Endocrinology Vol. 138, No. 7 2879-2885
Copyright © 1997 by The Endocrine Society
Hyperinsulinemia-Induced Hypoglycemia Is Enhanced by Overexpression of Connexin 431
Cristina Vozzi,
Domenico Bosco,
Emmanuel Dupont,
Anne Charollais and
Paolo Meda
Department of Morphology, University of Geneva Medical School,
CH-1211 Geneva 4, Switzerland
Address all correspondence and requests for reprints to: Paolo Meda, M.D., Department of Morphology, University of Geneva Medical School, 1 rue Michel-Servet, CH-1211 Geneva 4, Switzerland. E-mail:
PAOLO.MEDA{at}MEDECINE.UNIGE.CH
 |
Abstract
|
|---|
To assess whether cell to cell communications via connexins (Cx)
participate to insulin secretion in vivo, we studied
insulinoma cells (INS1) implanted in rats after stable transfection
with connexin 43 (Cx43). We found that compared to wild-type and
transfected cells, which in vivo express modest levels
of Cx43 and junctional communication, cells overexpressing Cx43
communicated extensively, featured decreased growth, and induced a much
higher hyperinsulinemia. As a result, rats with insulinomas made of
these cells became more severely hypoglycemic than rats implanted with
either wild-type, neomycin-transfected cells or cells transfected with
a Cx43 antisense complementary DNA. Rats implanted with transfected
cells that expressed modest level of Cx43 showed levels of circulating
insulin similar to those in rats implanted with wild-type INS1 cells.
The data show that overexpression of Cx43 influences the growth and
secretion of the implanted insulinoma cells, providing evidence for a
contribution of Cx-mediated cell to cell communication in the
functioning of insulin-producing cells in vivo.
|
Introduction
|
|---|
ALMOST ALL differentiated cell types are
connected by gap junctions, the membrane microdomains that concentrate
connexin (Cx)-made channels for cell to cell exchanges of cytoplasmic
ions and molecules (1, 2, 3). Experiments in adult systems have provided
evidence that Cx-mediated communication may contribute to multiple
functions, including electrical and mechanical synchronization of cells
in excitable tissues, control of cell growth in both normal and tumoral
systems, retention of differentiated characteristics, and secretion of
various products (1, 2, 3, 4). Evidence for a secretory involvement of Cx
channels has been derived from studies on the insulin-producing
ß-cells of the endocrine pancreas, which have indicated that gap
junctional communication significantly contributes to the control of
insulin biosynthesis and release in vitro (5, 6).
Few studies have investigated whether this contribution is relevant to
the in vivo functioning of insulin-producing cells, as it is
also the case, in fact, for most of the functions that have been
attributed to Cx channels in other systems (2, 7). Sustained exposure
to drugs, hormones, and conditions that affect insulin secretion have
been reported to modify in vivo the gap junctions of native
ß-cells and the cell to cell communications these structures ensure
within pancreatic islets (8, 9, 10, 11). As yet, however, we do not know
whether a primary alteration in the Cx channels of islet cells actually
results in changes in insulin secretion that would be relevant to the
in vivo control of blood glucose levels. As a first approach
to address this question, we have implanted normoglycemic rats with
insulinoma cells expressing different levels of connexin 43 (Cx43), a
gap junction protein that is expressed by native pancreatic ß-cells
(12, 13) and have monitored the growth of tumoral cells and their
effects on the circulating levels of glucose and insulin.
|
Materials and Methods
|
|---|
Cells
Cells of the INS1 line (14), which express no Cx and are
uncoupled in vitro (6), were used. Stable transfection of
the gene encoding Cx43 resulted in the selection of clones expressing
different levels of this Cx and coupling, and showing a different
secretion profile (6). Here, we have used three of these clones (e, i,
and n) that, compared to wild-type INS1 cells and to cells transfected
only for neomycin resistance, expressed high levels of Cx43, as judged
at gene, transcript, and protein levels (Fig. 1
). As
additional controls, we used, in a separate set of experiments, two
other clones (a and w) of transfected cells that in vitro
expressed much lower levels of Cx43 (6), as well as a sixth clone of
INS1 cells that had been transfected with a complementary DNA (cDNA)
comprising the sequence coding for Cx43 in an antisense orientation and
for resistance to phleomycin, which was used as the selection
antibiotic (not shown).
