| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Laboratory of Molecular Endocrinology, Massachusetts General Hospital, Howard Hughes Medical Institute, Harvard Medical School, Boston, Massachusetts 02114
Address all correspondence and requests for reprints to: Joel F. Habener, M.D., Laboratory of Molecular Endocrinology, Massachusetts General Hospital, 55 Fruit Street, WEL320, Boston, Massachusetts 02114. E-mail: jhabener{at}partners.org
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
70%) in Kir6.2 messenger RNA (mRNA)
levels in isolated rat pancreatic islets as well as in the INS-1
ß-cell line. This effect is reversible, because exposure to low
glucose reinduces Kir6.2 transcript levels. The cognate
KATP channel subunit SUR1 showed similar down-regulation at
high glucose concentration. The KATP channel activity of
INS-1 cells cultivated at high glucose was reduced by 33–51%. In
contrast, glucagon-like peptide-1 (GLP-1) induced Kir6.2
mRNA steady state levels and was able to prevent glucose-dependent
inhibition of Kir6.2 mRNA and KATP channel activity. To
provide further insight into the mechanisms by which glucose and
GLP-1 regulate ß-cell KATP channel genes, we
have cloned and initiated the characterization of the Kir6.2 gene
transcriptional regulatory regions contained within the entire 4.5 kb
flanked by the SUR1 and Kir6.2 genes. Transient transfection
experiments with five deletion constructs in a pancreatic ß-cell line
(INS-1) showed that the proximal 988 bp of the Kir6.2 promoter sequence
contributes only 25–30% to the total basal promoter activity. The
minimal promoter region -67/+140, also encompassing parts of the
5'-untranslated region, confers sensitivity to GLP-1,
which stimulates transcriptional activity of the Kir6.2 minigene by
about 2-fold. We propose that glucose- and GLP-1-dependent
regulation of KATP subunit genes may be important in the
adaptation of ß-cells to changes in secretory demands in
physiological and diseased states. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To ascertain a balanced expression of the two channel subunits in ß-cells, either transcription or protein synthesis needs to be regulated in a coordinated fashion. The coregulated expression of Kir6.2 and SUR1 messenger RNA (mRNA) described to date has been described only in mouse insulinoma cell lines ßTC3 and MIN6. A parallel decrease in both transcript levels was observed upon treatment of these cells with dexamethasone (14). Distinct activities of the Kir6.2 and SUR1 genes are nevertheless clearly demonstrated by tissue-specific differences in the expression levels of each subunit (3, 15, 16). A preliminary assessment of the 1.1-kb promoter region 5'-flanking the human SUR1 and Kir6.2 genes indicates that this region of DNA contains minimal regulatory cis elements involved in the transcription of the two genes (17), although further experiments are needed to clarify the role of distinct transcriptional regulatory elements.
Because of central role played by KATP channels in the regulation of insulin secretion, it is plausible that gene regulation of its subunits is a mechanism underlying adaptation of ß-cells to changes in insulin secretory demands. In the present study we have identified reciprocal and reversible effects of glucose and the incretin hormone glucagon-like peptide-1 (GLP-1) on KATP channel subunit mRNA levels in cultured islet cells. These effects on gene regulation are paralleled by changes in ß-cell KATP channel activity, suggesting that the genetic regulation of KATP channel subunit gene expression may be a relevant mechanism for the control of insulin secretion. To gain a further understanding of how KATP channel subunit genes are regulated, we initiated a characterization of the human Kir6.2 gene and have identified a novel positive cis-acting element as well as a GLP-1-responsive region in the Kir6.2 promoter.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cell culture
INS-1 cells at passage 99 were a gift from Dr. Claes B. Wollheim
(University of Geneva, Geneva, Switzerland) and were maintained in RPMI
1640 medium with 10% FBS, 50 µM ß-mercaptoethanol, 1
mM sodium pyruvate, 10 mM HEPES, 100 U/ml
penicillin G, and 100 µg/ml streptomycin (19). Passages
101–110 were used, and cells were incubated at 37 C in a 5%
CO2-95% air atmosphere with 85% relative
humidity.
