Available via license: CC BY 4.0
Content may be subject to copyright.
THE
JOURNAL
OF
BIOLOGICAL CHEMISTRY
(0
1992 by The American Society
for
Biochemistry and Malecular
Biology,
Inc
VOl.
267,
No.
31, Issue
of
November
5,
pp.
22401-22406,1992
Printed
in
U.S.A.
The Post-translational Processing and Intracellular Sorting
of
PC2
in
the Islets
of
Langerhans*
(Received for publication, June
15,
1992)
Paul C. Guest$, Susan D. Arden, Deborah
L.
Bennett, Anne Clark§, Nicholas G. Rutherford, and
John C. Hutton
From the University of Cambridge, Department of Clinical Biochemistry, Addenbrookes Hospital, Hills Road,
Cambridge, CB2
2QR
United Kingdom and the §University of Oxford, Nuffield Department of Clinical Medicine, Diabetes
Research Laboratories, Radcliffe Infirmary, Woodstock Road, Oxford,
OX2
6HE
United Kingdom
Proinsulin conversion in the insulin secretory gran-
ule
is
mediated by two sequence-specific endoproteases
related to the Kex2 homologues, PC2 and PC3 (Ben-
nett, D.
L.,
Bailyes,
E.
M.,
Nielsen,
E.,
Guest,
P.
C.,
Rutherford,
N.
G.,
Arden,
S.
D., and Hutton,
J.
C.
(1992)
J.
Biol.
Chem.
267, 15229-15236; Bailyes,
E.
M.,
Bennett, D.
L.,
and Hutton,
J.
C. (1992)
Enzyme,
in press). Radiolabeling studies using isolated rat islets
showed that PC2 was synthesized initially
as a
76-kDa
glycoprotein which was converted by limited proteol-
ysis
to the mature 64-66-kDa form. Conversion was
initiated approximately
1
h after synthesis and pro-
ceeded via intermediates of
71,68,
and
66
kDa with a
tIl2
of
140
min. Release
of
only the
66-
and 64-66-kDa
radiolabeled forms of PC2 was induced by glucose and
then only
at
times more than 2 h following synthesis.
Proinsulin conversion, by contrast, was more rapid
(delay
=
30
min,
tllz
=
60
min), and release commenced
as
soon as
1
h after synthesis with the secreted material
being comprised of the precursor, intermediate, and
mature forms of insulin. Ultrastructural analysis of
islet
6
cells showed that PC2 was concentrated in
se-
cretory granules. Subcellular fractionation combined
with immunoblot analysis showed that insulinoma
se-
cretory granules contained only the mature 64-66-
kDa form
of
PC2, whereas fractions enriched in Golgi
and endoplasmic reticulum contained
a
mixture of the
76-
and 66-kDa forms of the enzyme. These results
indicate that post-translational proteolysis of PC2
is
initiated before sorting into the regulated pathway of
secretion and that the relative proportions
of
proinsu-
lin and PC2 packaged into secretory granules will
change with physiological conditions.
Enzymological studies have shown that proinsulin conver-
sion requires two distinct endopeptidase activities (designated
type
1
and type
2)
which cleave on the C-terminal side of
Ar2’-Arg2 and Lyss4-ArgG5 (Davidson
et
al.,
1988) and an
exopeptidase (carboxypeptidase H) which removes the ex-
posed basic residues (Davidson and Hutton, 1987). All three
enzymes have a pH optimum of
5.5
which coincides with the
intragranular pH (Hutton, 1982). In addition, the endopepti-
dases have an absolute requirement for Ca2+ which would be
~ ~~ ~ ~ ~~ ~ ~~ ~~
*This work was supported by grants from the British Diabetic
Association, the Medical Research Council of Great Britain and
Wellcome Trust. The costs of publication of this article were defrayed
in part by the payment of page charges. This article must therefore
be hereby marked “aduertisement” in accordance with
18
U.S.C.
Section
1734
solely to indicate this fact.
$
To
whom correspondence should be sent.
met by the high Ca2+ concentration prevailing in mature
secretory granules (Hutton
et
al.,
1983).
Recent immunological and structural studies have identi-
fied the type
2
endopeptidase as PC2 (Bennett
et
al.,
1992), a
member of the mammalian Kex2-related family of proteases
(for review see Barr, 1991). Other members
of
the family
include the Golgi-localized furin (Breshnahan
et
al.,
1990),
PACE 4 (Kiefer
et
al.,
1991), PC3 (also known as PC1; Seidah
et
al.,
1991; Smeekens
et
aL,
1991), and PC4 (Nakayama
et
al.,
1992). Like the Kex2 gene product, all of the enzymes
appear to be involved in the post-translational processing of
proproteins in the secretory pathway and induce limited pro-
teolysis at sites preceded by basic amino acid sequences such
as Lys-Arg, Arg-Arg,
or
Arg-X-X-Arg. PC2 immunoreactivity
is present in pancreatic islets, the adrenal medulla, and the
anterior and neurointermediate lobes of the pituitary (Christie
et
al.,
1991; Kirchmair
et
al.,
1992; Bennett
et
al.,
19921,
suggesting that it has
a
general function in the conversion of
polypeptide hormone and neuropeptide precursors in cells of
the diffuse neuroendocrine system.
PC2 itself is subjected to post-translational proteolytic
processing during the course of its segregation to the site of
proinsulin conversion in secretory granules. This has been
shown by the presence of two molecular variants of the
enzyme in insulin granules which correspond to proteins with
N
termini commencing at amino acids 109 and 112
of
rat PC2
(Bennett
et
al.,
1992).
