Levels of the conversion endoproteases PC1 (PC3) and PC2 distinguish between insulin-producing pancreatic islet beta cells and non-beta cells

Abstract

PC1 (PC3) and PC2, members of the mammalian family of proprotein convertases homologous to the yeast Kex2 gene product, are both expressed in pancreatic islets of Langerhans. Recent studies have suggested that PC1 and PC2 are responsible for the conversion of proinsulin to insulin and connecting peptide (C-peptide) in the islet beta cells. However, the insulin-secreting beta cells are not the only cells present in these complex micro-organs, prompting us to evaluate the expression of PC1 and PC2 in islet beta and non-beta cells. Rat islet cells were sorted by autofluorescence-activated flow cytometry to separate beta cells from non-beta cells, and conversion endoprotease levels were analysed by Western blotting. The immunolabel ratio of PC1/PC2 in beta cells was 2.6. Non-beta cells displayed much lower levels of PC1 than beta cells, but twice as much PC2 (PC1/PC2 = 0.05). Post-translational modification of the convertases themselves was found to differ between the cell types. In particular, a 75 kDa precursor form of PC2 (pro-PC2) was found to accumulate in beta cells, whereas only the fully processed 67 kDa form was detected in the non-beta cells. Finally, the quantification of PC1 and PC2 and their precursor forms in transformed cells (insulin-producing beta-TC and glucagon-producing alpha-TC) showed that transformation appeared to be accompanied by unusually high levels of the precursors.

Full text

Biochem.
J.
(1994)
300,
57-61
(Printed
in
Great
Britain)
Levels
of
the
conversion
endoproteases
PCi
(PC3)
and
PC2
distinguish
between
insulin-producing
pancreatic
islet
cells
and
non-f
cells
Marguerite
NEERMAN-ARBEZ,
Vincenzo
CIRULLI
and
Philippe
A.
HALBAN
Laboratoires
de
Recherche
Louis
Jeantet,
University
of
Geneva
Medical
Centre,
1
rue
Michel
Servet,
1211
Geneva
4,
Switzerland
PCI
(PC3)
and
PC2,
members
of
the
mammalian
family
of
proprotein
convertases
homologous
to
the
yeast
Kex2
gene
product,
are
both
expressed
in
pancreatic
islets
of
Langerhans.
Recent
studies
have
suggested
that
PCI
and
PC2
are
responsible
for
the
conversion
of
proinsulin
to
insulin
and
connecting
peptide
(C-peptide)
in
the
islet
,8
cells.
However,
the
insulin-secreting
,3
cells
are
not
the
only
cells
present
in
these
complex
micro-organs,
prompting
us
to
evaluate
the
expression
of
PCI
and
PC2
in
islet
,/
and
non-,f
cells.
Rat
islet
cells
were
sorted
by
autofluorescence-
activated
flow
cytometry
to
separate
/3'cells
from'
no"n-fl
cells,
and
conversion
endoprotease
levels
were
analysed
by
Western
INTRODUCTION
Proprotein
cleavage
at
dibasic
amino
acids
is
characteristic
of
a
family
of
subtilisin-like
proteases
related
to
the
yeast
Kex2
gene
product
[1].
Several
mammalian
homologues
to
Kex2
have
recently
been
identified.
Amongst
these,
furin,
which
is
ubiquit-
ously
expressed,
is
thought
to
be
responsible
for
proprotein
processing
in
the
constitutive
pathway
[2,3],
while
PCI
(also
known
as
PC3)
and
PC2,
which
are
expressed
only
in
neuro-
endocrine
and
endocrine
cells,
are
thought
to
cleave
precursors
in
the
regulated
pathway
[4-10].
One
such
precursor,
proinsulin,
which
is
synthesized
in
pancreatic
islet
cells,
has
been
studied
in
some
detail.
This
propeptide
is
cleaved
at
two
distinct
sites
to
release
the
mature
insulin
molecule
from
the
connecting
peptide
(C-peptide)
[11]:
a
Type
1
activity
cleaves
between
the
B-chain
and
C-peptide,
and
a
Type
II
activity
cleaves
at
the
A-chain/C-
peptide
junction
[12].
Recent
studies
have
provided
evidence
to
suggest
that
PCI
is
responsible
for
Type
I
activity,
cleaving
after
Arg3l-Arg32
[13,14],
and
that
PC2
is
equivalent
to
the
Type
II
endoprotease,
specifically
cleaving
after
Lys64-Arg65
[15].
Although
both
PCI
and
PC2
have
been
shown
to
be
expressed
in
pancreatic
islets
[14],
insulin-secreting
,
cells
are
not
the
only
cells
present
in
these
complex
micro-organs.