Semiquantitative RT-PCR
Freshly isolated islets from 8-week-old Sprague Dawley rats were
incubated in 24-well plates (25/well) at either 5.6 or 16.7
mM glucose for 3 days. Total RNA was isolated (TRIzol,
Life Technologies, Inc., Gaithersburg, MD) and subjected
to first strand synthesis by oligo(deoxythymidine) priming and reverse
transcriptase (Superscript, Life Technologies, Inc.). RNA
samples were tested for the presence of contaminating genomic DNA by
omitting reverse transcriptase in the first strand synthesis step. Two
microliters of each RT reaction were used as input for
subsequent PCR for Kir6.2 or adenosine-phosphorybosil transferase
(APRT) as a reference to adjust for differences in complementary DNA
(cDNA) input. Primer oligonucleotides for Kir6.2 were
5'-GCACCTCCTACCTAGCTGACGAG-3' as sense primer and
5'-GCCTGAGGAACTGCAACTCAG-3' as antisense primer. The corresponding
primers for the APRT gene were 5'-TCCGAATCTGAGTTGCAGC-3' and
5'-CTGCACACATGGTTCCTCC-3'. After each of three consecutive PCR cycles,
equal aliquots were taken from each sample and analyzed by agarose gel
electrophoresis. Video images of ethidium bromide-stained gels were
quantified using ImageQuant software (Molecular Dynamics, Inc.,
Sunnyvale, CA). Only experiments in which linear amplification was seen
and the ratios of APRT to Kir6.2 signals were comparable in all three
consecutive cycles were used for quantification.
Northern blot
INS-1 cells were incubated for different lengths of time or at
various glucose concentrations and in the absence or presence of 10
nM GLP-1. Total RNA was isolated by the single
step guanidinium-isothiocyanate method with a commercial reagent
(TRIzol, Life Technologies, Inc.). Twenty to 30 µg total
RNA were run on 1% agarose gels containing 6% formaldehyde and
blotted overnight onto nylon membranes (Hybond
N+, Amersham Pharmacia Biotech,
Aylesbury, UK). The blots were dried in a vacuum oven at 80 C for
1 h and prehybridized for 30 min at 65 C in Rapid-Hy6 buffer
(Amersham Pharmacia Biotech). Hybridization was then
performed at the same temperature for 4–16 h with the addition of
random primed [32P]DNA probes prepared from
corresponding cDNAs using a commercial kit (RadPrime, Life Technologies, Inc.). The probe-hybridized membranes were washed
twice in 2 x SSC (standard saline citrate) and twice in 0.1
x SSC-0.1% SDS at 55 C. After exposure to film on an intensifying
screen at -70 C for 12–72 h, the signal intensities were measured
densitometrically using ImageQuant software (Molecular Dynamics, Inc., Sunnyvale, CA). The values were normalized to the
intensity of the ß-actin band.
Preparation of cDNA probes
A 1.16-kb full-length rat Kir6.2 cDNA probe was generated by RT
of INS-1 RNA with the Superscript preamplification system (Life Technologies, Inc.), and subsequent PCR amplification (TaKaRa,
Takara Shuzo Co., Otsu, Japan) with 5'-GCCATGCTGTCCCGAAAGGG-3' as
forward primer and 5'-GCAACTCAGGACAAGGAATCTGG-3' as reverse primer. A
0.63-kb rat ß-actin cDNA probe was amplified by PCR from INS-1 cDNA
using the primers 5'-TACAACCTCCTTGCAGCTCC-3' and
5'-GGATCTTCATGAGGTAGTCTGTC-3'. The rat SUR1 probe was obtained by RT of
INS-1 RNA and PCR amplification of a 1.05-kb fragment (forward,
5'-TTGCTGAAACTGTGGAAGGACTCAC-3'; reverse,
5'-TTCAGGACCATCACTAGGTCTGCAC-3').
Western blot
INS-1 cells were lysed in RIPA buffer (150 mM NaCl,
1% Igepal CA-630, 0.5% deoxycholate, 0.1% SDS, 50 mM
Tris-HCl (pH 8.0), 10 mM Na pyrophosphate, 10
mM NaF, 10 mM EDTA, 1 mM Na
orthovanadate), and 100 µg protein were separated on a 7.5%
SDS-polyacrylamide gel and electrophoretically transferred onto a
polyvinylidene difluoride membrane (Millipore Corp.,
Bedford, MA), and immunoreactivities were detected with the ECL Western
analysis system (Amersham Pharmacia Biotech). The signal
intensities were measured densitometrically using ImageQuant software
(Molecular Dynamics, Inc.). Anti-SUR1 antibody was raised
in a rabbit against the synthetic peptide (C)KDSVFASFVRADK-COOH,
corresponding to the 13 C-terminal amino acid residues of human and
murine SUR1.