It
is conceivable that the mature
peptides are produced by autoproteolysis
or
through cleavage
by another Kex2-related enzyme such as furin.
We report here an investigation of the localization and
biosynthesis of PC2 in rat pancreatic endocrine tissue. The
results show that PC2 is proteolytically processed and sorted
into the regulated pathway of secretion and that the cellular
site of post-translational modification and kinetics of proc-
essing and transport differ substantially from the insulin-
related peptides with which the enzyme is co-secreted.
EXPERIMENTAL PROCEDURES
Chemicals and Reagents-[”S]Methionine
(1,400
Ci/mmol) was
purchased from Amersham International (Little Chalfont, Bucking-
hamshire,
U.
K.),
N-Glycanase was from NSB Biologicals (Hatfield,
Hertfordshire,
U.
K.).
Dulbecco’s modified Eagle’s medium was ob-
tained from
Flow
Laboratories (Rickmansworth, Hertfordshire). Per-
oxidase-conjugated rabbit anti-guinea pig and swine anti-rabbit sera
were from Dakopaks (Copenhagen, Denmark). All other analytical
grade biochemicals were obtained from Sigma or BDH Chemicals
(both of Poole, Dorset,
U.
K.)
unless specified otherwise.
Antibodies-A mouse monoclonal antibody (3B7) raised
to
human
proinsulin (Gray et
al.,
1987) was partially purified by
(NH,),SO,
precipitation and coupled to CNBr-activated Sepharose by the
method
of
Axen et al. (1967). The resulting immunoadsorbent had a
binding capacity of approximately
100
pg of insulin/ml of swollen
22401
22402
PC2
Biosynthesis
gel. Antiserum to a glutathione S-transferase fusion protein incor-
porating amino acids 158-391 of the rat PC2 sequence was raised in
rabbits, and antibodies reacting with glutathione S-transferase were
removed by affinity chromatography
as
described previously (Bennett
et
aL,
1992).
Tissues-Islets of Langerhans were isolated from the pancreata of
10-12-week-old New England Deaconess Hospital rats by collagenase
digestion, as described previously (Guest
et
al.,
1989a). Islets were
collected into an incubation medium consisting of modified Krebs
bicarbonate buffer (120 mM NaC1, 5 mM KCI,
1
mM MgS04, 2.5 mM
CaC12, 24 mM NaHC03) containing
20
mM Hepes (pH 7.4), 0.1%
BSA,' and 16.7 mM glucose. Insulinoma tissue was propagated in
New England Deaconess Hospital rats as described previously (Sop-
with
et
al.,
1981), and subcellular fractions were prepared by Percoll
density gradient centrifugation according to the method of Hutton et
al.
(1982).
Radioisotopic Labeling-Batches of 100-200 islets were preincu-
bated in 0.2 ml of incubation medium for 45 min at 37 "C under
O,/
CO,
(19:l) in 1.5 ml-capacity microcentrifuge tubes (Alpha; Eastleigh,
Hampshire,
U.
K.).
Islets were recovered by centrifugation for
10
s
at
800
X
g
(Heraeus Sepatech microcentrifuge; Kalkberg, Germany) and
resuspended in the same prewarmed (37 "C) medium containing
0.2
mCi
of
["'S]methionine. Islets were recovered
as
above after
1
h and
incubated for various times in nonradioactive medium under the
conditions described in the figure legends. Incubations were termi-
nated by addition of
1
ml of ice-cold Krebs incubation medium
followed by centrifugation for
10
s
at
2,000
X
g.
Islets were washed
by two further cycles of resuspension and centrifugation and then
sonicated for 15 s at
3
microns (MSE Sonifier) in 0.4 ml of a lysis
buffer consisting of 25 mM Na2B4O7 (pH 9), 3% (w/v) BSA,
1%
(w/
v)
Tween 20,
1
mM phenylmethylsulfonyl fluoride, 0.1 mM E-64
(trans-epoxysuccinyl-~-leucyl-amido-(4-guanidino)butane),
1
mM
EDTA, and 0.1% NaN:1. The lysates were centrifuged for
3
min at
10,000
x
g
and the supernatants retained. Where release of radiola-
beled proteins into the medium was monitored, media were removed
from islets, clarified by centrifugation for 3 min
at
10,000
X
g,
and
the supernatants combined with 2
X
concentrated lysis buffer for
subsequent immunoprecipitation. Incorporation of [3sS]methionine
into islet lysate and medium proteins was determined by trichloroa-
cetic acid precipitation as described previously (Guest
et
al.,
1989a).
Immunoprecipitation
of
Proinsulin-related Peptides-Islet lysates
and media samples were immunoprecipitated with
100
p1 (packed gel)
of 3B7 immunoadsorbent as described previously (Guest
et
al.,
1989a).
The immunoadsorbent was recovered by centrifugation for
10
s
at
10,000
X
g
and the supernatant used for the immunoisolation of the
PC2-related peptides (see below). The immunoadsorbent was washed
and the bound protein eluted and subjected to alkaline urea-polyacryl-
amide gel electrophoresis (PAGE), and fluorographic analysis
as
described previously (Guest
et
al.,
1989a). Incorporation of radioac-
tivity into specific protein bands was determined by densitometric
scanning of suitably exposed fluorographs using a Chromoscan
111
densitometer (Joyce-Loebl, Gateshead, U.K.).