Non-fl
cells,
in-
cluding
principally
a
cells
producing
glucagon,
a
cells
producing
somatostatin
and
PP
cells
synthesizing
pancreatic
polypeptide,
are
found
at
the
periphery
of
the
islet
surrounding
a
core
of
cells
[16].
The
peptides
produced
by
non-fl
cells
are
also
initially
synthesized
as
larger
precursors
which
require
endoproteolytic
cleavages
(typically,
but
not
exclusively,
after
pairs
of
basic
residues)
to
produce
the
active
hormone.
We
have
separated
/8
and
non-,f
cells
by
autofluorescence-activated
cell
sorting
(FACS)
and
then
measured
the
relative
activities
of
PCI
and
PC2
in
the
two
cell
populations
by
quantitative
Western
blot
analysis.
The
data
show
differential
levels
of
expression
of
the
two
enzymes.
Native
(primary)
rat
cells
express
much
higher
levels
of
PCI
than
non-fl
cells,
but
only
half
as
much
PC2.
Two
mouse
cell
lines,
a-
and
,3-TC
cells,
secreting
glucagon
[17]
and
insulin
[18]
blotting.
The
immunolabel
ratio
of
PC1/PC2
in
,
cells
was
2.6.
Non-,f
cells
displayed
much
lower
levels
of
PCI
than
,
cells,
but
twice
as
much
PC2
(PC
l/PC2
=
0.05).
Post-translational
modi-
fication
of
the
convertases
themselves
was
found
to
differ
between
the
cell
types.
In
particular,
a
75
kDa
precursor
form
of
PC2
(pro-PC2)
was
found
to
accumulate
in
,
cells,
whereas
only
the
fully
processed
67
kDa
form
was
detected
in
the
non-f
cells.
Finally,
the
quantification
of
PCI
and
PC2
and
their
precursor
forms
in
transformed
cells
(insulin-producing
,3-TC
and
glu-
cagon-producing
a-TC)
showed
that
transformation
appeared
to
be
accompanied
by
unusually
high
levels
of
the
precursors.
respectively,
were
found
to
faithfully
reflect
their
primary
rat
cell
counterparts
in
terms
of
their
relative
levels
of
PCI
and
PC2.
MATERIALS
AND
METHODS
Materials
Standard
chemicals
were
from
Fluka
(Buchs,
Switzerland)
or
Sigma
(St.
Louis,
MO,
U.S.A.).
Islet
Isolation
and
FACS
of
islet
cells
Islets
were
obtained
from
the
pancreas
of
male
Sprague-Dawley
rats
weighing
200-250
g
by
collagenase
digestion
and
then
gently
digested
with
trypsin
in
order
to
obtain
individual
cells
as
described
previously
[19].
These
cells
were
sorted
according
to
their
FAD
autofluorescence
plotted
against
their
forward
light
scatter
using
a
FACStar
Plus
from
Becton-Dickinson
(Erembo-
degem,
Belgium).
The
sorting
procedure
has
been
described
in
detail
previously
[19].
Analysis
of
insulin
and
glucagon
immuno-
reactivity
by
double
antibody
cytochemistry
revealed
that
one
population
contained
more
than
930%
non-f
cells
(of
which
approx.
80
%
were
a
cells)
and
the
other
more
than
95
%
cells
[19].
Insulin-
and
glucagon-secreting
cell
lines
/3-TC
cells,
secreting
insulin
(from
Dr.
David
Gross,
Jerusalem,
Israel),
and
a-TC-6
cells
producing
glucagon
(from
Dr.
Edward
Leiter,
Bar
Harbor,
MN,
U.S.A.)
were
grown
in
Dulbecco's
modified
Eagle's
medium,
10
%
foetal
calf
serum,
15
mM
Hepes
and
16.7
mM
glucose.
Western
blot
analysis
of
PC1/3
and
PC2
Rabbit
antiserum
2B7
against
PCI,
recognizing
the
N-terminus
of
the
mature
enzyme
[20],
was
kindly
provided
by
Dr.
Iris
Lindberg,
New
Orleans,
LA,
U.S.A.
We
are
grateful
to
Dr.
Chris
Abbreviation
used:
FACS,
autofluorescence-activated
cell
sorting.
57
Biochem.
J.
(1
994)
300,
57-61
(Printed
in
Great
Britain)

58
M.
Neerman-Arbez,
V.
Cirulli
and
P.
A.
Halban
Rhodes,
Boston,
MA,
U.S.A.
for
the
gift
of
rabbit
antiserum
'Thumpa',
directed
against
the
last
15
residues
of
the
C-terminal
tail
of
PC2
[21],
and
to
Dr.
G.