Whole cell recording of ATP-sensitive K+currents
Cells were bathed in a standard extracellular solution
containing 138 mM NaCl, 5.6 mM KCl, 2.6
mM CaCl2, 1.2 mM
MgCl2, 10 mM HEPES (295 mosmol; pH
adjusted to 7.4 with NaOH, 4 mM), and 0.8 mM
D-glucose. Tolbutamide (100 µM in standard
extracellular solution) was applied to individual cells by focal
application from micropipettes using a PicoSpritzer II pressure
ejection system (General Valve, Fairfield, NJ). A gravity-fed bath
superfusion system was used to exchange and refresh bath solutions.
Tolbutamide was obtained from Sigma. Whole cell currents
were measured under voltage clamp using the whole cell configuration.
Patch pipettes pulled from borosilicate glass [Kimax-51 (Kimble Glass
Inc., Vineland, NJ); tip resistance, 2–4 M
] and
were fire polished. Pipettes were back-filled with K-pipette solution
containing 95 mM
K2SO4, 7 mM
MgCl2, 100 µM EGTA, 5
mM HEPES (pH adjusted to 7.4 with NaOH, 2
mM). The patch pipette was connected to a Heka Electronik
EPC-9 patch clamp amplifier interfaced with a PC computer running Pulse
software (Instrutech Corp., Mineola, NY). The series resistance and
cell capacitance were measured after break-in to the whole cell
recording mode, and currents were digitized and stored using Axoscope
software (Axon Instruments, Foster City, CA) for subsequent
analysis.
86Rb+ efflux
measurement
INS-1 cells were incubated at either 5 or 25 mM
glucose in the presence or absence of 10 nM
GLP-1. The medium was renewed every 24 h. Sixteen
hours before the experiment, the cells were split into 24-well plates
and adjusted for equal densities. After 72 h of incubation, the
medium was replaced by Krebs-Ringer buffer, containing 0.5% BSA and 40
µCi/ml 86RbCl (1.3 mCi/ml; Amersham Pharmacia Biotech). After 90 min, the metabolic inhibitors
oligomycin (2.5 mg/ml) and 2-deoxy-D-glucose (1
mM) were added, and the cells were further incubated for 15
min. The plate was then transferred on ice, and the cells were rinsed
quickly with ice-cold Krebs-Ringer buffer, 0.5% BSA, oligomycin,
2-deoxy-D-glucose, 100 µM ouabain, and either
100 µM tolbutamide or 50 µM diazoxide. All
chemicals were purchased from Sigma. Still on ice, washing
solution was collected and replaced every minute five times in a row
and analyzed for 86Rb radioactivity. Fractional
86Rb+ efflux within 5 min
is determined as the percentage of
86Rb+ content at the
beginning of the sampling.
86Rb+ efflux carried
through ATP-sensitive K+ channels was established
by calculating the difference of fractional
86Rb+ release in presence
of diazoxide and tolbutamide.
Cloning of the Kir6.2 gene promoter
An arrayed P1-phage human genomic library was screened by PCR
using oligonucleotide primers derived from the 3'-region of the human
SUR1 gene (5'-CGTGCACTGACCTTCTGTCCAGGGG-3' and
5'-AGGGAGAGGGGTGGGAAGAGTCCAAA-3'), essentially as previously described
(20). P1 clone 83c8 was shown to contain both SUR1 and
Kir6.2 sequences by Southern blot analysis using gene-specific probes,
and this was further confirmed by direct sequencing. PCR analysis
indicated that clone 83c8 lacks exon 1 of the SUR1 gene. The proximal
2.5-kb portion of the Kir6.2 promoter sequence was amplified by PCR
using the high fidelity Pfu polymerase
(Stratagene, La Jolla, CA) with clone 83c8 as template,
5'-GCCTGCCTCTGTTCCACTTA-3' as forward primer, and
5'-CTCGGTGGGCACCTTCTCACC-3' as reverse primer. The primers were
designed based on the published genomic sequence of the human PAC clone
pDJ239b22 encompassing the genes for SUR1 and Kir6.2 (GenBank accession
no. AC003969). The purified PCR product was cloned into pCR2.1
(Invitrogen, Carlsbad, CA). A second promoter fragment was
amplified by PCR, again using Pfu polymerase with sense
primer 5'-SpeI-CCTTCGTCCGTGCAGACAAGTGA-3' and
antisense primer 5'-ATGGAAGAACCCCTGACCATCTGC-3'. The 2.76-kb PCR
fragment and the pCR2.1 vector containing the 2.5-kb sequence were cut
with SpeI and religated to obtain a 4828-bp human Kir6.2
promoter construct corresponding to nucleotides 98,339–103,167 of the