Immunoprecipitation
of
PC2-related Peptides-The supernatants
obtained after precipitation of proinsulin-like peptides were incubated
overnight at 4 "C with 10 pl of anti-PC2 serum and 50 pl (packed gel)
of preswollen protein A-Sepharose. The protein A-Sepharose was
recovered by centrifugation and washed as above. Bound proteins
were eluted with 200
p1
of 20 mM HCI, freeze-dried, and reconstituted
in 50 pl of SDS sample loading buffer (125 mM Tris/HCl (pH 6.8)
containing
2%
(w/v) SDS, 0.25
M
sucrose, 5 mM EDTA, 65 mM
dithiothreitol, and 0.005% bromphenol blue). The samples were
heated for 5 min at 100 'C and subjected
to
SDS-PAGE on slab gels
polymerized from
10%
(w/v) acrylamide, 0.065% N,N'-methylenebis-
acrylamide, by using the discontinuous buffer system of Laemmli
(1970), and fluorography was performed
as
above. Molecular size
calibration was achieved with I4C-labeled lysozyme, 6-lactoglobulin,
cr-chymotrypsin, ovalbumin, BSA, and phosphorylase
b
(Bethesda
Research Laboratories, Paisley, Scotland).
N-Clycanase Treatment-Islets were sonicated for 15
s
at 3 microns
in 50
pl
of 100 mM Tris (pH 8.0) containing 0.5% SDS, 50 mM
P-
mercaptoethanol,
1
mM phenylmethylsulfonyl fluoride, and
1
mM
EDTA. The lysates were centrifuged for
3
min at
10,000
X
g
and the
supernatants heated for 5 min at 100 "C and then cooled for 5 min
on ice. The samples were incubated for
18
h at 37 "C in
a
final volume
of
150
pl containing 1.7% (w/v) Nonidet
P-40
and 1.5 units of
N-
'
The abbreviations used are: BSA, bovine serum albumin; PAGE,
polyacrylamide gel electrophoresis;
ER,
endoplasmic reticulum.
glycanase. The samples were then immunoprecipitated with anti-PC2
serum and the eluted proteins subjected to SDS-PAGE and fluorog-
raphy as above.
Ultrastructural Analysis-Ultrathin sections
of
human pancreatic
islets were fixed in 2.5% glutaraldehyde (pH 7.2) and embedded in
Spurr's resin. Sections were incubated for 2 h
at
room temperature
with a
1:lO
dilution of anti-PC2 serum in phosphate-buffered saline
(150 mM NaCl, 16 mM NaHzP04, 4 mM NazHP04) containing 1%
(w/v) ovalbumin and then exposed to protein A-gold (10 nm colloidal
gold) for
1
h at room temperature
as
described previously (Pow and
Clark, 1990). Ultrastructural details were visualized by staining for
10
min with saturated uranyl acetate and then
for
10 min with lead
nitrate. Sections were viewed with
a
Joel electron microscope. Control
experiments were performed by incubating sections with preabsorbed
antiserum (4 mg/ml PCZ/glutathione S-transferase fusion protein,
overnight at 4 "C).
Immunoblot Analysis-Insulinoma subcellular fractions were sus-
pended in SDS sample loading buffer, sonicated for 15
s
at
3
microns,
and heated for 5 min
at
100 "C. The samples were subjected
to
SDS-
PAGE either
as
described above (PC2-related peptides)
or
by using a
modification of the method of Schagger and von Jagow (1987) (proin-
sulin-related peptides) as described in Hutton et
al.
(1990). Gels were
subsequently equilibrated for
1
h at room temperature in 25% (v/v)
acetic acid, 10% (v/v) isopropyl alcohol and then for
1
h at room
temperature in 48 mM Tris, 39 mM glycine (pH 9.2) containing 0.1%
SDS and 20% (v/v) methanol. The electrophoresed proteins were
then transferred electrophoretically onto nitrocellulose paper and
subjected to immunoblot analysis as described previously (Guest
et
al.,
1991). Molecular size calibration was achieved with lysozyme,
0-
lactoglobulin, carbonic anhydrase, ovalbumin, BSA, and phosphoryl-
ase b (Bio-Rad).
Assays-Protein was determined by a dye-binding technique
(Bradford, 1976) with BSA as standard. Carboxypeptidase
H
activity
was determined by
a
radiometric assay
as
described previously (Guest
et
al.,
1989b).
P-N-Acetylglucosaminidase
was determined by the
fluorometric method of Barrett and Heath (1977) and galactosyl
transferase from the incorporation of [U-14C]UDP galactose into
soybean trypsin inhibitor (Bretz and Staubli, 1977). NADPH cyto-
chrome c reductase and alkaline phosphatase were determined by the
spectrophotometric methods of Sottocasa
et
al.
(1967) and Bowers
and McComb (1975), respectively.
RESULTS
Molecular Composition
of
35S-Labeled Proinsulin and PC2
Immunoprecipitates
in
Puke-Chase-labeled
Islets-Islets in-
cubated for
1
h in the presence of 16.7 mM glucose incorpo-
rated
1.2%
of the [35S]methionine from the labeling media
into trichloroacetic acid-precipitable material. Approximately
8%
of the incorporated radioactivity appeared in proinsulin-
related immunoprecipitates and
0.1%
in that of PC2. Under
these conditions only rat proinsulin I1 and not proinsulin
I
was labeled since the latter does not contain a methionine
residue (Smith, 1966).
