Download Inositol trisphosphate and calcium oscillations

BY0074-0001.indd Page 1
12/18/06
4:29:24 PM elhi /Volumes/ju108/POIN002/symbosia_indd%0
1
Biochem. Soc. Symp. 74, 1– 7
(Printed in Great Britain)
© 2007 The Biochemical Society
Inositol trisphosphate and
calcium oscillations
Michael J. Berridge1
The Babraham Institute, Babraham, Cambridge, CB2 4AT, U.K.
Abstract
InsP3 has two important functions in generating Ca2+ oscillations. It releases
Ca from the internal store and it can contribute to Ca2+ entry. A hypothesis
has been developed to describe a mechanism for Ca2+ oscillations with particular
emphasis on the way agonist concentration regulates oscillator frequency. The
main idea is that the InsP3 receptors are sensitized to release Ca2+ periodically
by cyclical fluctuations of Ca2+ within the lumen of the endoplasmic reticulum.
Each time a pulse of Ca2+ is released, the luminal level of Ca2+ declines and has
to be replenished before the InsP3 receptors are resensitized to deliver the next
pulse of Ca2+. It is this loading of the internal store that explains why frequency
is sensitive to external Ca2+ and may also account for how variations in agonist
concentration are translated into changes in oscillation frequency. Variations
in agonist-induced entry of external Ca2+, which can occur through different
mechanisms, determine the variable rates of store loading responsible for
adjusting the sensitivity of the InsP3 receptors to produce the periodic pulses
of Ca2+. The Ca2+ oscillator is an effective analogue-to-digital converter in that
variations in the concentration of the external stimulus are translated into a
change in oscillator frequency.
2+
Introduction
Ca2+ is a highly versatile second messenger capable of regulating many
different cellular processes [1]. More often than not, this Ca2+ signal is presented
as a brief transient. In some cases, as in muscle and neurons, each Ca2+ transient
is produced ‘on demand’ in that it has to be activated by a brief membrane
depolarization. In many other cells, however, that respond to stimuli that persist
for much longer periods, Ca2+ is still presented as brief transients that occur
repetitively in the form of a Ca2+ oscillation. The mechanism responsible for
1
email [email protected]
1
BY0074-0001.indd Page 2 12/5/06 11:05:33 AM elhi /Users/elhi/Desktop/5 Dec_Bijender/indd
2
M.J. Berridge
setting up these oscillations remains a subject of debate. In this chapter I shall
explore how the Ca2+-mobilizing second messenger InsP3 may function to
generate these oscillations.
Oscillatory cell types
There are many examples of Ca2+ oscillations, which occur throughout the
life history of a typical cell. Oscillations occur early during development in that
they appear in immature oocytes [2] and play a critical role in oocyte activation
at the time of fertilization [3]. Once cells have differentiated, oscillations
continue to play a role in regulating many different cell types such as liver cells
[4], smooth muscle cells [5], astrocytes [6], airway epithelial cells [7], duck salt
gland [8] and insulin-secreting β-cells [9,10]. For many of these oscillations, it
is clear that InsP3 plays a central role, but exactly how it functions to set up
the repetitive transients remains to be determined. Before considering possible
mechanisms, it is necessary to establish the main properties of this oscillatory
activity.
Properties of cellular oscillators
Oscillations of Ca2+ usually occur at low agonist concentrations, which
normally correspond to the narrow range responsible for activating physiological
responses. In the case of the insect salivary gland, for example, the range of 5-HT
(5-hydroxytryptamine) concentrations that induced oscillations corresponded
exactly to those that stimulated fluid secretion [11,12]. At threshold doses of
5-HT, oscillation frequency was very slow, but this increased markedly over
the course of the dose–response curve for the activation of fluid secretion. This
observation provided the first indication that cells might employ FM (frequency
modulation) as a mechanism to encode information. Such an FM mode of
signalling has been described in many other cell types such as airway epithelial
cells [7], liver cells [4], smooth muscle cells [9] and insulin-secreting β-cells [10].