View larger version (53K):
[in this window]
[in a new window]
|
Figure 1. a, A cDNA probe detected the endogenous gene
coding for Cx43 in genomic DNA (6 ) of both wild-type (INS1) and
Cx43-transfected clones (INS1Cx43-e, -i, and -n). In contrast, the
transfected gene, which was identified by a mobility similar to that of
the plasmid insert (VECTOR), was detected only in the latter clones. b,
Analysis of total RNA revealed the presence of a native Cx43 transcript
(Cx43-N) only in the heart extract that was used as control. In
contrast, high levels of another transcript (Cx43-T), which had a
slightly higher mobility due to the construction of the transfected
insert, were detected in the three clones of Cx43-transfected cells
(INS1Cx43-e, -i, and -n). No transcript was detected in samples of
liver, which was used as negative control, or in wild-type cells
(INS1). Lanes were loaded with 5 µg (heart) or 10 µg total RNA (all
other lanes). c, Antibodies to Cx43 revealed this protein in membranes
of heart and of the three Cx43-transfected clones. The two
immunoreactive bands detected in Cx43-transfected cells did not have
the same mobility as those in heart, presumably indicating different
levels of phosphorylation of the gap junction protein in the two
systems. In contrast, no Cx43 was detected in samples of liver and
wild-type cells. All lanes were loaded with 50 µg protein.
|
|
Analysis of Cx43 expression by INS cells
Cell clones to be injected were tested for Cx43 expression at
the gene, transcript, and protein levels. Genomic DNA was extracted
from cultures in 10 mM Tris-HCl (pH 8) supplemented with
0.1 M EDTA, 20 µg/ml pancreatic ribonuclease, and 0.5%
SDS. Southern blots of DNA onto nylon membranes were prehybridized 30
min at 65 C and hybridized at this temperature with 105
cpm/cm2 32P-labeled cDNA probe to Cx43, prepared as
previously described (13). Washed membranes were exposed to Kodak XAR-5
films (Eastman Kodak, Rochester, NY) for 1–7 days.
For identification of Cx43 transcript, cell cultures were homogenized
in 2.5 ml 0.1 M Tris-HCl, pH 7.4, containing 2
M ß-mercaptoethanol and 4 M guanidium
thiocyanate. After the addition of solid CsCl (0.4 g/ml), the
homogenate was layered on a 2-ml 5.7 M CsCl-0.1
M EDTA (pH 7.4) cushion and centrifuged for 20 h at 20
C at 150,000 x g, using a SW55 rotor and a Beckman
L8–70M ultracentrifuge (Beckman, Fullerton, CA). Pelleted RNA was
resuspended in 300 µl 10 mM Tris-HCl, pH 8.1,
supplemented with 5 mM EDTA and 0.1% SDS, extracted twice
with phenol-chloroform, precipitated in ethanol, and resuspended in
water. Probes for Cx43 were constructed and used as previously
described (13). For Northern blots, total cellular RNA was denatured
with glyoxal, electrophoresed in a 1% agarose gel (5–10 µg total
cellular RNA/lane) and transferred overnight onto nylon membranes
(Hybond N, Amersham International, Aylesbury, UK). Filters were baked
under vacuum at 80 C for 2 h, exposed for 30 sec to 302 nm light,
stained with methylene blue, and prehybridized for 30 min at 65 C in
5 x SSC (standard saline citrate) containing 0.1% SDS, 5 x
Denhardt’s solution, and 250 mg/ml salmon sperm DNA. Filters were then
hybridized for 18 h at 65 C with 105 cpm/cm2
32P-labeled probe and washed twice at room temperature in 2
x SSC containing 0.1% SDS, followed by two washings at 65 C in
0.5 x SSC and 0.1% SDS. Filters were then exposed to film
(XAR-5, Eastman Kodak Co.) between intensifying screens at -80 C.
Samples of total cellular RNA were similarly extracted from heart and
liver, and used as internal controls in all blots.
