1. No category

From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
1992 79: 2349-2355
Preferred sequence requirements for cleavage of pro-von Willebrand
factor by propeptide-processing enzymes
A Rehemtulla and RJ Kaufman
Updated information and services can be found at:
http://www.bloodjournal.org/content/79/9/2349.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
Preferred Sequence Requirements for Cleavage of Pro-von Willebrand Factor by
Propeptide-Processing Enzymes
By Alnawaz Rehemtulla and
Randal J. Kaufman
Maturation of pro-von Willebrand factor (vWF) to its active
form requires proteolytic processing after a pair of dibasic
amino acids (-LysArg-) at residue 763. By coexpression of
vWF and various propeptide processing enzymes in COS-I
cells, we here demonstrate that vWF is preferentially processed by the paired dibasic amino acid-cleaving enzyme
PACE (furin). Processing of vWF by the yeast homologue of
PACE, Kex2, was inefficient and not specific for the authentic
site. Two additional recently identified mammalian propeptide-processing enzymes PC2 and PC3 had no detectable
vWF-processing activity. The inability of PC2 and PC3 t o
cleave vWF was apparently not due t o the absence of a
transmembrane domain, since deletion of the transmembrane domain from PACE resulted in a secreted form which
retained its propeptide processing activity within the secretory apparatus. The inability of PC2 and PC3 t o process
wild-type vWF or any of the vWF mutants described suggests
different members of subtilisin-related propeptide-processing enzyme family have evolved t o selectively recognize and
cleave specific sets of substrates. In addition t o paired
dibasic residues at the propeptide cleavage site, many proteins, including vWF, also contain an arginine at the P4
position. We have generated mutant vWFs with substitutions at the P2 lysine and/ or the P4 arginine t o investigate
their significance in substrate specificity. A conservative
substitution of the P4 arginine by lysine resulted in a decrease in vWF processing by PACE, as did a nonconservative
substitution t o alanine. Substitution of the P2 lysine t o
aspartic acid decreased processing and little or no processing was detected when both the P4 and P2 were mutated t o
lysine and aspartic acid, respectively. These data indicate
that both the P4 arginine and the P2 lysine play an important
role in substrate recognition by PACE.
0 1992by The American Society of Hematology.
MANY
pro-vWF at its natural site suggests it may be involved in
processing vWF in endothelial cells. Consistent with this
hypothesis is the detection of both PACE mRNA and
protein in endothelial cells,1° while PC2 and PC3 expression is restricted to neuroendocrine tissues.“ By cotransfecting vWF with various processing enzymes, we here demonstrate that PACE and not PC2 or PC3 can significantly
process vWF, suggesting that the processing of vWF in
endothelial cells is specifically mediated by PACE. Using
vWF as a model substrate, we here demonstrate that
processing by PACE requires both a paired dibasic amino
acid motif and an arginine residue at P4.
PROTEINS, including plasma proteins, hormones, neuropeptides, growth factors, and viral
glycoproteins, require posttranslational proteolytic processing to generate biologically active molecules. These cleavages frequently remove propeptides that are essential for
protein maturation. Recently, putative enzymes involved in
propeptide processing have been cloned. Functional activities have been reported for several of these cloned products, most notably PACE/furin,’ PC2, and PC3.2-5These
enzymes belong to the family of subtilisin-like serine
proteases that are homologous to the yeast protease Kex2,
which is involved in the processing of the prohormone
a-mating factor.6 In addition to the subtilisin-like catalytic
domain, PACE and Kex2 contain a putative transmembrane domain at the carboxyl terminus, whereas PC2 and
PC3 do not. The K e d cytoplasmic tail and transmembrane
domain have been implicated to play a role in targeting the
enzyme to the Golgi ~ o m p l e x .Although
~
potential substrates have been identified for PACE, PC2, and PC3, little
is known about the primary amino acid sequence requirements of any particular substrate for cleavage by any of
these enzymes.
The yeast a-mating factor is one well-characterized
substrate for the Kex2 gene product that requires paired
dibasic residues at the amino terminal side of the cleavage
site.8 Examination of amino acid sequences around the
cleavage site of many mammalian precursor polypeptides
shows a common motif, typically -LysArg- or -ArgArg-.
