1: \documentclass[11pt,titlepage]{article}
2:
3: \usepackage{setspace}
4: \doublespacing
5:
6: \usepackage{amsmath}
7: \usepackage{graphicx}
8: \usepackage[round,numbers,sort&compress]{natbib}
9:
10: \renewcommand{\bibnumfmt}[1]{#1.}
11: \newcommand{\BibTeX}{\textsc{Bib}\TeX}
12:
13: \makeatletter
14: \long\def\@makecaption#1#2{%
15: \vskip\abovecaptionskip
16: \sbox\@tempboxa{#1: #2}%
17: \ifdim \wd\@tempboxa >\hsize
18: #1: #2\par
19: \else
20: \global \@minipagefalse
21: \hb@xt@\hsize{\box\@tempboxa\hfil}%
22: \fi
23: \vskip\belowcaptionskip}
24: \makeatother
25:
26:
27: % Running head
28: \pagestyle{myheadings}
29: \markright{Heterogeneity induced bursting action potential}
30:
31: \begin{document}
32:
33: \begin{titlepage}
34: %\noindent
35: %This un-edited manuscript has been accepted for publication in Biophysical Journal
36: %and is freely available on BioFast at http://www.biophysj.org. The final copyedited version
37: %of the paper may be found at http://www.biophysj.org.
38: %\vspace{1cm}
39:
40: \noindent
41: \textbf{\large How Noise and Coupling Induce Bursting Action Potentials
42: in Pancreatic $\boldsymbol{\beta}$-cells}
43: \vspace{2cm}
44:
45: \noindent
46: \textbf{Junghyo Jo,$^{\star}$ Hyuk Kang,$^{\star}$ Moo Young Choi,$^{\star \dagger}$
47: and Duk-Su Koh$^{\ddagger}$}
48: \vspace{1cm}
49:
50: \noindent
51: $^{\star}$Department of Physics, Seoul National University, Seoul 151-747, Korea;
52: $^{\dagger}$Korea Institute for Advanced Study, Seoul 130-722, Korea;
53: and $^{\ddagger}$Department of Physics, Pohang University of Science and Technology,
54: Pohang 790-784, Korea
55: \vspace{1cm}
56:
57: \noindent
58: E-mail: mychoi@snu.ac.kr
59:
60: \noindent
61: Corresponding author's present address:
62: Department of Physics, Seoul National University, Seoul 151-747, Korea;
63: and Korea Institute for Advanced Study, Seoul 130-722, Korea
64: \end{titlepage}
65:
66:
67: % 200 words max Abstract
68: \abstract{
69: Unlike isolated $\beta$-cells, which usually produce continuous spikes
70: or fast and irregular bursts, electrically coupled $\beta$-cells are
71: apt to exhibit robust bursting action potentials. We consider the noise
72: induced by thermal fluctuations as well as that by channel gating
73: stochasticity and examine its effects on the action potential behavior
74: of the $\beta$-cell model. It is observed numerically that such noise
75: in general helps single cells to produce a variety of electrical activities.
76: In addition, we also probe coupling via gap junctions between neighboring cells,
77: with heterogeneity induced by noise, to find that it enhances regular bursts.
78:
79: \emph{Key words:} Thermal fluctuation; Channel gating stochasticity; Heterogeneity; Gap junction}
80:
81: % New page
82: \clearpage
83:
84: \section*{Introduction}
85: Bursting action potentials, which are characterized by rapid
86: firing interspersed with quiescent periods in pancreatic
87: $\beta$-cells, play a central role in the secretion of insulin,
88: the hormone for glucose homeostasis. It has been reported that
89: isolated $\beta$-cells actually show continuous spikes or fast and
90: irregular bursts \citep{Falke,Kinard,Smith} while $\beta$-cells in a
91: cluster or in an intact islet produce regular bursting action
92: potentials \citep{Andreu,Dean,Sanchez-Andres,Valdeolmillos}.
93: As for the correlations between the electrical activity on the cell
94: membrane and insulin secretion \citep{Henquin}, the
95: robust bursts appear more effective in maintaining glucose homeostasis
96: than continuous spikes, since coupled $\beta$-cells can control insulin
97: release better than isolated $\beta$-cells \citep{Bosco,Halban,Pipeleers}.
98:
99: However, the question as to whether bursting is an endogenous
100: property of individual $\beta$-cells or of a cluster still remains
101: to be answered, which has attracted a number of investigations.
102: Among proposed explanations is the channel-sharing hypothesis,
103: which postulates that current fluctuations arising from channel
104: gating stochasticity prevent single cells, originally capable of
105: bursting, from bursting, but when they are electrically coupled,
106: the perturbing effects are shared by neighbors and the regular
107: bursting is recovered \citep{Aguirre,Chay1,Sherman1}.
108: In contrast to this hypothesis of negative
109: effects of noise, recent research \citep{Lee,Longtin,Pei,Pikovsky,Zaikin}
110: has established that noise can play a constructive role
111: in many biological systems including $\beta$-cell bursting \citep{Vries1}.
112: The heterogeneity hypothesis, providing another explanation,
113: was also postulated by the same group. According to it, when heterogeneous cells,
114: each of which produces continuous spikes or bursts depending upon
115: such cell parameters as the size, channel density, etc., are coupled,
116: those cells in the cluster exhibit more pronounced bursts.
117: This gives a useful insight into the functioning of heterogeneous
118: cell populations \citep{Smolen}.
119:
120: In this study, we expand the concept of heterogeneity and probe
121: how such general heterogeneity enhances bursting. It is proposed that noise
122: induces heterogeneity in otherwise homogeneous individual $\beta$-cells,
123: which in turn assists the $\beta$-cells to produce robust bursts
124: when they are coupled.
125: Existing studies have mostly focused on the synchronizing role of
126: coupling \citep{Sherman2,Vries3};
127: the slow dynamics, which has a period about 10 to 60 seconds,
128: is synchronized successfully between adjacent cells.
129: In contrast, we focus here on the fact that rapid firing in the
130: active phase of bursting is asynchronous between neighbors \citep{Sherman3}
131: and these fluctuating currents
132: through the gap junction act like noise, enhancing the robust
133: bursting action potential.
134: It is also presented that various action potentials
135: of single $\beta$-cells are embodied with optimal noise induced by thermal
136: fluctuations or by ionic channel gating stochasticity.
137: In particular, noise stimulates occasionally itself to produce
138: {\it fast bursts} in a single cell.
139:
140: There are four sections in this paper: In the second section
141: the mathematical model for $\beta$-cells is introduced and the
142: simulation method is described. The third section is devoted to
143: the effects of random noise in currents and of voltage-dependent
144: noise in single cells while the fourth section examines how coupling
145: between cells influences the electrical activity of a cell.
146: Finally, main results are summarized and discussed in the last section.
147:
148:
149: \section*{Model and Methods}
150: \subsection*{Mathematical model for a $\beta$-cell}
151: As the Hodgkin-Huxley model \citep{Hodgkin} describes the
152: electrical activity on the cell membrane with ion channels, a few
153: mathematical models for $\beta$-cells, based on the
154: electrophysiological data \citep{Ashcroft,Gopel,Rorsman} of the ion channels in
155: $\beta$-cells, have been proposed. Although there are simple
156: models using two-dimensional maps \citep{Vries2,Rulkov1,Rulkov2},
157: we consider the Sherman model, which allows direct physical
158: interpretation \citep{Vries1,Sherman4}.
159:
160: The model is described by the current balance equation between
161: capacitive and ionic currents:
162: \begin{eqnarray}
163: C_M \frac{dV}{dt} &=& - I_{Ca}(V) - I_K(V,N) - I_{K(ATP)}(V,P)\nonumber \\
164: && - I_S(V,S), \label{V}
165: \end{eqnarray}
166: where $C_M$ and $V$ denote the membrane capacitance and the
167: membrane potential, respectively. The activation variable $N$ and
168: the slow variable $S$ are governed by
169: \begin{eqnarray}\label{N}
170: \tau_N \frac{dN}{dt} & = & N_{\infty}(V) - N \nonumber \\
171: \tau_{S} \frac{dS}{dt} & = & S_{\infty}(V) - S
172: \end{eqnarray}
173: with appropriate relaxation times $\tau_N$ and $\tau_S$, which are taken to be
174: constants for simplicity.