There is increasing evidence that cells use FM to encode information about the
level of the external stimulus. Oscillation models must thus account for this
analogue-to-digital conversion, which is the basis for this FM mode of signal
transmission.
One of the properties of Ca2+ oscillations that is difficult to explain is that
frequency can sometimes be very low. For example, in the case of mammalian
oocytes undergoing fertilization, the Ca2+ transients can have periodicities of
the order of 4 min. Similarly, the repetitive intercellular Ca2+ waves that sweep
through the developing zebrafish embryo at fertilization have a periodicity as
long as 10 min [13]. These long intervals present a major challenge for any model
attempting to explain such Ca2+ oscillations.
Another key feature of this oscillatory activity is its sensitivity to the
concentration of external Ca2+. However, this property can be somewhat
variable. In those cells that are very efficient at recycling their internal Ca2+,
© 2007 The Biochemical Society
BY0074-0001.indd Page 3 12/5/06 11:05:33 AM elhi /Users/elhi/Desktop/5 Dec_Bijender/indd
Inositol trisphosphate and calcium oscillations
3
oscillator frequency is relatively independent of external Ca2+. However, other
cell types are very sensitive to external Ca2+ and oscillator frequency responds
immediately to any change in the level of external Ca2+. In these cases, the entry
of external Ca2+ is a key feature of many oscillators and may play a critical role
as part of the mechanism for analogue-to-digital conversion. These different
properties of Ca2+ oscillations have to be taken into account when considering
possible oscillatory mechanisms.
Ca2+ oscillation models
A number of models have been proposed to account for these cytosolic Ca2+
oscillations [14]. They fall into two main groups; some are based on feedback
loops between Ca2+ and the enzymes that either form or inactivate InsP3, such
as PLC (phospholipase C) or InsP3 kinase respectively, that set up oscillations
in InsP3 that then drive the Ca2+ oscillations [15–17]. Some evidence for such
oscillations has been reported using a GFP (green fluorescent protein)-tagged
PH (pleckstrin homology) domain that translocates from the membrane to
the cytoplasm when InsP3 is formed [18]. In a separate series of experiments
using the same technique, InsP3 oscillations were recorded in CHO (Chinese
hamster ovary) cells following stimulation of the mGluR5a receptors but not
the M3-muscarinic receptors [19,20]. It is possible therefore, that there may be
different mechanisms for generating Ca2+ oscillations, which may or may not
depend upon InsP3 oscillations. With regard to these published measurement,
it should be stressed that the GFP-tagged PH domain is not measuring InsP3
directly, but is monitoring the breakdown of the precursor lipid PtdIns(4,5)P2.
Also, the time resolution of the traces is not sufficiently precise to determine
whether the proposed InsP3 responses precede or follow the Ca2+ transients.
Whether or not oscillations in InsP3 drive Ca2+ oscillations is thus still an open
question.
In order to determine whether the level of InsP3 oscillates during the course
of a typical Ca2+ oscillation, I carried out an experiment on the insect salivary
gland where it is possible to monitor both Ca2+ oscillations and the formation
of InsP3 indirectly by measuring the accumulation of InsP1 using the lithium
amplification technique [21]. First, I determined the 5-HT concentration that
produced low frequency oscillations that could then be accelerated to the same
extent either by the addition of higher doses of 5-HT or by increasing the external level of Ca2+. I then measured the accumulation of InsP1 under these two
conditions. If InsP3 formation is driving each transient, the expectation would
be for the turnover of inositol phosphates to increase for both conditions.
However, what was found was that the level of InsP1 was increased only when
the oscillator was accelerated by an increase in the agonist but not when it was
accelerated by an elevation of external Ca2+ (M. J. Berridge, unpublished work).
This experimental observation suggests that periodic increases in InsP3 are not
responsible for driving Ca2+ oscillations, at least in the case of some cell types.