human PAC clone (GenBank accession no. AC003969). The entire fragment
was then cut out by XhoI and limited digestion with
HindIII and ligated into the pXP2 vector containing
the coding sequence of firefly luciferase cDNA (21). The
proximal 2.5-kb promoter sequence was subcloned from pCR2.1 into pXP2
in a identical fashion. The two 5'-deletions constructs, -67/+140 and
-348/+140, were obtained by digesting the 2.5-kb PCR fragment with
TaqI and NaeI, respectively, blunted, and ligated
into pXP2/SmaI. The promoter constructs -656/+140 and
-1651/+140 were obtained by digesting the 2.5-kb PCR fragment with
BamHI and DraI, respectively, which were cloned
into pCR2.1, cut again by XhoI and HindIII, and
ligated into pXP2. The -988/+140 construct was generated by PCR
(Pfu polymerase), using the 4.8-kb clone as template and
5'-GCCTAGCCCAGGTCGGTCTCC-3' and 5'-CTCGGTGGGCACCTTCTCACC-3' as forward
and reverse primers, respectively. The product was cloned into pCR2.1,
digested by XhoI and HindIII, and religated into
pXP2. The orientation and length of each construct were confirmed by
sequencing analysis. The search for possible transcription
factor-binding sites was carried out by applying the WWW-based
program Mat-Inspector program version 2.2 (22).
Transfections
Full-length promoter-pXP2, 5'-deletion constructs, and
promoterless vector pXP2 (each 1 µg/well) were transiently
transfected into INS-1 cells grown in six-well tissue culture plates by
liposomal DNA transfer (Lipofectamine, Life Technologies, Inc.). Transfection was performed according to the
manufacturer’s recommendations, and the cells were kept in serum-free
medium and 5 mM glucose. After 5 h, the medium was
changed according to the experimental setting, and the cells were
further incubated for another 48 h. In cases of GLP-1
treatment, the hormone was added freshly at 0 and 24 h.
Recombinant GLP-1-(7–36)amide was obtained from
Peninsula Laboratories, Inc. (Belmont, CA), and was used
at a 10-nM concentration unless otherwise stated. The
solvent for GLP-1 was cell culture medium. Luciferase
expression was determined by the luciferase assay system (Promega Corp., Madison, WI), and values were normalized to protein
content in each extract.
Statistical analysis
Data are expressed as the mean ± SEM.
P values were calculated using Student’s t test
for unpaired values.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). When
normalized to APRT signals, Kir6.2 levels were reduced by 70% in
islets cultured at high glucose (16.7 mM) as opposed to
islets kept in conditions of low glucose (5.6 mM; Fig. 1B).
Thus, exposure of rat islets to high glucose causes a decrease in the
steady state mRNA levels of Kir6.2.
|
). The decline in steady state levels
of Kir6.2 mRNA occurred at glucose concentrations above 10
mM, as demonstrated in studies of INS-1 cells during a 72-h
incubation (Fig. 2B).
|
, 10 nM
GLP-1 induced a time-dependent 3-fold enhancement of the
steady state levels of Kir6.2 mRNA over a 72-h course. This finding may
reflect changes in either gene transcription and/or mRNA turnover.
Loss of Kir6.2 mRNA due to high glucose is paralleled by changes in
SUR1 mRNA and can be reversed by low glucose or GLP-1
INS-1 cells were incubated for 72 h at 25 mM
glucose, after which Kir6.2 mRNA levels were reduced to 36 ± 10%
(mean ± SEM) compared with those in cells kept at 5
mM glucose (100%; Fig. 3A).
Interestingly, mRNA levels of the other constituent of the
ATP-sensitive potassium channel, the sulfonylurea receptor 1 (SUR1),
was reduced to a similar extent (26 ± 6%). The values did not
change after an additional 24 h of incubation under the same
conditions. However, lowering the glucose concentration from 25 to 5
mM reversed the effect of high glucose, and mRNA levels
recovered to 121 ± 3% for Kir6.2 and to 95 ± 5% for SUR1.
Under these rescue conditions, the presence of 10 nM
GLP-1 had no obvious additional effect. At 5
mM glucose, administration of 10 nM
GLP-1 for 72 h raises the transcript levels of Kir6.2
and SUR1 to 288 ± 58% and 311 ± 119%, respectively.