175: The fraction $P$ of open K(ATP) channels
176: may also be regarded as a constant for the moment [see Eq.~\ref{P}].
177: Ionic currents here are fast voltage-dependent L-type Ca$^{2+}$ current $I_{Ca}$,
178: delayed-rectifier K$^+$ current $I_K$, ATP-blockable K$^+$ current
179: $I_{K(ATP)}$, and very slow inhibitory potassium current $I_S$:
180: \begin{eqnarray}
181: \label{I}
182: I_{Ca}(V) & = & g_{Ca} M_{\infty}(V)(V - V_{Ca}) \nonumber\\
183: I_K(V, N) & = & g_K N (V - V_K) \nonumber \\
184: I_{K(ATP)}(V, P) & = & g_{K(ATP)} P (V - V_K) \\
185: I_S(V, S) & = & g_S S (V - V_K).\nonumber
186: \end{eqnarray}
187: $I_{Ca}$ and $I_K$ are responsible for generating action
188: potentials; $I_{Ca}$ is assumed to respond instantaneously to a
189: change in the membrane potential, whereas $I_K$ is governed by the
190: dynamics of the activation variable $N$ via Eq.~\ref {N}.
191: $I_{K(ATP)}$ is the background current with voltage-independent
192: conductance $g_{K(ATP)}$; this determines the plateau fraction,
193: i.e., the ratio of the active phase duration to the burst period.
194: For example, as $g_{K(ATP)}$ decreases under high glucose
195: concentration, there are only active phases without silent phases.
196: $I_S$ is a phenomenological current representing slow dynamics in
197: the bursting action potential.
198: This model thus assumes that single $\beta$-cells originally
199: contain the slow dynamics, which works just under the appropriate
200: condition. Biological candidates for such slow dynamics include
201: slow free Ca$^{2+}$ dynamics \citep{Chay2} and ATP
202: metabolism \citep{Keizer}. Finally, $M_{\infty}$,
203: $N_{\infty}$, and $S_{\infty}$ of the voltage-dependent activation
204: are defined to be
205: \begin{equation}
206: X_{\infty}(V) = \frac{1}{1 + \exp{[(V_X - V) / \theta_X]}},
207: \end{equation}
208: where $X$ denotes $M$, $N$, or $S$.
209:
210: This set of coupled nonlinear differential equations in
211: Eqs.~\ref{V} - \ref{I} has been analyzed in
212: detail \citep{Keener,Rinzel}. There it is noted
213: that $S$ responds on a much slower time scale than $V$ and $N$
214: because $\tau_S$ has the time scale of several seconds compared
215: with the milli-second time scale in firing. Then $S$ is regarded
216: just as a parameter, and the dynamics of the fast subsystem on the
217: two-dimensional phase space of $V$ and $N$ is analyzed. Furthermore,
218: after eliminating one degree of freedom by substituting $N_{\infty}$
219: to $N$, the whole behavior of this model may be analyzed
220: approximately with fast variable $V$ and slow variable $S$.
221:
222: \subsection*{Numerical details}
223: Integration of differential equations including noise demands some
224: caution, and is commonly achieved via the Euler method. For better
225: efficiency, we employ the Euler method for integrating the noise
226: term, combined with the second-order Runge-Kutta method for other
227: terms. In order to be concrete, we consider the one-variable
228: problem
229: \begin{equation}
230: \frac{dx}{dt} = f(x) + \xi(t),
231: \end{equation}
232: where $f(x(t))$ is a (nonlinear) function of $x$, the variable of
233: concern, and $\xi(t)$ is the white noise with zero mean and
234: delta-function correlations
235: \begin{eqnarray}
236: \langle \xi(t)\rangle & = & 0, \nonumber \\
237: \label{GWN} \langle \xi(t) \xi(t')\rangle & = & 2 D\delta(t-t').
238: \end{eqnarray}
239: Taking the time step of size $\Delta t$, we obtain from the
240: equation of motion the value of $x$ at time $t+\Delta t$:
241: \begin{equation}
242: x(t+\Delta t) = x(t) + \frac{f(x(t)) + f(\bar{x})}{2} \Delta t +
243: \xi(t){\Delta t},
244: \end{equation}
245: where $\bar{x} \equiv x(t) + f(x(t)) \Delta t + \xi(t){\Delta t}$ \citep{Batrouni}.
246: Although there is no gurantee that this algorithm should converge in general,
247: it works fine here since the noise term does not depend on the variable $x$ \citep{Kloeden}.
248:
249: The white noise $\xi$ of variance $D$ is produced by the gaussian random
250: numbers with the variance $\sigma^2$ determined by
251: \begin{equation}
252: \langle\xi(t)^2\rangle
253: = \int_{-\infty}^{\infty} d\xi \frac{1}{\sqrt{2 \pi} \sigma}
254: e^{-\xi^2/2\sigma^2} \xi^2 = \frac{2D}{\Delta t},
255: \end{equation}
256: where the Dirac delta function has been represented by ${\Delta t}^{-1}$
257: within the numerical accuracy.
258: We thus have the relation $\sigma =\sqrt{2 D/\Delta t}$.
259:
260: In our simulations, we take $\Delta t = 1$\,ms,
261: which turns out to be small enough, and integrate the set of equations for current
262: balance. This gives the time evolution of the action potential,
263: from which the power spectrum is computed through the use of the
264: fast Fourier transform technique.
265:
266:
267: \section*{Results and Discussion}
268: \subsection*{Noise effects}
269: Before explaining the coupling effects, we first probe the role of
270: noise, either the usual (additive) random noise or the
271: (multiplicative) voltage-dependent one. Comparison of the effects
272: of such noise helps us to understand better the coupling effects.
273:
274: \subsubsection*{Random noise}
275: Among many kinds of noise on the cell membrane, the simplest case
276: is the random noise, which may come from thermal
277: fluctuations (see below). When such random current fluctuations
278: are present on the membrane, the current balance equation in
279: Eq.~\ref{V} is generalized to
280: \begin{eqnarray}
281: \label{V2}
282: C_M \frac{dV}{dt} & = & - I_{ion}(V,N,S) - \xi(t),
283: \end{eqnarray}
284: where $I_{ion}$ represents all the ionic currents on the
285: right-hand side in Eq.~\ref{V}, and the noise current $\xi (t)$
286: satisfies Eq.~\ref{GWN} with the variance denoted by $D_{\xi}$.
287:
288: Figure~\ref{Fig:rd_noise} exhibits the solution of the set of
289: coupled differential equations in Eqs.~\ref{N}
290: and~\ref{V2} under various strengths of the random noise.
291: It is observed that single $\beta$-cells produce various
292: electrical activities according to the value of $\tau_N$ in
293: Eq.~\ref{N}, which lies in the narrow range 4 to
294: 11\,ms depending on the membrane potential~(27).
295: When the time constant $\tau_N$ of delayed-rectifier K$^+$ channel
296: activity exceeds 11.0\,ms, the $\beta$-cell produces regular spiking
297: action potentials in Fig.~\ref{Fig:rd_noise} \textit{A},
298: while for $\tau_N$ below 10.0\,ms faster repolarization does not
299: allow enough time for the slow variable $S$ to decrease, yielding
300: bursting action potentials [see Fig.~\ref{Fig:rd_noise} \textit{C}].
301: In the intermediate regime of $\tau_N$=10.2\,ms,
302: Fig.~\ref{Fig:rd_noise} \textit{B} shows that spiking action potentials
303: are generated but the bursting property is resident.
304: As an appropriate amount of noise comes into play,
305: in particular, the regular spikes in Figs.~\ref{Fig:rd_noise} \textit{A}
306: and \textit{B} and bursts in \textit{C} change into fast bursts
307: in \textit{E}, irregular spikes in \textit{G}, or irregular bursts
308: in \textit{H} and \textit{I}.