In agreement with previous studies (see Docherty and Stei-
ner, 1982; Hutton
et
al.,
1987; Davidson
et al.,
1988) proinsulin
to insulin conversion commenced approximately 30 min after
synthesis via intermediates which co-migrated with the des-
31,32- and des-64,65-proinsulins and proceeded with apparent
first order kinetics equivalent to a
loss
of 0.83%
of
the initial
proinsulin/min
(tl,z
=
60 min) (Figs.
1A
and
2).
In contrast,
PC2 processing occurred more slowly. By completion
of
the
1-h labeling period, approximately 95%
of
the immunoprecip-
itable PCB-related radioactivity appeared as a protein of 76
kDa (Fig. 1B). After this time, the 76-kDa component de-
creased over 4-5 h with a
loss
of
0.35%/min
(tl/z
=
140
min)
(Figs. 1B and 2). Proteins of 71 and 68 kDa appeared by the
end of the
1st
h of chase incubation and together comprised
23% of the total immunoprecipitable radioactivity. These
forms disappeared between
1
and 3 h of the chase period to
be replaced
by
a 66-kDa protein. This in turn was converted
between
3
and 7 h to a diffuse 64-66-kDa form which ac-
PC2
Biosynthesis
22403
IIIII
6
98
-
43
-
FIG.
1.
Molecular composition
of
35S-labeled proinsulin and
PC2 immunoprecipitates in pulse-chase-labeled islets.
Islets
were labeled for
1
h in Krebs incubation medium containing 16.7 mM
glucose and 0.2 mCi ['"Slmethionine and then resuspended in non-
radioactive Dulhecco's modified Eagle's medium containing
4.5
mM
glucose and
10%
fetal calf serum. The cellular forms of
A,
proinsulin
and
H,
PC2 were immunoprecipitated upon termination of the incu-
hations at the indicated times and subjected to PAGE and fluorog-
raphy as described under "Experimental Procedures." The migration
of
molecular size standards, proinsulin
(PI),
proinsulin conversion
intermediates
(IN?'),
and insulin
(INS)
is
indicated. Similar results
were obtained on three separate occasions.
02468'
time
(h)
FIG. 2.
Time course of proinsulin and PC2 conversion.
The
extent of conversion
of
proinsulin
(0)
and the 76-kDa form of PC2
(0)
was derived from Fig.
1
by densitometric scanning of the fluoro-
graphs. The results are the percentage of precursor remaining and
are expressed relative to the total immunoprecipitable radioactivity
at the indicated times (h).
counted for approximately 91% of the radioactivity immuno-
precipitated at 17 h.
N-Glycanase Treatment of "S-Labeled PC2-N-Glycanase
treatment resulted in a molecular size reduction of 7 kDa for
both the 76- and 64-66-kDa forms of PC2, producing peptides
of
69 and 57-59 kDa, respectively (Fig.
3).
A similar reduction
in size has been reported previously for the native insulin
secretory granule enzyme following N-glycanase treatment
(Bennett et
al.,
1992) and is consistent with the presence of
three consensus sequences for N-linked glycosylation in rat
PC2 (Hakes et
al.,
1991).
Secretion of Proinsulin- and PC2-related Peptides from
"S-
Labeled Islets-Islets were labeled for
1
h and the proinsulin-
and PCB-related peptides immunoprecipitated from media
collected over eight successive intervals
(1
h) of chase incu-
bation in the presence of 16.7 mM glucose. The proinsulin-
12
34
IIII
98
-
h
0
2
66-
v
c
s
m
.
1
0
."
c
&
43
-
N-glycanase
-
+-
+
FIG.
3.
N-Glycanase treatment of '%-labeled PC2.
Islets
were pulse-chase labeled as in the legend for Fig. 1, lysed, and either
mock-treated
or
suhjected to N-glycanase treatment as described
under "Experimental Procedures." PC2 immunoprecipitates are from
a
1-h pulse
(lanes
I
and
2)
and
a
1-h pulse plus 17-h chase
(lanes
3
and
4).
Similar results were obtained on two separate occasions.
123.I567X
A
IIIIIIII
PI-.."
8u88a
S
B
98
-
-
2
66-
x
r
e
-
8.
43
-
FIG.
4.
Secretion of proinsulin- and PC2-related peptides
from pulse-chase-labeled islets.
Islets were labeled for 60 min in
Krebs medium containing 16.7 mM glucose and 0.2 mCi of
[""S]
methionine and then resuspended in nonradioactive Dulbecco's mod-
ified Eagle's medium containing 16.7 mM glucose and 10% fetal calf
serum. The medium was again replaced
at
1-h intervals
for
up to
8
h
of chase incubation, and the molecular forms of
A,
proinsulin and
R,
PC2 were immunoprecipitated and subjected to PAGE and fluorog-
raphy. Similar results were obtained on four separate occasions. The
abbreviations are
as
in Fig.
1.
related peptides appeared initially in the medium within 60
min of chase incubation and were comprised of a mixture of
proinsulin, intermediates, and insulin with a shift to increas-
ing proportions of insulin at later time intervals (Fig.
4A).
Approximately 90% of the initial islet proinsulin-related im-
munoreactivity was released over the 8-h chase period with a
tIl2 of
3.3
h (Fig.
5).
By contrast, release of the PCB-related
peptides was not detected until the 2nd h of chase incubation.
After this time the quantity of PC2-related material secreted
(approximately 70% of the cellular content) and the rate of
release
(
t1r2
=
3.2
h) were comparable to that of insulin (Figs.
