| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Medicine B, Westfälische Wilhelms-Universität, Münster;
Institute of Neurobiology, Heinrich-Heine-Universität, Düsseldorf, Germany; and
Cell Communication Group, Department of Biological Sciences, University of Central Lancashire, Preston, England
3Correspondence: Department of Medicine B, Westfälische Wilhelms-Universität, Albert-Schweitzer-Str. 33, D-48129 Münster, Germany. E-mail: markus.lerch{at}uni-muenster.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: acetylcholine • cholecystokinin-octapeptide (CCK) • fura-2 (AM) • ion transport • magfura-2 (AM) • pancreatic acinar cells
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3
4)
, the
importance of intracellular magnesium changes is largely unknown.
Although Mg2+ is a known cofactor for multiple
enzymatic reactions and its intracellular concentration has been shown
to respond to hormonal stimuli—albeit in a much slower fashion than
the corresponding Ca2+ concentrations—its role
in regulating physiological or pathophysiological processes has
remained speculative. Important similarities and differences regarding
the physiology of the two divalent cations calcium and magnesium have
been established. Intracellular calcium concentrations are generally
kept at very low levels, and these low intracellular calcium levels (of
100 nM) not only require the active maintenance of a 10,000-fold
calcium gradient across the plasma membrane, but are also a
prerequisite for the intracellular messenger role of calcium
(5)
. Conversely, whereas magnesium has been shown to be
involved in the modulation of various ion channels and transporters
(6
, 7)
and this modulating effect appears to be of great
pathophysiological and clinical relevance (8)
, its
gradient across the plasma membrane is very small. It has been assumed
that the intracellular magnesium concentration is, analogous to the
intracellular calcium concentration, tightly controlled and regulated
by agonist-dependent signaling events. However, only recent advances in
the development of microanalytical methods, such as
Mg2+-sensitive fluorescent probes
(9)
and Mg2+-selective
microelectrodes (10)
, have made investigations into the
intracellular Mg2+ regulation possible
(11
12
13
14)
and enabled experiments on
Mg2+ changes in response to hormone agonists in
either culture cell systems (15
, 16)
or primary cell
isolates (17
18
19)
.
In this study we have investigated the biological role of
Mg2+ in the exocrine pancreas, an organ system in
which rapid intracellular Ca2+ changes determine
the onset and course of physiological (1)
as well as
pathophysiological (20)
events. Using isolated mouse
pancreatic acini—intact secretory units of living exocrine cells—we
used Mg2+-sensitive fluorescent indicators as
well as intracellular ion-selective microelectrodes to investigate the
spatial and temporal changes in
[Mg2+]i in response to
physiological and pathological stimuli. We have further determined
their correlation to and interference with the intracellular calcium
changes in pancreatic acinar cells. Our results indicate that
intracellular Mg2+ concentrations in the exocrine
pancreas are regulated in a stimulus-controlled manner. Moreover, they
can counteract effects in the Ca2+ signal
transduction pathway; the ‘slower’ uptake and release of
Mg2+ from intracellular, most probably
ER-associated stores, occur in a manner that is completely antiparallel
to the ‘rapid’ Ca2+ release and reuptake in
acinar cells.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Preparation of pancreatic acini
All animal experiments were approved by and conducted under the
guidelines of the animal welfare committee of Münster University.
Adult male white mice (NMRI) weighing between 20 and 25 g were
obtained from Charles River (Sulzfeld, Germany), kept in Nalgene shoe
box cages in a 12 h:12 h light:dark cycle with unlimited access to
standard chow and water, and adjusted to laboratory conditions over the
course of 1 wk. Pancreatic acini were prepared by a modified
collagenase protocol as previously reported (21
, 22)
. In
brief, after a 12 h fast the animals were killed and the pancreas
was rapidly removed, minced into small pieces, and placed into buffer
(pH 7.4) containing NaCl (130 mM); KCl (5 mM); HEPES (10 mM);
KH2PO4 (1.2 mM);
CaCl2 (1 mM); MgSO4 (1 mM);
glucose (10 mM); and collagenase (100 U/ml Type V, Sigma,
München, Germany). After 10 min incubation at 37°C under
continuous shaking (120 cycles/min), the digested tissue was washed 3
times in 10 ml buffer without the presence of collagenase and again
shaken 10 times to dissociate the acini. Acini were filtered through
muslin gauze, centrifuged at 400 rpm for 3 min, and washed twice more
in buffer solution containing 4% bovine serum albumin (BSA). After the
second wash, acini were suspended in 8 ml of buffer containing soybean
trypsin inhibitor (0.1 mg/ml) and BSA (0.2% w/v). A stock suspension
of acini was kept on ice for up to 4 h without significant
reduction in cell viability (>95%) as assessed by trypan blue
exclusion.
Loading of acini with ion-sensitive probes
For microfluorometric experiments acini were loaded with 5 µM
of the ester of either magfura-2 or fura-2 at room temperature for 20
min for measurements of either
[Mg2+]i or
[Ca2+]i, respectively.