Gabbiani,
Geneva,
Switzerland,
for
the
anti-actin
antibody.
SDS/PAGE
was
performed
ac-
cording
to
Laemmli
[22].
All
cell
types
were
extracted
in
sample
buffer
(62.5
mM
Tris/HCl,
pH
6.8,
20%
SDS,
100%
glycerol,
0.01
%
Bromophenol
Blue),
boiled
for
5
min
and
loaded
on
a
4
%
stacking
gel/7.5
%
resolving
gel.
Approx.
0.2
x
106
cells
were
loaded
per
lane.
After
having
been
run
overnight
at
7
mA,
the
gels
were
electrotransferred
for
5
h
at
30-50
V
on
to
nitrocellulose
(Schleicher
and
Schuell).
Immunodetection
was
performed
using
the
ECL
detection
procedure
from
Amersham
International
(Amersham,
Bucks.,
U.K.).
Antibody
dilutions
and
incubations
were
as
previously
described
[14].
Densitometry
measurements
were
performed
by
scanning
the
films
using
a
flat-bed
scanner
(Macintosh)
and
quantifying
the
bands
of
interest
using
the
Image
1.33g
program
(Macintosh).
Band
density
was
shown
to
be
linearly
related
to
antigen
quantity
in
a
control
experiment
in
which
increasing
amounts
of
cell
extracts
(50
000
to
250000
cells)
were
immunodetected
with
the
three
specific
antisera:
sorted
,-
cells
were
used
for
detection
of
PC
1
and
actin,
whereas
a-TC
cells
were
used
for
PC2
blotting.
RESULTS
PC1
and
PC2
levels
in
Islet
cell
types
Rat
islet
cells
were
sorted
by
FACS
as
described
in
the
Materials
and
methods
section.
One
population
('/,
cells')
contained
more
than
950%
,
cells,
whereas
the
other
('non-fl
cells')
contained
more
than
93
%
non-fl
cells,
of
which
some
80
%
were
glucagon-
producing
a
cells
[19].
The
levels
of
the
conversion
endoproteases
PC
1(3)
and
PC2
were
determined
by
Western
blotting.
Figure
1
is
a
representative
blot
of
three
independent
experiments
showing
the
expression
of
PCI
and
PC2
in
transformed
a
(a-TC)
and
,3
(/3-TC)
cells
from
the
mouse,
sorted
non-,8
cells,
sorted
,/
cells
and
whole
rat
islets.
High
levels
of
expression
of
the
mature
forms
of
both
PC2
(67
kDa)
and
PCI
(66
kDa)
were
found
in
whole
rat
islets
(lane
5)
in
addition
to
significant
levels
of
higher-molecular-
mass
forms
(approx.
75
kDa
for
PC2
and
87
kDa
for
PCI).
The
Pc1
PC2
1
2
3 4 5
1
2
3
4
5
(kDal
97.4
_
_
6aW
*66.2
Actin
-+1
*
42.7
Figure
1
Expression
of
endoproteases
PC1
and
PC2
in
primary
(rat)
and
transformed
(mouse)
islet
cell
types
Western
blot
analysis
was
performed
as
described
in
the
Materials
and
methods
section.
Approx.
0.2
x
106
cells
were
loaded
per
lane.
The
positions
of
the
molecular
size
markers
are
shown
on
the
right
and
that
of
the
43
kDa
band
detected
using
the
anti-actin
antiserum
is
shown
on
the
left.
Lane
1,
a-TC;
lane
2,
fl-TC;
lane
3,
sorted
non-fl
cells;
lane
4,
sorted
,
cells;
lane
5,
whole
rat
islets.
2500
2000
1500
1000
500
0
2000
:..
1500
c
500-
0
0
2000
1500
1000
500
0
0
5
10
15
20
104
x
Cell
number
25
30
Figure
2
Linear
relationship
between
band
density
and
quantity
of
antigen
(cell
number)
for
PC1,
PC2
and
actin
antisera
Western
blot
analysis
was
performed
as
described
in
the
Materials
and
methods
section
with
increasing
amounts
of
cell
extracts.
Sorted
,
cells
were
used
for
PC1
and
actin
detection;
a-
TC
cells
were
used
for
PC2
immunoblotting.
The
band
density
was
measured
(in-arbitrary
units)
after
scanning
the
films
with
a
Macintosh
flat-bed
scanner
using
the
Macintosh
Image
1.33g
program.
Four
different
exposures
of
the
same
blot
yielded
similar
results.
distribution
of
these
two
convertases
between
non-flcells
(lane
3)
and
cells
(lane
4)
was
quite
striking.
The
cells
expressed
much
higher
levels
of
PCl
than
non-fl
cells,
in
the
face
of
lower
levels
of
PC2.