4B
and
5).
Also in contrast to the proinsulin-related peptides,
the precursor form of PC2 was not released (Fig. 4B). The
secreted material was comprised initially of only the 66-kDa
intermediate and then increasing proportions of the 64-66-
kDa form at later intervals. Secretion of both the proinsulin-
22404
PC2
Biosynthesis
0
2
4
6
8
Time
(h)
FIG.
5.
Time course
of
secretion
of
labeled proinsulin- and
PC2-related peptides.
The amount of released radioactivity asso-
ciated with proinsulin-related
(0)
and PC2-related
(0)
peptides was
derived from Fig.
4
by densitometric scanning of the fluorographs.
The results are cumulative radioactivities released hy the indicated
times and are expressed
as
a percentage of the initial islet content at
the end of the labeling period,
as
determined in
parallel
incubations.
and PCP-related peptides occurred by regulated release of
secretory granule contents and was not caused by cell damage
or constitutive secretion, as neither of the proteins was de-
tected in significant quantity in the medium of control islets
incubated in the presence of a nonstimulatory concentration
of glucose
(3.3
mM) (data not shown).
Ultrastructural Analysis-Ultrastructural labeling of pan-
creatic
/3
cells by the protein A-gold technique showed that
PC2 immunoreactivity was localized predominantly in secre-
tory granules (26 gold particles/pm2) with the majority of
labeling occurring over the granule core (Fig. 6).
A
low level
of
protein A binding was also associated with smaller electron-
lucent vesicles and cytoplasmic elements (2 gold particles/
pm’) but not with lysozomes and mitochondria.
No
labeling
of
sections was observed using antiserum absorbed previously
with 4 mg/ml of the PCZ/glutathione S-transferase fusion
protein (data not shown).
Molecular Characterization
of
Proinsulin and PC2-related
Peptides in Insulinoma Subcellular Fractions-Analysis of the
molecular forms of proinsulin and PC2 in subcellular com-
partments of the secretory pathway was facilitated by using a
transplantable rat insulinoma. This tissue is available in gram
quantities (Chick et al., 1977) and synthesizes and stores
insulin (Gold et al., 1984) and other granule proteins such as
betagranin (Hutton et al., 1987) and PC22 with kinetics similar
to those observed in rat islets. Fractionation of the insulinoma
by Percoll density gradient centrifugation produced six frac-
tions designated 1-6 in order of increasing buoyant density.
Marker protein analysis showed that ER and Golgi elements
were concentrated in fractions
1
and 2, secretory granules in
fractions
3-5,
and lysozomes in fraction
6,
whereas plasma
membranes were distributed across the gradient with the
highest levels observed in fraction
1
(Table
I).
Immunoblot analysis of the fractions showed that insulin
and PC2 immunoreactivities were most abundant in secretory
granules with lower levels in Golgi/ER and lysozomes. By
adjusting sample loadings to give immunoblots of similar
intensity, it became apparent that the molecular forms of the
insulin- and PCB-immunoreactive peptides present in each
fraction varied (Fig. 7). Golgi/ER-enriched fractions con-
tained proteins of equivalent size to the precursor (76-kDa)
and mature (64-66-kDa forms of PC2 but only the precursor
form of insulin (proinsulin). Secretory granules, by contrast,
P. C. Guest,
S.
D.
Arden,
D.
L.
Bennett, A. Clarke,
N.
G.
Rutherford, and
.J.
C. Hutton, unpublished findings.
FIG.
6.
Ultrastructural labeling
of
an islet
@
cell.
An ultrathin
human islet section containing
a
B
cell was labeled by the protein
A-
gold technique
as
described under “Experimental Procedures.” Insulin
granules
(I),
mitochondria
(M),
and lysozomes
(L)
are indicated.
Bar
=
0.5
pm. Similar results were obtained on at least three separate
occasions.
contained only the mature form
of
PC2 and yet significant
amounts of proinsulin along with insulin.
DISCUSSION
The post-translational processing of many proteins enter-
ing the secretory pathway appears to be achieved by a highly
conserved mechanism involving initial cleavage after se-
quences containing basic amino acids by members of a family
of endoproteases related to the bacterial subtilisins (see Doch-
erty and Steiner, 1982; Thomas et al., 1988; Hutton, 1990;
Lindberg and Hutton, 1991). The majority of proproteins
which are secreted in a regulated manner appear to undergo
relatively slow conversion in dense core secretory vesicles,
whereas constitutively secreted proteins and membrane-
bound proreceptors are cleaved rapidly during brief transit
through the trans-Golgi network. Effective regulation of the
converting activities might be achieved by a combination of
compartmentalization of the enzymes and kinetic mecha-
nisms related to substrate concentration, structural features
of the cleavage site, or regulation by the prevailing ionic
conditions (e.g. Ca’+ and pH) of the compartment where
activity is expressed. Further, the processing enzymes them-
selves are likely to be produced as proproteins inasmuch as
all of those identified to date contain cleavage site consensus
sequences. Synthesis as a pro-form may afford protection of
the enzyme from denaturation in early segments of the secre-
tory pathway or provide a mechanism for limiting activity or
targeting the enzyme to specific intracellular compartments.