Wherever calcium and magnesium measurements were performed
simultaneously, magfura-loaded acini were subjected to a second
incubation with the calcium-sensitive probe fluo-3 (in combination with
0.001% Pluronic) for an additional 35 min at 35°C. Acini were then
washed twice by centrifugation and preincubated for 30 min to allow
complete de-esterification of the probes. Loaded cells were plated onto
glass coverslips coated with Cell-Tak (1
µl/cm2) and kept at room temperature for 30 min
before use. In experiments where the uptake of magfura-2 into
intracellular calcium stores was quantified, the nonesterified dextran
derivative of magfura-2 was microinjected into individual acinar cells
using an Eppendorf microinjection instrument.
Measurements of cytosolic free [Mg2+]i,
[Ca2+]i and pHi
Plated cells were mounted in a perfusion chamber (internal
volume 0.28 ml) and placed on the stage of a Nikon Diaphot-TMD inverted
microscope equipped with a Fluor X 100 oil immersion objective. Acini
were continuously perfused with medium at a flow rate of 3 ml
min-1 at room temperature. Fluorochrome
excitation (340 and 380 nm for magfura-2 and fura-2, 488 nm for fluo-3,
495 and 440 nm for BCECF) was achieved using a Xenon lamp in series
with two chopper-linked monochromators that were coupled to the
microscope via fiber optics. Emitted light was collected behind a
band-pass filter (509 nm for magfura-2 and fura-2, 535 nm for fluo-3,
538 nm for BCECF) by either a photomultiplier or an intensified CCD
camera. In control background, fluorescence was found to be negligible.
Ratios obtained from the dual excitation wavelengths probes magfura-2
and fura-2 were converted into magnesium and calcium concentrations by
the formula given by Grynkiewicz et al. (23)
:
[Mg2+]i and
[Ca2+]i = (R - R
min)/(R max - R) x KD x (I ion-free/I
ion-saturated)380 nm, in which R min and R max
are the minimal and maximal ratios obtained in either ion free or ion
saturated solutions, respectively. KD is the
dissociation constant of magfura-2 for magnesium or of fura-2 for
calcium, and I is the fluorescence intensity at the 380 nm wavelength
of the free dye divided by the saturated ion-dye complex. All
calibration parameters were obtained in separate experiments.
In a different approach,
[Mg2+]i was determined
using a magnesium-sensitive intracellular microelectrode. The
preparation of Mg2+-sensitive microelectrodes was
carried out as described by Hintz et al. (24)
. In brief,
double-barreled microelectrodes were pulled from theta-style
borosilicate glass (Clark, Reading, England). One of the two channels
was silanized (40 min exposure to hexamethyldisilazan vapor at 40°C,
followed by 2 h baking at 200°C); the second channel was
prevented from being silanized by continuous perfusion with compressed
air (1.5 bar). The tip of the silanized channel was filled with the
Mg2+ sensor based on the neutral carrier ETH 5214
(Fluka, Buchs, Switzerland) and then backfilled with 100 mM
MgCl2; the other channel was filled with 3 M KCl
and served as an intracellular reference electrode. All potentials were
measured against the potential of an extracellular reference electrode
(agar bridge and Ag/AgCl cell), using a two-channel voltmeter with
differential amplifier (FD 223, WPI, Berlin, Germany). All output
signals were AD-converted and recorded continuously on a PC. Before and
after an experiment, the microelectrodes were calibrated in solutions
containing 10, 2.5, 0.5, and nominally 0 mM MgCl2
at an ionic background simulating intracellular conditions (in mM: KCl,
110; NaCl, 10; HEPES, 10, pH adjusted to 7.3 with KOH). Calibration
curves were fitted with the Nicolsky-Eisenman equation, as described by
Günzel and Schlue (25)
.
Statistical analysis
Data shown in the figures are representative of 6 or more
individual experiments in each group. Wherever applicable, data points
represent means ± standard errors of the means (SE).
Differences between groups were compared using the Student‘s
t test and P values <0.05 were considered
significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, B
). On the other hand, when acini were preincubated for 60
min in the presence of different extracellular magnesium
concentrations, CCK evoked a profound effect on the frequency of
calcium transients. Under near physiological
[Mg2+]e of 1.0 mM (Fig. 2A
) calcium oscillations in response to CCK were recorded at
a frequency of 1 min-1. A preincubation in
magnesium-rich buffer
([Mg2+]e, 5 mM) resulted
in a decrease of this frequency to 0.6 min-1
(Fig. 2C
); a further increase of
[Mg2+]e to 10 mM reduced
the frequency of calcium oscillations to 0.3 spikes
min-1 (Fig. 2D
) and often resulted in
a complete termination of the oscillations. In contrast, a decrease in
the [Mg2+]e to 0.2 mM
increased the frequency of calcium oscillations to 1.1
min-1 (Fig. 2B
).