Similar results were obtained with 10 µM forskolin,
suggesting that cAMP mediates the effect of GLP-1 on
KATP mRNA expression. At 25 mM
glucose, GLP-1 was less effective at raising
KATP subunit transcript levels, but did prevent
the loss of Kir6.2 and SUR1 mRNA usually observed under such high
glucose concentrations (98 ± 27% and 102 ± 33%,
respectively).
|
). A 72-h incubation of INS-1 cells in 25 mM glucose
decreased SUR1 protein levels in INS-1 cells to 36 ± 2% compared
with those in cells incubated at 5 mM glucose. Similar to
observations made with Northern blots, 10 nM
GLP-1 attenuated the effects of high glucose on SUR1
protein expression levels, which were brought back to 130 ± 16%.
Corresponding to observations on mRNA levels in 25 mM
glucose, the administration of 10 µM forskolin mimics the
effect of GLP-1 by normalizing SUR1 protein levels (Fig. 3C). This suggests that the expression of SUR1 protein occurs at least
in part through the generation of cAMP. However, when applied at low
glucose conditions (5 mM), forskolin has no further
stimulatory effect on the expression of SUR1 protein, although SUR1
mRNA levels are increased (Fig. 3A). SUR1 protein expression was not
affected by the addition of 20 mM mannitol to low glucose
medium (Fig. 3C). Likewise, no changes in Kir6.2 mRNA steady state
levels were observed upon administration of 20 mM mannitol
(data not shown), ruling out a generalized effect on gene expression
due to osmotic stress or changes in cell volume. Due to the lack of a
suitable antibody, we were unable to confirm our Northern blot data
showing that these observations also hold true for Kir6.2 protein
expression. Nevertheless, it is conceivable that both
KATP subunits are coordinately expressed, as only
the concerted expression and association of Kir6.2 and SUR1 subunits
lead to the formation of fully functional KATP
channels.
Exposure to high glucose and GLP-1 results in changes
in KATP channel activity in INS-1 cells
It is conceivable that variations in Kir6.2 and/or SUR1 gene
expression will affect KATP channel activity and
therefore influence the whole cell membrane conductance.
Tolbutamide-sensitive KATP currents were measured
using whole cell recording of voltage clamped INS-1 cells and were
normalized to the cell capacitance (Fig. 4A). INS-1 cells that have been kept at
25 mM glucose for 3 days exhibited a 33% decrease in
KATP current compared with control cells
incubated at 5 mM glucose (Fig. 4B). An alternative method
to study the activity of potassium channels is to measure the efflux of
the K+ analog 86Rb. After
incubating INS-1 cells at different glucose concentrations for 3 days,
the cells were loaded with the radioisotope 86Rb.
After 86Rb incubation, the cells were washed, and
86Rb efflux into the medium was determined within
the first 5 min. These experiments were performed in the presence of
diazoxide or tolbutamide to specifically establish
86Rb efflux through the ATP-sensitive potassium
channel. The fractional release of 86Rb in INS-1
cells at 25 mM glucose was only 51% of that from control
cells kept at 5 mM glucose (Fig. 4C).
KATP activity could be partially restored (75%)
by adding 10 nM GLP-1 to the high glucose
medium, supporting our previous observation of the stimulatory effect
of GLP-1 on Kir6.2/SUR1 expression. Thus, ß-cell
KATP activity is dependent on regulation of
KATP channel subunit genes.
|
). Several 5'-truncation constructs
were made to study transcriptional activity in given promoter segments
(Fig. 5B). INS-1 cells were transfected with each construct, and
luciferase activity was determined after 48 h of incubation. The
promoter activity is expressed as fold activity compared with that of
the empty reporter vector. Basal activity ranges from 40-fold for the
shortest construct (-67/+140) to 1200 fold for the longest construct
(-4689/+140; Fig. 5C). These results suggest that multiple controlling
elements reside within the entire length of the 4.57-kb promoter.
Particularly the very proximal 5'-flanking region (-67/+140) and two
regions between nucleotides -656/-988 and nucleotide -1651/-2377
contain important stimulatory elements that contribute considerably to
the overall basal promoter activity. Noteworthy is also the region
between -348/-67, which contains an Alu repeat and appears to exert
an inhibitory effect on gene transcription (17).
|
).