These include coagulation factors IX and VII, protein C,
von Willebrand factor (vWF), human immunodeficiency
virus (HIV) gp 160 glycoprotein, pro-opiomelanocortin,
pro-insulin, the insulin receptor, and p-nerve growth factor.
Interestingly, proteins that may serve as substrates for
PACE, namely p-nerve growth factor9 and vWF,’ contain a
conserved arginine at the P4 position (the amino acid four
residues amino-terminal from the cleavage site) (Table 1).
In contrast, all the known substrates for PC2 or PC3 do not
contain the P4 arginine; examples include pro-opiomelanocortin and pro-insulin.2 The ability of PACE to process
Blood, Vol79, No 9 (May 1). 1992: pp 2349-2355
MATERIALS AND METHODS
Recombinant plasmids and transfections. PACE cDNA in the
expression plasmid pMT3 and vWF cDNA in the expression
plasmid pMT2 have been described previously.’ Kex2 cDNA
(kindly provided by P. Barr, Chiron Corp, Emeryville, CA) was
introduced into pMT2. PC2 and PC3 cDNAs were introduced into
the EcoRV site of the expression vector pMT3SV2,12 and were
kindly provided by S. Smeekens (HHMI, University of Chicago,
IL). COS-1 monkey kidney cells were cultured and transfected as
described previously.12Cotransfections were performed using equal
amounts of the respective plasmids.
Analysis of expressed proteins. Forty hours posttransfection,
cells were radiolabeled with 35S-methionine (250 FCiImL, > 8,000
Ci/mmol, New England Nuclear, Boston, MA) in methionine-free
media containing 2% dialyzed fetal calf serum for 1 hour, and
chase was performed in complete medium containing 10% fetal
calf serum for 4 hours. Conditioned media samples were harvested,
From the Genetics Institute, Cambridge, MA.
Submitted September 23, 1991; accepted December 30, 1991.
Address reprint requests to Randal J. Kaufman, PhD, 87 Cambridgepark Dr, Cambridge, MA 02140.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
0 1992 by The American Society of Hematology.
0006-4971I9217909-0023$3.00/0
2349
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
2350
REHEMTULLA AND KAUFMAN
Table 1. Examples of Proteins That Contain a Conserved P4 Arginine
P4
Protein S
Profactor VI1
Profactor IX
Factor X
Prothrombin
HIV gp160
Insulin receptci
c4
c3
Pro-NGF*
Pro-vWF*
Pro-TGF-pl*
R
R
R
R
R
R
I
H
N
Q
Q
Q
r
S
T
A
H
H
S
R
R
P
A
V
E
R
R
R
R
R
R
P2
P1
R
R
K
K
R
K
R
R
R
R
R
R
R
R
R
R
R
R
K
K
R
S
S
H
R
K
R
K
K
R
A
R
Y
S
A
A
S
N
S
S
S
A
Sequences shown were derived from the following references:factor
VI1 (211, factor IX (22). factor X (23). prothrombin (24), HIV gp160 (25).
insulin receptor (26), C4 (27). protein S (28), and C3 (29).
*These substrates have been shown to be processed by PACE.
Pro-NGF (9), Pro-vWF (l),
and Pro-TGF-p1 (30).
and soybean trypsin inhibitor (1 mg/mL), phenylmethylsulfonyl
fluoride (PMSF; 1 mmol/L, and aprotinin (0.2 mg/mL) were
added. Immunoprecipitation with vWF-specific antiserum (Diagnostica Stago, France) was as described previously.' Immunoprecipitates were electrophoresed on 8% sodium dodecyl sulfatepolyacrylamide gels (SDS-PAGE) in the presence of reducing
agent and visualized following fluorography in ENHANCE (DuPont, MA). Relative intensities of specific bands were determined
using an LKB (Piscataway, NJ) ultroscan XL laser densitometer
and accompanying software.
Western blot analysis was performed by electroblotting SDSPAGE-resolved proteins onto nitrocellulose membranes, and
PACE-specific bands were identified using a rabbit anti-PACE
antiserum described previously' (kindly provided by P. Barr,
Chiron).