309:
310: To explain these phenomena, we note two thresholds of the slow
311: variable $S$: One is the upper threshold above which the membrane
312: potential is falling into the resting potential; the other is the
313: lower threshold above which the membrane potential begins to fire.
314: At the moment that fluctuations take negative values, they may
315: assist the repolarizing membrane potential to remain above the
316: lower threshold before the membrane repolarizes completely and then,
317: depolarizes slowly to the lower threshold. This induces
318: occasionally consecutive firing in Fig.~\ref{Fig:rd_noise} \textit{G}
319: or even fast bursts in Fig.~\ref{Fig:rd_noise} \textit{E} for the
320: $\beta$-cell in the critical parameter range, i.e.,
321: $\tau_N$=10.2\,ms. Such consecutive firing raises the average
322: membrane potential for a while, compared with the case of regular
323: spikes. Hence the value of $S_{\infty}$ becomes large, and
324: consequently $S$ grows with the delay represented by the time
325: constant $\tau_S$. When it goes over the upper threshold, the
326: membrane potential returns to the resting potential. At the same
327: time, $S_{\infty}$ now becomes small and $S$ reduces to the lower
328: threshold. During this period of $S$ varying from the upper
329: threshold to the lower one, the membrane potential stays in the
330: silent phase. When $S$ comes to the lower threshold, the membrane
331: potential starts to depolarize and fire. Repetition of these
332: processes simply constitutes the fast bursts.
333: As the noise level is raised further, the slow variable $S$ may
334: start to increase before it reaches the lower threshold,
335: assisted by the fluctuations taking negative values.
336: Similarly it may start to decrease before it reaches the upper
337: threshold due to positive fluctuations.
338: In consequence, irregular bursts in Figs.~\ref{Fig:rd_noise} \textit{H}
339: and \textit{I} can thus be induced.
340: When fluctuations become sufficiently strong and dominant, such a role
341: of noise, turning on the slow dynamics of $S$, is concealed and
342: the membrane potential appears noisy.
343: Here it is notable that under optimal fluctuations, there exists
344: the critical parameter range in which the difference between the
345: upper and lower thresholds is small and the dramatic effect of fast bursts
346: is produced; similar results were obtained in a recent study \citep{Aguirre}.
347:
348: It is revealing to examine the power spectra of the obtained
349: action potentials, computed through the use of the fast Fourier
350: transform technique for various noise levels and displayed in
351: Fig.~\ref{Fig:rd_power}.
352: In particular, Figure~\ref{Fig:rd_power} \textit{B} manifests
353: that the regular spiking action potential of frequency 2\,Hz
354: in the absence of noise has changed into fast bursts containing
355: oscillations of 0.2\,Hz and 5\,Hz at moderate noise levels.
356:
357: To characterize the positive/negative role of noise in bursting,
358: we define the bursting tendency according to
359: $\cal{B} \equiv \log [\cal{P}($$f$$_B)/\cal{P}($$0)]$,
360: where $\cal{P}($$f_B)$ is the power spectrum at the bursting frequency $f_B$
361: and $\cal{P}($$0)$ is the background intensity at 0\,Hz.
362: Figure~\ref{Fig:rd_summary} shows the behavior of the bursting tendency $\cal{B}$
363: with the noise level, manifesting the noise effects on bursting.
364:
365: Finally, one may ask whether thermal fluctuations known to
366: generate white noise are enough to induce the fast bursts,
367: irregular bursts or spikes, observed in our simulations. In
368: simulations, the variance $D_{\xi}$ is taken in the range
369: $10^{-29}$\,J/$\Omega$ $\sim$ $10^{-27}$\,J/$\Omega$. In reality,
370: noise currents due to thermal fluctuations can be estimated via
371: the fluctuation-dissipation theorem: $D_{\xi} = {k_B T}/{R}$. This
372: gives $D_{\xi} \sim 10^{-29}$\,J/$\Omega$
373: when $R$ is taken to be a few giga ohms (G$\Omega$) or less.
374: Accordingly, thermal fluctuations alone may not be enough
375: to induce irregular spikes or bursts.
376: Nevertheless, it appears possible that thermal fluctuations
377: actually expedite the emergence of fast bursts when the cell lies
378: in the critical parameter regime.
379:
380: \subsubsection*{Voltage-dependent noise}
381: As another simple type of noise, one can consider the voltage-dependent
382: fluctuations, which are closely related to the channel gating
383: stochasticity (see below). In the presence of such multiplicative noise,
384: the current balance condition in Eq.
385: \ref{V} takes the form
386: \begin{eqnarray}
387: \label{V3}
388: C_M \frac{dV}{dt} & = & - I_{ion}(V,N,S) -\eta(t)(V-V_K),
389: \end{eqnarray}
390: where $I_{ion}$ also represents all the ionic currents in
391: Eq.~\ref{V}, and $\eta(t)$ is the Gaussian white noise, again
392: satisfying Eq.~\ref{GWN} with variance $D_{\eta}$. Solving
393: numerically the coupled differential equations given by
394: Eqs.~\ref{N} and~\ref{V3} at various noise levels with
395: $\tau_N$ set equal to 11\,ms, we obtain the results, which are
396: illustrated in Fig.~\ref{Fig:vd_noise}. Note the overall similarity
397: to the case of random (additive) noise shown in
398: Figs.~\ref{Fig:rd_noise} \textit{D} and \textit{G}.
399:
400: When the voltage-dependent noise stimulates the cell membrane,
401: irregular spikes arise, similarly to the case of random noise, if
402: its amplitude multiplied by the voltage difference $(V{-}V_K)$ is
403: comparable to the amplitude of random noise. In fact,
404: voltage-dependent noise may be regarded simply as the noise
405: weighted more in the active phase of the membrane potential than
406: in the silent phase. When taking negative values, therefore,
407: fluctuations boost firing more effectively in the active phase and
408: contribute less to the erratic evolution of the resting potential
409: in the silent phase.
410: Such voltage-dependent (multiplicative) noise may arise from ion
411: channel gating stochasticity, since currents through channels
412: depend upon the membrane potential difference. If the number of
413: channels is sufficiently large, the channel stochasticity can be
414: described by a Langevin equation \citep{DeFelice,Fox1,Fox2}.
415: %
416: Specifically, the stochasticity of K(ATP) channels has been
417: considered \citep{Vries1}. In the expression for the
418: ATP-dependent $K^+$ current, $I_{K(ATP)} = g_{K(ATP)} P (V - V_K)$,
419: the opening ratio $P$, which is no more constant, evolves
420: according to
421: \begin{eqnarray}
422: \label{P} \frac{dP}{dt} = \frac{\gamma_1}{\tau_{P}} (1 - {P}) -
423: \frac{\gamma_2}{\tau_{P}} P + \bar{\xi} (t),
424: \end{eqnarray}
425: where ${\gamma_1}/{\tau_{P}}$ and ${\gamma_2}/{\tau_{P}}$
426: represent the rates for a closed channel to switch to the open state
427: and vice versa, respectively.
428: Note that $\gamma_1$ and $\gamma_2$ thus determine the equilibrium ratio
429: between the open state and the closed one.
430: Fluctuations in the opening ratio are described by the Gaussian
431: white noise $\bar{\xi}(t)$ satisfying Eq.~\ref{GWN} with the variance
432: \begin{equation}
433: D_{\bar{\xi}} = \frac{\gamma_1 (1 - P) + \gamma_2 P}{2 \tau_{P} N_{K(ATP)}}
434: \approx \frac{\gamma_1 \gamma_2}{\tau_{P} N_{K(ATP)} (\gamma_1 +
435: \gamma_2)},
436: \label{Dbarxi}
437: \end{equation}
438: where $N_{K(ATP)}$ is the total number of ATP-dependent $K^+$
439: channels in a $\beta$-cell \citep{Fox1}.
440:
441: Solving Eq.~\ref{P}, we obtain that $P$ fluctuates around the equilibrium
442: value $P_0$, taken to be $0.5$ in our simulations: $P(t) = P_0 + \bar{\eta}(t)$.