For Western blotting, cell cultures were lysed and homogenized by
sonication in 0.02 M Tris-HCl, pH 8.0, supplemented with 20
mM EDTA, 1 µg/ml pepstatin A, 1 µg/ml antipain, 1
mM benzamidine, 200 kallikrein inhibitor units/ml
aprotinin, 2 mM phenylmethylsulfonylfluoride, and 1
mM diisoprophyl fluorophosphate. After a 10-min
centrifugation of the sonicate at 3,000 x g and 4 C,
the supernatant was collected and centrifuged for 60 min at
100,000 x g and 4 C. Pelleted material was resuspended
in a 0.1 M Tris buffer, pH 7.0, containing 20% SDS, 10
mM EDTA, and 2.5% ß-mercaptoethanol and stored at -80
C. The protein content was measured by a detergent-compatible protein
assay kit (500–0-116, Bio-Rad Laboratories, Glattbrugg, Switzerland).
Samples of crude membrane preparations (50 µg protein/lane) were
fractionated by electrophoresis in a 12.5% polyacrylamide gel and
immunoblotted as previously described (12). To this end,
electrophoresed samples were transferred onto Immobilon membranes
(Millipore, Wolketswil, Switzerland) for 18 h at a constant
voltage of 25 V in the presence of 0.02% SDS. After checking for
efficient transfer by Ponceau S staining, the nitrocellulose membranes
were saturated at room temperature in BLOTTO solution (40
mM Tris-HCl, 0.1% Tween-20, and 4% dry milk) and then
incubated for 60 min with a mouse monoclonal antibody against heart
Cx43 (Zymed Laboratories, San Francisco, CA), diluted 1:500. After
repeated rinsing in BLOTTO, the nitrocellulose immmunoblots were
incubated for 60 min at room temperature with a biotinylated serum
(Jackson ImmunoResearch Laboratories, West Grove, PA) against mouse Ig
(diluted 1:500). Filters were incubated for 45 min at room temperature
with alkaline phosphatase-labeled streptavidin (Amersham International,
Little Chalfont, UK) diluted 1:5,000, rinsed repeatedly in 40
mM Tris-HCl supplemented with 0.5 M NaCl and
0.1% Tween-20, and eventually processed for detection of alkaline
phosphatase activity using bromochloroindolyl phosphate-nitro blue
tetrazolium as substrate. Heart and liver samples were processed in a
similar way and used as controls.
Animals
Rats (200–300 g BW) of the inbred NEDH strain were used because
they are immunologically adequate recipients for implantation of the
RIN-derived INS cells (15). The animals were briefly anesthetized by
inhalation of 5% Ethrane (Abbott Laboratories, Cham, Switzerland) and
injected sc in one thigh with 106 INS1 cells that had been
freshly trypsinized from 1-week-old cultures. In all experiments, one
or two rats were injected with each of the six different types of INS1
cells tested (control, neomycin-transfected, Cx43
antisense-transfected, and three different clones of Cx43-transfected
cells). The animals were then returned to standard housing conditions
with free access to water and food. All animal manipulations were
conducted according to the rules of our institutional committee on
animal experiments.
Blood measurement
Samples of venous blood were obtained from the tail of each rat
just before cell implantation and thereafter at regular intervals up to
the time of tumor removal. Small aliquots of blood were analyzed for
glucose concentrations using an Accutrend glucometer (Boehringer
Mannheim, Mannheim, Germany). The remaining blood was collected in
heparinized tubes and centrifuged at 4 C to prepare plasma, for
measurement of insulin concentrations. To this end, we used a RIA with
a charcoal separation step and rat insulin as standard (16).
Tumor sampling
Macroscopically detectable tumors developed exclusively at the
site of injection. All tumors were surgically removed 5 weeks after
cell injection. After measurement of tumor weight and volume, fragments
of each tumor were sampled for light and electron microscopy,
immunoidentification of Cx, total RNA extraction, and evaluation of dye
coupling.
Histology
Fragments of the tumors were fixed in Bouin’s solution and
processed according to standard histological methods. The volume
density of endocrine cells was assessed on sections by computerized
planimetry, using a Leica Quantimet 500 (Leica, Cambridge, UK).
