mono-oxygenase dopamine-13-hydroxylase (Njus et al., 1983). Among the cytochromes, cytochrome b561 is unique since it is localized to a subset of secretory vesicles and is expressed only in peptide or catecholamine secreting cells. Cytochrome b561 also appears to be unusual in its enzymatic mechanism. Unlike other protein electron carriers, it does not directly interact with any other protein for its oxidation-reduction cycle. Instead, ascorbic acid serves as the extravesicular electron donor and semidehydroascorbic acid as the intravesicular electron acceptor, where after
Cytochrome b561 was isolated from purified chromaffin granule membranes in a single step by reverse phase chromatography in formic acid (Figure 1). Compared to the established procedure of chromatofocusing for the purification of cytochrome b561 (Wakefield et al., 1984), reverse phase chromatography allowed rapid generation of sufficient amounts of pure protein to enable us to sequence this extremely hydrophobic protein (see below). The protein is denatured by this scheme and tends to aggregate into multimers after treatment with organic solvents but not when directly solubilized with detergents (Figure 1, insert). The identity of the purified cytochrome b561 was confirmed by its cross-reactivity with a monoclonal antibody against the bovine protein (Pruss, 1987). In agreement with other reports (Duong and Fleming, 1982), the amino terminus of the protein was found to be blocked independently of the purification method used.
©IRL Press Limited, Oxford, England
2697
M.S.Perin et al.
-i40 1
3
/2
4
-130 0
a
0.2
20 c, 0 U) /i
._
__
0--
0
CO\01
110 i
i.
10 mi
i
1_
20 grad ient
i AM.
i
I
30
0
Fig. 1. Purification of cytochrome b561 by reverse phase chromatography. Stripped chromaffin granule membranes solubilized in 60% formic acid loaded onto a Pharmacia ProRPC column and eluted with a 0-40% isopropanol gradient. The position where cytochrome b561 elutes as the last major peak is indicated by the arrow. The insert shows a silver-stained 12% SDS-polyacrylamide gel of cytochrome b561 purified by this method (lane 2) or by chromatofocusing (lane 3) and of the starting chromaffin granule membrane fraction (lane 4). Cytochrome b561 purified by reverse phase chromatography strongly aggregates after removal of formic acid (lane 2). Lane 1 shows mol. wt standards (rabbit muscle phosphorylase [97 400], bovine serum albumin [66 200], ovalbumin [42 700], bovine carbonic anhydrase [31 000], soybean trypsin inhibitor [21 500] and lysozyme [14 400]).
were
Limited internal amino acid sequences were therefore obtained from the purified protein after cleavage with trypsin and cyanogen bromide (CNBr) and used to design complementary oligonucleotides. Screening of a bovine brain cDNA library with these probes resulted in the isolation of two overlapping clones containing cDNA inserts of 2.10 and 2.05 kb. The nucleotide sequences of these cDNAs were determined as depicted in Figure 2A and their complete sequence is shown in Figure 2B. One of the clones (pCyt6a) appears to represent a full length copy of the cytochrome b561 as judged by the presence of an in-frame stop codon in the 5' untranslated sequence. This clone contained a much shorter 3' untranslated region than the second clone (pCytlb), which was determined not to be full length. The nucleotide sequences of the two clones are identical in the region of overlap. The 3' end of pCyt6a has a poly(A) tail after a canonical polyadenylation signal (AATAAA), while pCytlb's sequence continues 562 nt beyond this point before ending in a poly(A) tract (Figure 2B). Translation of the nucleotide sequence reveals the presence of a single open reading frame (ORF) of 273 amino acids, encoding a protein of 30 061 daltons mol. wt. This value is within the range of published size estimates of cytochrome b561 by SDS -PAGE or amino acid analysis (Duong and Fleming, 1982). The coding region is followed by a very long 3' untranslated region of 1692 bp. 2698
All the amino acid sequences that we experimentally determined from the purified protein are contained in the predicted ORF (underlined in Figure 2B, peptides P1-5). Computer-assisted searches of nucleic acid and protein sequence databanks showed no significant homology to any sequence contained in these banks and, in particular, no similarity to known cytochrome structures. At the amino terminus, the reading frame contains two potential initiator methionines. The sequence context of the second methionine codon agrees more closely with the consensus sequences for initiator methionines than does that of the first methionine (Kozak, 1987). However, the first methionine is usually used in eukaryotic messages. Furthermore, although the N terminus of the cytochrome was blocked, a peptide sequence beginning at the second methionine was found after CNBrcleavage of the protein (P1 in Figure 2B). This result strongly suggests that the second methionine is contained within the mature protein and that the first methionine codon is used for initiation. The amino acid sequence of cytochrome b561 as translated from the nucleotide sequence predicts a very hydrophobic protein. Plotting of the average residue hydrophobicity as a function of residue number reveals the presence of six internal regions of high average hydrophobicity (Figure 3). These hydrophobic sequences are separated by hydrophilic sequences that frequently contain clustered positive charges.