Mutagenesis. Mutagenesis of the vWF cDNA was performed by
first subcloning a 1.3-kb internalXho1 fragment (that encompasses
the propeptide cleavage site) into pBluescript (Stratagene, La
Jolla, CA). The uracil incorporation methodI3 was used to modify
the coding sequence, and the mutated fragment was reintroduced
into pMT2VWF to reconstruct a complete cDNA. Similarly, the
PACE cDNA was mutated in pBluescript and the 2.4-kb EcoRISal1 fragment was reintroduced into pMT3. The DNA sequence of
all mutagenized fragments was confirmed by dideoxy nucleotide
sequencing using the Sequenase system (U.S. Biochemical, Cleveland, OH).
The mutagenic oligonucleotide primers used for the different
mutants were as follows: R760K, CCCCTGTCTCATAAAAGCAAACGATCGCTATCCTGTCGG; R760A, CCCCTGTCTCATGCTAGCAAAAGGAGCC; K762D, TCTCATCGCAGCGATCGAAGCTTATCCTGTCGG; and RXKRIKXDR,
CCCTGTCTCATAAAAGCGATCGGAGCCTATCCTG.
The soluble PACE (Sol PACE) mutant was constructed by deleting
sequences coding for the predicted transmembrane domain and
cytoplasmic tail using the oligonucleotide CCTCACACCTGCCTGAGTGATGAGCCCACTGCCCAC, and the active site mutant
of PACE (PACE S/A) was constructed by substituting the activesite serine residue 368 to an alanine using the oligonucleotide
GGCAGAGGCTGCGGTACCCGTGTGAGA.
RESULTS
PACE requires an Aig at P4 and basic residue at P2 for
efficient processing. To investigate the substrate requirements of the family of recently described propeptide-
processing enzymes, we here use vWF as a model substrate.
vWF is synthesized as a 2,813 residue proprotein and
undergoes extensive posttranslational processing that includes N- and 0-linked glycosylation, sulfation, and proteolytic processing to release a propeptide. Correct processing of vWF is essential for its ability to bind and stabilize
factor VIII.14Transfection of the wild-type vWF expression
plasmid into COS-1 cells resulted in secretion of vWF. A
majority, approximately 80%, of the vWF secreted from
COS-1 cells was unprocessed (Fig 1, lane 1). The small
amount that of mature vWF is attributed to the presence of
an endogenous COS-1 cell-processing enzyme. In contrast,
when the vWF expression vector pMT2vWF was cotransfected with the PACE expression vector pMT3PACE, the
majority of the secreted vWF comigrated with mature
processed vWF (Fig 1, lane 6). These results are consistent
with previously reported data.'
Since all known putative substrates of PACE have an
arginine at the P4 position and a lysine at P2, we have
generated mutant vWFs in expression vectors to investigate
the significance of these residues in substrate recognition.
P4 position mutants have a conservative substitution of
arginine to lysine (R760K) or a nonconservative substitution of arginine to alanine (R760A). A P2 position mutant
has a nonconservative substitution of lysine to aspartic acid
(K762D). A double mutant contains both the P4 arginine to
lysine and P2 lysine to aspartic acid (RXKR/KXDR).
These mutants were transfected into COS-1 cells alone or
cotransfected with expression vectors encoding wild-type
PACE, an active site serine to alanine mutant of PACE
(S/A), PC2, PC3, or Kex2.
To first evaluate the ability of the endogenous COS-1 cell
enzyme to process wild-type vWF and mutant vWFs, the
respective expression vectors were transfected into COS-1
cells and the secreted, metabolicaly labeled vWF was
analyzed by immunoprecipitation and SDS-PAGE. In contrast to wild-type vWF, processing of the R760K, K762D,
R760A, and the RXKR/KXDR mutants was not detectable
(Fig 1, lanes 2,3,4, and 5, respectively), whereas cotransfection of wild-type vWF with PACE resulted in secretion of
100% mature vWF (Fig 1, lane 6). vWF that contained a
conservative Arg to Lys substitution at the P4 position
yielded 80% mature vWF (Fig 1, lane 7). A nonconservative
substitution at P4 Arg to Ala yielded 65% mature vWF (Fig
1, lane 9). A nonconservative substitution of the P2 Lys to
Asp resulted in secretion of vWF that was 33% mature (Fig
1, lane 8). Substitution of both the P4 Arg and the P2 Lys to
Lys and Asp, respectively, did not yield detectable mature
vWF in the presence of PACE cotransfection (Fig 1, lane
10).