PC2
Biosynthesis
22405
TARLE
I
Charactwization of insulinoma subcellular fractions
Suhcellular fractions were prepared from rat insulinoma, and protein markers for secretory granules (carboxypeptidase
H),
lysozomes
(8-
N-acetylglucosaminidase),
ER
(NADPH-cytochrome c reductase), Golgi (galactosyl transferase), and plasma membranes (alkaline phospha-
tase)
were assaved
as
described under “Exaerimental Procedures.” ND. not detected.
Protein Carboxypeptidase S-N-Acetyl- NADPH-
glucosamini- cytochrome
c
dase reductase
H
Galactosyl Alkaline
transferase phosphatase
m:: nmol/min/mg nrbitrary units nmol/min/mg nmol/min/mg nmol/min/rng
Homogenate 620 0.20 7.5 4.1 ND 0.37
Fraction
1
33
0.57
7.2 16.2 0.60 1.30
Fraction 2 39 0.48
8.5
11.9 0.34 0.29
Fraction
:3
63
2.54 26.0 6.2 0.13 0.68
Fraction 4
7.5
4.15 38.4 ND
0.05
0.64
Fraction
5
8.0 5.90 83.2 ND ND 0.35
Fraction
G
5.5
1.72 100 ND ND 0.21
I:r:laion
A
123456
Illlll
PI
-“
ISS
-
v-
B
9x
-
z
66-
-
“
-
t
r
,.
..
43
-
FIG.
7.
Molecular composition of proinsulin- and
PC2-re-
lated peptides in insulinoma subcellular fractions.
Aliquots of
insulinoma subcellular fractions were electrophoresed and transferred
to
nitrocellulose paper, which was reacted with either
A,
guinea pig
anti-bovine insulin serum (1:500) (Guest
et al.,
1991)
or
R,
anti-PC2
serum (1:500),
as
described under “Experimental Procedures.” Load-
ings were
as
follows: fractions
1
and 2,
250
pg;
fractions 3-6,
50
pg.
Similar results were obtained on three separate occasions. The
ab-
hreviations
are
as
in Fig.
1.
The present studies show that the proinsulin-converting
enzyme, PC2, is synthesized initially as
a
76-kDa precursor
which undergoes limited post-translational processing to gen-
erate the mature 64-66-kDa form of the protein. N-Glycanase
treatment showed that both the precursor and mature forms
of
PC2 contained approximately 7 kDa contributed by
N-
linked carbohydrate units, consistent with the view that the
difference in size between the forms is
a
result of proteolysis.
The mobility of the mature form on SDS-PAGE was indistin-
guishable from that of the native insulin granule enzyme
which has been shown to be comprised of two molecular
variants (Bennett
et
al.,
1992). One of these commences at a
position equivalent to amino acid 109 of the deduced rat
sequence (Hakes
et
al.,
1991) which implicates cleavage after
the tetrabasic sequence
Arg’05-Lys’”6-Lys1”’-Arg’oR.
The sec-
ond has an
N
terminus beginning at residue 112 which might
be the result of endoproteolytic cleavage after Arg’”8-Gly1”9-
Tyr’”’-Arg”’. Both of these sequences fit the Arg-X-X-Arg
minimal motif thought to be recognized by membrane-asso-
ciated Golgi-resident proteases such as the proalbumin-con-
verting enzyme (Brennan and Peach, 1991) and furin (Molloy
et
al.,
1992). Cleavage after either sequence would result in
removal of a nonglycosylated N-terminal pro-sequence of 9-
10 kDa, which approximates the size difference observed
between the precursor and mature forms in the present pulse-
chase radiolabeling studies. The transient appearance of the
71-, 68-, and 66-kDa immunoreactive peptides during conver-
sion suggests that these may have been intermediates in the
overall process generated by cleavage after other basic sites
which occur within the pro-sequence of the precursor.
The ultrastructural localization of PC2 to secretory gran-
ules and the finding that the newly synthesized enzyme is
secreted from islets in response to glucose stimulation dem-
onstrate that it is segregated into the regulated pathway of
secretion. The time course of its synthesis and packaging into
granules however was significantly slower than that of insulin.
Proinsulin to insulin conversion commenced following an
initial delay of
30
min and thereafter proceeded with a
tlr2
of
approximately 60 min. Regulated release of the newly synthe-
sized proinsulin-related peptides (proinsulin, conversion in-
termediates, and insulin) began after
1
h and proceeded with
a
tllP
of
3.3
h. The finding that the precursor and intermediate
forms of insulin were secreted is consistent with previous
studies which have shown that proinsulin conversion occurs
mainly within secretory granules (Orci
et
al.,
1985, 1987;
Davidson
et
al.,
1988). Further evidence for this was provided
by the present immunoblot studies which showed that insu-
linoma fractions enriched in Golgi/ER elements contained
only proinsulin, whereas both proinsulin and insulin were
present in secretory granules.
Proteolytic cleavage of the PC2 precursor was significantly
slower than that of insulin with an initial delay of approxi-
mately
1
h and a
tlrz
of 140 min. In addition, glucose activated
the release of only the 66- and 64-66-kDa forms of PC2 and
then only
at
times more than 2 h after synthesis. After this
time it was released with a
tl12
similar to that of insulin. This
indicated that delivery of PC2 to secretory granules is delayed
by approximately
1
h compared with insulin but that, once
sorted, PC2 and insulin are released in response to secreta-
gogue stimulation with similar kinetics. There was no evi-
dence for secretion of PC2 via the constitutive pathway as it
was not released in the presence of a substimulatory concen-
tration of secretagogue. The finding that only the 66- and 64-
66-kDa forms of the enzyme were released suggested that
proteolysis was initiated in a pre-granule compartment. This
is consistent with the observation that insulinoma Golgi/ER
fractions contained
a
mixture of the 76- and 66-kDa forms of
PC2, whereas secretory granules contained only the mature
enzyme.