Secretory
A
SacIl
pCyt6a
SacI
XhoI
vesicle electron
transport protein
SailI
SacI
An .
Sac I
pCyt1b 1
200
0
400
800
600
SacI
XhoI
1000
1400
1200
1800
1600
SacI
Sal I I
2000
2400
2200
B
An
2600
p
*
TCTCTGTATTTCACCCATGAACCCGACCTGAGAAAACCCAGGGCCTCGTGAAACAAAGCACTGCCGTGGGGCCCTCCTGCTCTCCTCAGAATGTGGAGGACCTTGACTTGCACC M
W
R
T
L
T
C
T
CTCACTGCAGGTGGAAGTGCAGCCAGTAGCCGCCTCAGCATGGAGGGCCCTGCCAGCCCAGCACGTGCCCCCGGGGCGCTGCCCTACTACGTGGCCTTCTCCCAGCTGCTGGGC L T A G G S A A S S R L S M E G P A S P A R A P G A L P Y Y V A F S Q L L G -
P1
-
114 8
228 46
-
342 84
CTGATCGTGGTGGCCATGACTGGCGCCTGGCTTGGTATGTACCGAGGCGGCATTGCCTGGGAGAGTGCCCTGCAGTTCAACGTGCACCCCCTCTGCATGATCATAGGCCTGGTC L I V V A M T G A W L G M Y R G G I A W E S A L Q F N V H P L C M I I G L V
TTCCTGCAGGGAGATGCCCTGCTGGTTTACCGCGTCTTCAGGAATGAGGCCAAACGCACCACCAAAGTCCTGCACGGGCTGCTGCACGTCTTCGCCTTCGTCATCGCCTTGGI'G 456 F
L
Q
G
D
A
L
L
V
Y
R
F
V
R
E
N
A
K
R
T
T
K
L
V
H
G
L
L
H
V
F
A
F
I
V
A
L
V
122
GGCCTGGTGGCGGTGTTCGAGCACCACAGGAAGAAGGGCTACGCCGACCTGTACAGCCTGCACAGCTGGTGCGGCATCCTGGTCTTCGCGCTCTTCTTTGCGCAGTGGCTCGTG G L V A V F E H H R K K G Y A D L Y S L H S W C G I L V F A L F F A Q W L V
570 160
GGCTTTAGCTTCTTCCTGTTCCCGGGCGCCTCGTTTTCTCTGCGGAGCCGCTACCGCCCGCAGCA%-IGTCTTCTTCGGCGCCGCCATCTTCCTGCTCTCGGTGGCCACCGCCCTG
684 198
G
F
S
F
F
L
F
P
G
A
S
F
S
L
R
S
R
Y
R
P
Q
H
F
V
F
G
A
A
I
F
L
L
S
V
A
T
A
L
CTGGGCCTGAAGGAGGCGCTGCTGTTCGAGCTCGGGACCA7GTACAGCATGTTCGAGCCCGAGGGCGTCCTGGCCAACGTGCTGGGCCTGCTGCTGGCCACCTTTGCCACGGTC798 V
236
ATAOTCTACATCCTGACCCGCGCCGACTGGAAGCGGCCCCTCCAGGCGGAAGAGCAAGCGCTCTCCATGGACTTCAAGACGCTGACGGAGGGCGACAGCCCCAGCTCCCAGTGA I L Y I L T R A D W K R P L Q A E E Q A L S M D F K T L T E G D S P S S Q *
273
L
G
L
K
E
A
L
L
F
E
L
T
G
K
Y
S
M
F
E
P
E
G
L
V
1
1
1f
A
N
V
L
G
L
L
L P2
A
T
F
A
T
912
It4p3 p5 I, P 1
CGGGCGCCCCATCCTGGCCCCTCCTGTTCCGGGGGCCCAGGCAGGTGTCCGTCCCGCAACCGGCAACCGCCGCTGAGGGTCCGGGCAGGGCTGGGCACAGCGGAGGTAGGGGGG 1026 AAGTCCTCTCTGAGCGCAGATTTTTGGGATTCAGGCCCCCTCCGCTTCCCGCTCCTCGCTGCTGCTGCTGTCGGTTGTTAGTCCTGAGCACGCGTTCCGGCCCCCCCGCCCCTG 1140 TTGCCACCCGACTTCCTCCCCAAGATGGAGCAGCTTTGGGAAAGGCTCTGGGCCTAAATGGGTACGAGAGTTGGGGGCGCTCCCTTCTCGAGGCCGCAGCGAACGTGAGACTTG 1254
CTTTGTGTGACCCCGAACTCCTGCCAGGAGTCTTGCGGCCCGAAGGACCGGCTCTCTGAAAGGGCCCAGCCTTTGGCAGCGGAGCTAGTGACATTATATGTGAAATATGCTGGT 1368 GTCAGGGCCAGGTTCCCCAGGAGAGGGGAGGGACTCCCTGGGACTCTCTGGACCCAGGCTTTGGCACAGCTGCGTTCTGGAGCCTTCCCTTTCTTTAACCCTTTAGCCCCGAGG 1482