K a 2 , PC2, and PC3 cannot efficiently process pro-vWF.
Cotransfection of vWF in the presence of the yeast protease
Kex2 expression vector resulted in the secretion of three
forms of vWF (Fig 2, lane 4). One form comigrated with
unprocessed pro-vWF, a second form comigrated with
PACE-processed mature vWF, and a third, faster migrating
form (vWF*), likely resulted from cleavage at an alternate
site. In contrast to PACE, the P4 substitutions had a
minimal effect on processing by Kex2 (Fig 2, lanes 7 and
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
SPECIFICITY OF PROPEPTIDE-PROCESSING ENZYMES
2351
+
PACE
I
Y
LL
3
>
0
Q
N
(D
(D
a
Y
b
b
-200
Fig 1. Effect of P2 and P4 substitutions at the vWF
propeptide cleavage site on processing in the absence or presence of PACE. COS-1 cells were transfected with various VWFSalone (lanes 1 through 5) or
in the presence of PACE (lanes 6 through 10). Lanes 1
and 6, wild-type vWF; lanes 2 and 7, mutant vWF that
contains a P4 Arg to Lys substitution; lanes 3 and 8,
mutant vWF that contains a PZ Lys to Asp substitution; lanes 4 and 9, mutant vWF that contains a P4
Arg to Ala substitution; lanes 5 and 10, mutant vWF
t h a t contains a P4 Arg to Lys as well as a P2 Lys to
Asp substttution.
-97
-65
I
2
3 4 5
13). On the other hand, substitution of the P2 lysine
resulted in a much more significant decrease in processing
by K e d (Fig 2, lane 10). Substitution of both the P2 and the
P4 (RXKR/KXDR) positions resulted in a substrate that
was negligibly processed by PACE or Kex2 (Fig 2, lanes 15
and 16). Unexpectedly, cotransfection of vWF with PACE
SIA resulted in decreased processing by the endogenous
COS-I cell enzyme (Fig 2, lane 3). The ability of PACE S/A
B
A
w t vWF
200 kDo
6 7 8 9 1 0
to decrease processing by the endogenous COS-1 cell
enzyme has also been observed with other substrates such
as transforming growth factor-f3l (data not shown).
Cotransfection of the PACE homologues PC2 and PC3
with vWF did not increase the amount of mature vWF
secreted (Fig 3, lanes 1 and 6). In addition, there was no
increase in mature vWF when either PC2 or PC3 were
cotransfected with vWF mutants that had P4 or P2 muta-
C
R760K
K762D
7
8 9 0
D
E
R760A RXKIKXD
1
2
3 4
5 6
I1 1213 1415 16
Fig 2 Effect of P2 and P4 subatltutlom at the vWF propeptide cleavage site on processing by PACE, PACE S/A mutant, and Kex2. COS-1 cells
were transfected with wild-type PACE (wt, lanes 2,6,9,12, 15), the active site mutant (serineto alanine) of PACE (SA, lanes 3,5,8,11,14) or Kex2
in the presence of (A) wild-type vWF (wt vWF, lanes 1 through 4). (B) the P4 Arg to Lyr mutant (R760K. lanes 5 through 7). (C) the P2 Lys to Asp
mutant (K762D. lanes 8 through 10). (D) the P4 Arg to Ala mutant (R760A, 11-13), or a mutant containing a P2, as well as a P4 substitution
(RXK/KXD, lanes 14 through 16). Radlolabeled vWF was immunoprecipitatedfrom conditioned media and analyzed by SDS-PAGE followed by
autoradiography. Lane 1 representsvWF secreted from cells that were transfected with vWF alone.