Sections were also processed for insulin and glucagon immunostaining,
as previously described (17).
Analysis of Cx43 expression and coupling in tumors
Northern blots of insulinoma RNA were performed as detailed for
INS cell cultures. Immunofluorescence labeling of cryostat sections of
unfixed tumor fragments was carried out using monoclonal antibodies
against Cx43 (Zymed Laboratories, San Francisco, CA), diluted 1/50.
Sections of rat heart were similarly processed and used as positive
controls. Negative controls were provided by omitting the anti-Cx
antibodies during the first incubation.
For assessment of junctional coupling, fragments of the tumors were
incubated for 30 min at 37 C with 0.8 mg/ml Collagenase 1 (Serva,
Heidelberg, Germany) under continuous shaking. The resulting cell
suspension was diluted in ice-cold Hanks’ solution supplemented with
0.3% BSA and passed through two nylon filters. After repeated rinsing,
2 x 106 cells/ml RPMI 1640 medium were plated in
culture dishes. Immediately after attachment, individual cells were
microinjected with 4% Lucifer Yellow CH (Sigma Chemical Co., St.
Louis, MO), as described previously (12). The incidence of coupling was
determined by scoring whether each injection resulted in cell to cell
transfer of Lucifer Yellow. The extent of coupling was determined by
scoring the number of cells labeled by Lucifer Yellow (including the
injected cells) on photographs taken immediately at the end of each
injection.
|
Results
|
|---|
When studied in vitro, wild-type INS1 cells and cells
transfected with a neomycin resistance gene did not express Cx43 (Fig. 1
), Cx26, or Cx32 and were essentially uncoupled, as judged both by
Lucifer Yellow injections and dual patch-clamp electrophysiology (not
shown). Stable transfection of the gene encoding Cx43 resulted in the
selection of clones expressing different levels of this Cx (6). Here,
we studied the in vivo behavior of three independent clones
of INS1 cells (clones i, n, and e) that stably expressed high levels of
Cx43, as judged at gene, transcript, and protein levels (Fig. 1). This
behavior was compared to that of INS1 cells that expressed levels of
Cx43, which were either undetectable in vitro (wild-type
cells and cells transfected only with a neomycin resistance gene; Fig. 1) or were much lower than those expressed by clones i, n, and e
[Cx43-transfected clones a and w (6) and cells transfected with a Cx43
cDNA in an antisense orientation].
For 2 weeks after the sc injection of these INS1 cell types in the
thigh of control rats, no change was detectable at the site of
injection or in the circulating levels of glucose and insulin (Fig. 2). In contrast, from the third week onward, a tumor was
detectable at the site of injection, and the animals showed persistent
hyperinsulinemia, which caused a progressive drop in their circulating
glucose levels (Fig. 2). From this time on, rats bearing tumors made of
Cx43-transfected cells of clones i, n, and e had higher plasma insulin
and lower blood glucose levels than the rats injected with all other
types of INS1 cells tested (Fig. 2) and displayed smaller tumors.
View larger version (20K):
[in this window]
[in a new window]
|
Figure 2. a, Five weeks after cell injection, all
insulinoma-bearing rats (T-CONT, T-NEO, T-e, T-i, and T-n) had lower
blood glucose levels than nonimplanted rats (CONT). In addition, the
insulinomas made by Cx43-transfected cells (T-e, T-i, and T-n) induced
greater hypoglycemia than those formed by either wild-type (T-CONT) or
neomycin-transfected (T-NEO) cells. Values are the mean ±
SEM for the number of rats indicated. b, Corresponding
insulin levels were significantly higher in the rats injected with
Cx43-transfected cells (T-e, T-i, and T-n). Values are the mean ±
SEM for the number of rats indicated. c, Up to 2 weeks
after cell injection, all rats featured normal circulating levels of
glucose and insulin. Thereafter, the level of glucose decreased with
time, whereas that of insulin increased. Also, rats injected with
Cx43-transfected cells (open squares show the mean
± SEM for T-e, T-i, and T-n cells, four rats per cell
type) became more hypoglycemic and hyperinsulinemic than rats injected
with other INS1 cells (solid squares show the mean
± SEM for T-CONT and T-NEO cells, four rats per cell
type).
|
|
These differences became larger with time (Fig. 2), and the experiment
was terminated in the fifth week, when rats implanted with cells of
clones i, n, and e showed symptoms of profound hypoglycemia (Fig. 2).