When applied to 5'-truncated promoter constructs, the luciferase
reporter assay showed similar fold effects of 10
nM GLP-1 on all of the different
constructs, but had no effect on the empty basic vector (Fig. 6B). This
suggests that a GLP-1-responsive element resides within
the very proximal 200 bp of the 5'-flanking promoter region including
the first 140 nucleotides of the 5'-untranslated region.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
One potential site involved in the regulation of glucose responsiveness is the modulation of the expression of ATP-sensitive K+ channel subunits. Rat pancreatic islets exposed to 16.7 mM glucose during 24 h show signs of decreased KATP channel function as measured by 86Rb efflux (29). The Zucker diabetic fatty rat showed a decreased expression of Kir6.2 mRNA compared with lean control rats (23), and Kir6.2 mRNA levels dropped by 40% in pancreatectomized, highly hyperglycemic rats, but were nearly unaffected in only mildly hyperglycemic rats (32). In the present study we demonstrate in an in vitro experiment a decrease in Kir6.2 transcript levels in rat islets exposed to high glucose concentrations (16.7 mM), suggesting that high glucose per se, as opposed to other metabolic derangements associated with diabetes, is a contributory factor to the impaired islet cell function described in the previously mentioned in vivo studies. To explore this relationship in more detail, we used the rat insulinoma cell line INS-1. Culturing INS-1 cells at high glucose concentrations (25 mM) resulted in a marked decrease in Kir6.2 transcript levels. This effect is pronounced at glucose levels above 10 mM and becomes evident after 24 h of incubation. Transfection experiments at different glucose concentrations with a luciferase reporter construct did not reveal a significant effect on Kir6.2 promoter activity. One can speculate that luciferase mRNA stability masks differences in transcriptional activity or that a glucose-responsive element may be located outside the cloned promoter region. Actinomycin-treated INS-1 cells did not show differences in Kir6.2 mRNA turnover under different glucose concentrations (data not shown), suggesting that glucose has an inhibitory effect on Kir6.2 promoter activity rather than on mRNA stability. A decrease in Kir6.2 mRNA levels upon cultivation in high glucose is reversible by reverting to low glucose and is prevented in the presence of 10 nM GLP-1. The effect of GLP-1 can be mimicked by forskolin, indicating that GLP-1 regulates Kir6.2 promoter activity via the second messenger cAMP. However, we cannot rule out a second, cAMP-independent signaling pathway leading to enhanced KATP expression by the action of GLP-1. The stimulating effect of GLP-1 on Kir6.2/SUR1 mRNA levels was not inhibited by coincubation with the adenylate cyclase inhibitor MDL-12,330A. Recently, a protein kinase A/cAMP-independent mechanism for the regulation of the insulin gene promoter activity has been proposed, and this may also apply for the transcriptional control of the KATP genes (33). Northern blot analysis further indicates that SUR1 mRNA levels parallel the changes in Kir6.2 mRNA levels induced by glucose or GLP-1. This observation suggests a coordinated expression of both genes in ß-cells, allowing the formation of a functional entity by the assembly of both subunits with a stoichiometry of 1:1. We further demonstrate that changes in SUR1 transcript levels induced by glucose or GLP-1 are reflected by parallel changes in SUR1 protein levels.
The mechanism(s) by which high glucose suppresses the expression of the KATP channel subunits is unknown. It is tempting to speculate, however, that the transcription factor C/EBPß induced in response to glucotoxic stress in ß-cells and shown to inhibit insulin gene expression (34, 35) may be involved in suppressing the expression of the KATP channel subunit genes. It was shown previously that C/EBPß interferes with the transcriptional activator E47/E12 in its interactions on the glucose-responsive enhancer (E2-A3/4) in the promoter of the rat insulin gene 1 (34).
Reduced expression of KATP channel subunits in islet cells preexposed to high glucose also reduced KATP channel current amplitude, as determined by patch clamp studies and 86Rb efflux measurements. The presence of 10 nM GLP-1 in high glucose medium partially restored KATP activity, which supports the idea that GLP-1 stimulates Kir6.2/SUR1 gene expression. GLP-1 is known not only to augment insulin release in response to glucose, but also to up-regulate mRNAs of insulin, Glut-1, Glut-2, hexokinase I, and glucokinase (24, 36). All of these changes are beneficial to ß-cell function and involve mainly components of the glucose-sensing machinery. Our results suggest, therefore, that GLP-1 promotes the expression of KATP channels, representing another essential ß-cell constituent involved in the secretory pathway.