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
2352
REHEMTULIA AND KAUFMAN
PC3
PC2
a
X
0
X
rant:
t pro-vWF
200 kDa -
M
I
2
3
4
5
6 7 8 9 1 0
Flg 3. Processingof wlld-type vWF and mutant v w h by PC2 and PC3. COS1 cells were transfected with wild-type vWF (lanes 1 and 61, a P4
Arg to Lys mutant (R76OK. lanes 2 and 7). a P2 Lys to h p mutant (K762D. lanes 3 and 8). a P4 Arg to Ala mutant (R76OA. lanes 4 and 9) or a P4, P2
double mutant (RXK/KXD, lanes 5 and 10) in the presence of PC2 (lanes 1 through 5 ) or PC3 (lanes 6 through 10). Condkioned media were
prepared after labeling the cells with radiolabeled methionine and vWF was immunoprecipitated. Immunoprecipitates were analyzed by
SDS-PAGE and autoradiography.
tions (Fig 3, lanes 2 to 5 and 7 to 10). Expression of PC2 and
PC3 was easily detected by analysis of total, labeled,
conditioned medium (Fig4). Therefore, the lack of processing was not due to inefficient expression of these enzymes.
Membrane anchoring is not required for PACE activiv.
We next investigated if the lack of processing of vWF by
Y
0
0
W
O
a
e
N
N
o
a
fox
o w
e x
200 97 65
-
lack transmembrane domains and are secreted from COS-1
cells, in contrast to PACE which contains a transmembrane
domain and is cell-associated.A mutant form of PACE (Sol
PACE) that lacks a transmembrane domain was constructed by deleting residues 716 to 794, which code for the
transmembrane domain and cytoplasmic tail. Western blot
analysis of conditioned media and cell extracts prepared
from COS-1 cells transfected with expression vectors encoding Sol PACE or wild-type PACE (Fig 5A) showed that a
large amount of Sol PACE was secreted into the conditioned medium (Fig SA, lane 2), while none of the wild-type
PACE was detected in the medium (Fig 5A, lane 1).
Cotransfection experiments of Sol PACE with vWF demonstrated that Sol PACE efficiently processed vWF to its
mature form (Fig SB). Mixing of conditioned media demonstrated that the secreted PACE was not active (data not
shown). Thus, anchoring of PACE into the membrane is
not required for functional activity. The absence of a
transmembrane domain on PC2 and PC3 is likely not the
reason the enzymes cannot recognize and cleave vWF.
DISCUSSION
45 -
I
2
3
4
5
5 4* Exp-ion
and secretIanof PC2 and PC3* cos-1
t”fected
with no DNA (lane 11, PACE (lane 2). PC2 (lane 3). PC3
(lane 4). or Kex2 (lane 5 ) and radiolabeled conditioned media were
DreDaredasdescrlbed in the Methods. Medium samdes 1100 uLI were
analyzed by SDS-PAGE and autoradiography.
. .
PC2 or PC3 was due to the fact that both these enzymes
. . ..
Transfection of the vWD cDNA into COS-1 cells results
in the secretion of vWF protein that is predominantly
pro-vWF due to inefficient processing by the endogenous
COS-1 cell-processing enzyme. Cotransfection of vWF with
PACE resulted in secretion of only processed mature vWF
and demonstrated the ability of PACE to recognize and
process vWF as a substrate. Amino-terminal sequence
analysis has previously shown that processing of vWF in the
presence of PACE occurred at the authentic site.’ Cotransfection of vWF and an active site mutant of PACE (the
active site serine was mutated to alanine) resulted in
secretion of vWF that was predominantly unprocessed.
This apparent inhibition of the endogenous pro-vWF processing activity in c o s - 1 cells by overexpression Of the
inactive PACE mutant was reproducible and unexpected.
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
2353
SPECIFICITY OF PROPEPTIDE-PROCESSINGENZYMES
-Cell
Conditioned
Medium
Extracts
A
- 200
-97.1
I
3
2
P
- 67
-pro v W F
L--
1
4
VWF
2
Flg 5. (A) Western blot analysis of conditloned media and cell
lysates preparedfrom COS1 wlls transfectedwith PACE (lanes 1 and
3) or Sol PACE (lanes 2 and 4). Samples were analyzed by SDS-PAGE
followed by electroblotting to nitrocellulose and detected using a
PACE-specific antiserum as described above. (6)Functional analysis
of the soluble mutant of PACE. COS-1 cells were transfectedwith vWF
alone (lane 1) or in the presence of Sol PACE (lane 2) and the secreted
vWF was analyzed as described above.
Several possibilities may explain the dominance of the
active site mutant over the endogenous enzyme. The two
enzymes may compete for substrate binding or for a cellular
compartment required for PACE activity. Alternatively,the
overexpression of mutant enzyme may directly inhibit the
activity of endogenous enzyme. An alternate explanation
may be that the active site mutant inhibits cleavage of the
amino terminal propeptide, which is required for functional
activation of many propeptide cleaving enzymes.