|
|
Because Mg2+ can potentially interact with a variety of Ca2+ transport mechanisms at either the plasma membrane or at the membrane of intracellular calcium stores, it was necessary to elucidate in what manner it may affect calcium oscillations. Calcium oscillations are assumed to result from a rapid emptying and refilling of intracellular calcium stores. The filling state of these stores is therefore the net result of calcium leakage from the stores and the activity of the sarcoplasmic reticulum Ca2+-ATPase. To determine whether Mg2+ was affecting this filling state of calcium stores, acini were incubated with the calcium ionophore ionomycin in a calcium-free solution. This results in a rapid leakage of Ca2+ from the stores and allows for the quantification of stored calcium. A preincubation in the presence of the different Mg2+ concentrations that had resulted in marked changes in Ca2+ oscillations, however, had no effect on the ionomycin-induced release of Ca2+ from the cells (data not shown).
One of the alternative mechanisms for the interference of
Mg2+ with calcium oscillations, that of a calcium
influx into acinar cells via the plasma membrane, was tested using the
manganese influx technique. For this assay the manganese-induced
quenching of fura-2 fluorescence at the isosbestic wavelength of 360 nm
was used as an indicator of calcium movements across the plasma
membrane (26)
. A reduction of
[Mg2+]e in the
preincubation buffer from 1 mM to 0.2 mM enhanced the manganese-induced
quenching at 360 nm significantly (Fig. 3
), whereas an increase in
[Mg2+]e to 10 mM greatly
reduced manganese bleaching. This result indicates that one mechanism
by which magnesium interacts with calcium signaling involves inhibition
of the capacitive calcium influx pathway in pancreatic acinar cells.
|
The question remained as to whether changes in
[Mg2+]e or changes in
[Mg2+]i could account for
this effect. The observation that an extended period of preincubation
was required before increased Mg2+ levels in the
buffer affected CCK-induced calcium oscillations suggests that the
latter is the case. To study the effect of
[Mg2+]i we used the
magnesium-sensitive probe magfura-2. Preincubation for 60 min in buffer
containing either 0.2 mM, 1.2 mM, or 10 mM magnesium did indeed result
in different intracellular magnesium concentrations (n=22)
of 0.41 ± 0.07 mM, 0.56 ± 0.08 mM, or 1.25 ± 0.29 mM,
respectively. Within a preparation of acini that was preincubated at
the identical [Mg2+]e,
the [Mg2+]i of individual
cells still varied to some degree, probably due to differences between
individual acinar cells in regulating magnesium uptake. We made use of
these individual differences between acinar cells and directly
correlated the basal
[Mg2+]i with the
frequency of calcium oscillations after loading cells simultaneously
with a calcium-sensitive probe (fluo-3) together with a
magnesium-sensitive probe (magfura-2) and exposing the cells to
different extracellular magnesium concentrations for 60 min. We found
an almost linear correlation between the intracellular magnesium
concentration and the suppression of CCK-induced calcium oscillations
(Fig. 4
).
|
To study whether agonist-evoked changes in intracellular calcium
concentrations were paralleled by comparable intracellular magnesium
changes, we used the same magfura-2-based technique to quantify
[Mg2+]i in individual
acinar cells. Addition of CCK (100 pM) to isolated acinar cells
resulted in an initial increase of the magfura-2 fluorescence ratio,
followed by a sustained decrease, with a return of the ratio to
prestimulatory values after CCK was removed from the medium (Fig. 5A
). This time course of the magfura-2 fluorescence signal
could suggest that an initial increase in
[Mg2+]i is followed by a
decrease in the free magnesium concentration and a return to basal
levels after withdrawal of the stimulus. When, instead of using maximal
CCK concentrations, we exposed acinar cells to a submaximal
concentration of CCK (10 pM) that produced Ca2+
oscillations, we found that in 30–40% of the cells the decrease in
magfura-2 ratio progressed in a stepwise manner, and the frequency of
these steps corresponded to the frequency of agonist-evoked calcium
oscillations (Fig. 5B
). This observation suggests that the
magfura-2 signal could have reflected not only changes in
[Mg2+]i, but also a
detection of calcium transients, which would have been accounted for by
the known low-calcium affinity of magfura-2. The y-axes in Figs. 5
6
7B
, 11
, and 12
therefore denote the magfura ratio rather
than [Mg2+]i in order to
document potential interferences of the signal by calcium. Because of
this potential codetection of calcium signals by magfura-2 (which
although unlikely in the low calcium conditions of resting cells, would
have compromised our
[Mg2+]i measurements in
the presence of the micromolar calcium concentrations reported in
agonist-stimulated cells; ref 27
28
29
), we have validated
our methods to determine intracellular magnesium in a number of ways.
|
|
|
|
|
To differentiate between the detection of calcium and magnesium by
magfura-2, we used chelating agents with different affinities to the
two cations. Whereas BAPTA binds to calcium with higher affinity than
to magnesium, EDTA chelates both ions with comparable affinities.