443: Here $\bar{\eta}(t)$ is colored noise, characterized by the variance
444: \begin{eqnarray}
445: \langle\bar{\eta}(t)\bar{\eta}(t')\rangle = D_{\bar{\eta}}
446: [{\gamma}e^{-\gamma |t-t'|} - {\gamma}e^{-\gamma (t +t')}]
447: \end{eqnarray}
448: with $\gamma \equiv (\gamma_1 + \gamma_2)/\tau_p$ and
449: $D_{\bar{\eta}} \equiv D_{\bar{\xi}}/\gamma^2$ (see Appendix for details).
450: Note that the firing time scale is comparable to the correlation time $\gamma^{-1}$
451: of the noise $\bar{\eta}(t)$ (see Fig.~\ref{Fig:corr}).
452: Consequently this colored noise is more effective to induce several consecutive firings,
453: which resemble irregular burst, than the white noise.
454: In particular, the modules of several spikes are observed to become longer
455: as the correlation time $\gamma^{-1}$ is increased.
456: Figure~\ref{Fig:co_noise} shows the behaviors in the presence of the
457: channel-gating noise $\bar{\xi}(t)$ for two different channel numbers.
458: In this case of multiplicative colored noise, modules of spikes arise
459: more efficiently than in the case of mutiplicative white noise shown in Fig.~\ref{Fig:vd_noise}.
460: Further, it is also found that stronger gating fluctuations from less channels
461: ($N_{K(ATP)}$=500) in Fig.~\ref{Fig:co_noise} \textit{B} give rise to modules
462: of more rapid spikes, compared with the case $N_{K(ATP)}$=2500 in Fig.~\ref{Fig:co_noise} \textit{A}.
463:
464: Similar results can be obtained with fluctuations in the
465: Ca$^{2+}$ channels and in the delayed-rectifier K$^+$ channels
466: although they act somewhat differently from
467: the fluctuations in the ATP-blockable K$^+$ channels (data not shown).
468:
469: It is thus concluded that noise generates diverse firing patterns in single $\beta$-cells.
470: In a real (physiological) islet, however, $\beta$-cells are not isolated but coupled with each other,
471: making it desirable to consider coupled $\beta$-cells and to investigate effects of noise together with
472: those of coupling. This will be the subject of the next section.
473:
474: \subsection*{Coupling effects}
475: We consider two cells coupled with each other via a
476: gap junction. With the coupling incorporated, Eq.~\ref{V} is
477: extended to the coupled equations:
478: \begin{eqnarray}
479: \label{V4}
480: C_{M} \frac{dV_1}{dt}&=& - I_{ion}(V_1,N_1,S_1,P_1) -g_C (V_1 - V_2) \nonumber \\
481: C_{M} \frac{dV_2}{dt}&=& - I_{ion}(V_2,N_2,S_2,P_2) -g_C (V_2 -
482: V_1),
483: \end{eqnarray}
484: where the subscripts $1$ and $2$ are the cell indices, $I_{ion}$
485: again denotes all the ionic currents, and $g_C$ is the coupling
486: conductance. Note that the heterogeneity between both cells is
487: accommodated in the K(ATP) channel opening ratio $P$. Namely, the
488: noise associated with channel gating stochasticity induces
489: continuously heterogeneity between the cells.
490:
491: We thus have eight coupled differential equations, which consist
492: of Eqs.~\ref{N} and~\ref{P} for each cell and Eq.~\ref{V4},
493: for eight variables ($V, N, S$, and $P$ for each cell).
494: Integration of these coupled equations yields the results displayed
495: in Fig.~\ref{Fig:couple}, for the channel-gating noise of variance
496: $D_{\bar{\xi}}= 4 \times 10^{-4}$\,s$^{-1}$ given by Eq.~\ref{Dbarxi}
497: and for three values of the coupling conductance:
498: $g_C = 50$\,pS, $110$\,pS, and $200$\,pS.
499: Revealed is the optimal coupling strength for longer bursting periods:
500: While weak coupling is not enough to couple individual cells and
501: to generate consecutive firing, too strong coupling tends to make the
502: cluster behave as a single large cell \citep{Sherman2}.
503:
504: Robust bursts emerge as a consequence of the competition
505: between heterogeneity and coupling \citep{Vries3}.
506: On one hand, the coupling term in Eq.~\ref{V4} helps the two cells
507: to act synchronously; on the other hand, it also plays the role of
508: stimulating noise, which acts strongly on the two cells
509: with asynchronous phases.
510: The perfect asynchrony results from the harmony of coupling
511: to be similar and heterogeneity to be different (see Fig.~\ref{Fig:phase}).
512: Namely, the coupling currents between asynchronous neighboring cells give
513: rise to consecutive firing; this in turn increases the upper
514: threshold of the slow variable $S$ above which firing disappears.
515: As $S$ grows up toward the increased upper threshold, it takes
516: longer to reduce down to the lower threshold.
517: This larger rising and falling divides more clearly the active and silent
518: phases in the membrane potential, and accordingly induces
519: robust bursting action potentials with periods longer than 20\,s.
520: Note that in the absence of coupling we have not been able to observe
521: bursting periods longer than 10\,s
522: (see Figs.~\ref{Fig:rd_noise}-\ref{Fig:co_noise})
523: (Parameter values different from those in Table~I may
524: yield bursting periods somewhat longer than 10\,s even in a single cell.
525: In this case, the coupling gives rise to robust bursting of even longer
526: periods, say, 30\,s, still demonstrating its crucial role in generating
527: regular bursts.)
528:
529: In the two-cell model here the optimal value of the coupling conductance
530: is observed to be $g_C = 110$\,pS. As the number of cells is increased,
531: however, more heterogeneity is introduced, which should be matched by
532: stronger coupling to generate robust bursts with longer periods.
533: Although the detailed investigation is beyond our computing capacity,
534: we have performed multi-cell simulations, which indeed confirms such
535: an increase of the optimal coupling conductance.
536: For example, the optimal conductance in the system of 1000 cells
537: turns out to be 100 to 300\,pS (data not shown),
538: which coincides with experimental results of the gap junctional conductance
539: \citep{Perezarmendariz}.
540:
541: These features of the coupled cells do not change much in the presence of
542: the voltage-dependent noise instead of the channel-gating noise,
543: except that the channel-gating noise is more efficient for robust bursting
544: than the voltage-dependent one, as shown in Fig.~\ref{Fig:couple_vd} for
545: $D_{\eta}= 10^{-24}$\,J/$\Omega\cdot$V$^2$ and $D_{\bar{\xi}}=0$.
546: Note also that the coupled cells depicted in Figs.~\ref{Fig:couple} and
547: ~\ref{Fig:couple_vd} do not burst in the absence of noise-induced heterogeneity.
548:
549: Recall that in the emergence of robust bursts, the asynchrony from the
550: heterogeneity induced by noise plays an important role,
551: which has also been addressed in a very recent study \citep{Pedersen}.
552: Similar to such noise-induced heterogeneity,
553: the cell-to-cell heterogeneity associated with variations of the cell
554: parameters among the cells is also expected to play for robust bursts \citep{Smolen}.
555: To check this, we allowed variations of the membrane capacitance $C_M$
556: related to the cell size as well as of the channel conductance $g_{K(ATP)}$
557: and examine the resulting behavior:
558: Shown in Fig.~\ref{Fig:hetero} \textit{A} and \textit{B}
559: are bursts generated in the case of 20\,\% variation
560: of $C_M$ (5.0\,pF, 6.3\,pF) and in the case of 10\,\% variation of $g_{K(ATP)}$
561: (1000\,pS, 1100\,pS), respectively.
562: Specifically, a spiking cell (with $C_M =6.3$\,pF) is coupled with a bursting cell
563: (with $C_M =5.0$\,pF) in Fig.~\ref{Fig:hetero} \textit{A},
564: which results in that both cells are bursting synchronously with
565: a longer bursting period than that of a single cell (5.0\,pF).