CGGCTTTGCCTTGGGGGCCGTAGACTTGGGACAGACTTGCCGTGAGCCGCTTCAGAGGGCAGCCCAGAGAAGAGGCTGGAGGTGGGCCTGGGTGGGCGCTGGGCCACCTGCAGG 1596 GGCTTCTGTCCCCACTGGGCCAGGCCTCCGCTCCTGACCCTCAGAGCTCACGGGAGGCTGCTCTGACAAGAGGGCACAGGCCTTTTCTCATTAAACAAAAGGCTGCCACTAGAA 1710
AAGGATAAGGTGCCCCCTTTCGATTTCCATGAGGAGGGTTCGGGAGGACAGTCTCCTCTCCTACCTGCGGCCGGTCTTTGCAGGAAATTGGGACCAAGACGGAATCAGATTAAT 1824 CCCCCCACAACCCTGTGAGTGTGTAACGCTGTTAGCTGAGTGCTGTTTGCTTCGGTCTAGAGATAACGTGATTAGAGCATTGATCTGCGCTTCTTGGACTTGGCAAATGCTCTC 1938
TCCTTGTTTTCCTCATTTTCCGGGCTTTGAGGAACATGACGCAGTTTCAGACAAGACAGAGTCGACTTAAGACATTAATAAAATTTCCAGTTCTGACTACCGTTTTTCAGCAGC 2052 TGCCTCTCCCAGCTAAATGCTGCCTTCGAGAAAAGGGGCTCATCTCCTCTAATACTCGGATTCCTACAACTCGGCCTTTGCCGTGGGTCCCGCCTCAGGCGATGCTGCTCCTCA 2166
GGGCAGGTGGCGGAGGTGGGCGCCCTCCCCAGCAAGGCTGCCTGCTCCCCATCCCACCCCGTGGCCGCTCTGAGCTCTGCTGCGGCTGGCCACCCCCGCACACCACCGCAAGCC 2280
GTTTACACTGTTCTTTGATCGCCTGGTATAGTTGATGTTTACACGTGGGAAACTCCACCTTAGACAAACGACCAGAGTAGGATTGTGTATCTCTCTTGCCAGAAGGGCACAGAG 2394
AGAAACAGAACCTTGGATTTGCAAAGAAAAAAAAAAGTCTGAACACTGGCCTGTTTGCCAAATCATATGCTGGATMACTAATAAGCCAAAAAGGCACCCCCCTCCCCACCT 2508
GCCCCAGCCAAGCTGATCCGTGGTGGCTTGTTTTGCCTGTGGCCAATCTTCGATGAAGTCCAGTGTAGTGCTGGTACAACAGATGATTCAATAAATAA
AMAAAAAAAAAAA
2622
Fig. 2. A. Restriction endonuclease maps of bovine cytochrome bS61 cDNAs and DNA sequencing strategy. B. Nucleotide and amino acid sequence of cytochrome b561. The translated amino acid sequence is shown in single letter code below the nucleotide sequence. Asterisks above the sequence denote the location of the two potential initiator methionines. The in-frame stop codon S' of the first methionine is underlined. Pepfide sequences --A determined Inderline and lable P1PSnf these, P1,2ad4wr -Aaie afte CNBr-cleavage and] IP3 and PS afte-r trypsi;n cleuavage The broken line in P1 indicates unidentified amino acid residues. Potential polyadenylation signals are underlined and the start of the poly(A) tract in pCyt6a at position 2043 is shown by the vertical arrow. pCytlb starts at nt 573. Nucleotide and amino acid sequence numbers are given on the _._
right.
2699
M.S.Perin et al.
soluble domain of the protein. A short sequence after the amino-terminal end of the protein is repeated (6/10 match) near the cytoplasmic terminus: 4-TLTAGGSAAS- 14 Residue # 263-T LT E G DS P S S-273 Some of the loops between the proposed six transmembrane regions of cytochrome b561 are very polar. However, the segment between the first and second transmembrane regions is quite hydrophobic. It is quite possible that some of the