At this time, rats bearing tumors generated by these Cx43-transfected
cells had levels of circulating insulin that were 6–9 times those
measured in all other rats. Animals that had been implanted with INS1
clones a and w, which expressed modest levels of Cx43, had levels of
circulating insulin (20, 6 ± 3.2 ng/ml; n = 11) that were in
the high range (P < 0.08) of those observed in rats
implanted with wild-type cells (12.0 ± 1.4ng/ml; n = 6).
Tumors generated by clones i, n, and e were consistently half as large
as those made of either wild-type or neomycin-transfected cells (Fig. 3). Histology revealed that all tumors comprised cords
of endocrine cells, surrounded by abundant vessels and contained within
a thick connective capsule (Figs. 4 and 5). Morphometric measurements revealed that the
endocrine cells had similar volume densities and contained comparable
amounts of immunolabeled insulin in all tumors (Fig. 4). However,
tumors generated by Cx43-transfected cells expressed significantly
higher levels of this Cx, as judged at both transcript and protein
levels, than tumors formed by either wild-type or neomycin-transfected
cells (Fig. 5).
View larger version (56K):
[in this window]
[in a new window]
|
Figure 3. a, All insulinomas removed 5 weeks after cell
injection were solid tumors enclosed by a capsule of connective tissue.
Tumors formed by wild-type (first from left) and
neomycin-transfected cells (second from left) were
similar in size and consistently larger than tumors formed by
Cx43-transfected cells (T-n, T-i, and T-e tumors are first,
second and third from right, respectively). b, On the average,
insulinomas formed by Cx43-transfected cells (T-e, T-i, and T-n) were
half the volume of those formed by either wild-type (T-CONT) or
neomycin-transfected (T-NEO) cells. c, In contrast, the volume density
(Vv) of endocrine cells was comparable in all tumors. Values are the
mean ± SEM for the number of insulinomas indicated.
The bar represents 1 cm.
|
|
View larger version (120K):
[in this window]
[in a new window]
|
Figure 4. Immunostaining for insulin revealed the
cytoplasmic accumulation of the hormone in the endocrine cells of all
insulinomas (A and C), but not in their connective capsule (c). No
obvious difference in insulin content was observed between tumors made
of wild-type (A) and Cx43-transfected (C) cells. B and D are phase
contrast views of A and C, respectively. The bar
represents 20 µm.
|
|
View larger version (101K):
[in this window]
[in a new window]
|
Figure 5. Upper panel, Northern blot analysis
of all insulinomas (T-CONT, -NEO, -e, -i, and -n) revealed the presence
of a single transcript for endogenous Cx43 (Cx43-N), that was much less
abundant than in the heart extract used as a control. In addition,
insulinomas made of Cx43-transfected cells (T-e, T-i, and T-n) also
expressed high levels of a second, transfection-induced transcript of
slightly higher mobility (Cx43-T). Lower panel,
Immunolabeling with antibodies to Cx43 resulted in a punctuate labeling
of cells in the connective capsule (arrows) of all
insulinomas (A and C). In addition, the endocrine cell cords that
formed the bulk of the tumors were immunolabeled in the tumors formed
by Cx43-transfected cells (C), but were not detectable in those formed
by wild-type or neomycin-transfected cells (A). B and D are phase
contrast views of A and C, respectively. The barrepresents 12 µm.
|
|
Comparison of Figs. 1 and 5 also shows that whereas in vitro
wild-type INS1 cells did not express detectable levels of Cx43 (Fig. 1), this was not the case in vivo. Hence, the endogenous
transcript encoding for the gap junction protein was detected in
extracts of all tumors, including those made of wild-type cells (Fig. 5). Immunolabeling revealed that Cx43 was highly expressed by the
connective cells making up the capsule of every tumor as well as by the
insulin-containing cells of tumors generated by Cx43-transfected clones
(Fig. 5). In contrast, under the immunofluorescence conditions used
here, the gap junction protein was usually not detected in the
insulin-containing cells of tumors generated by either wild-type or
neomycin-transfected cells (Fig. 5). In a few instances, however, a
minimal punctate staining attributable to Cx43 was observed on the
membranes of tumors made of wild-type cells (not shown).