It is possible that cells have different mechanisms for generating oscillations,
with some using InsP3 oscillations and others not. This possibility is supported
© 2007 The Biochemical Society
BY0074-0001.indd Page 4 12/5/06 11:05:33 AM elhi /Users/elhi/Desktop/5 Dec_Bijender/indd
4
M.J. Berridge
by a method based on studying the effect of applying pulses of InsP3 at specific
points during the oscillatory cycle, which concluded that the oscillatory
mechanism in cells may vary with regard to whether or not the level of InsP3
oscillates [22]. In addition to the mathematical models described above, which
convincingly demonstrate that it is oscillations in InsP3 that drive Ca2+ oscillations,
there are equally convincing models showing that oscillations in Ca2+ can occur
at constant levels of InsP3. Some of these oscillation models propose that Ca2+
oscillations are driven by feedback mechanisms that operate on the InsP3
receptor and can occur at a constant level of InsP3. Many of the mechanisms
depend upon a reversible desensitization of the InsP3 receptor [23–27]. The idea
is that there is an alternation between two processes: a Ca2+-induced activation
of the InsP3 receptor followed by inactivation of the Ca2+-sensitive state of the
InsP3 receptor. What is not clear for many of these desensitization models is
whether they can account for agonist-dependent frequency modulation.
All of the models described above are based on feedback loops operating
within the cytoplasm. There has been little attention paid to the possible role
played by luminal Ca2+ in setting up oscillatory activity. A luminal loading
Ca2+ oscillation model belongs to the group of models that predict that Ca2+
oscillations can occur at constant InsP3 levels. This luminal loading model can
also provide an explanation for the oscillations that have very long periods (i.e.
4–10 min) and it can also account for frequency modulation [28,29].
Luminal loading Ca2+ oscillation model
The main feature of the luminal loading Ca2+ oscillation model is that the
timing of each Ca2+ spike is determined by the periodic loading of the endoplasmic reticulum with Ca2+ (Figure 1). The best way of describing how this
oscillator might function is to follow what happens when cells are stimulated
with the low agonist concentrations that set up Ca2+ oscillations. The agonist
usually has two effects. It acts to stimulate the formation of InsP3, but this is not
sufficient at low agonist concentrations to stimulate release directly. In addition,
the agonist also promotes an inward Ca2+ current (ICa) by activating one of the entry
mechanisms that may or may not depend on InsP3. The result of this increased
entry is that the endoplasmic reticulum begins to load up with Ca2+. Since the
InsP3 receptors are sensitive to the concentration of Ca2+ in the lumen, the gradual
build up of luminal Ca2+ will sensitize the receptors to the low ambient levels
of InsP3 and Ca2+ and these are then able to activate release. Once release terminates, the cytosolic Ca2+ is dealt with by three mechanisms (Figure 1). Some is
extruded from the cell by the PMCA (plasma-membrane Ca2+-ATPase) pump,
some of it is returned to the endoplasmic reticulum by the SERCA (sarcoplasmic/endoplasmic-reticulum Ca2+-ATPase) pump and some of it is taken up by
the mitochondria. This mitochondrial Ca2+ is then gradually returned to the
cytoplasm during the interspike interval and much of it may be transferred back
to the endoplasmic reticulum to help reload the store. This reloading process
repeats itself to give rise to the next spike. Oscillation frequency is determined
© 2007 The Biochemical Society
BY0074-0001.indd Page 5 12/5/06 11:05:33 AM elhi /Users/elhi/Desktop/5 Dec_Bijender/indd
Inositol trisphosphate and calcium oscillations
5
Figure 1 A luminal loading Ca2+ oscillation model. The agonists that
induce oscillations have two important actions. During the initial phase following
agonist addition, there is a latent (L) phase during which the agonist activates
PLC (phospholipase C) to produce a local microdomain of InsP3, which diffuses
into the cell to sensitize the InsP3 receptor (InsP3R). Agonists also stimulate
the entry of external Ca2+, which may or may not depend on InsP3. The main
purpose of this entry mechanism is to increase the inward Ca2+ current (ICa)
that then begins to load the store. The gradual increase in the concentration of
Ca2+ within the lumen of the ER then sensitizes the InsP3Rs to the low ambient
levels of InsP3 and Ca2+. When the InsP3Rs are fully sensitized, they begin to
release Ca2+. The corresponding fall in the luminal level of Ca2+ may contribute
to the termination of the release phase. During the recovery phase, Ca2+ is
dealt with by three processes: some is extruded from the cell by the PMCA,
some is returned to the ER (endoplasmic reticulum) by the SERCA pump and
some is taken up by the mitochondria before being returned to the ER. The
period (P) to the next spike depends upon the time it takes to reload the
store. See Figure 2 for further details of the four main events during a typical
oscillatory cycle.