transmembrane helices are tilted or bent and that a loop may be partially embedded in the bilayer. Not enough sequence is present in the cytochrome b561 structure to accommodate more than six membrane spanning regions. The model suggests that most of the hydrophilic sequences of the cytochrome are contained in the amino and carboxy termini and that these reach into the cytoplasm. This prediction agrees very well with biochemical experiments which suggest the presence of a 3 kd protease sensitive cytoplasmic domain in cytochrome b561 (Abbs and Phillips, 1980). To determine the tissues in which cytochrome b561 is expressed, we performed RNA blotting experiments with single-stranded probes derived from the cytochrome cDNA. As shown in Figure 5, a single message of 2.6 kb was found in RNA purified from bovine adrenal medulla (lanes 10, 11 and 13) and various bovine brain regions (lanes 3-9). No signal was found with RNA from bovine liver (lane 1) or heart (lane 12) even though equivalent amounts of RNA were blotted as judged by the hybridization with an actin probe
However, no consensus sites for N-linked glycosylation (Asn-x-Ser/Thr) are observed. In Chou-Fasman predictions, the internal hydrophobic sequences are predicted to be a-helical (data not shown) and therefore they may be transmembrane regions. Neither of the two potential initiator methionines is followed by a hydrophobic sequence with the characteristics of a signal sequence (Von Heijne, 1984). A model for the transmembrane structure of cytochrome b561 based on this sequence information is shown in Figure 4. This model predicts a cytoplasmic amino terminus consistent with the lack of a signal sequence and a blocked amino terminus. There are six transmembrane regions that are separated by relatively small interconnecting loops. The carboxy terminus, again cytoplasmic, makes up the largest 3.0 2.0
6
51.0 0
%L0 1
0 -1.0 -2.0 90 120 150 180 Amino Acid Number
60
30
210 240 270
Fig. 3. Hydrophobicity plot of the cytochrome b561 sequence. Average hydrophobicities were calculated using the algorithm of Kyte and Doolittle (1982) with a window size of 15 amino acids.
I A
G ..
.
.
.
Y
.
W®S
A L Q
H
.........
M
N
l..
.
..
.
.
..
P
M
G
L
F
A
F
T
Q A
G
V
L
S
V
A
VR
V
S
P
L
L
y I
G
3T T
......
........
G
G®M
V
S
Y
SL(jSE]Y[1PQ
S u ~~~~~~S
..............
u
S
F
JA P S A
A A
A A
I
F
L
~~~~NA (
p
T
V
F F.... L T--F V5 .... T P F
P N .......V ®...TY........ .....