In agreement with these findings, endocrine cells freshly dispersed
from tumors generated by clones i, e, and n showed a significantly
higher incidence and extent of junctional communications than these
dispersed from tumors made of wild-type and neomycin-transfected cells,
as assessed by microinjection of Lucifer Yellow (Fig. 6). The overall extent of coupling, evaluated by
multiplying coupling incidence by coupling extent, was 6–7 times
higher in the tumors made of Cx43-transfected cells (329.5, 308.3, and
373.8 arbitrary units in tumors made by cells e, i, and n,
respectively) than in those made of either wild-type or
neomycin-transfected cells (56.7 and 51.8 arbitrary units,
respectively).
View larger version (20K):
[in this window]
[in a new window]
|
Figure 6. Upper panel, Junctional coupling
was detected in 25% of the injections performed in either wild-type
(T-CONT) or neomycin-transfected (T-NEO) insulinoma cells and in 90%
of the injections performed in cells overexpressing Cx43 (T-e, T-i, and
T-n). This difference was highly significant (P <
0.001). Lower panel, After the injections that revealed
junctional coupling, Lucifer Yellow was found transferred between an
average of 2.2 cells in T-CONT and T-NEO tumors and between 3–4 cells
in T-e, T-i, and T-n tumors. The difference reached statistical
significance (P < 0.02–0.03). Shown are the
mean ± SEM of 20 injections for each cell type. *,
P < 0.001;  , P < 0.03
(T-i); P < 0.02 (T-e and T-n).
|
|
|
Discussion
|
|---|
We found that three independent clones of INS1 cells that showed
extensive junctional communication after stable transfection with a
Cx43 cDNA (6) generated tumors with fewer insulin-containing cells, but
still induced much greater hyperinsulinemia than the tumors formed by
wild-type cells or by cells transfected with either a Cx43-unrelated
(neomycin resistance) or antisense sequence. All of these tumors were
comparable in terms of histological organization and immunolabeling for
insulin. We also observed that two other clones of transfected INS1
cells, which expressed much lower levels of Cx43 (6), did not result in
such enhanced hyperinsulinemia, even though they slightly raised the
levels of circulating insulin over those observed after implantation of
wild-type INS1 cells. The two sets of data imply that the changes in
growth and secretion we observed were closely if not causally related
to the transfection-induced increase in Cx expression and gap
junctional communication. They also suggest that the amplitude of the
secretory changes was dependent on the abundance of Cx channels.
Previous observations have documented that junctional communication of
pancreatic ß-cells affects the biosynthesis, storage, and release of
insulin in vitro and have indicated that an enlargement of
this communication is required for the up-regulation of insulin release
(4, 5, 6, 12, 16, 17, 18, 19). The present findings provide evidence that
junctional communication is also operative and influential under the
pathophysiological conditions encountered in vivo as a
result of the development of an insulinoma. However, this study has
also documented marked differences in the communication and secretory
behavior of tumoral cells in vitro and in vivo.
Thus, INS1 cells do not express detectable levels of Cx43 (or, in fact,
of two other Cx that are prominent in secretory cells), are essentially
uncoupled, and synthesize and secrete insulin rather poorly in
two-dimensional cultures (6, 14). In contrast, under in vivo
conditions, these cells rapidly acquire the ability to express Cx43,
but not Cx32 or Cx26 (6, 12, 13), and to establish sizable levels of
junctional communication (6, 12). Noticeably, these changes are also
associated with a marked increase in the cell’s ability to
biosynthesize and release insulin (14, 15). Even though the cause of
this parallel differentiation has not yet been elucidated, the data
suggest that junctional coupling is necessary, if not obligatory, for
the proper control of insulin output in vivo.