This
differential
expression
was
also
seen
in
mouse
fl-TC
and
a-TC
cells
(lanes
1
and
2).
To
ensure
that
band
density
was
proportional
to
the
amount
of
antigen
in
the
experimental
conditions
used
for
this
study,
increasing
amounts
of
cell
extracts
(50000
to
250000
cells)
were
blotted
with
the
specific
antisera
against
PCI,
PC2
and
actin.
Figure
2
shows
band
density
as
measured
using
the
Macintosh
Image
1.33g
program,
expressed
as
a
function
of
cell
number
for
all
three
antisera.
Similar
graphs
were
obtained
for
four
different
exposures
of
the
same
blot.
The
values
used
to
calculate
PC1/PC2
ratios
were
in
linear
parts
of
Pc1
Actin
,
* .
I
.-
I
.
I
.
I
.
.
.
. .
.
*
. *
-
E

Conversion
endoprotease
levels
in
islet
cell
types
59
Table
1
PC1
and
PC2
expression
in
whole
islets,
in
primary
fi
and
non-fl
Post-translational
processing
of
PC1
and
PC2
in
islet
cells
calls.
and
In
B-TC
and
ae-TC
cealls
--
-
I
--
.
.--
-
Densitometric
analysis
of
Western
blots
was
performed
as
described
in
the
Materials
and
methods
section.
The
data
are
represented
as
means
+
S.E.M.
for
three
independent
experiments
and
are
expressed
in
arbitrary
units
normalized
for
actin
content.
PC1
PC2
PC1/PC2
Rat
islets
fi
cells
Non-,f
cells
fl-TC
a-TC
1.3
+
0.3
1.3
+
0.1
0.05
+
0.05
1.24
+
0.4
0.03
+
0.01
0.9
+
0.3
0.5+
0.08
1.1
+
0.08
0.4
+
0.01
1.36
+
0.2
1.4
+
0.1
2.6
+
0.2
0.05
+
0.04
2.9+
0.7
0.01
+
0
In
addition
to
their
differential
levels
of
expression
of
the
two
convertases,
the
two
islet
cell
populations
seemed
to
process
the
conversion
enzymes
themselves
differently.
Figure
3
represents
data
obtained
from
three
independent
observations
in
which
the
relative
contributions
of
the
precursor
forms
(87
kDa
for
PC1,
75
kDa
for
PC2)
and
the
fully
processed
enzymes
(66
kDa
for
PC1,
67
kDa
for
PC2)
to
the
total
immunoreactivity
obtained
for
each
endoprotease
were
determined.
There
was
a
striking
ac-
cumulation
of
the
PC2
75
kDa
precursor
in
the native
,
cells
(32.2
%
±
10.9
of
total
PC2
immunoreactivity),
a
form
which
is
present
at
very
low
levels
in
sorted
non-,f
cells
(3.60%
±
1.9)
(Figure
3b).
More
generally,
the
transformed
mouse
cells
accumu-
lated
larger
amounts
of
the
precursor
molecules
than
of
their
native
counterparts,
and
this
was
observed
not
only
for
PC2
but
also
for
PCI
(Figure
3a).
B
Precursor
120
100
80
60
->
40
0
0
20
c
E
0
E
g
120
0
.
100
+
80
0
60
U
Mature
form
401ET
20
Non-,
f3-TC
a-TC
Figure
3
Processing
of
PC1
and
PC2
precursors
by
islet
cell
types
The
relative
contributions
of
pro-PCi
(87
kDa)
and
mature
PC1
(66
kDa)
to
total
PC1
immunoreactivity
(a),
and
of
pro-PC2
(75
kDa)
and
PC2
(67
kDa)
to
total
PC2
immunoreactivity
(b),
were
calculated
by
densitometric
analysis
of
Western
blots
as
already
described.
The
data
are
expressed
as
means+
S.E.M.
for
three
independent
experiments.
the
curves,
except
for
those
cells
(oc-TC
and
sorted
non-,f
cells)
for
which
PCI
immunoreactivity
was
hardly
detectable.
Table
1
shows
the
densitometric
analyses
performed
on
blots
obtained
from
three
independent
experiments.
All
immuno-
positive
bands
for
each
enzyme
were
summed
and
normalized
for
actin
content,
ensuring
that
the
amount
of
cells
loaded
for
immunodetection
by
anti-PCI
and
anti-PC2
was
the
same.
In
summary,
PCI
levels
were
much
higher
in
than
in
non-fl
cells,
whereas
PC2
levels
were
lower.
The
transformed
cells
faithfully
reflected
their
primary
counterparts.