A
F
F
L G L
y
F
L L
G
Q
A T
L
V
A
A
S
L
VESICLE MEMBRANE
L
V
W
F
G
L
L
A
V
L
L
G
L
I
Q
L
G
L
F
V
V
A
A
V
A G p
A N
L
V
L
V L
L
V
L
F
() ....*.........................
L
A G
G
I
-EnFG
L
T
L
.....
S M F i)
L
I
G
AA
KY
L
A V
I
F
.............. ....
V
N
W A
W
.
G T L
L
C
C
L
..
A®L
INTRAVESICULAR SPACE
YS
F
L
G
Y
sMS LA
Q®A
Q L p
CYTOPLASMIC SPACE
2HN-M WJT L T A G G S A A S T L
T®G®S
P S S Q-COOH
Fig. 4. Model of the transmembrane structure,of cytochrome b561. The sequence is represented in single letter amino acid code. Positively charged amino acids are boxed and negatively charged residues are circled. The histidine that may serve as the fifth ligand for the heme iron and the methionine that may be the sixth ligand are framed by triangles. 2700
Secretory vesicle electron transport protein
as control (shown on the bottom of Figure 5). The mRNA appeared to be very abundant in adrenal medulla, but only moderately abundant in brain. All brain regions were found to contain at least some cytochrome b561 mRNA, including the cerebral cortex and cerebellum, with the highest concentrations present in caudate (lane 9). Prolonged exposure (Figure 5A) or short exposure (Figure SB) to Xray film reveals only a single sized hybridizing message on RNA blots employing either total or poly(A)+-enriched RNA. This result is surprising since the difference between the 3' ends of our cDNA clones (562 nt) is large enough to be resolved on the agarose gels employed here. To test if the observed mRNA corresponded to the long or the short 3' untranslated regions of our cDNA clones, RNA blots were hybridized with probes derived exclusively from the 3' end of the longer clone (Figure SC). Results identical to those obtained with coding regions probes were obtained, suggesting that the mRNA that we detect corresponds to the longer cDNA clone. Finally, primer extension experiments were performed. These suggested that our more 5' clone starts within 20 nt of the start site of the mRNA (data not shown), indicating little or no heterogeneity at the 5' end of the cytochrome b561 mRNA.
Discussion This report describes the isolation, amino acid sequencing and molecular cloning of bovine cytochrome b561. The protein is shown to consist of 273 amino acids with a high average hydrophobicity. A model based on the hydrophobicity profile is presented wherein cytochrome b561 is envisioned to span the membrane six times, with the majority of the hydrophilic sequences being cytoplasmic. The
structure of cytochrome b561 is not homologous to other electron transport proteins such as mitochondrial cytochromes and thus represents a new class of electron carriers, as already suggested by its unique biochemical behavior. RNA blotting experiments indicate that the protein is expressed in brain and neuroendocrine cells but not in liver or heart ventricles. Evidence that our cloned cDNA encodes cytochrome b561 can be summarized as follows: (i) the protein that we purified is the same size as previously purified cytochrome b561 and crossreacts with a characterized monoclonal antibody against cytochrome b561; (ii) the sequence is compatible with the known properties of cytochrome b561, including the predicted mol. wt, the extreme hydrophobicity, the absence of glycosylation and the presence of a 3 kd cytoplasmic hydrophilic C terminus; and (iii) the distribution of the mRNA corresponds to the proposed function of cytochrome b561. The transmembrane model of cytochrome b561 presented here suggests that the N terminus is cytoplasmic. This topology appears to be the case for many proteins containing multiple transmembrane domains, especially transport proteins (e.g. Kopito and Lodish, 1985; Gros et al., 1986; Noda et al., 1986; Paul, 1986). The model predicts that most of the hydrophilic sequences of cytochrome b561 are on the cytoplasmic face of the vesicles, as has been suggested from biochemical experiments (Abbs and Phillips, 1980). Since cytochrome b561 appears to be enzymatically symmetric, these cytoplasmic domains may have no relation to electron transport but could be involved in the sorting of the protein into separate vesicles. With the availability of molecular clones this can now be tested using protein engineering and expression techniques.