Moreover, the on rate of EDTA is slower than that of BAPTA, which
facilitates the diffusion of calcium ions. Loading acinar cells with
EDTA-AM resulted in an agonist-evoked initial increase of the magfura-2
signal that was of lower amplitude than in controls (Fig. 6A
) and, subsequently, in a slower and less pronounced
decrease of the magfura-2 ratio. Loading acini with BAPTA-AM, on the
other hand, completely abolished the initial increase of magfura-2
fluorescence as well as the stepwise reduction of the magfura-2 ratio,
which in turn was replaced by a precipitous decline (Fig. 6B
). The result of this experiment suggested that the
magfura-2 signal in the presence of the more calcium-selective chelator
BAPTA specifically reflected an intracellular magnesium signal. An
alternative explanation could still have been a decrease of free
calcium within intracellular stores because magfura-2 had been used as
a low-affinity indicator to detect calcium in intracellular stores
(30
31
32)
. In the respective experiments, however, longer
loading periods and higher loading concentrations of magfura-2-AM were
used in order to saturate intracellular esterases and thus allow the
magfura-2-ester to leave the cytosol intact and enter calcium stores.
Since our aim was to detect cytosolic magnesium rather than
store-confined calcium, we loaded cells with magfura-2 at lower
concentrations and for shorter periods of time.
To test what proportion of magfura-2 had remained cytosolic rather than
being taken up into intracellular stores, we permeabilized the plasma
membrane with 5 µM digitonin and quantified dye fluorescence at 360
nm. This low digitonin concentration has been found to permeabilize the
plasma membrane without affecting the membrane of intracellular ion
stores (33)
. We found that >95% of magfura-2 had
remained in the cytosol and only a negligible amount of the probe had
entered intracellular stores (Fig. 7A
). In another attempt to rule out the detection of a calcium
leak from intracellular stores rather than a decrease of cytosolic
magnesium by the magfura-2-fluorescence, we used the dextran derivative
of magfura-2, which is nonpermeant for the plasma membrane or the
calcium-store membrane, and microinjected it directly into the cytosol
of acinar cells. The CCK-induced magfura-2 signal in microinjected
cells (Fig. 7B
) was identical to the signal obtained from
cells that had been loaded by incubation with magfura-2-AM (Fig. 5A
). This finding again indicates that the magfura-2 signal
reflected cytosolic magnesium concentrations rather than the calcium in
intracellular stores. We finally used an electrophysiological approach
to confirm the nature of the magfura-2 signal in response to CCK
stimulation and in the presence of BAPTA. Ion-selective microelectrodes
were inserted into single acinar cells as well as tissue fragments in
order to record changes in
[Mg2+]i simultaneous with
the membrane potential.
The basal mean [Mg2+]i as
determined by the magnesium-sensitive electrode (n=45) was
0.58 mM (pMg 3.24±0.26). After CCK stimulation
[Mg2+]i decreased by
0.11 ± 0.12 mM and, after withdrawal of the stimulus, returned to
baseline levels (Fig. 8
). By using this independent approach to quantify
[Mg2+]i in pancreatic
acini, we were not only able to validate the magfura-2 ratio as a
reliable indicator for intracellular magnesium measurements, but found
that recordings in the presence of the
Ca2+-chelator BAPTA most accurately reflected
absolute [Mg2+]i. All
Mg2+ measurements in the ensuing experiments
therefore used the magfura-2 technique in BAPTA-AM-loaded acini unless
indicated otherwise. The basal
[Mg2+]i in pancreatic
acinar cells obtained by this method (n=62) was 0.59 ±
0.13 mM and the corresponding decrease in response to 100 pM CCK was
0.22 ± 0.1 mM. These results are in good agreement with the
electrophysiological data obtained in this study.
|
To determine whether the changes in intracellular magnesium
concentrations depended on the specific activation of the CCK receptor
and its second messenger pathway, we tested other calcium mobilizing
agents. In response to a maximal stimulating concentration of the
acetylcholine analog carbachol (100 µM),
[Mg2+]i decreased in a
manner completely parallel to that observed after CCK stimulation and
returned to baseline levels after the withdrawal of carbachol. Neither
the addition of carbachol to CCK-stimulated cells nor the addition of
CCK to carbachol-stimulated cells caused a further decrease in
[Mg2+]i or changed the
kinetics of the recovery (data not shown). A nonreceptor-dependent
manner in which intracellular calcium mobilization can be affected is
by inhibiting the Ca2+-ATPase activity of calcium
stores with, for example, thapsigargin. Thapsigargin therefore depletes
intracellular calcium stores and this calcium depletion is followed by
a subsequent Ca2+ influx into acinar cells.