The present studies suggest that the delay in transport of
PC2 to secretory granules is caused by retention of the un-
processed precursor in
a
pre-granule compartment such as the
ER. Such retention might be a result of interactions with ER-
22406
PC2
Biosynthesis
resident proteins such as binding protein, which is thought to
promote protein assembly, and protein disulfide isomerase,
which catalyzes thiol oxidation and disulfide exchange reac-
tions (see Pelham, 1989). Retention in the ER might also
result from slow removal of glucosyl residues from N-linked
oligosaccharides, which has been shown in the case of some
secretory proteins to be necessary for transport to the Golgi
complex (Lodish and Kong, 1984).
The observation that the transfer of proinsulin and
PC2
from the ER to secretory granules occurs with different ki-
netics has important implications for secretory granule func-
tion. The results suggest that
at
the point of secretory granule
assembly, the
PC2
which is incorporated is synthesized ap-
proximately
1
h earlier than the condensing prohormone.
Since insulin mRNA translation undergoes rapid and marked
changes in response to ambient glucose concentrations, it
follows that the ratio of
PC2
to proinsulin in the nascent
granule will vary. This will occur irrespective of any effects
glucose may have on
PC2
biosynthesis. Since the
V,,,
of the
proinsulin endoproteases appear to be rate limiting, it follows
that the relative proportions of precursor, processing inter-
mediates, and insulin released might respond to changes in
glucose concentration. Other contributory factors will include
the time required
for
newly formed granules to become com-
petent for exocytosis, the pool size of stored granules, and the
intensity and duration of the stimulus. Reports that glucose
can accelerate proinsulin conversion (Nagamatsu
et
al.,
1987)
and promote preferential release of the newly synthesized
hormone (Sando
et
al.,
1972; Rhodes and Halban, 1987; Arvan
et
al.,
1991) are
of
interest in this context and warrant further
investigation. The abnormalities in circulating levels of proin-
sulin and processing intermediate which occur in persons with
glucose intolerance, noninsulin-dependent diabetes,
or
insu-
linomas are conceivably related to such phenomena (Heaton
et
al.,
1987; Ward
et
al.,
1987).
Acknowledgment-We thank
Dr.
E.
M.
Bailyes for helpful discus-
sion
and
editing
of
the manuscript.
REFERENCES
Arvan, P., Kuliawat, R., Prabakaran, D., Zavacki, A,, Elahi, D., Wang,
S.,
and
Pilkev. D.
(1991) J.
Bid.
Chem.
266.
14171-14174
Axen,
R.;
Porath, J., and Ernback,
S.
(1967)
Nature
215, 1302-1304
Bailyes,
E.
M., Bennett, D. L., and Hutton,
J.
C.
(1992)
Enzyme,
in press
Barr, P.
J.
(1991)
Cell
66, 1-3
Barrett, A. J., and Heath,
M.
F.
(1977)
Lysozomes:
A
Laboratory Handbook
(Dinele.
J.
T..
ed)
DD. 9-145.
North Holland Publishine Co., Amsterdam
Benne$,
D.
L.,’B&&es, E.
M.,
Nielsen, E., Guest,
P.
c.,
Rutherford, N. G.,
Arden.
S.
D.. and Hutton.
J.
C.
(1992) J. Biol.
Chem.
267. 15229-15236
Bowers,
b.
N.,’Jr, and McComb,
R.
B.
(1975) Clin.
Chem.
2’1, 1988-1995
Brennan,
S.
O., and Peach, R.
J.
(1991)
J.
Biol.
Chem.
266,21504-21508
Bradford, M.
M.
(1976)
Anal. Biochem.
72, 248-254
Breshnahan,
P.
A,, Leduc, R., Thomas, L., Thorner, J., Gibson, H. L., Brake,
Bretz, R., and Staubli,
W.
(1977)
Eur.
J.
Biochem.
77,
181-192
Chick, W. L., Warren,
S.,
Chute, R. N., Like, A. A,, Lauris, V., and Kitchen,
Christie, D. L., Batchelor, D. C., and Palmer, D.
J.
(1991)
J.
Biol.
Chem.
266,
Davidson, H.
W.,
and Hutton,
J.
C.
(1987)
Biochem.
J.
245, 575-582
Davidson, H. W., Rhodes,
C.
J.,
and Hutton,
J.
C.
(1988)
Nature
333, 93-96
Docherty,
K.,
and Steiner, D. F.
(1982)
Annu.
Reu.
Physiol.
44,625-638
Gold, G., Gishizky, M. L., Chick,
W.
L., and Grodsky,
G.
M.
(1984)
Diabetes
Gray,
I.
P., Siddle, K., Frank, B. H., and Hales, C. N.
(1987)
Diabetes
36,684-
Guest, P. C., Rhodes, C. J., and Hutton,
J.
C.
(1989a)
Biochem.
J. 257, 431-
Guest,
P.
C., Pipeleers, D., Rossier, J., Rhodes, C.
J.,
and Hutton,
J.
C.
(1989b)
Guest,
P.
C.,
Ravazzola, M., Davidson, H.
W.,
Orci, L., and Hutton,
J.
C.
(1991)
Hakes. D.
J..
Birch. N. P.. Mezev. A,. and Dixon.