B.
A. -1
C.
7
_
-9.4 kb 7.4 kb - 4.4kb
Cytochrome b561
-
2.3kb
-1.4kb
ii,
Actin
-
*
-0
Fig. 5. Blotting analysis of the expression of cytochrome b561 mRNA in bovine tissues. Lanes as indicated by the numbers above the blot were loaded with RNA from heart ventricles (lane 1); liver (lanes 2, 12); medulla oblongata (lane 3); pons (lane 4); mesencephalon (lane 5); cerebellum (lane 6); cerebral cortex (lane 7); thalamus and hypothalamus (lane 8); caudate (lane 9); adrenal medulla (lanes 10, 11 and 13). 20 itg of total RNA (lanes 1, 3-10) or 2 ytg of poly A+ RNA (lanes 2, 11-13) were electrophoresed and the blots were hybridized with a mixed probe spanning the whole coding region (panels A and B), or a probe complementary to nt 2322-2473 in Figure 2B (panel C). After exposure to X-ray film at -70°C with an intensifying screen for seven days (panel A), 21 h (panel B) or 16 h (panel C), filters were stripped of the cytochrome b561 probe, rehybridized with an actin probe, and exposed for 16 h to allow estimation of the relative amounts of RNA in each lane (lower portion of figure). Numbers on the right give mol. wt of the RNA size standards used during electrophoresis.
2701
M.S.Perin et al.
Clusters of positively charged amino acids are observed in the hydrophilic sequence loops of cytochrome b561 on both sides of the membrane. These charges could serve as counterions to the negatively charged phospholipid headgroups of the membrane, although these are usually concentrated on the cytoplasmic face (Von Heijne, 1986). Alternatively, these positive charges may serve to facilitate the interaction of the protein with the negatively charged ascorbic acid and the subsequent reaction of this cofactor with the buried heme group of the cytochrome. Biochemical evidence has shown that the heme group in cytochrome b561 is not covalently bound (Apps et al., 1980). This result is supported by the absence of the consensus sequence for a covalent heme attachment site [CysX-Y-Cys-His (Mathews, 1985)] in the primary structure of cytochrome b561 (Figure 4). In all cytochromes studied, a histidine side chain provides the fifth coordination site for the heme iron, while the first four ligands are derived from the heme. The sixth ligand can be supplied by another histidine or a methionine residue or it can remain vacant (Mathews, 1985). Inspection of the cytochrome b561 sequence identifies the histidine at position 113 as a potential fifth ligand since it is the only histidine situated in the middle of the membrane bilayer. The sixth ligand may be supplied by the methionine at position 52. The assignment of this ligand constellation is supported by systematic considerations which indicate that cytochromes with relatively high midpoint oxidation-reduction potentials such as cytochrome b561 (120-140 mV) preferably have a methionine as a sixth ligand (Mathews, 1985). In addition to its structural enzymatic properties, cytochrome b561 is also unique in that it is localized to a specialized subcellular organelle, the large dense-core secretory vesicle. This unique localization correlates with its restricted tissue distribution and its importance in the biosynthesis of biogenic amines and peptides. The structure of cytochrome b561 reported here represents the first sequence of an intrinsic membrane protein from the large dense-core secretory vesicle. The structure of only one other intrinsic membrane protein (synaptophysin) from secretory vesicles is known (Leube et al., 1987; Sudhof et al., 1987a). The comparative study of these two proteins should provide novel insights into structure-function relationships in secretory vesicle proteins and into the cell biology of regulated secretion.