The yeast propeptide processing enzyme K e d processed
pro-vWF at two sites to yield a form that comigrates with
mature vWF and a second, faster migrating form. The
presence of numerous dibasic residues (-LysArg- and -ArgArg-) within the vWF coding sequence might explain the
additional cleavage by Ked. This is consistent with the
observation that Kex2 is able to cleave its natural substrates
pro-a-mating factor and pro-killer toxin after multiple
dibasic residues that do not contain any obviously conserved adjacent sequences. The inability of PACE to cleave
at this alternate site might reflect a requirement for greater
specificity at the authentic PACE cleavage site. The detectable presence of pro-vWF when vWF was cotransfected
with K e d suggested inefficient processing by the K e d gene
product. Since complete processing of pro-albumin: proopi~melanocortin,~~
and pro-protein C16by K e d can occur
in mammalian cells, we believe that the inability of Kex2 to
efficiently process vWF was not due to the unnatural
mammalian cell environment, but rather that processing by
a specific protease requires substrate recognition signals in
addition to accessibilityof dibasic residues.
Analysis of processing of vWF and vWF mutants by
PACE showed that substitution of the P4 arginine with a
conserved basic residue lysine reduced processing by 30%,
while substitution with nonconserved alanine resulted in a
35% decrease in processing. Substitution of the P2 lysine
with aspartic acid decreased processing by 67%. The
efficiency by which PACE can process the P4 and P2
mutants is probably overestimated in this analysis, since the
transfected enzyme is overexpressed to such a large degree.
Our estimation by Western blot analysis suggests PACE is
overexpressed at least 100-fold in transfected cells. The
endogenous processing enzyme in COS cells was not
detected by our analysis. We do not know whether the
COS-1 cell enzyme activity results from PACE or a related
product. Assuming that the conformation of the cleavage
site in the mutants is favorable, these data suggest a
requirement of both the P4 arginine and the paired dibasic
motif for processing by PACE, while the presence of only a
paired dibasic sequence may be sufficient for cleavage by
Ked. Recent reports17 of substrate requirements for the
processing of the insulin receptor in Chinese hamster ovary
(CHO) cells suggest that the P4 arginine is required,
although this study seemed to suggest that the substitution
of the P2 lysine did not effect processing. Processing of the
insulin receptor in the CHO cells studied could have been
mediated by an unidentified enzyme, which may have a
different specificity compared with PACE. The ability
of the endogenous CHO cell enzyme to cleave monobasic substrates is intriguing, since coagulation factor X,
which contains a P4 arginine and a monobasic residue
(-ArgValThrArg-), is efficiently cleaved in CHO cells (D.
Pittman, unpublished observations).
The importance of the P4 arginine in substrate recognition is emphasized by the observation that several hemophilia B patients have decreased factor IX activity attributable to substitution of the P4 arginine.18J9 Identification of
the mutation in two families shows nonconservative glutamic acid substitutions for the P4 arginine. According to
the data presented here, these mutant proteins would make
very unfavorable substrates. Such patients contain significant levels of circulating factor IX antigen, although the
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
2354
REHEMTULLA AND KAUFMAN
protein is nonfunctional due to inefficient propeptide
processing.
The inability of PC2 and PC3 (which typically cleave
substrates that do not contain a P4 arginine) to cleave
wild-type vWF or vWF mutants that contain P4 and P2
substitutions suggests that the presence of the P4 arginine is
not the restricting factor in the ability of these enzymes to
cleave vWF. In addition, the lack of a transmembrane
domain in PC2 and PC3 does not account for the inability
for these enzymes to process vWF, since Sol PACE, which
lacks a transmembrane domain, is able to process vWF.