by the rate of store reloading, which may be the primary event responsible for
the process of analogue-to-digital conversion.
The four main events of the oscillatory cycle are illustrated in Figure 2.
The agonist-dependent entry of external Ca2+ is responsible for store loading.
The build up of luminal Ca2+ begins to sensitize the InsP3 receptors and leads to
spike initiation, often accompanied by the appearance of puffs during the final
stages of the pacemaker phase [30]. The development of the spike depends upon
the regenerative process of CICR (Ca2+-induced Ca2+ release) and often results
in the appearance of an intracellular Ca2+ wave. Indeed, the rise time of the Ca2+
spike can provide an estimate of the rate of wave propagation. Once release is
complete, spike recovery begins and Ca2+ is removed by the three mechanisms
© 2007 The Biochemical Society
BY0074-0001.indd Page 6 12/5/06 11:05:34 AM elhi /Users/elhi/Desktop/5 Dec_Bijender/indd
6
M.J. Berridge
Figure 2 Summary of the four main phases of the luminal loading Ca2+
oscillation model. The basis of the model is that the agonist generates both a
constant level of InsP3 and a constant rate of Ca2+ entry.The latter is responsible
for loading the store (bottom panel).When the endoplasmic reticulum begins to
load with Ca2+, it sensitizes the InsP3 receptors such that they begin to generate
puffs as part of a spike initiation process.The puffs then trigger the development
of a spike during which a wave of Ca2+ spreads through the cell through a process
of CICR (Ca2+-induced Ca2+ release). During the spike recovery phase, some of
the Ca2+ is pumped out of the cell and the process of store loading begins again
as a prelude to the next spike.
described earlier (Figure 1). The process then loops back to the store reloading
phase to set the stage for the next spike.
The period (P) between spikes will depend on how quickly the store can
reload, which in turn, depends upon the rate at which Ca2+ enters the cell across
the plasma membrane. This model predicts that the concentration of Ca2+ within
the lumen of the endoplasmic reticulum must decline rapidly during the course
of the spike and then build up gradually towards the onset of the next spike,
which is exactly what is found experimentally when the level of luminal Ca2+ is
measured in gonadotrophs [31] and in pancreatic acinar cells [32]. An important
feature of the model is the sudden fall in the luminal level of Ca2+ during the
course of the spike, which is followed by a gradual phase of reloading during the
period leading up to the next spike as shown diagrammatically in Figure 1.
Conclusions
The luminal loading Ca2+ oscillation model was developed to explain how
variations in agonist concentration are transformed in to digital Ca2+ spikes with
© 2007 The Biochemical Society
BY0074-0001.indd Page 7 12/5/06 11:05:34 AM elhi /Users/elhi/Desktop/5 Dec_Bijender/indd
Inositol trisphosphate and calcium oscillations
7
variable frequencies, which is the basis of frequency modulated Ca2+ oscillations.