In this perspective, it remains to be established by which mechanism a
6- to 7-fold increase in Cx-mediated communications between INS1 cells
results in an enhancement of insulin release, which is large enough to
dramatically perturb the regulation of blood glucose levels in the
intact animal. Conceivably, such a sizable increase in Cx channels
could result in a faster or more widespread cell to cell exchange of
second messengers that are critical for insulin secretion (4, 5, 20, 21, 22, 23). It is also possible that a greater abundance of Cx channels
could significantly favor the coupling of insulin-producing cells
exhibiting markedly different sensitivities to secretagogue stimulation
(19). In this case, coupling could favor the recruitment of larger
numbers of actively secreting cells and/or decrease their threshold
level of stimulation (4, 5, 24, 25). Also unclear at this time is how
the cell to cell exchanges of current-carrying ions and low mol wt
tracers via Cx channels are integrated with the other mechanisms that
depend on signaling via components of the extracellular matrix,
neurotransmitters, and hormones to help coordinate insulin release from
multicellular assemblies (5). The animal model we describe here should
facilitate further investigations of these matters.
Another central question obviously concerns the relevance of these
findings for the physiological control of insulin secretion by
pancreatic ß-cells. Previous observations have documented that
different physiological, pathological, and pharmacological conditions
that affect insulin secretion in vivo are associated with
changes in the expression of gap junctions, Cx43, and/or junctional
coupling of native ß-cells (8, 9, 10, 11, 12). The present findings for tumoral
cells extend these data by showing that a primary alteration in the
level of Cx-mediated communication is sufficient in vivo to
induce major changes in insulin secretion. Together, the available data
suggest that Cx channels are required to ensure a proper response of
the endocrine pancreas to a sustained increase in the demand for
insulin. Direct experimental testing of this hypothesis awaits the
development of transgenic animals in which the pattern of ß-cell Cx
could be selectively (26) and conditionally modified. Analogous
experiments are needed in other systems, as the consistent expression
of Cx43 in all endocrine glands investigated to date (13) raises the
possibility that specific characteristics of junctional communication,
made possible through channels formed by Cx43, may represent a
widespread mechanism to regulate the in vivo output of
several hormones (4).
|
Acknowledgments
|
|---|
We thank L. Burkhardt, K. Casada, F. Cogne, J.-P. Gerber, and E.
Sutter for excellent technical assistance.
|
Footnotes
|
|---|
1 This work was supported by grants from the Swiss National Science
Foundation (32–34086.95), the Juvenile Diabetes Foundation
International (195077), and the European Union (BMH4-CT96–1427). 
Received December 20, 1996.
|
References
|
|---|
-
Kumar N, Gilula NB 1996 The gap junction
communication channel. Cell 84:381–388[CrossRef][Medline]
-
Paul D 1995 New functions for gap junctions. Curr
Opin Cell Biol 7:665–672[CrossRef][Medline]
-
Bennett MVL, Barrio LC, Bargiello TA, Spray DC,
Hertzberg Sáez JC 1991 Gap junctions: new tools, new
answers, new questions. Neuron 6:305–320[CrossRef][Medline]
-
Meda P 1996 The role of gap junction membrane
channels in secretion and hormonal action. J Bioenerg Biomembr 28:369–377[CrossRef][Medline]
-
Meda P 1995 Junctional coupling of pancreatic
ß-cells. In: Huizinga JD (ed) Pacemaker Activity and Intercellular
Communication. CRC Press, Boca Raton, pp 275–291
-