To further characterize the regulated expression of the Kir6.2 gene, we
have initiated the analysis of its transcriptional control
cis elements. The human genes for SUR1 and Kir6.2 are both
located on chromosome 11p15.1, separated by only a small stretch of 4.5
kb of DNA (2). It is therefore tempting to assume the
presence of common regulatory elements within this short DNA segment,
allowing the coordinated expression of the two channel subunit genes.
On the other hand, independent expression of Kir6.2 and SUR1 is
required in other tissues such as cardiac and skeletal muscle, where
instead of SUR1, SUR2A, encoded by a separate gene, is expressed and
associates with Kir6.2 to produce KATP channels
with different pharmacological properties (16). Thus, it
seems reasonable to believe that the individual promoters for both
KATP channel subunit genes can be subject to
common control elements that lend both coordinated transcription and
tissue-specific independent transcription. Ashfield et al.
(17) described the proximal 1.1 kb of the promoters for
both the Kir6.2 and SUR1 genes. In contrast to the SUR1 promoter,
maximal transcriptional activity of the Kir6.2 gene requires at least 1
kb of the proximal promoter sequence. By measuring reporter gene
expression in transiently transfected INS-1 cells, we demonstrate that
the entire sequence between the human SUR1 and the Kir6.2 gene is
necessary for maximal basal Kir6.2 promoter activity. In agreement with
previous observations, we were able to confirm the silencing effect of
the 5'-proximal 300-bp Alu repeat, and also identified a
putative enhancer element between -656 and -988 bp upstream of the
transcriptional start site that contains several SP1 sites and one
E-box. Further upstream, the region between nucleotides -1651 and
-2377 confers additional transcriptional activity by doubling the
promoter activity. Analysis of that specific stretch of DNA reveals
several putative transcription factor-binding sites, such as an E box,
C/EBP
, C/EBPß, and other consensus sites with unknown associations
to pancreatic ß-cell function.
Although we did not address the question of tissue specificity in detail, the 4.8-kb promoter construct is up to 100-fold more active in ß-cells (INS-1 and HIT T15) than in non-ß-cells such as HepG2 or NIH-3T3 (data not shown). This suggests that elements conferring tissue specificity are probably located within the DNA region between the SUR1 and Kir6.2 genes. Based on transient transfection assays with 5'-truncated promoter constructs, it becomes obvious that the proximal 1 kb of the promoter adds only 25–30% of the basal activity of the complete promoter. We therefore conclude that, unlike the majority of genes, Kir6.2 requires at least 2.5 kb of promoter sequence for maximal basal transcriptional activity. A regulatory element, which allows GLP-1 to stimulate Kir6.2 transcription in a dose-dependent manner, is apparently located between nucleotides -67/+140 in the proximal promoter or in the 5'-untranslated region of the Kir6.2 gene. Transcription consensus site analysis revealed a CRE half-site at nucleotides -58/-55, and a potential activating protein-2-binding site between nucleotides +70/+75 in the first exon. Further analysis of these sites will be required to demonstrate their relative contribution to the GLP-1-mediated activation of Kir6.2 gene expression.
The exact in vivo functional consequences of modulation of KATP channel subunit gene expression remain to be established. However, the analysis of both gain of function and loss of function KATP channel mutations has provided important information concerning the in vivo effects of decreasing or increasing KATP channel activity. Disruption of Kir6.2 gene in humans and mice results in depolarization of the resting potential and, consequently, elevation of the intracellular Ca2+ concentration, leading to increased secretion in the presence of low extracellular glucose concentrations (37). The same situation holds true in mice overexpressing a dominant negative form of the Kir6.2 subunit, in which deranged KATP activity leads to increased basal intracellular Ca2+ and, as a consequence, unregulated insulin secretion and hypoglycemic hyperinsulinemia in neonates (38). In a later phase these mice become hyperglycemic due to accelerated apoptosis of ß-cells. In contrast, transgenic mice expressing a Kir6.2 mutant with reduced ATP sensitivity display increased channel activity, resulting in decreased insulin secretion and diabetes (13). Further, a clonal cell line (NES2Y) established from a patient with PHHI displays chronic depolarization that can be restored to near normal by the transfection of the genes encoding the subunits of the KATP channel (39).