DISCUSSION
The
recent
discovery
of
a
mammalian
family
of
endoproteases
homologous
to
the
yeast
Kex2
protease
and
responsible
for
the
cleavage
of
proproteins
and
prohormones
at
dibasic
sites
[23-25]
has
led
to
the
identification
of
the
enzymes
involved
in
the
processing
of
a
number
of
precursors
including
proinsulin
[13,15].
While
furin,
a
transmembrane
endoprotease
localized
in
the
Golgi
apparatus,
is
ubiquitously
expressed,
other
members
of
this
family,
such
as
PCI
and
PC2,
have
been
found
only
in
neuroendocrine
and
endocrine
tissues
equipped
with
the
regu-
lated
secretory
pathway.
One
example
of
such
a
tissue
is
the
pancreatic
islet
of
Langerhans,
which
expresses
both
proteases
[14].
It
has
been
thought
up
to
now,
based
on
analysis
of
transformed
cells,
that
,
cells
show
higher
levels
of
expression
of
PC2
than
PC1
[6,26].
Since
on
the
one
hand
islets
consist
of
several
different
endocrine
cell
types
aside
from
a
cells,
and
on
the
other
transformed
,
cells
may
differ
in
their
properties
from
primary
cells,
we
wished
to
determine
the
levels
of
expression
of
these
two
endoproteases
in
primary
islet
fl
and
non-,f
cells.
To
this
end,
rat
islet
cells
were
sorted
into
two
populations
('fi'
and
'non-,f'),
and
levels
of
PCI
and
PC2
were
monitored
by
quantitative
Western
blot
analysis.
For
comparison
with
the
primary
cell
populations,
two
well
differentiated
cell
lines,
fl-TC
and
a-TC
cells
secreting
insulin
and
glucagon
respectively,
were
included
in
the
study.
The
measurement
of
PCI
and
PC2
by
Western
blotting
using
two
unrelated
enzyme-specific
antisera
does
not
allow
for
com-
parison
of
the
absolute
amounts
of
each
enzyme
in
a
given
cell
type.
It
must,
furthermore,
be
assumed
that
the
level
of
PCI
or
PC2
protein
(as
measured
by
Western
blotting)
faithfully
reflects
that
of
the
corresponding
enzymic
activity.
Despite
these
reserv-
ations,
it
is
perfectly
valid
to
compare
the
levels
of
PCI
or
PC2
and
their
ratios
in
different
cell
types,
and
such
a
comparison
reveals
unexpected
differences
between
fi
and
non-,8
cells.
The
level
of
PCI
in
f8
cells
was
>
20-fold
higher
than
in
non-fl
cells.
By
contrast,
PC2
levels
in
non-,f
cells
were
approximately
twice
those
found
in
,
cells.
This
differential
expression
was
found
also
in
the
transformed
a-TC
and
,-TC
cells
from
the
mouse,
indicating
that,
at
least
in
terms
of
endoprotease
expression
(and
assuming
that
the
direct
comparison
of
mouse
and
rat
cells
is
valid
in
this
context),
the
mouse
,-TC
cell
line
is
more
rep-
resentative
of
the
native
f-cell
than
are
INS
[14]
and
RIN-m5F
cells
(M.
Neerman-Arbez
and
P.
A.
Halban,
unpublished
work)
derived
from
rat
insulinomas.
The
antisera
used
in
this
study
are
able
to
recognize
both
the
precursor
and
mature
forms
of
their
cognate
enzymes.
Thus
pro-
PCI
is
initially
processed
by
removal
of
the
first,
N-terminal,
83

60
M.
Neerman-Arbez,
V.
Cirulli
and
P.
A.
Halban
residues
and
thereafter
by
a
C-terminal
truncation
[20,27].
The
antiserum
used
in
this
study
is
directed
towards
the
N-terminal
region
that
is
common
to
both
processed
molecules
and
present
as
an
internal
domain
of
the
precursor.
Pro-PC2
processing
in
other
cell
types
has
been
suggested
to
involve
only
removal
of
the
N-terminal
pro-sequence,
with
no
C-terminal
truncation
[27].
In
keeping
with
this,
in
rat
islets,
pulse-chase
experiments
have
shown
processing
to
a
mature
64-67
kDa
form
of
PC2
[21,28],
which
appears
to
be
recognized
both
by
the
antiserum
used
in
this
study
(raised
against
the
last
15
amino
acids
of
the
C-
terminal
tail)
[21]
and
by
an
antiserum
directed
towards
the
catalytic
domain
(residues
158-391)
[21,28].
Based
upon
these
results,
it
is
thus
assumed
that
in
the
present
study
pro-PC2
and
mature
PC2
will
be
equally
well
recognized.