Materials and methods Protein purification and sequencing Chromaffin granule membranes were isolated from highly purified bovine chromaffin granules as described (Smith and Winkler, 1967; Sudhof et al., 1983). Membranes were stripped of extrinsic proteins by incubation in 15 mM NaOH, 1 mM EDTA, 1 mM PMSF for 1 h at 0°C. Stripped membranes were pelleted (250 000 g for 40 min) and resuspended in 20 mM Tris-HCl pH 7.4, 10 mM KCl, 1 mM EDTA. For purification of cytochrome b561, stripped membranes (1-2 mg protein) were solubilized in 60% formic acid, centrifuged at 150 000 g for 40 minutes to separate unsolubilized material, and then chromatographed on a Pharmacia FPLC ProRPC reverse phase column. Proteins were eluted with a 0-40% isopropanol gradient in 60% formic acid. Cytochrome b561 is the last major protein that elutes from the column (Figure 1). Column fractions were analyzed by PAGE followed by silver staining or immunoblotting. Cytochrome b561 was also isolated by isoelectric focusing chromatography (Wakefield et al., 1984). The identity of the cytochrome isolated was confirmed by immunoblotting experiments with a monoclonal antibody against cytochrome b561 (kind gift of Dr R.Pruss; Pruss, 1987).
2702
Purified cytochrome was subjected to N-terminal sequencing and found to be blocked to Edman degradation. The protein was chemically cleaved with CNBr or enzymatically cleaved with trypsin. Peptides were isolated by both gel electrophoresis followed by blotting onto Immobilon PVDF membranes [Millipore CA, Bedford, MA; (Matsudaira, 1987)], and by HPLC, and were subjected to automatic amino acid sequence analysis as described (Wang and Fried, 1987; Sudhof et al., 1988). cDNA cloning and nucleotide sequencing Seven redundant oligonucleotide probes were designed from peptide sequences P1 and P3 in Figure 2B and used to screen a bovine caudate complementary DNA library (gift of Dr Paul Greengard, Rockefeller University) as described (Sudhof et al., 1987a). Screening of 200 000 colonies resulted in the isolated of two different cDNA clones. cDNA fragments were subcloned into bacteriophage M13 vectors and sequenced by the didexoy chain termination method using [35S]dATP (Sanger et al., 1977; Sudhof et al., 1985). Nucleotide and protein sequences were analyzed with the Beckman Microgenie program on an IBM PC-AT. GenBank (release no. 52) and NBRF (release no. 13) databanks were searched. RNA blotting and primer extension analysis Total RNA and poly(A)+ RNA were purified from bovine tissues (Aviv and Leder, 1972; Sudhof et al., 1987b). RNA blotting experiments were performed with 32P-labeled single stranded DNA probes after gyoxal denaturation and agarose gel electrophoresis of the RNA (Lehrman et al., 1987). RNA size was determined by comparison with RNA ladder standards (Bethesda Research Lab., Gaithersburg, MD). Primer extension reactions were performed as described (Sudhof et al., 1987b).
Acknowledgements We wish to thank Drs Petra Kuster, David W.Russell, Reinhard Jahn, Patricia A.Buck, Johann Deisenhofer, Joseph L.Goldstein and Michael S.Brown for advice and critical review of the manuscript. P.Barjon and I.Leznicki provided excellent technical assistance. The technical help of K.Orth, C.Moomaw, M.Ando and A.Bell in peptide isolation and amino acid sequencing is gratefully acknowledged. This work was supported in part by Cancer Center Support (CORE) Grant CA21765 and a Research Development Award RR05584 (to VAF) from the National Institutes of Health and the American Lebanese Syrian Associated Charities (ALSAC).