566: In Fig.~\ref{Fig:hetero} \textit{B}, on the other hand, two spiking cells
567: (with $g_{K(ATP)}=1000$\,pS and $1100$\,pS) are coupled with each other,
568: and both are bursting.
569: Therefore heterogeneity is in general important for bursting in coupled cells,
570: no matter whether it is cell-to-cell heterogeneity or induced by noise.
571:
572:
573: \section*{Conclusions}
574: We have probed whether noise and coupling serve as an appropriate
575: stimulus for inducing the bursting action potential in pancreatic
576: $\beta$-cells, and found that they effectively call into action
577: the inherent slow dynamics in individual cells. Fast bursts,
578: irregular spikes or bursts in single $\beta$-cells have been
579: observed as the results of the noise effects. In particular the
580: emergence of regular bursts assisted by an appropriate amount of
581: noise [see Figs.~\ref{Fig:rd_noise} \textit{E} and~\ref{Fig:rd_power} \textit{B}]
582: is reminiscent of {\it coherence resonance} \citep{Lee,Longtin,Pei,Pikovsky,Zaikin}.
583: In view of physiology, the consecutive firing induced by
584: fluctuations gives rise to relative depolarization for a while,
585: which is followed by the activation of the slow potassium channel
586: lasting until the slow variable reaches the upper threshold. At
587: this time the slow $K^+$ channel opens fully, and the outflux of
588: cytosolic potassium ions gets very large, thus hindering
589: depolarization. Accordingly, the membrane potential is compelled
590: to stay in the silent phase, and the slow $K^+$ channel in turn
591: starts to be inactivated. In consequence, the membrane can become
592: depolarized as the outflux of $K^+$ ions reduces. Finally, firing
593: occurs again, and consecutive firing also happens by the help of
594: appropriate stimulation. As candidates for the stimulus, both the
595: (additive) random noise coming from fluctuating currents and the
596: (multiplicative) voltage-dependent noise from the channel gating
597: stochasticity have been considered.
598:
599: In particular, coupling between cells has turned out essential for
600: attaining regular bursts with longer periods compared with the
601: fast bursts. The coupling term, proportional to the potential
602: difference between two cells, operates in a similar manner to the
603: voltage-dependent noise: It increases with the potential
604: difference and thus becomes large for the cells in active phases,
605: stimulating the cells like noise. On the other hand, it is small
606: for perfectly synchronized cells in silent phases. The coupling
607: also increases the upper threshold of $S$ and induces robust
608: regular bursts.
609:
610: In the analysis, the heterogeneity has been found to play an important role
611: in inducing strong fluctuations during active phases, which may cause robust bursts.
612: Namely, bursting in general results from the interplay of
613: coupling and heterogeneity.
614: This allows us to interpret the fact that large cell clusters (up to
615: the critical size) show more regular bursts \citep{Vries1,Sherman2}:
616: Assuming a cubic islet, we have considered $\beta$-cells
617: arranged into an $L^3$ cube, under free boundary conditions.
618: Adopting physiological gap junction conductance, $g_C=200$\,pS~(42),
619: we have found that the bursting period and duration first increases with
620: the size $L$ but tends to saturate beyond $L=5$ (data not shown).
621: Such saturation behavior may be explained as follows:
622: Via the coupling through gap junctions, the number of nearest neighbors
623: in the three-dimensional space is limited, e.g., to six or so;
624: this suggests that the cluster above some critical size
625: can get no more advantage of the heterogeneity from
626: neighboring cells through given coupling strength.
627:
628: The Langerhans islet, however, consists of several endocrine cells
629: in addition to $\beta$-cells. Other endocrine cells in an islet
630: have been studied recently \citep{Kanno,Nadal},
631: and it will be of interest to study the coupling effects between
632: originally different $\alpha$-, $\beta$-, and $\delta$-cells,
633: coupled via hormones or neurotransmitters \citep{Moriyama}.
634: This might give a clue to understanding the size of a Langerhans
635: islet in the pancreas, which is left for further study.
636:
637: \section*{Appendix}
638: Equation~\ref{P} can be solved to give the time evolution of the
639: opening ratio $P$:
640: \begin{equation}
641: P(t) = P_0 + [ P(0) - P_0 ] e^{-\gamma t} + \int_{0}^{t}
642: e^{-\gamma (t-t')} \bar{\xi}(t') dt' \nonumber
643: \end{equation}
644: with $P_0 \equiv \gamma_1 / (\gamma_1 + \gamma_2 )$ and
645: $\gamma \equiv (\gamma_1 + \gamma_2)/\tau_p$,
646: where $P(0)$ is the initial value of $P$.
647: After sufficiently long time, we thus have $P$
648: fluctuating around the equilibrium ratio $P_0$: $P(t) = P_0 +
649: \bar{\eta}(t)$, where the noise $\bar{\eta}(t)$ is given by
650: \begin{eqnarray}
651: \bar{\eta}(t) \equiv \int_{0}^{t} e^{-\gamma (t-t')} \bar{\xi}(t') dt'.
652: \nonumber
653: \end{eqnarray}
654: From the above definition of the noise $\bar{\eta}(t)$, it is
655: straightforward to derive its characteristics:
656: \begin{eqnarray*}
657: \langle\bar{\eta}(t)\bar{\eta}(t')\rangle
658: &=& \int_0^t d\tau e^{\gamma(\tau-t)} \int_0^{t'} d\tau' e^{\gamma(\tau' -t')}
659: \langle \bar{\xi}(\tau) \bar{\xi}(\tau') \rangle \\
660: &=& 2D_{\bar{\xi}} e^{-\gamma (t+t')} \int_0^{\bar{t}} d\tau e^{2\gamma\tau},
661: \end{eqnarray*}
662: where we have used the relation
663: $\langle \bar{\xi}(\tau) \bar{\xi}(\tau') \rangle = 2D_{\bar{\xi}} \delta (\tau -\tau')$
664: and $\bar{t}$ denotes the smaller one between $t$ and $t'$.
665: We thus obtain the correlations of the noise $\bar{\eta}$ at different times
666: \begin{equation*}
667: \langle\bar{\eta}(t)\bar{\eta}(t')\rangle
668: = D_{\bar{\eta}}
669: [{\gamma}e^{-\gamma |t-t'|} - {\gamma}e^{-\gamma (t +t')}]
670: \end{equation*}
671: with $D_{\bar{\eta}} \equiv D_{\bar{\xi}}/\gamma^2$,
672: which manifests the colored nature.
673:
674: \vspace{2cm}
675: %/acknowledgment
676: This work was supported in part by KOSEF through Grant
677: No. 01-2002-000-00285-0 and by the MOST (KOSEF) through
678: National Core Research Center for Systems Bio-Dynamics, as well as
679: by the BK21 Program.
680: Helpful reprints from the Laboratory of Biological Modeling
681: in NIDDK of NIH are also gratefully acknowledged.
682:
683: %\bibliography{beta}
684: \begin{thebibliography}{46}
685: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
686:
687: \bibitem[{Falke et~al.(1989)Falke, Gillis, Pressel, and Misler}]{Falke}
688: Falke, L.~C., K.~D. Gillis, D.~M. Pressel, and S.~Misler. 1989.
689: \newblock `Perforated patch recording' allows long-term monitoring of
690: metabolite-induced electrical activity and voltage-dependent Ca$^{2+}$
691: currents in pancreatic islet $\beta$-cells.
692: \newblock \emph{FEBS Lett.} 251:167--172.
693:
694: \bibitem[{Kinard et~al.(1999)Kinard, de~Vries, Sherman, and Satin}]{Kinard}
695: Kinard, T.~A., G.~de~Vries, A.~Sherman, and L.~S. Satin. 1999.
696: \newblock Modulation of the bursting properties of single mouse pancreatic
697: $\beta$-cells by artificial conductances.
698: \newblock \emph{Biophys. J.} 76:1423--1435.
699:
700: \bibitem[{Smith et~al.(1990)Smith, Ashcroft, and Rorsman}]{Smith}
701: Smith, P.~A., F.~M. Ashcroft, and P.~Rorsman. 1990.