It
cannot,
however,
be
totally
excluded
from
the
earlier
studies
that
some
limited
truncation
at
the
extreme
C-terminus
of
PC2
may
occur,
and
in
this
event
such
a
processed
form
would
not
be
detected
by
the
antiserum
used
in
this
study.
A
more
detailed
analysis
of
the
molecular
forms
detected
by
the
anti-PCI
and
-PC2
antisera
in
the
various
cell
types
revealed
interesting
differences
in
the
post-
translational
modifications
of
the
endoproteases
themselves.
As
a
general
observation,
the
transformed
cells
accumulated
larger
amounts
of
the
precursor
forms
of
the
conversion
enzymes
(87
kDa
for
pro-PCI
and
75
kDa
for
pro-PC2).
In
fact,
for
the
fl-TC
cells,
pro-PC2
represented
more
than
76%
of
the
total
immunoreactivity.
Amongst
the
native
sorted
cells,
8f-cells
accumulated
larger
amounts
of
pro-PC2
than
did
the
non-,8
cells.
It
is
not
yet
clear
what
these
differences
in
convertase
processing
reflect.
The
biosynthesis
and
processing
of
PC2
has
been
studied
in
some
detail
in
rat
islets
[28],
and
it
appears
that
only
the
mature
66-67
kDa
form
is
present
in
secretory
granules
and
released
after
stimulation
of
regulated
exocytosis
by
glucose,
whereas
both
the
75
kDa
and
67
kDa
forms
are
found
in
endoplasmic
reticulum-
and
Golgi-enriched
fractions
[28].
The
post-translational
processing
of
this
precursor
is
relatively
slow
[27,28]
and
is
believed
to
reflect
an
intrinsic
property
of
pro-PC2
[27].
Non-fl
cells
may
thus
provide
a
better
environment
and/or
a
more
active
enzyme
(the
enzyme
responsible
has
not
yet
been
identified,
although
furin
seems
to
be
excluded
[27])
for
pro-PC2
processing
than
fl-cells.
It
is
perhaps
relevant
to
note
in
this
context
that
f
cells
do
seem
able
to
process
pro-PC1
quite
efficiently,
raising
in
turn
the
possibility
that
the
processing
machinery
for
the
two
endoproteases
is
not
identical
(as
indeed
suggested
by
the
striking
differences
in
processing
kinetics
[27]).
It
must
be
stressed,
however,
that
measuring
the
steady-state
levels
of
the
various
forms
of
PCI
and
PC2
in
cells
does
not
provide
any
information
on
the
precise
kinetics
of
conversion
of
the
endoproteases.
The
observation
of
much
higher
levels
of
PCI
in
fi
than
in
non-
,f
cells
supports
the
hypothesis
that
this
enzyme
is
important
for
proinsulin
processing
[14].
This
hypothesis
was
based
upon
work
by
ourselves
and
others.
The
study
of
rat
proinsulin
I
conversion
in
COS
cells
cotransfected
with
proinsulin
and
conversion
endoproteases
has
shown
that
PCI
is
able
to
convert
proinsulin
to
fully
processed
insulin,
whereas
PC2
cleaves
only
at
the
C-
peptide/A-chain
junction
[26].
Note,
however,
that
COS
cells
release
proteins
only
via
the
constitutive
pathway
and
thus
provide
an
unusual
setting
for
these
enzymes,
which
are
normally
restricted
in
their
expression
to
cells
with
the
regulated
secretory
pathway.
In
addition,
we
have
demonstrated
that
in
transformed
,f
(INS)
cells
which
show
an
abnormally
low
level
of
PCI,
rat
proinsulin
conversion
is
significantly
impaired
[14],
whereas
it
is
rapid
and
efficient
in
AtT20
cells
[29],
which
have
very
high
levels
of
PCi
and
vanishingly
low
quantities
of
PC2
[30].
The
situation
might
be
different,
however,
for
human
proinsulin,
as
studies
in
vitro
have
shown
that
PCI
cannot
cleave
the
C-peptide/A-chain
junction
[13],
whereas
PC2
favours
this
site
(albeit
with
a
preference
for
des-31,32-split
proinsulin
rather
than
intact
pro-
insulin
as
its
substrate
[31]).
In
human
f8
cells,
which
remain
to
be
characterized
with
regard
to
their
endoprotease
levels,
PC2
might
play
a
more
essential
role
in
the
cleavage
of
this
particular
junction.
One
difference
between
human
and
rat
proinsulins
is
the
presence
of
a
basic
residue
(Arg62)
in
a
-4
position
preceding
the
C-peptide/A-chain
junction
of
only
the
two
rat
proinsulins
[32].
This
basic
residue
may
affect
cleavage
at
this
junction
[32,33].