Incubation of acini with 1 µM thapsigargin resulted in a slow and
only moderate decrease in
[Mg2+]i. A subsequent
stimulation with CCK, however, induced an additional and rapid decrease
in [Mg2+]i (Fig. 9A
). When we reversed the sequence of incubation, we found
that exposure to thapsigargin did not induce any further decrease in
[Mg2+]i in addition to
that evoked by CCK (Fig. 9B
). However, in both types of
experiments, thapsigargin treatment prevented the recovery of the
magfura-2 signal after withdrawal of the stimulus.
|
The rapid disappearance of significant amounts of magnesium from the
acinar cell cytosol might be caused by two alternative mechanisms. The
first would involve a magnesium efflux from acinar cells across the
acinar cell membrane whereas the second would involve a magnesium
uptake into intracellular stores. Provided a magnesium efflux from
acinar cells had occurred, this efflux would most likely be affected by
gradient changes across the plasma membrane of either magnesium itself
or other ions. To test this possibility, we changed the ion
concentrations in the extracellular buffer in a number of ways. Neither
an increase to high magnesium concentrations (10 mM, not shown) nor
incubation in magnesium-free buffer (Fig. 10A
) had any effect on intracellular magnesium changes in
response to CCK stimulation.
|
Because a sodium-magnesium antiporter has been suggested to be
operative in a variety of cells, we tested its potential role in
regulating [Mg2+]i in
acini by replacing sodium with N-methyl-D-glucamine in the perfusion
buffer. In addition, we omitted Ca2+ from the
buffer to avoid an expected Ca2+ increase in
response to the sodium depletion, which has been shown to be mediated
by a sodium-calcium exchanger. Again, neither the presence nor the
absence of sodium and calcium in the extracellular fluid had any
significant effect on
[Mg2+]i changes in
response to external stimuli (Fig. 10B
). Similarly, neither
the changes in membrane potential caused by exposure to high potassium
concentrations (Fig. 10C
) nor the complete omission of
magnesium and calcium from the extracellular medium had any noticeable
effect on the CCK-induced
[Mg2+]i changes in
pancreatic acinar cells (Fig. 10D
).
Our use of the calcium chelator BAPTA during magfura-2 measurements
could potentially have masked the effect of calcium on intracellular
magnesium movements because BAPTA could have served as calcium
reservoir, releasing its bound calcium after the period of CCK
stimulation had ended. To exclude this hypothesis, we performed a
series of experiments in which acini were not loaded with BAPTA, but
exposed to calcium-free buffer for 100 s before CCK stimulation
during CCK stimulation and for 500 s after CCK had been withdrawn.
In fura-2 calcium experiments we observed that this treatment is
sufficient to deplete intracellular calcium. In magfura-2 experiments,
however, this manner of calcium depletion in the absence of BAPTA
affected neither the basal
[Mg2+]i nor the magnesium
changes in response to CCK (Fig. 11A
). In a different approach to rule out the influence of
calcium on intracellular magnesium changes, we used lanthanum, which is
known to block calcium entry into acinar cells when used in micromolar
concentrations. Exposure of acinar cells to 250 µM lanthanum in the
absence of BAPTA again had no significant effect on
[Mg2+]i changes in
response to CCK (Fig. 11B
).
To test whether CCK stimulation is at all associated with a magnesium
efflux from acinar cells, we measured
[Mg2+]e in acinar cell
preparations of high cell density (cytocrit 35%) after exposure to
CCK, using a magnesium-sensitive electrode. Even in the presence of low
extracellular magnesium concentrations (0.35 mM), which were chosen to
better detect magnesium efflux from the cells, the increase in
[Mg2+]e after 20 min of
incubation (0.01±0.001 mM, n=16) was the same in
CCK-stimulated as in control cells.
Once magnesium movements across the acinar cell membrane had been ruled
out as a mechanism involved in regulating
[Mg2+]i changes in
response to secretory stimuli, we attempted to characterize potential
intracellular magnesium buffers. One of the known intracellular
buffering mechanisms that would have explained the disappearance of
magnesium from the cytosol of acinar cells is its binding to
intracellular ligands such as ATP, ADP, phosphates, and calmodulin. If,
however, such a binding of magnesium were involved in the CCK-induced
[Mg2+]i changes, this
mechanism would be affected by intracellular pH alterations
(34)
. When we used magfura-2-loaded cells in the absence
of BAPTA and determined the intracellular pH with the pH-sensitive
fluorescent probe BCECF, the addition of ammonium chloride to the
acinar preparation induced a rapid intracellular alkalinization, which
was followed by a reacidification after removal of
NH4Cl (Fig. 12A
). The NH4Cl-induced
alkalinization/acidification sequence, however, had no effect on
intracellular magnesium concentrations (Fig. 12B
).
Continuous incubation with either amiloride (0.1 mM, data not shown) or
N-methyl-D-glucamine (see above), which both alter intracellular pH
(35)
, had no effect on CCK-induced
Mg2+ changes.