Vozzi C, Ullrich S, Charollais A, Philippe J, Orci L,
Meda P 1995 Adequate connexin-mediated coupling is required for
proper insulin production. J Cell Biol 131:1561–1572[Abstract/Free Full Text]
-
Goodenough D, Musil LS 1993 Gap junctions and
tissue business: problems and strategies for developing specific
function reagents. J Cell Sci 17:133–138
-
Meda P, Perrelet A, Orci L 1979 Increase of gap
junctions between pancreatic B-cells during stimulation of insulin
secretion. J Cell Biol 82:441–448[Abstract/Free Full Text]
-
Meda P, Halban P, Perrelet A, Renold AE, Orci L 1980 Gap junction development is correlated with insulin content in the
pancreatic B cell. Science 209:1026–1028[Abstract/Free Full Text]
-
Michaels RL, Sorenson RL, Parsons JA, Sheridan JD 1987 Prolactin enhances cell-to-cell communication among ß-cells in
pancreatic islets. Diabetes 36:1098–1103[Abstract]
-
Sheridan JD, Anaya PA, Parsons JA, Sorenson RL 1988 Increased dye coupling in pancreatic islets from rats in late-term
pregnancy. Diabetes 37:908–911[Abstract]
-
Meda P, Chanson M, Pepper M, Giordano E, Bosco D, Traub
O, Willecke K, El Aoumari A, Gros D, Beyer EC, Orci L, Spray DC 1991 In vivo modulation of connexin 43 gene expression and
junctional coupling of pancreatic B-cells. Exp Cell Res 192:469–480[CrossRef][Medline]
-
Meda P, Pepper MS, Traub O, Willecke K, Gros D, Beyer E,
Nicholson B, Paul D, Orci L 1993 Differential expression of gap
junction connexins in endocrine and exocrine glands. Endocrinology 33:2371–2378
-
Asfari M, Janjic D, Meda P, Li G, Halban PA, Wollheim
CB 1992 Establishment of 2-mercaptoethanol-dependent
differentiated insulin-secreting cell lines. Endocrinology 130:167–178[Abstract]
-
Chick WL, Warren S, Chute RN, Like AA, Lauris V, Kitchen
KC 1977 A transplantable insulinoma in the rat. Proc Natl Acad Sci
USA 74:628–632[Abstract/Free Full Text]
-
Meda P, Bosco D, Chanson M, Giordano E, Vallar L,
Wollheim C, Orci L 1990 Rapid and reversible secretion changes
during uncoupling of rat insulin-producing cells. J Clin Invest 86:759–768
-
Stefan Y, Meda P, Neufeld M, Orci L 1987 Stimulation of insulin secretion reveals heterogeneity of pancreatic B
cells in vivo. J Clin Invest 80:175–183
-
Philippe J, Giordano E, Gjinovci A, Meda P 1992 Cyclic adenosine monophosphate prevents the glucocorticoid-mediated
inhibition of insulin gene expression in rodent islet cells. J
Clin Invest 90:2228–2233
-
Bosco D, Meda P 1991 Actively synthesizing
ß-cells secrete preferentially after glucose stimulation.
Endocrinology 129:3157–3166[Abstract]
-
Cao D, Lin G, Westphale EM, Beyer EC, Steinberg TH 1997 Mechanisms for the coordination of intercellular calcium signaling
in insulin-secreting cells. J Cell Sci 110:497–504[Abstract]
-
Gylfe E, Grapengiesser E, Hellman B 1991 Propagation of cytoplasmic Ca2+ oscillations in clusters of
pancretic ß-cells exposed to glucose. Cell Calcium 12:229–240[CrossRef][Medline]
-
Atwater I, Rosario L, Rojas E 1983 Properties of
the Ca-activated K+ channel in pancreatic ß-cells. Cell
Calcium 4:451–461[CrossRef][Medline]
-
Kohen E, Kohen C, Thorell B, Mintz DH, Rabinovitch
A 1979 Intercellular communication in pancreatic islet monolayer
cultures: a microfluorometric study. Science 204:862–865[Abstract/Free Full Text]
-
Sorenson RL, Parsons JA 1985 Insulin secretion in
mammosomatotropic tumor-bearing and pregnant rats. Diabetes 34:337–41[Abstract]
-
Bosco D, Meda P 1994 Individual cell-to-cell
contacts rapidly recruit pancreatic B-cells for glucose-induced
insulin-secretion. Acta Anat. 149:148
-
White TW, Chanson M, Huarte J, Meda P 1996 Generation of transgenic mice expressing connexin 32 in pancreatic
ß-cells. Mol Biol Cell [Suppl] 7:461A
Top
Abstract
Introduction