It is proposed that the analogue detector is the Ca2+ entry process that responds
to changes in agonist concentration by varying the amount of Ca2+ entering the
cell. The digital converter is the InsP3 receptor that is sensitive to the level of
Ca2+ within the lumen of the endoplasmic reticulum. The variable rate of Ca2+
entry (the analogue response) will determine oscillator frequency by adjusting
the rate of store loading and hence the sensitivity of the InsP3 receptor (the
digital response). This model can also explain why the oscillator is so sensitive
to variations in the concentration of external Ca2+ because this has a direct effect
on the rate of store loading.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
Berridge, M.J., Bootman, M.D. and Roderick, H.L. (2003) Nat. Rev. Mol. Cell Biol. 4,
517–529
Carroll, J., Swann, K., Whittingham, D. and Whitaker, M. (1994) Development 120,
3507–3517
Marangos, P., FitzHarris, G. and Carroll, J. (2003) Development 130, 1461–1472
Robb-Gaspers, L.D., Rutter, G.A., Burnett, P., Hajnóczky, G., Denton, R.M. and Thomas,
A.P. (1998) Biochim. Biophys. Acta 1366, 17–32
Filosa, J.A., Bonev, A.D. and Nelson, M.T. (2004) Circ. Res. 95, e73–e81
Nett, W.J., Oloff, S.H. and McCarthy, K.D. (2002) J. Neurophysiol. 87, 528–537
Zhang, L. and Sanderson, M.J. (2003) J. Physiol. 546, 733–749
Crawford, K.M., Stuenkel, E.L. and Ernst, S.A. (1991) Am. J. Physiol. 261, C177–C184
Iino, M., Kasai, H. and Yamazawa, T. (1994) EMBO J. 13, 5026–5031
Schöfl, C., Rössig, L., Leitholf, H., Mader, T., von zur Mühlen, A. and Brabant, G. (1996)
Cell Calcium 19, 485–493
Rapp, P.E. and Berridge, M.J. (1981) J. Exp. Biol. 93, 119–132
Galione, A. and Berridge, M.J. (1988) Biochem. Soc. Trans. 16, 988–990
Gilland, E., Miller, A.L., Kerpus, E., Baker, R. and Webb, S.E. (1999) Proc. Natl. Acad.
Sci. U.S.A. 96, 157–161
Schuster, S., Marhl, M. and Höfer, T. (2002) Eur. J. Biochem. 269, 1333–1355
Meyer, T and Stryer, L. (1988) Proc. Natl. Acad. Sci. U.S.A. 85, 5051–5055
Cuthbertson, K.S. and Chay, T.R. (1991) Cell Calcium 12, 97–109
Politi, A., Gaspers, L.D., Thomas, A.P. and Höfer, T. (2006) Biophys. J. 90, 3120–3133
Hirose,K., Kadowaki, S., Tanabe, M., Takeshima, H. and Iino, M. (1999) Science 284,
1527–1530
Nash, M.S., Young, K.W., Callis, R.A, J. and Nahorski, S.R. (2001) Nature 413, 381–382
Young, K.W., Nash, M.S., Challis, R.A. and Nahorski, S.R. (2003) J. Biol. Chem. 278,
20753–20760
Berridge, M.J., Downes, C.P. and Hanley, M.R. (1982) Biochem. J. 206, 587–595
Sneyd, J., Tsaneva-Atanasov, K., Reznikov, V., Bai, Y., Sanderson, M.J. and Yule, D.I. (2006)
Proc. Natl. Acad. Sci. U.S.A. 103, 1675–1680
Atri, A., Amundson, J., Clapham, D. and Sneyd, J. (1993) Biophys. J. 65, 1727–1739
Li, Y-X. and Rinzel, J. (1994) J. Theor. Biol. 166, 461–473
Oancea, E. and Meyer,T. (1996) J. Biol. Chem. 271, 17253–17260
Dupont, G. and Swillens, S. (1996) Biophys. J. 71, 1714–1722
Hajnóczky, G. and Thomas, A.P. (1997) EMBO J. 16, 3533–3543
Berridge, M.J. (1993) Nature 361, 315–325
Berridge, M.J. (1994) Biochem. J. 302, 545–550
Bootman, M.D., Berridge, M.J. and Lipp, P. (1997) Cell 91, 367–373
Tse, F.W., Tse, A. and Hille, B. (1994) Proc. Natl. Acad. Sci. U.S.A. 91, 9750–9754
Park, M.K., Petersen, O.H., and Tepikin, A.V. (2000) EMBO J. 19, 5729–5739
© 2007 The Biochemical Society