An intracellular organelle capable of storing large amounts of
magnesium is the mitochondrion, and magnesium is involved in a number
of enzymatic processes that are confined to mitochondria (36
, 37)
. To test whether mitochondria represent the critical
magnesium stores responsible for the uptake of cytosolic magnesium, we
interfered with the mitochondrial metabolism in various ways. Neither
uncoupling of the mitochondrial respiratory chain by incubation with
carbonyl cyanide p-(trifluoromethoxy)-phenylhydrazone (FCCP)
(Fig. 13A
), inhibition of the mitochondrial transition pore with
cyclosporin (Fig. 13B
), nor incubation with phenylarsine,
which is known to release mitochondrial magnesium most likely via the
transition pore (not shown), had any effect on stimulus-evoked
[Mg2+]i changes in
pancreatic acinar cells.
|
To further characterize the nature and site of intracellular magnesium
stores, we performed localization experiments using fluorescence
imaging of ion-specific probes. When we used the calcium-specific dye
fura-2, we found 76% of cells that were stimulated with 100 pM CCK to
respond with an increase of cytoplasmic calcium that was initiated at
the apical or luminal pole of the acinar cell, which contains the
zymogen granules, and progressed from there to the basolateral portion
of the respective cell (Fig. 14A
). This result agrees with earlier publications in which
this characteristic and spatially distinct pattern of
Ca2+ mobilization in response to CCK was
initially reported (27
, 28)
. The temporal delay between
the initiation of the Ca2+ signal at the apical
pole and the subsequent Ca2+ increase in the
basolateral region was calculated to be 0.38 ± 0.04 s after
CCK stimulation and 0.29 ± 0.01 s after carbachol
stimulation. In 20% of cells, no regional differences in the
release of calcium were detectable. On the other hand, when we used
magfura-2 and BAPTA-loaded cells to image regional magnesium changes by
fluorescence microscopy, the cytosolic magnesium signal began to
decrease in the basolateral region of 67% of acinar cells, and this
decrease progressed to the apical pole (Fig. 14B
). The
temporal delay from the initial basolateral decrease to the subsequent
apical decline of [Mg2+]i
was calculated to be 0.49 ± 0.12 s after CCK stimulation
(n=15) and 0.60 ± 0.12 s after carbachol
stimulation (n=16). A spatially inhomogeneous disappearance
of the Mg signal from the cytosol, indicated by a more prominent
decrease of the magfura-2 fluorescence from the basolateral rather than
from the apical region, was found in >90% of all imaged acinar cells.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, others found that high
Mg2+ concentrations can impair the cellular
calcium response (39
40
41
42
43)
. Our present data, which were
obtained from isolated pancreatic acini and by single-cell
measurements, agree with those of the latter investigators because they
indicate that increased Mg2+ concentrations have
a direct effect on the frequency and amplitude of
Ca2+ oscillations in response to CCK or carbachol
stimulation. An interference of high
[Mg2+]i with fura-2, the
fluorescent indicator we used to determine calcium concentrations, can
be ruled out because Mg2+ does not affect the
KD of fura-2 over a wide range of concentrations
(44)
. Moreover, the manganese influx measurements in our
study provide direct evidence that the capacitive calcium entry is the
principal target of Mg2+ ions and that
Mg2+ blocks this pathway by interference with
Ca2+ permeation. An analogous interference of
Mg2+ with the entry of Ca2+
into the cell has previously been shown for L-type calcium channels in
myocytes (45
, 46)
and for a ligand-gated calcium influx
pathway in rat pituitary lactotropes (47)
. Our current
data further suggest that these magnesium effects depend on
[Mg2+]i changes rather
than [Mg2+]e alterations,
because acute extracellular magnesium changes failed to affect
intracellular calcium signaling and an extended period of incubation
(
30 min) at high extracellular magnesium concentrations was required
before any effects were detectable. By making use of the fact that
individual cells from the same acinar preparation take up magnesium at
different rates—a phenomenon not unlike the heterogeneous
[Mg2+]i loading rates
previously found in tracheal gland cells (48)
—we could
correlate individual calcium signals with individual cytosolic
magnesium concentrations, and have found a linear relationship between
the two. This must be regarded as further evidence for a direct effect
of [Mg2+]i on the
regulation of calcium signaling in acinar cells.