J.
E.
(1991)
Endocrinolon
A.
J., Barr,
P.
J., and Thomas, G.
(1990) J.
Cell
Biol.
11 1, 2851-2859
K. C.
(1977)
Proc. Natl. Acad. Sei.
U.
S.
A.
74, 628-632
15679-15683
33,556-561
688
437
Biochem.
J.
264,503-508
Endocrinology
129, 734-740
129,305313063
’
and Leslie. R. D. G.
(1987)
Br.
Med.
J.
294,
145-146
_I I
..
”
Heaton, D. A,, Millward, B. A,, Gray,
I.
P.,
Tun,
Y.,
Hales, C. N., Pyke, D.
A.,
Hutton,
J.
C.
(1982)
Biochem.
J.
204, 171-178
Hutton,
J.
C.
(1990)
Curr.
Opinion
Cell
Biol.
2, 1131-1142
Hutton,
J.
q.,
Penn,
E.
J., and Peshavaria,
M.
(1982)
Diabetologia
23, 365-373
,.,
Penn, E. J., and Peshavaria, M.
(1983)
Biochem.
J.
210, 297-
Hutton,
J.
C.,
Davidson, H.
W.,
Grimaldi, K. A,, and Peshavaria, M.
(1987)
Hutton,
J.
C., Bailyes, E. M., Rhodes,
C.
J., and Guest,
P.
C.
(1990)
in
Peptide
Biochem.
J.
244,449-456
Hormone Secretion: A Practical Approach
(Hutton,
J.
C., and Siddle, K., eds)
pp.
309-336.
Oxford University Press, Oxford,
U.
K.
Kiefer, M. C., Tucker,
J.
E., Joh, R., Landsberg, K. E., Saltman, D., and Barr,
P.
J.
(1991)
DNA Cell
Biol.
10, 757-769
Kirchmair, R., Gee, P., Hogue-Angeletti, R., Laslop, A,, Fischer-Colbrie, R.,
and Winkler, H.
(1992)
FEBS Lett.
297,302-305
Laemmli, U. K.
(1970)
Nature
227, 680-685
Lindberg,
I.,
and Hutton,
J.
C.
(1991)
in
Peptide Biosynthesis and Processing
Lodisb, H. F., and Kong, N.
(1984)
J. Cell
Biol.
98, 1720-1729
(Fricker, L. D., ed) pp.
141-173,
CRC Press, Boca Raton, FL
Molloy,
S. S.,
Breshnahan,
P.
A,, Leppla,
S.
H., Klimpel, K. R., and Thomas,
Nagamatsu,
S.,
Bolaffi,
J.
L., and Grodsky, G.
M.
(1987)
Endocrinology
120,
Nakayama, K., Kim, W.-S., Torii,
S.,
Hosaka,
M.,
Nakagawa, T., Ikemizu, J.,
Ora.
L..
Ravazzola. M.. Amherdt. M.. Madsen.
0..
Vassalll, J.-D.. and Perrelet,
G.
(1992)
Proc. Natl. Acad. Sci.
U.
S.
A,,
in press
1225-1231
Baba, T., and Murakami, K.
(1992)
J.
Biol. Chem.
267!
5897-5900
A.’fi965)
Cell
42,671-681
Perrelet, A.
(1987)
Cell
49,
865-868
.. ..
Orci, L., Ravazzola, M., Storch, M.-J., Anderson, R. G. W., Vassalli, J.-D., and
Pelham, H. R. B.
(1989)
Annu.
Reu.
Cell
Biol.
5, 1-23
Pow,
D.
V., and Clark, A.
(1990)
in
Peptide Hormone Secretion: A Practical
Amroach
(Hutton.
J.
C., and Siddle, K., eds)
pp.
189-210,
Oxford University
Pi&, Oxford,
U.
K.
~~
Rhodes, C. J., and Halban,
P.
A.
(1987)
J.
Cell
Biol.
105, 145-153
Sando, H., Borg, J., and Steiner, D. F.
(1972) J.
Clin.
Inuest.
51, 1476-1485
Schagger, H., and von Jagow, G.
(1987)
Anal. Biochem.
166,368-379
Seidah,
N.
G., Marcinkiewicz, M., Benjannet,
S.,
Gaspar,
L.,
Beaubien, G.,
Mattei,
M.
G.,
Lazure, C., Mbikay,
M.,
and Chretien,
M.
(1991)
Mol. Endo-
crinol.
5, 111-122
Smeekens,
S.
P., Avruch, A.
S.,
LaMendola,
J.,
Chan,
S.
J., and Steiner,
D.
F.
(1991)
Proc.
Natl. Acad. Sci.
U.
S.
A.
88,340-344
Smith,
L.
F.
(1966)
Am.
J.
Med.
40,662-666
Sopwith, A: M., Hutton,
J.
C., Naber,
S.
P.,
Chick,
W.
L., and Hales, C. N.
Sottocasa,
G.
L., Kuylenstierna, B., Ernster, L., and Bergstrand, A.
(1967)
J.
Thomas.
G.,
Thorne, B. A,, and Hruby,
D.
E.
(1988)
Annu.
Reu.
Physiol.
50,
(1981)
Dlabetologia
21, 224-229
Cell
Biol.
32, 415-438
Ward, W.
K.,
LaCava, E. C., Paquette,
T.
L.,
Beard,
J.
C., Wallum,
B.
J., and
323-332
Porte, D., Jr.
(1987)
Diabetologra
30, 698-702