Mixing experiments demonstrated that the soluble secreted
form of PACE was active within the secretory pathway, but
was not active in the conditioned medium. We speculate
that the enzyme is inactivated upon secretion into the
medium. One explanation for the inability of PC2 and PC3
to cleave vWF may be that COS-1 cells do not contain the
proper intracellular environment. Cell lines and tissues of
neuroendocrine origin contain secretory granules that constitute the regulated secretory pathway. It has recently been
demonstrated that PC2 and PC3 may localize in these
secretory granules.I1 However, the inability of PC2 and PC3
to cleave pro-vWF in COS-1 cells may not simply be due to
an inappropriate intracellular environment, since expres-
sion of PC2 in COS-1 cells can elicit cleavage of other
substrates such as pro-glucagon and pro-insulin (S. Smeekens, personal communication). The lack of pro-vWF processing activity by PC2 or PC3 suggests that these processing
enzymes have a defined range of substrates for which
specificity is determined by the paired dibasic amino acid
residues. On the other hand, PC2 and PC3 may be involved
in processing of substrates that typically do not contain a P4
arginine and whose secretion is regulated, such as proinsulin and pro-opiomelanocortin. One might speculate
that PACE, which is ubiquitously expressed,M may be
involved in processing of constitutively secreted proteins
such as $-nerve growth factor, the insulin receptor, and the
coagulation factors IX, VII, and vWF, all of which contain a
P4 arginine. Experiments are in progress to determine if
PACE expressed in endothelial cells enters the regulated
pathway and is stored in weibel palade bodies.
ACKNOWLEDGMENT
We acknowledge A. Dorner, P. Barr, and B. Wise for helpful
discussion and support, and S. Smeekens and D. Steiner for
providing cDNAs for PC2 and PC3. We also acknowledge K. Kerns
and D. Pittman for providing information prior to publication.
REFERENCES
1. Wise RJ, Barr PJ, Wong PA, Kiefer MC, Brake AJ, Kaufman
RJ: Expression of a human proprotein processing enzyme: Correct
cleavage of the von Willebrand factor precursor at a paired basic
amino acid site. Proc Natl Acad Sci USA 87:9378,1990
2. Benjannet S, Rondeau N, Day R, Chreitien M, Seidah NG:
PC1 and PC2 are proprotein convertases capable of cleaving
proopiomelanocortin at distinct pairs of basic residues. Proc Natl
Acad Sci USA 88:3564,1991
3. Smeekens SP, Steiner DF: Identification of a human insulinoma cDNA encoding a novel mammalian protein structurally
related to the yeast dibasic processing protease Kex2. J Biol Chem
265:2997,1990
4. Smeekens SP, Avruch AS, LaMendola J, Chan SJ, Steiner
D F Identification of a cDNA encoding a second putative prohormone convertase related to PC2 in AtT20 cells and islets of
Langherhans. Proc Natl Acad Sci USA 88:340,1991
5. Seidah NG, Gaspar L, Mion P, Marcinkiewicz M, Mbikay M,
Chretien M: cDNA sequence of two distinct pituitary proteins
homologous to K e d and Furin gene products: Tissue specific
mRNAs encoding candidates for prohormone processing proteinases. DNA Cell Biol9:415, 1990
6. Fuller RS, Steme RE, Thorner J: Enzymes required for
prohormone processing. Ann Rev Physiol50:345,1988
7. Redding K, Holcomb C, Fuller R: Immunolocalization of
Kex2 protease identifies a putative late golgi compartment in the
yeast Succhromyces cervisiue. J Cell Biol113527, 1991
8. Fuller RS, Brake AJ, Thorner J: Intracellular targeting and
structural conservation of a prohormone-processing endoprotease.
Science 246:482, 1989
9. Bresnahan PA, Leduc R, Thomas L, Thorner J, Gibson HL,
Brake AJ, Barr PJ, Thomas G: Human fur gene encodes a yeast
Ked-like endoprotease that cleaves pro-P-NGF in-vivo. J Cell Biol
111:2851,1990
10. Wise RJ, Barr PJ, Jacobson BC, Ewenstein BM, Kaufman
RJ: Expression of a human proprotein processing enzyme that
correctly cleaves the von Willebrand factor precursor at its dibasic
amino acid recognition site. Blood 76:443a, 1990 (suppl, abstr)
11. Smeekens SP, Albiges-Rizzo C, Phillips LA, Montiag A,
Shwin M, Swift H, Chutkow W, Benig M, Steiner D F Precursor
processing by PC2 and PC3. J Cell Biochem 136,1991 (suppl15G)
12. Kaufman R J Vectors used for expression in mammalian
cells, in Goeddel D (ed): Methods in Enzymology, vol 185. San
Diego, CA, Academic, 1990, p 487
13. Kunkel T A A method for mutagenking DNA. Proc Natl
Acad Sci USA 82488,1975
14. Wise JW, Dorner AJ, Krane M, Pittman DD, Kaufman RJ:
The role of von Willebrand factor multimers and propeptide
cleavage in binding and stabilizing of factor VIII. J Biol Chem
266:21948,1991
15. Thomas G, Thorne BA, Thomas L, Allen RG, Hruby DE,
Fuller R, Thomer J: Yeast K e d endopeptidase corectly cleaves a
neuroendocrine prohormone in mammalian cells. Science 241:226,
1988
16. Foster DC, Sprecher CA, Holly RD, Gambee JE, Walker
KM, Kumar AA: Endoproteol tic processing of the dibasic cleavage site in the human protein precursor in transfected mammalian cells: Effects of sequence alterations on efficiency of cleavage.