However, before a significant role of magnesium in the physiology of an
exocrine gland like the pancreas can be assumed, it remains to be shown
that intracellular magnesium concentrations respond to cellular status
changes such as hormonal stimulation in a dynamic manner. Although
attempts to elucidate the role of
[Mg2+]i in pancreatic
acinar cell physiology have been made in the past (16
, 17)
, they were often impeded by the unavailability of
single-cell imaging techniques or the limitations of
Mg2+-specific probes. Even today, the most
specific Mg2+-sensing fluorescent probe
available, magfura-2, has a residual low affinity to calcium, with a
KD of 20 µM (49)
, and
intracellular calcium concentrations above 1 µM have been shown to be
detectable using magfura-2. This cross-reactivity of magfura-2 with
calcium should not affect the accurate determination of the basal
intracellular magnesium concentrations under resting conditions, but
might affect magnesium measurements in the presence of high calcium
concentrations, i.e., after an agonist-evoked release of
[Ca2+]i from
intracellular stores (27
, 28)
. An additional source of
interference by the low calcium affinity of magfura-2 consists in the
accumulation of the probe in intracellular compartments such as calcium
stores. Some authors have made use of these properties of magfura-2 to
investigate calcium storage mechanisms (30
31
32)
. In this
type of experiment longer incubation periods and higher loading
concentrations were used to saturate the intracellular esterases that
would have converted the fluorochrome-ester magfura-AM to
membrane-impermeant magfura and would have prevented its uptake into
calcium stores. In addition, cells were usually permeabilized to
release the cytosolic portion of magfura and thus reduce background
fluorescence (32)
. In our experiments where we used
magfura-2 as a high-affinity probe for magnesium, we prevented a
potential cross-reactivity with calcium by several means. To test our
experimental and loading conditions, we either permeabilized acini with
digitonin or injected the membrane-impermeant dextran derivative of
magfura-2 directly into the cells.
The results from these experiments indicated that >95% of magfura-2 was localized in the acinar cell cytosol and no cross-reactivity with calcium from intracellular stores had interfered with our magnesium measurements. To rule out interference by cytosolic calcium, we used ion chelators with different affinities to calcium and magnesium to differentiate between their cytosolic binding to magfura-2. In experiments where we used BAPTA-AM, a chelator that has a higher affinity to calcium than to magnesium, we found that the prominent initial rise in the magfura-2 ratio as well as oscillation-like changes during the subsequent decline of the ratio could be accounted for by magfura-2 detection of cytosolic calcium. The rapid decrease of the magfura-2 ratio in response to secretory stimuli, however, and the return of the ratio to prestimulatory levels after the stimulus was withdrawn appeared to reflect a disappearance of magnesium from the cytosol. These observations provided us with a tool to study dynamic [Mg2+]i changes under different experimental conditions. As a control for our magnesium recordings in magfura-2 and BAPTA-AM-loaded acini, we used an electrophysiological approach that is completely independent of the limitations of fluorescent ion-sensitive probes and found that, indeed, our magfura-2 ratio recordings accurately reflected cytosolic magnesium concentrations.
We have thus confirmed by independent means that stimulation of
pancreatic acinar cells with Ca2+ mobilizing
agonists like CCK or carbachol elicits a rapid decrease of cytosolic
Mg2+, and that withdrawal of the stimulus is
followed by a restitution of the initial cytosolic
Mg2+ concentration. Though these observations are
in contrast with results obtained from other cell types that were
treated with a variety of stimuli (15
, 50
, 51)
, they agree
with studies on the sublingual mucous gland (52)
. The
decrease in [Mg2+]i
observed in pancreatic acinar cells in this study might theoretically
be caused either by an export of magnesium across the plasma membrane
or by a shift of magnesium from the cytosol into cellular stores. In
subsequent experiments we found no evidence of a
Mg2+ extrusion across the plasma membrane,
because the cytosolic magnesium concentrations under resting conditions
or in response to hormonal stimuli were not affected by changes of the
electrochemical gradient for Mg2+ or by changes
in membrane potential induced by diminishing the potassium equilibrium.
In particular, the absence of any influence of magnesium-free buffer on
the recovery of [Mg2+]i
after the withdrawal of the stimulus illustrates clearly that the
intracellular magnesium movements are independent of the presence or
absence of extracellular magnesium.
We could further rule out the involvement of a sodium-magnesium
exchanger, although it has been suggested that this antiporter is
involved in magnesium movements across the membrane of pancreatic
acinar cells (53
, 54)
. Those experiments, however, were
done in the absence of a calcium chelator, and the perfusion of acinar
cells with sodium-deficient medium resulted in a large increase of
intracellular calcium via action of the
Na+/Ca2+ exchanger. This,
in turn, would be expected to affect magfura-2 ratio recordings in a
non-magnesium-specific manner (55)
. When we avoided the
effect of sodium depletion on calcium movements by using a sodium- and
calcium-deficient medium in addition to chelating calcium with BAPTA,
we could solve the seemingly incompatible differences between the two
studies. In additional experiments in which we measured the release of
magnesium into the perfusion medium directly, we could confirm that
agonist stimulation does not evoke a significant extrusion of magnesium
from isolated acini. The experiments presented here therefore appear to
reflect the mechanisms involved in intracellular magnesium regulation
more accurately than earlier attempts when slices of pancreas were used
(54)
.
Accordingly, we must assume that the decrease in
[Mg2+ ]i in response to
CCK or carbachol is caused by an intracellular mechanism. A significant
buffering of cytosolic magnesium by intracellular ligands such as ATP
and calmodulin, or phospholipids like phosphatidylserine and
phosphatidylcholine, was ruled out as a potential explanation because
[Mg2+]i was not affected
by changes in intracellular pH. A second potential mechanism, that of a
magnesium uptake into mitochondria (56)
, was ruled out in
experiments in which we uncoupled oxidative phosphorylation or impaired
the function of the mitochondrial transition pore, a large conductance
ion channel that is known to permit Mg2+
movements between the cytosol and the mitochondrial matrix.