When
compared
with
f8
cells,
PC2
is
the
dominant
regulated
pathway
conversion
endoprotease
expressed
in
non-,f
cells.
Although
the
non-fl
population
obtained
by
FACS
is
composed
of
a
mixture
of
a,
a
and
PP
cells,
the
majority
(80%)
are
glucagon-producing
a
cells.
It
therefore
remains
possible
that
the
low
levels
of
PCI
detected
in
the
non-fl
cell
population
may
reflect
expression
limited to
just
one
non-a/non-fl
cell
subtype.
The
high
levels
of
expression
of
PC2
in
glucagon-,
somatostatin-
and
pancreatic
polypeptide-producing
cells
are
consistent
with
a
possible
role
for
this
enzyme
in
the
post-
translational
processing
of
the
corresponding
precursors.
In
most
cases,
the
conversion
sites
consist
of
Lys-Arg
sequences,
which,
at
least
in
the
proinsulin
molecule,
is
the
preferred
dibasic
cleavage
site
for
PC2
[15,26].
It
is
interesting
to
note
that,
for
proglucagon,
only
those
sites
which
must
be
cleaved
to
release
glucagon
itself
present
the
putative
PC2
consensus
substrate
sequence,
thereby
possibly
accounting
for
the
production
of
this
particular
hormone
in
the
a
cells
of
pancreatic
islets
[34].
The
generation
of
other
active
peptides
from
the
proglucagon
mol-
ecule,
notably
of
glucagon-like
peptide
1
from
the
major
pro-
glucagon
fragment
[34],
involves
cleavage
at
sites
which
are
probably
less
suitable
for
PC2,
i.e.
Arg-Arg
sites.
It
will
be
interesting
to
see
whether
PC1
or
another
related
enzyme
indeed
dominates
in
the
intestinal
L-cells
responsible
for
the
secretion
of
this
interesting
peptide.
If
so,
this
would
be
a
new
example
of
different
endoprotease
levels
being
responsible
for
tissue-specific
differential
processing
of
prohormones.
In
conclusion,
rat
islet
f
and
non-fl
cells
display
differential
levels
of
PCI
and
PC2.
The
data
confirm
that
PCI
is
important
for
proinsulin
conversion,
with
PC2
probably
assuring
the
processing
of
the
other
islet
prohormones,
notably
the
tissue-
specific
conversion
of
proglucagon
to
glucagon.
This
laboratory
is
a
member
of
the
Geneva
Diabetes
Group.
We
thank
Dr.
I.
Lindberg
and
Dr.
C.
Rhodes
for
providing
antibodies
to
PC1
and
PC2
respectively.
This
work
was
supported
by
grant
no.
DK
35292
from
the
National
Institutes
of
Health,
by
Hoechst
AG
and
by
a
Fellowship
(V.C.)
from
the
Juvenile
Diabetes
Foundation
International.
REFERENCES
1
Fuller,
R.
S.,
Sterne,
R.
E.
and
Thorner,
J.
(1988)
Annu.
Rev.
Physiol.
50,
345-362
2
van
de
Ven,
W.
J.,
Voorberg,
J.,
Fontijn,
R.,
Pannekoek,
H.,
van
den
Ouweland,
A.
M.,
van
Duijnhoven,
H.
L.,
Roebroek,
A.
J.
and
Siezen,
R.
J.
(1990)
Mol.
Biol.
Rep.
14,
265-275
3
van
de
Ven,
W.
J.
M.,
Creemers,
J.
W.
M.
and
Roebroek,
A.
J.
M.
(1991)
Enzyme
45,
257-270
4
Seidah,
N.
G.,
Gaspar,
L.,
Mion,
P.,
Marcinkiewicz,
M.,
Mbikay,
M.
and
Chretien,
M.
(1990)
DNA
Cell
Biol.
9,
415-424
5
Smeekens,
S.
P.
and
Steiner,
D.
F.
(1990)
J.
Biol.
Chem.
265,
2997-3000
6
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
7
Nakayama,
K.,
Hosaka,
M.,
Hatsuzawa,
K.
and
Murakami,
K.
(1991)
J.
Biochem.
(Tokyo)
109,
803-806
8
Seidah,
N.
G.,
Marcinkiewicz,
M.,
Benjannet,
S.,
Gaspar,
L,
Beaubien,
G.,
Mattel,
M.
G.,
Lazure,
C.,
Mbikay,
M.
and
Chr6tien,
M.
(1991)
Mol.
Endocrinol.
5,
111-122

Conversion
endoprotease
levels
in
islet
cell
types
Benjannet,
S.,
Rondeau,
N.,
Day,
R.,
Chretien,
M.
and
Seidah,
N.
G.
(1991)
Proc.
Natl.
Acad.