References Abbs,M.T. and Phillips,J.H. (1980) Biochim. Biophys. Acta, 595, 200-221. Apps,D.K., Pryde,J.G. and Phillips,J.H. (1980) Neuroscience, 5, 2279-2287. Aviv,H. and Leder,P. (1972) Proc. Natl. Acad. Sci. USA, 69, 1408-1412. Bradbury,A.F., Finnie,M.D.A. and Smyth,C.G. (1982) Nature, 298, 686-688. Duong,L.T. and Fleming,P.J. (1982) J. Biol. Chem., 257, 8561-8564. Flatmark,T. and Terland,O. (1971) Biochim. Biophys. Acta, 253, 487-491. Gros,P., Croop,J. and Housman,D. (1986) Cell, 47, 371-380. Hortnagl,H., Winkler,H. and Lochs,H. (1973) J. Neurochem., 20, 977-985. Kent,U.M. and Fleming,P.J. (1987) J. Biol. Chem., 262, 8174-8178. Kopito,R.R. and Lodish,H.F. (1985) Nature, 316, 234-238. Kozak,M. (1987) Nucleic Acids Res., 15, 8125-8148. Kyte,J. and Doolittle,R. (1982) J. Mol. Biol., 157, 105-132. Lehrman,M.A., Russell,D.W., Goldstein,J.L. and Brown,M.S. (1987) J. Biol. Chem., 262, 3354-3361. Leube,R.E., Kaiser,P., Seiter,A., Zimbelmann,R., Werner,W.F., Rehm,H., Knaus,P., Prior,P., Betz,H., Reinke,H., Beyreuther,K. and Weidenmann,B. (1987) EMBO J., 6, 3261-3268. Mathews,F.S. (1985) Prog. Biophys. Mol. Biol., 45, 1-56. Matsudaira,P. (1987) J. Biol. Chem., 262, 10035-10038. Murthy,A.S.N., Mains,R.E. and Eipper,B.A. (1986) J. Biol. Chem., 261, 1815-1822. Njus,D., Knoth,J., Cook,C. and Kelley,P.M. (1983) J. Biol. Chem., 258, 27-30. Noda,M., Ikeda,T., Kayano,T., Suzuki,H., Takeshima,H., Kurasaki,M., Takahashi,H. and Numa,S. (1986) Nature, 320, 188-192. Paul,D.L. (1986) J. Cell. Biol., 103, 123-134. Pruss,R.M. (1987) Neurosci., 22, 141-147. Russell,J.T., Levine,M. and Njus,D. (1985) J. Biol. Chem., 260, 226-231. Sanger,F., Nicklen,S. and Coulson,A.R. (1977) Proc. Natl. Acad. Sci. USA, 74, 5463-5467.
Secretory vesicle electron transport protein Smith,A.D. and Winkler,H. (1967) Biochem. J., 103, 480-482. Srivastava,M., Duong,L.T. and Fleming,P.J. (1984) J. Biol. Chem., 259, 8072 -8075. Sudhof,T.C., Walker,J.H. and Obrocki,J. (1983) EMBO J., 1, 1167-70. Sudhof,T.C., Goldstein,J.L., Brown,M.S. and Russell,D.W. (1985) Science, 228, 815-822. Sudhof,T.C., Lottspeich,F., Greengard,P., Ehrenfried,M. and Jahn,R. (1987a) Science, 238, 1142-1144. Sudhof,T.C., Russell,D.W., Brown,M.S. and Goldstein,J.L. (1987b) Cell, 48, 1061-1069. Sudhof,T.C., Slaughter,C.A., Leznicki,I., Barjon,P. and Reynolds,G.A. (1988) Proc. Natl. Acad. Sci. USA, 85, 664-668. Terland,O., Silsand,T. and Flatmark,T. (1974) Biochim. Biophys. Acta, 359, 253-256. Von Heijne,G. (1984) EMBO J., 3, 2315-2318. Von Heijne,G. (1986) EMBO J., 5, 3021 -3027. Wakefield,L.M., Casse,A.E.G. and Radda,G.K. (1984) J. Biochem. Biophys. Methods, 9, 331-341. Wakefield,L.M., Casse,A.E.G. and Radda,G.K. (1986a) J. Biol. Chem., 261, 9739-9745. Wakefield,M.L., Casse,A.E.G. and Radda,G.K. (1986b) J. Biol. Chem., 261, 9746-9752. Wang,S.S. and Fried,V.A. (1987) J. Biol. Chem., 262, 6357-6364.
Received on May 24, 1988
Note added in proof The primary structure of the 22-kd light chain of human neutrophil cytochrome b was recently reported [Proc. Natl. Acad. Sci. USA (1988), 85, 3319-3323.]. Although there is no overall sequence homology between this cytochrome and cytochrome b561, there is a notable stretch of 10 amino acids where a 7 out of 10 match can be observed: cytochrome b561 VLHGLLHVFA (residues 107 to 116)
1 11 1 1
22 kd light chain VL HL L L S VP A (residues 92 to 101) Since this stretch includes a conserved histidine residue (His 109), it is possible that this histidine actually provides the fifth ligand for the heme iron.
2703