702: \newblock Simultaneous recordings of glucose dependent electrical activity and
703: atp-regulated K$^+$-currents in isolated mouse pancreatic $\beta$-cells.
704: \newblock \emph{FEBS Lett.} 261:187--190.
705:
706: \bibitem[{Andreu et~al.(1997)Andreu, Soria, and S\'anchez-Andr\'es}]{Andreu}
707: Andreu, E., B.~Soria, and J.~V. S\'anchez-Andr\'es. 1997.
708: \newblock Oscillation of gap junction electrical coupling in the mouse
709: pancreatic islets of langerhans.
710: \newblock \emph{J. Physiol.} 498:753--761.
711:
712: \bibitem[{Dean and Matthews(1968)}]{Dean}
713: Dean, P.~M., and E.~K. Matthews. 1968.
714: \newblock Electrical activity in pancreatic islet cells.
715: \newblock \emph{Nature} 219:389--390.
716:
717: \bibitem[{S\'anchez-Andr\'es et~al.(1995)S\'anchez-Andr\'es, Gomis, and
718: Valdeolmillos}]{Sanchez-Andres}
719: S\'anchez-Andr\'es, J.~V., A.~Gomis, and M.~Valdeolmillos. 1995.
720: \newblock The electrical activity of mouse pancreatic $\beta$-cells recorded in
721: vivo shows glucose-dependent oscillations.
722: \newblock \emph{J. Physiol.} 486:223--228.
723:
724: \bibitem[{Valdeolmillos et~al.(1996)Valdeolmillos, Gomis, and
725: S\'anchez-Andr\'es}]{Valdeolmillos}
726: Valdeolmillos, M., A.~Gomis, and J.~V. S\'anchez-Andr\'es. 1996.
727: \newblock In vivo synchronous membrane potential oscillations in mouse
728: pancreatic $\beta$-cells: lack of co-ordination between islets.
729: \newblock \emph{J. Physiol.} 493:9--18.
730:
731: \bibitem[{Henquin and Meissner(1984)}]{Henquin}
732: Henquin, J.~C., and H.~P. Meissner. 1984.
733: \newblock Significance of ionic fluxes and changes in membrane potential for
734: stimulus-secretion coupling in pancreatic $\beta$-cells.
735: \newblock \emph{Experientia} 40:1043--1052.
736:
737: \bibitem[{Bosco et~al.(1989)Bosco, Orci, and Meda}]{Bosco}
738: Bosco, D., L.~Orci, and P.~Meda. 1989.
739: \newblock Homologous but not heterologous contact increases the insulin
740: secretion of individual pancreatic $\beta$-cells.
741: \newblock \emph{Exp. Cell Res.} 184:72--80.
742:
743: \bibitem[{Halban et~al.(1982)Halban, Wollheim, Blondel, Meda, Niesor, and
744: Mintz}]{Halban}
745: Halban, P.~A., C.~B. Wollheim, B.~Blondel, P.~Meda, E.~N. Niesor, and D.~H.
746: Mintz. 1982.
747: \newblock The possible importance of contact between pancreatic islet cells for
748: the control of insulin release.
749: \newblock \emph{Endocrinology} 111:86--94.
750:
751: \bibitem[{Pipeleers et~al.(1982)Pipeleers, Veld, Maes, and Winkel}]{Pipeleers}
752: Pipeleers, D., P.~I. Veld, E.~Maes, and M.~V.~D. Winkel. 1982.
753: \newblock Glucose-induced insulin release depends on functional cooperation
754: between islet cells.
755: \newblock \emph{Proc. Natl. Acad. Sci. USA} 79:7322--7325.
756:
757: \bibitem[{Aguirre et~al.(2004)Aguirre, Mosekilde, and Sanju\'an}]{Aguirre}
758: Aguirre, J., E.~Mosekilde, and M.~A.~F. Sanju\'an. 2004.
759: \newblock Analysis of the noise-induced bursting-spiking transition in a
760: pancreatic $\beta$-cell model.
761: \newblock \emph{Phys. Rev. E} 69:041910.
762:
763: \bibitem[{Chay and Kang(1988)}]{Chay1}
764: Chay, T.~R., and H.~S. Kang. 1988.
765: \newblock Role of single-channel stochastic noise on bursting clusters of
766: pancreatic $\beta$-cells.
767: \newblock \emph{Biophys. J.} 54:427--435.
768:
769: \bibitem[{Sherman et~al.(1988)Sherman, Rinzel, and Keizer}]{Sherman1}
770: Sherman, A., J.~Rinzel, and J.~Keizer. 1988.
771: \newblock Emergence of organized bursting in clusters of pancreatic
772: $\beta$-cells by channel sharing.
773: \newblock \emph{Biophys. J.} 54:411--425.
774:
775: \bibitem[{Lee et~al.(1998)Lee, Neiman, and Kim}]{Lee}
776: Lee, S.~G., A.~Neiman, and S.~Kim. 1998.
777: \newblock Coherence resonance in a hodgkin-huxley neuron.
778: \newblock \emph{Phys. Rev. E} 57:3292--3297.
779:
780: \bibitem[{Longtin(1997)}]{Longtin}
781: Longtin, A. 1997.
782: \newblock Autonomous stochastic resonance in bursting neurons.
783: \newblock \emph{Phys. Rev. E} 55:868--876.
784:
785: \bibitem[{Pei et~al.(1996)Pei, Wilkens, and Moss}]{Pei}
786: Pei, X., L.~Wilkens, and F.~Moss. 1996.
787: \newblock Noise-mediated spike timing precision from aperiodic stimuli in an
788: array of hodgkin-huxley-type neuron.
789: \newblock \emph{Phys. Rev. Lett.} 77:4679--4682.
790:
791: \bibitem[{Pikovsky and Kurths(1997)}]{Pikovsky}
792: Pikovsky, A.~S., and J.~Kurths. 1997.
793: \newblock Coherence resonance in a noise-driven excitable system.
794: \newblock \emph{Phys. Rev. Lett.} 78:775--778.
795:
796: \bibitem[{Zaikin et~al.(2003)Zaikin, Garc\'{\i}a-Ojalvo, B\'ascones, Ullner,
797: and Kurths}]{Zaikin}
798: Zaikin, A., J.~Garc\'{\i}a-Ojalvo, R.~B\'ascones, E.~Ullner, and J.~Kurths.
799: 2003.
800: \newblock Doubly stochastic coherence via noise-induced symmetry in bistable
801: neural models.
802: \newblock \emph{Phys. Rev. Lett.} 90:03061.
803:
804: \bibitem[{de~Vries and Sherman(2000)}]{Vries1}
805: de~Vries, G., and A.~Sherman. 2000.
806: \newblock Channel sharing in pancreatic $\beta$-cells revisited: enhancement of
807: emergent bursting by noise.
808: \newblock \emph{J. Theor. Biol.} 207:513--530.
809:
810: \bibitem[{Smolen et~al.(1993)Smolen, Rinzel, and Sherman}]{Smolen}
811: Smolen, P., J.~Rinzel, and A.~Sherman. 1993.
812: \newblock Why pancreatic islets burst but single beta cells do not. the
813: heterogeneity hypothesis.
814: \newblock \emph{Biophys. J.} 64:1668--1680.
815:
816: \bibitem[{Sherman and Rinzel(1991)}]{Sherman2}
817: Sherman, A., and J.~Rinzel. 1991.
818: \newblock Model for synchronization of pancreatic $\beta$-cells by gap junction
819: coupling.
820: \newblock \emph{Biophys. J.} 59:547--559.
821:
822: \bibitem[{de~Vries and Sherman(2001)}]{Vries3}
823: de~Vries, G., and A.~Sherman. 2001.
824: \newblock From spikers to bursters via coupling: help from heterogeneity.
825: \newblock \emph{Bull. Math. Biol.} 63:371--391.
826:
827: \bibitem[{Sherman and Rinzel(1992)}]{Sherman3}
828: Sherman, A., and J.~Rinzel. 1992.
829: \newblock Rhythmogenic effects of weak electrotonic coupling in neuronal
830: models.