Sci.
U.S.A.
88,
3564-3568
Shennan,
K.
I.
J.,
Smeekens,
S.
P.,
Steiner,
D.
and
Docherty,
K.
(1991)
FEBS
Lett.
284,
277-280
Steiner,
D.
F.,
Kemmler,
W.,
Tager,
H.
S.
and
Peterson,
J.
D.
(1974)
Fed.
Proc.
Fed.
Am.
Soc.
Exp.
Biol.
33,
2105-2115
Davidson,
H.
W.,
Rhodes,
C.
J.
and
Hutton,
J.
C.
(1988)
Nature
(London)
333,
93-96
Bailyes,
E.
M.,
Shennan,
K.
I.
J.,
Seal,
A.
J.,
Smeekens,
S.
P.,
Steiner,
D.
F.,
Hutton,
J.
C.
and
Docherty,
K.
(1992)
Biochem.
J.
285,
391-394
Neerman-Arbez,
M.,
Sizonenko,
S.
V.
and
Halban,
P.
A.
(1993)
J.
Biol.
Chem.
268,
16098-16100
Bennett,
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
Orci,
L.
and
Unger,
R.
(1975)
Lancet
ii,
1243-1244
Powers,
A.
C.,
Efrat, S.,
Mojsov,
S.,
Spector,
D.,
Habener,
J.
F.
and
Hanahan,
D.
(1990)
Diabetes
39,
406-414
D'Ambra,
R.,
Surana,
M.,
Efrat,
S.,
Starr,
R.
G.
and
Fleischer,
N.
(1990)
Endocrinology
(Baltimore)
126,
2815-2822
Rouiller,
D.
G.,
Cirulli,
V.
and
Halban,
P.
A.
(1990)
Exp.
Cell
Res.
191,
305-312
Vindrola,
0.
and
Lindberg,
I.
(1992)
Mol.
Endocrinol.
6,1088-1094
Alarcon,
C.,
Lincoln,
B.
and
Rhodes,
C.
J.
(1993)
J.
Biol.
Chem.
268,
4276-4280
22
23
24
25
26
27
28
29
30
31
32
33
34
Laemmli,
U.
K.
(1970)
Nature
(London)
227,
680-685
Steiner,
D.
F.
(1991)
Curr.
Biol.
1,
375-377
Halban,
P.
A.
and
Irminger,
J.
C.
(1994)
Biochem.
J.
299,
1-18
Seidah,
N.
G.,
Day,
R.,
Marcinkiewicz,
M.
and
Chretien,
M.
(1993)
Ann.
N.Y.
Acad.
Sci.
680,
135-146
Smeekens,
S.
P.,
Montag,
A.
G.,
Thomas,
G.,
Albiges-Rizo,
C.,
Carroll,
R.,
Benig,
M.,
Phillips,
L.
A.,
Martin,
S.,
Ohagi,
S.,
Gardner,
P.,
Swift,
H.
H.
and
Steiner,
D.
F.
(1992)
Proc.
Natl.
Acad.
Sci.
U.S.A.
89,
8822-8826
Benjannet,
S.,
Rondeau,
N.,
Paquet,
L.,
Boudreault,
A.,
Lazure,
C.,
Chretien,
M.
and
Seidah,
N.
G.
(1993)
Biochem.
J.
294,
735-743
Guest,
P.
C.,
Arden,
S.
D.,
Bennett,
D.
L.,
Clark,
A.,
Rutherford,
N. G.
and
Hutton,
J.
C.
(1992)
J.
Biol.
Chem.
267,
22401-22406
Gross,
D.
J.,
Villa-Komaroff,
L.,
Kahn,
C.
R.,
Weir,
G.
C.
and
Halban,
P.
A.
(1989)
J.
Biol.
Chem.
264,
21486-21490
Day,
R.,
Schafer,
M.
K.,
Watson,
S.
J.,
Chretien,
M.
and
Seidah,
N.
G.
(1992)
Mol.
Endocrinol.
6,
485-497
Rhodes,
C.
J.,
Lincoln,
B.
and
Shoelson,
S.
E.
(1992)
J.
Biol.
Chem.
267,
2271
9-22727
Sizonenko,
S.
V.
and
Halban,
P.
A.
(1991)
Biochem.
J.
278,
621-625
Sizonenko,
S.,
Irminger,
J.
C.,
Buhler,
L.,
Deng,
S.,
Morel,
P.
and
Halban,
P.
A.
(1993)
Diabetes
42,
933-935
Orskov,
C.
(1992)
Diabetologia
35,
701-711
Received
20
September
1993/22
November
1993;
accepted
7
December
1993
9
10
11
12
13
14
15
16
17
18
19
20
21
61