Support for an uptake into intracellular magnesium stores came
from imaging studies in which we recorded the spatial and temporal
resolution of magfura-2 ratio changes in response to agonist
stimulation. The data from these studies have indicated that the
disappearance of magfura-2 from the cytosol is initiated in the
basolateral portion of acinar cells and progresses from there to the
apical pole. Because this localized shift from the basolateral to the
secretory compartment constitutes a completely antiparallel event to
the known calcium release from intracellular stores, we studied calcium
and magnesium changes in the same setup. In these imaging studies we
have confirmed the initial calcium rise at the apical pole, which was
followed by a progression to the basolateral portion of acinar cells as
previously reported (27
, 28)
, whereas the magfura-2
magnesium signal disappeared in the opposite direction. Moreover, this
temporal polarity was paralleled by a sustained spatial gradient of
cytosolic Mg2+ and
[Mg2+]i was found to be
consistently lower close to the basolateral membrane. The initiation of
the Mg decrease in the basolateral portion of acinar cells, where most
of the endoplasmic reticulum is located, makes it seem most likely that
[Mg2+]i is predominantly
taken up by stores that are located in the ER. Data that have indicated
a spatial heterogeneity of
[Mg2+]i as well as the
presence of cytosolic magnesium gradients have previously been reported
from studies of the human tracheal gland (48)
. Additional
support for the general existence of ER-based magnesium stores can be
seen in studies that have investigated the bee photoreceptor and the
rat sublingual gland (52
, 57)
.
Our experiments indicate further that particularly the recovery of cytosolic magnesium concentrations after withdrawal of the secretory stimulus is independent of the presence or absence of calcium. It depends, however, on the activity of an ATPase as indicated by the fact that this release is thapsigargin sensitive.
The exact mechanisms by which Mg2+ enters and
leaves the intracellular store require further study. One possible
entry pathway might be the intracellular calcium release channels.
Despite the ability of Mg2+ to inhibit
intracellular calcium release channels, it may also pass through these
channels. This kind of retrograde movement ought to be favored by the
electrochemical gradient for Mg2+ across the ER
membrane during secretagogue stimulation and has been shown to occur
for other divalent cations such as Mn2+ in
permeabilized hepatocytes (58)
.
The intracellular signaling events involved in the CCK-induced decrease
of [Mg2+]i in pancreatic
acinar cells are currently under investigation. One critical process
that appears to participate in this signaling cascade is the activation
of protein kinase C. Another issue that requires further study is the
role of agonist-evoked magnesium changes in the physiology and
pathophysiology of the exocrine pancreas. Ongoing studies in which the
early events involved in the onset of acute pancreatitis are being
explored have already shown that the modulating effect of magnesium on
intracellular calcium signaling can be therapeutically used to
attenuate the severity of the disease in an animal model
(59)
.
In conclusion, the results of this study have indicated that intracellular Mg2+ concentrations in the exocrine pancreas are regulated in a stimulus-controlled manner. Changes in [Mg2+]i can counteract the effects in the Ca2+ signal transduction pathway, and the ‘slower’ uptake and release of Mg2+ from intracellular, most probably ER-associated stores, occur in a manner that is completely anti-parallel to the ‘rapid’ Ca2+ release and reuptake in acinar cells. These observations suggest that the role of Mg2+ in the physiology as well as the pathophysiology of the exocrine pancreas and its diseases warrants a completely new evaluation.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Current address: Institut für Sportmedizin der Westfälischen Wilhelms-Universität Münster,
Horstmarer Landweg 39, 48149 Münster, Germany.
Received for publication June 30, 2000. Revised for publication August 31, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. Wiesenberger, K. Steinleitner, R. Malli, W. F. Graier, J. Vormann, R. J. Schweyen, and J. A. Stadler Mg2+ Deprivation Elicits Rapid Ca2+ Uptake and Activates Ca2+/Calcineurin Signaling in Saccharomyces cerevisiae Eukaryot. Cell, April 1, 2007; 6(4): 592 - 599. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Mahurkar, M M Idris, D N Reddy, S Bhaskar, G V Rao, V Thomas, L Singh, and G R Chandak Association of cathepsin B gene polymorphisms with tropical calcific pancreatitis Gut, September 1, 2006; 55(9): 1270 - 1275. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Ch. Mooren, V. Hlouschek, T. Finkes, S. Turi, I. A. Weber, J. Singh, W. Domschke, J. Schnekenburger, B. Kruger, and M. M. Lerch Early Changes in Pancreatic Acinar Cell Calcium Signaling after Pancreatic Duct Obstruction J. Biol. Chem., March 7, 2003; 278(11): 9361 - 9369. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