831: \newblock \emph{Proc. Natl. Acad. Sci. USA} 89:2471--2474.
832:
833: \bibitem[{Hodgkin and Huxley(1952)}]{Hodgkin}
834: Hodgkin, A.~L., and A.~F. Huxley. 1952.
835: \newblock A quantitative description of membrane current and its application to
836: conduction and excitation in nerve.
837: \newblock \emph{J. Physiol.} 117:500--544.
838:
839: \bibitem[{Ashcroft and Rorsman(1989)}]{Ashcroft}
840: Ashcroft, F.~M., and P.~Rorsman. 1989.
841: \newblock Electrophysiology of the pancreatic $\beta$-cell.
842: \newblock \emph{Prog. Biophys. Mol. Biol.} 54:87--143.
843:
844: \bibitem[{G\"opel et~al.(1999)G\"opel, Kanno, Barg, Galvanovskis, and
845: Rorsman}]{Gopel}
846: G\"opel, S., T.~Kanno, S.~Barg, J.~Galvanovskis, and P.~Rorsman. 1999.
847: \newblock Voltage-gated and resting membrane currents recorded from
848: $\beta$-cells in intact mouse pancreatic islets.
849: \newblock \emph{J. Physiol.} 521:717--728.
850:
851: \bibitem[{Rorsman and Trube(1986)}]{Rorsman}
852: Rorsman, P., and G.~Trube. 1986.
853: \newblock Calcium and delayed potassium currents in mouse pancreatic
854: $\beta$-cells under voltage-clamp conditions.
855: \newblock \emph{J. Physiol.} 374:531--550.
856:
857: \bibitem[{de~Vries(2001)}]{Vries2}
858: de~Vries, G. 2001.
859: \newblock Bursting as an emergent phenomenon in coupled chaotic maps.
860: \newblock \emph{Phys. Rev. E} 64:051914.
861:
862: \bibitem[{Rulkov(2001)}]{Rulkov1}
863: Rulkov, N.~F. 2001.
864: \newblock Regularization of synchronized chaotic bursts.
865: \newblock \emph{Phys. Rev. Lett.} 86:183--186.
866:
867: \bibitem[{Rulkov(2002)}]{Rulkov2}
868: Rulkov, N.~F. 2002.
869: \newblock Modeling of spiking-bursting neural behavior using two-dimensional
870: map.
871: \newblock \emph{Phys. Rev. E} 65:041922.
872:
873: \bibitem[{Sherman(1996)}]{Sherman4}
874: Sherman, A. 1996.
875: \newblock Contributions of modeling to understanding stimulus-secretion
876: coupling in pancreatic $\beta$-cells.
877: \newblock \emph{Am. J. Physiol.} 271:362--372.
878:
879: \bibitem[{Chay and Keizer(1983)}]{Chay2}
880: Chay, T.~R., and J.~Keizer. 1983.
881: \newblock Minimal model for membrane oscillations in the pancreatic beta-cell.
882: \newblock \emph{Biophys. J.} 42:181--190.
883:
884: \bibitem[{Keizer and Magnus(1989)}]{Keizer}
885: Keizer, J., and G.~Magnus. 1989.
886: \newblock ATP-sensitive potassium channel and bursting in the pancreatic beta
887: cell. a theoretical study.
888: \newblock \emph{Biophys. J.} 56:229--242.
889:
890: \bibitem[{Keener and Sneyd(1998)}]{Keener}
891: Keener, J., and J.~Sneyd. 1998.
892: \newblock Mathematical Physiology.
893: \newblock Springer-Verlag, New York, 188--215.
894:
895: \bibitem[{Rinzel(1987)}]{Rinzel}
896: Rinzel, J. 1987.
897: \newblock A formal classification of bursting mechanisms in excitable systems.
898: \newblock \emph{In} Mathematical Topics in Population Biology, Morphogenesis,
899: and Neurosciences, E.~Teramoto, and M.~Yamaguti, editors. Springer-Verlag,
900: New York, 267--281.
901:
902: \bibitem[{Batrouni et~al.(1985)Batrouni, Katz, Kronfeld, Lepage, Svetitsky, and
903: Wilson}]{Batrouni}
904: Batrouni, G.~G., G.~R. Katz, A.~S. Kronfeld, G.~P. Lepage, B.~Svetitsky, and
905: K.~G. Wilson. 1985.
906: \newblock Langevin simulations of lattice field theories.
907: \newblock \emph{Phys. Rev. D} 32:2736.
908:
909: \bibitem[{Kloeden et~al.(1994)Kloeden, Platen, and Schurz}]{Kloeden}
910: Kloeden, P.~E., E.~Platen, and H.~Schurz. 1994.
911: \newblock Numerical Solution of SDE through Computer Experiments.
912: \newblock Springer-Verlag, Berlin, 151.
913:
914: \bibitem[{DeFelice and Isaac(1992)}]{DeFelice}
915: DeFelice, L.~J., and A.~Isaac. 1992.
916: \newblock Chaotic states in a random world: relationship between the nonlinear
917: differential equations of excitability and the stochastic properties of ion
918: channels.
919: \newblock \emph{J. Stat. Phys.} 70:339--354.
920:
921: \bibitem[{Fox and Lu(1994)}]{Fox1}
922: Fox, R.~F., and Y.~N. Lu. 1994.
923: \newblock Emergent collective behavior in large numbers of globally coupled
924: independently stochastic ion channels.
925: \newblock \emph{Phys. Rev. E} 49:3421--3431.
926:
927: \bibitem[{Fox(1997)}]{Fox2}
928: Fox, R.~F. 1997.
929: \newblock Stochastic versions of the hodgkin-huxley equations.
930: \newblock \emph{Biophys. J.} 72:2068--2074.
931:
932: \bibitem[{P\'erez-Armendariz et~al.(1991)P\'erez-Armendariz, Roy, Sparay, and
933: Bennett}]{Perezarmendariz}
934: P\'erez-Armendariz, M., C.~Roy, D.~C. Sparay, and M.~V.~L. Bennett. 1991.
935: \newblock Biophysical properties of gap junctions between freshly dispersed
936: pairs of mouse pancreatic beta cells.
937: \newblock \emph{Biophys. J.} 59:76--92.
938:
939: \bibitem[{Pedersen(2005)}]{Pedersen}
940: Pedersen, M.~G. 2005.
941: \newblock A comment on noise enhanced bursting in pancreatic $\beta$-cells.
942: \newblock \emph{J. Theor. Biol.} 235:1--3.
943:
944: \bibitem[{Kanno et~al.(2002)Kanno, G\"opel, Rorsman, and Wakui}]{Kanno}
945: Kanno, T., S.~O. G\"opel, P.~Rorsman, and M.~Wakui. 2002.
946: \newblock Cellular function in multicellular system for hormone-secretion:
947: electrophysiological aspect of studies on $\alpha$-, $\beta$- and
948: $\delta$-cells of the pancreatic islet.
949: \newblock \emph{Neurosci. Res.} 42:79--90.
950:
951: \bibitem[{Nadal et~al.(1999)Nadal, Quesada, and Soria}]{Nadal}
952: Nadal, A., I.~Quesada, and B.~Soria. 1999.
953: \newblock Homologous and heterologous asynchronicity between identified alpha-,
954: beta- and delta-cells within intact islets of langerhans in the mouse.
955: \newblock \emph{J. Physiol.} 517:85--93.
956:
957: \bibitem[{Moriyama and Hayashi(2003)}]{Moriyama}
958: Moriyama, Y., and M.~Hayashi. 2003.
959: \newblock Glutamate-mediated signaling in the islets of langerhans: a thread
960: entangled.
961: \newblock \emph{TRENDS in Pharmacol. Sci.} 42:511--517.