Biochemistry 29:347,1990
17. Yoshimasa Y, Paul JI, Whittaker J, Steiner D F Effects of
amino acid replacements within the tetrabasic clevage site on the
processing of the human insulin receptor precursor expressed in
Chinese hamster ovary cells. J Biol Chem 265:17230,1990
18. Bentley AK, Rees DJG, Rizza C, Brownlee GG: Defective
propeptide processing of blood clotting factor IX caused by
mutation of arginine to glutamine at position -4. Cell 45:343,1986
19. Sugimoto M, Miyata T, Kawabata S, Yoshioka A, Fukui H,
Iwanaga S: Factor IX Kawachinagano: Impaired function of the
Gla-domain caused by attached propeptide region due to substitution of arginine by glutamine at position -4. Br J Haematol72:216,
1989
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
SPECIFICITY OF PROPEPTIDE-PROCESSING ENZYMES
20. Roebroek AJM, Schalken JA, Leunissen JAM, Onnekink C,
Bloemers HPJ, Van de Ven WJM Evolutionary conserved close
linkage of the c-feslfbs proto-oncogene and genetic sequences
encoding a receptor-like protein. EMBO J 5:2197,1986
21. Hagen FS, Gray CL, O’Hara PJ, Grant FJ, Saari GC,
Woodbury RG, Hart CE, Insley M, Kisiel W, Kurachi K, Davie
E W Characterization of a cDNA coding for human factor VII.
Proc Natl Acad Sci USA 83:2412,1986
22. Kurachi K, Davie EW: Isolation and cloning of cDNA for
human factor IX. Proc Natl Acad Sci USA 79:6461,1982
23. Fung MR, Hay CW, MacGillivray R T A Characterization of
an almost full length cDNA coding for human blood coagulation
factor X. Proc Natl Acad Sci USA 82:3591,1985
24. Degan SJF, MacGillavray RTA, Davie EW: Nucleotide
sequence of the gene for human prothrombin. Biochemistry
22:2087, 1983
25. McCune JM, Rabin LB, Feinberg MB, Lieberman M, Kosek
JC, Reyes GR, Weissman I L Endoproteolytic cleavage of gp160 is
2355
required for the activation of human immunodeficiency virus. Cell
2355,1988
26. Ullrich A, Bell JR, Chen EY, Herrera R, Petruzzeli M, Dull
TJ, Gray A, Coussens L, Liao Y-C, Tsubokawa M, Mason A,
Seeburg PH, Grunfeld C, Rosen OM, Ramachadran J: Human
insulin receptor and its relationship to the tyrosine kinase family of
oncogenes. Nature 313:756,1985
27. Belt KT, Carrol MC, Porter RR: The structural basis of the
multiple forms of human complement component C4. Cell 36:907,
1984
28. Dahlback B, Lundwall A, Stenflo J: Primary structure of
bovine vitamin K dependent protein S. Proc Natl Acad Sci USA
83:4199,1986
29. DeBruijn MHL, Fey GH: Human complement component
C3: cDNA coding sequence and derived primary structure. Proc
Natl Acad Sci USA 82708,1985
30. Rehemtulla A, Kaufman RJ: The propeptide cleaving enzyme PACE is a zymogen and requires proteolytic activation at the
RTKR cleavage site. J Cell Biol115:301a, 1991 (abstr)