In keeping with findings from these extreme genetic modifications, moderately reduced expression of KATP channels could result in elevation of ß-cell membrane potential, which is expected to result in increased insulin secretion. This could represent a mechanism underlying physiological hyperinsulinemic compensation for mild insulin resistance. Eventually, ß-cells could fail in certain individuals due to insulin exhaustion or could even undergo apoptotic cell death due to constant membrane depolarization and Ca2+ overloading. This, in turn, could provide an additional explanation for the beneficial effects of long-term treatment of type 2 diabetic subjects with GLP-1 or exendin-4, a potent agonist of the GLP-1 receptor. Besides the already known effects of GLP-1 on biosynthesis of insulin and of the glucose-sensing machinery, up-regulation of KATP expression is predicted to restore membrane resting potential of ß-cells and hence improve cell function and perhaps survival. It is also noteworthy in this regard that polymorphisms within the SUR1/Kir6.2 locus have been associated with human type 2 diabetes or obesity in several independent studies, although the exact sequence variants presumably involved remain elusive (40, 41, 42, 43, 44, 45, 46, 47).
In conclusion, this report presents data showing an inhibitory effect of high glucose concentrations on Kir6.2/SUR1 expression leading to reduced KATP channel activity. These observations provide a possible explanation for the occurrence of reactive hyperinsulinemia in the early phases of the pathogenesis of type 2 diabetes. The presence of GLP-1 attenuates the effects of high glucose by partially restoring Kir6.2/SUR1 expression levels. This observation supports the idea that the antidiabetic drug GLP-1 acts at different levels to improve glucose homeostasis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Current address: Department of Surgery D-Lab, University Hospital
of Zürich, Rämistrasse 100, 8091 Zürich, Switzerland.
3 Current address: Endocrinologia, Hospital Clinic, Villarroel 170,
Barcelona 08036, Spain.
4 Investigator with the Howard Hughes Medical Institute.
Received June 6, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Polakof, S. Panserat, E. Plagnes-Juan, and J. L. Soengas Altered dietary carbohydrates significantly affect gene expression of the major glucosensing components in Brockmann bodies and hypothalamus of rainbow trout Am J Physiol Regulatory Integrative Comp Physiol, October 1, 2008; 295(4): R1077 - R1088. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Kang, C. A. Leech, O. G. Chepurny, W. A. Coetzee, and G. G. Holz Role of the cAMP sensor Epac as a determinant of KATP channel ATP sensitivity in human pancreatic {beta}-cells and rat INS-1 cells J. Physiol., March 1, 2008; 586(5): 1307 - 1319. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Acosta-Martinez and J. E. Levine Regulation of KATP channel subunit gene expression by hyperglycemia in the mediobasal hypothalamus of female rats Am J Physiol Endocrinol Metab, June 1, 2007; 292(6): E1801 - E1807. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Allagnat, D. Martin, D. F. Condorelli, G. Waeber, and J.-A. Haefliger Glucose represses connexin36 in insulin-secreting cells J. Cell Sci., November 15, 2005; 118(22): 5335 - 5344. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Minami, M. Yokokura, N. Ishizuka, and S. Seino Normalization of Intracellular Ca2+ Induces a Glucose-responsive State in Glucose-unresponsive beta -Cells J. Biol. Chem., July 5, 2002; 277(28): 25277 - 25282. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Nakazaki, M. Kakei, H. Ishihara, N. Koriyama, H. Hashiguchi, K. Aso, M. Fukudome, Y. Oka, T. Yada, and C. Tei Association of upregulated activity of KATP channels with impaired insulin secretion in UCP1-expressing insulinoma cells J. Physiol., May 1, 2002; 540(3): 781 - 789. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Nakazaki, M. Kakei, H. Ishihara, N. Koriyama, H. Hashiguchi, K. Aso, M. Fukudome, Y. Oka, T. Yada, and C. Tei Association of upregulated activity of KATP channels with impaired insulin secretion in UCP1-expressing MIN6 cells J. Physiol., March 28, 2002; (2002) 2001013048. [Abstract] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
, P < 0.02). C, 86Rb efflux
measurement. INS-1 cells were incubated for 72 h with 5 or 25
mM glucose or with 25 mM glucose and 10
nM GLP-1. Cells were then exposed to 0.3
mM 86RbCl (40 mCi/ml) and washed, and
86Rb release was determined in the presence of the
metabolic inhibitors 2-deoxy-D-glucose and oligomycin and
the Na+,K+-adenosine triphosphatase inhibitor
ouabain. 86Rb efflux carried through ATP-sensitive
K+ channels was determined by measuring the release of
86Rb in the presence of KATP activator
diazoxide or inhibitor tolbutamide. Data represent the mean ±
SEM from three independent experiments (*,
P < 0.001).