962:
963: \end{thebibliography}
964: \clearpage
965:
966: \section*{Table}
967: \begin{table}[h]
968: \caption{Standard parameter values}
969: \begin{tabular}{cc}
970: \hline
971: $C_M = 6.3$\,pF & $g_{Ca} = 3000$\,pS \\
972: $g_{K} = 4000$\,pS & $g_{K(ATP)} = 1000$\,pS \\
973: $g_{S} = 3000$\,pS & $g_C = 110$\,pS \\
974: $V_{Ca} = 25$\,mV & $V_{K} = -75$\,mV \\
975: $V_{M} = -20$\,mV & $\theta_{M} = 12$\,mV \\
976: $V_{N} = -17$\,mV & $\theta_{N} = 5.6$\,mV \\
977: $V_{S} = -22$\,mV & $\theta_{S} = 8.0$\,mV \\
978: $\tau_N = 1.1\times 10^{-2}$\,s & $\tau_S = 20$\,s \\
979: $\tau_P = 0.50$\,s & $N_{K(ATP)} = 2500$\, \\
980: $\gamma_1 = 1$ & $\gamma_2 = 1$ \\
981: \hline
982: \end{tabular}
983: \label{tab:table1}
984: \end{table}
985:
986:
987: % Figure legends
988: \clearpage
989: \section*{Figure Legends}
990: \subsubsection*{Figure~\ref{Fig:rd_noise}.}
991: Action potential $V$ and slow channel activity
992: $S$ in single $\beta$-cells at the noise level $D_{\xi}=0, 10^{-29}$,
993: and $10^{-27}$\,J/$\Omega$ under several values of time constant
994: $\tau_N$ of delayed-rectifier K$^+$ channel activity $N$. All
995: simulations have been performed under the standard parameter
996: values in Table~I except $\tau_N$, the values of
997: which are given above.
998:
999: \subsubsection*{Figure~\ref{Fig:rd_power}.}
1000: Power spectra of the action potentials for
1001: the random noise levels in Fig.~\ref{Fig:rd_noise}. The time constant
1002: $\tau_N$ of the activation variable $N$ is (\textit{A}) 11.0\,ms and
1003: (\textit{B}) 10.2\,ms. Observed in the power spectra are main peaks
1004: together with their harmonics. The peak at 1\,Hz, indicated by the asterisk
1005: in (\textit{B}), reflects the tendency to form dimerization of spikes.
1006: Each power spectrum has been obtained from the average over
1007: 1000 samples, each having a time sequence of $132$ seconds.
1008:
1009:
1010: \subsubsection*{Figure~\ref{Fig:rd_summary}.}
1011: Bursting tendency $\cal{B}$ of $\beta$-cells versus the noise level
1012: for several values of $\tau_N$, corresponding to different firing patterns
1013: in the absence of noise.
1014:
1015: \subsubsection*{Figure~\ref{Fig:vd_noise}.}
1016: Action potential $V$ and slow channel activity $S$
1017: in single $\beta$-cells at two values of the voltage-dependent noise.
1018: Again parameter values in Table~I have been used.
1019:
1020: \subsubsection*{Figure~\ref{Fig:corr}.}
1021: Correlations between the action potential and
1022: multiplicative colored noise due to channel-gating stochasticity.
1023: The correlation time $\gamma^{-1}$ is taken to be (\textit{A}) 25\,ms,
1024: (\textit{B}) 250\,ms, and (\textit{C}) 2500\,ms. Note that each figure
1025: has a different time scale. Their corresponding power spectra are shown in (\textit{D}).
1026: Parameter values in Table~I have been used except $\tau_P$.
1027:
1028: \subsubsection*{Figure~\ref{Fig:co_noise}.}
1029: Action potential $V$ and slow channel activity $S$
1030: in single $\beta$-cells at two values of the channel gating stochasticity:
1031: (\textit{A}) and (\textit{B}) correspond to the channel number
1032: $N_{K(ATP)}= 2500$ and $500$, respectively.
1033: Note that Figs.~\ref{Fig:corr} \textit{B} and \ref{Fig:co_noise} \textit{A}
1034: represent the same sample path, but with different variables plotted.
1035: Other parameter values have been taken from Table~I.
1036:
1037: \subsubsection*{Figure~\ref{Fig:couple}.}
1038: Figure~\ref{Fig:phase}. Enlarged view of the interval between 10\,s to 11\,s in
1039: Fig.~\ref{Fig:couple} \textit{B}, disclosing the detailed behavior of the two
1040: membrane potentials $V_1$ (\textit{solid line}) and $V_2$ (\textit{dashed line}).
1041:
1042: \subsubsection*{Figure~\ref{Fig:couple_vd}.}
1043: Bursting action potential induced
1044: by cell coupling via the gap junction of conductance $g_C =110$\,pS
1045: under the voltage-dependent noise of strength
1046: $D_{\eta}=10^{-24}$\,J/$\Omega\cdot$V$^2$.
1047: Parameter values in Table~I have been used.
1048:
1049: \subsubsection*{Figure~\ref{Fig:hetero}.}
1050: Bursting action potential induced by cell coupling,
1051: with the cell-to-cell heterogeneity due to variations of the membrane capacitance $C_M$
1052: and of the ATP-blockable K$^+$ channel conductance $g_{K(ATP)}$:
1053: (\textit{A}) 20\,\% variation of $C_M$ (5.0\,pF, 6.3\,pF);
1054: (\textit{B}) 10\,\% variation of $g_{K(ATP)}$ (1000\,pS, 1100\,pS).
1055: Other parameter values have been taken from Table~I.
1056:
1057: % Figures, one per page (fig_1.eps and fig_1.pdf files must be present
1058: % in the document directory)
1059: \clearpage
1060: \begin{figure}
1061: \begin{center}
1062: \includegraphics*[width=1.2\textwidth]{fig1}
1063: \caption{}
1064: \label{Fig:rd_noise}
1065: \end{center}
1066: \end{figure}
1067:
1068: \clearpage
1069: \begin{figure}
1070: \begin{center}
1071: \includegraphics*[width=1.2\textwidth]{fig2}
1072: \caption{}
1073: \label{Fig:rd_power}
1074: \end{center}
1075: \end{figure}
1076:
1077: \clearpage
1078: \begin{figure}
1079: \begin{center}
1080: \includegraphics*[width=0.6\textwidth]{fig3}
1081: \caption{}
1082: \label{Fig:rd_summary}
1083: \end{center}
1084: \end{figure}
1085:
1086: \clearpage
1087: \begin{figure}
1088: \begin{center}
1089: \includegraphics*[width=1.2\textwidth]{fig4}
1090: \caption{}
1091: \label{Fig:vd_noise}
1092: \end{center}
1093: \end{figure}
1094:
1095: \clearpage
1096: \begin{figure}
1097: \begin{center}
1098: \includegraphics*[width=1.2\textwidth]{fig5}
1099: \caption{}
1100: \label{Fig:corr}
1101: \end{center}
1102: \end{figure}
1103:
1104: \clearpage
1105: \begin{figure}
1106: \begin{center}
1107: \includegraphics*[width=1.2\textwidth]{fig6}
1108: \caption{}
1109: \label{Fig:co_noise}
1110: \end{center}
1111: \end{figure}
1112:
1113: \clearpage
1114: \begin{figure}
1115: \begin{center}
1116: \includegraphics*[width=1.2\textwidth]{fig7}
1117: \caption{}
1118: \label{Fig:couple}
1119: \end{center}
1120: \end{figure}
1121:
1122: \clearpage
1123: \begin{figure}
1124: \begin{center}
1125: \includegraphics*[width=0.6\textwidth]{fig8}
1126: \caption{}
1127: \label{Fig:phase}
1128: \end{center}
1129: \end{figure}
1130:
1131: \clearpage
1132: \begin{figure}
1133: \begin{center}
1134: \includegraphics*[width=0.6\textwidth]{fig9}
1135: \caption{}
1136: \label{Fig:couple_vd}
1137: \end{center}
1138: \end{figure}
1139:
1140: \clearpage
1141: \begin{figure}
1142: \begin{center}
1143: \includegraphics*[width=1.2\textwidth]{fig10}
1144: \caption{}
1145: \label{Fig:hetero}
1146: \end{center}
1147: \end{figure}
1148:
1149:
1150: % closing statement, nothing below matters
1151: \end{document}
1152:
1153: