Finite Volume Model to Study the Effect of

ER flux on Cytosolic Calcium Distribution

in Astrocytes
Brajesh Kumar Jha, Neeru Adlakha and M. N. Mehta
Abstract Most of the intra-cellular events involved in the initiation and propagation phases of this process has now been
identified astrocytes. The control of the spread of intracellular calcium signaling has been demonstrated to occur at several le-
vels including IP3 receptors, intracellular Ca
2+
stores like endoplasmic reticulum (ER) . In normal and pathological situations that
affect one or several of these steps can be predicted to influence on astrocytic calcium waves. In view of above a mathematical
model is developed to study interdependence of all the important parameters like diffusion coefficient and influx over [Ca
2+
] pro-
file. Model incorporates the ER fluxes like ,
leak Pump Chan
J J and J . Finite volume method is employed to solve the problem.
A program has been developed using in MATLAB 7.5 for the entire problem and simulated on an AMD-Turion 32-bite machine
to compute the numerical results. In view of above a mathematical model is developed to study calcium transport between cyto-
sol and ER.
Key words Ca
2+
profile, ER flux, Astrocytes, FVM.

1 INTRODUCTION
ARGEandlonglastingcytosoliccalciumsignallingin
astrocytes have been described in cultured cells and
acute slice preparations. The mechanisms that give
risetothesecalciumeventshavebeenextensivelystudied
in vivo. However, their existence and functions in the in
tactbrainareunknown[1],[6],[7][11].About20yearsago
astrocytes have been considered to mediate supportive
and protective functions in the central nervous system
becauseoftheirstrategicplacementrelativetothevascu
lature, and because they lack fast sodium action poten
tials. NowAstrocytes are controlling the dynamics of the
neuronalnetworksinthecentralnervoussystem[8],[10].
In cultured astrocytes response to many physio
logical and pharmacological manipulations, for example
mechanical stimulation, membrane potential depolariza
tion, and activation of metabotropic glutamate receptors
etc.[2],[5],[11].
ThesesloweventsaremediatedbyreleaseofCa
2+
from
intracellular stores (Charles et al. 1993)[4].The ER is the
major Ca
2+
storage organelle in most cells. ER membrane
Ca
2+
ATPases accumulate Ca
2+
in the ER lumen to quite
high levels. Because the ER lumen contains high concen
trationsofCa
2+
bindingproteins,duetothistotalamount
ofCa
2+
inthelumenmaybe>1mM[7].Theconcentration
of Ca
2+
in the cytoplasm of unstimulated cells is 34 or
ders of magnitude lower than in the ER lumen. This low
concentrationismaintainedbyCa
2+
pumpsandotherCa
2+

transporters located in the ER, as well as plasma, mem

branes.DuetobindingofInsP3,itprovidesapathwayfor
Ca
2+
to diffuse from the ER lumen to cytoplasm. Ca
2+
in
the cytoplasm moves by passive diffusion. The Ca
2+
con
centrationadjacenttotheopenchannelmaybe100Mor
more,whereasconcentrationsascloseas12mfromthe
channel pore may be below 1 M [7],[9],[12]. Thus Ca
2+

signalling appears to provide the most versatile and cru

cialmechanismsforsuchanAstrocytescontribution.
Astrocytespossesaspecificandsignificantroute
forCa
2+
entrywhichiscontrolledbythefillingstateofthe
ER calcium stores. The De YoungKeizer model is the
most general model that incorporates a scheme for the
gatingkineticsofthechannels.Theeffectofcalciumbind
ing to the IP3receptors was ignored on the calcium dy
namics. OthmerTang et. Al. Simplify the De Young
Keizer model which exhibits both excitability and fre
quencyencoding.
Inviewofabovea mathematicalmodelisdevel
opedtostudycytosoliccalciumprofileforAstrocytes.To
derive the calcium dynamics equation we follow the De
YoungKeizermodel.Inproposedmodeltheonlycalcium
exchange taken into account is the one across the endop
lasmicreticulum(ER).Themodelhasbeendevelopedfor
aonedimensionsteadystatecase.Thefinitevolumeme
thod [2],[8] is employed to obtain the solution. Calcium
sequestration by the buffering proteins like calmodulin
and exchanges across the mitochondria are not consi
dered.Also,thereisnocalciumfluxbetweenthecelland
the extra cellular medium, which implies that the total
calcium concentration inside the cell is constant
[5],[13],[14].

2 Mathematical Formulation

Theconservationequationforcalciumprofile[14]
Ca ER Cyt
q q q = + (1)

- Brajesh Kumar Jha is with the S. V. National Institute of Technology,
Surat, India-395007
- Neeru Adlakha. is with the S. V. National Institute of Technology, Surat,
India-395007
- M.N.Mehta is with the S. V. National Institute of Technology, Surat,
India-395007

L
JOURNAL OF COMPUTING, VOLUME 3, ISSUE 11, NOVEMBER 2011, ISSN 2151-9617
HTTPS://SITES.GOOGLE.COM/SITE/JOURNALOFCOMPUTING
WWW.JOURNALOFCOMPUTING.ORG 74

Where
Ca
q ,
ER
q and
Cyt
q represent the number of
moles of calcium inside the cell, the ER and the cytosol.
Dividingbothsidesbythecytosolicvolume
Cyt
V weget
Cyt
Ca ER ER
Cyt ER Cyt Cyt
V
V V V V
q
q q
= + (2)
Theconservationequationisgivenas
2 2
1
[ ]
o i
ER
C C Ca Ca
+ +
( = +

(3)
2
0 1
2
, , [ ]
[ ]
Ca ER ER
ER
Cyt Cyt ER
Cyt
i
Cyt
V
C C Ca
V V V
and Ca
V
q q
q
+
+
= = =
=
(4)

Where
0
C istotalcalciumconcentration(withrespectto
thecytosolicvolume),
1
C isthevolumeratiobetweenthe
ER and the cytosol and
2
[ ]
i
Ca
+
is the cytosolic concen
tration. The three fluxes are incorporated for all the cal
ciumfluxesbetweentheERandcytosolare
leak
J ,
Chan
J
and
Pump
J .

. . .
i
leak ER Chan ER Pump Cyt
dn
J V J V J V
dt
= + (5)

Where the different fluxes are expressed in

1
.sec M

.
The material balance equation in terms of concentration
usingnotation(4)iswrittenas
| |
2
1
[ ]
i
leak chan Pump
d Ca
C J J J
dt
+
= + (6)
This is the general form of the calcium dynamics equa
tion.Hereitisalsoassumedthatthecalciumfluxesoccur
solelybetweentheERandthecytosol.Theoutwardflux
es(
leak
J and
chan
J )areobtainedfromFickslawofdiffu
sion.
( )
( )
2 2
2
2 0
1
2 0
1
1 1
[ ] [ ]
[ ]
[ ]
1 [ ]
1
leak leak ER i
i
leak i
leak
i
J D Ca Ca
C Ca
D Ca
C
D C
C Ca
C C
+ +
+
+
+
=
| |
=
|
\ .
| |
= +
|
+
\ .
(7)
( )
2 2
. . . [ ] [ ]
*
Chan ER ER i
Channels
J P V Ca Ca
unitvolER unittime
+ +
=

(8)
orequivalently,

( )
( )
2 2
2 0
1
1 1
. . [ ] [ ]
1 [ ]
1
Chan Chan ER i
Chan
i
J P D Ca Ca
D C
C Ca
C C
+ +
+
=
| |
= +
|
+
\ .
(9)
Where
leak
D is the constant of diffusivity associated to
theleakand
Chan
D isthechannelconductanceexpressed
insec
1
.Pistheopenchannelprobability(proportionalto
thenumberofchannelsopen).
The inward flux
Pump
J is modelled by a Michaelis
MentenexpressionwithaHillcoefficient2
( )
2
2
max
2 2
2
i
Pump R
M
R
i
Ca
J P
Ca K
+
+
(

=
( +

(10)
Where
max
R
P is the maximal pump rate expressed in
1
.sec M

, and
M
R
K is the MichaelisMenten constant
in M . Finally the general form of the calcium dynam
icsequationusingFickiandiffusioncanbestatedas
( )( )
( )
2 2 2
2 0
1 2
1
2
2
max
2 2
2
[ ] [ ]
1 . . [ ]
1
i i
Ca leak Chan i
i
R
M
R
i
Ca Ca C
D C D PD Ca
t x C
Ca
P
Ca K
+ +
+
+
+
| | c c
= + + +
|
c c +
\ .
(

( +

(11)
for a steady state case the equation (8) is reduced in the
form:
( )( )
2
0
1 2
1
2
max
2 2
1 . .
1
0
Ca leak Chan
R
C d C
D C D P D C
dx C
C
P
C K
| |
+ + +
|
+
\ .
=
+
(12)
Alongwithboundaryconditionsas:
2
0
lim
Ca Ca
x
d Ca
D
dx
o
+

| | (

| =
|
\ .
(13)
2
lim 0.1
x
Ca M
+

( =

(14)
Tohandlethenonlineartermtwocasesareconsidered
CASEI
When K C >>
Then
2 2
2 2 2
C C C
C K K K
< <
+

Thus equation (12) reduced as
JOURNAL OF COMPUTING, VOLUME 3, ISSUE 11, NOVEMBER 2011, ISSN 2151-9617
HTTPS://SITES.GOOGLE.COM/SITE/JOURNALOFCOMPUTING
WWW.JOURNALOFCOMPUTING.ORG 75

( )( )
2
0
1 2
1
max
1 . .
1
0
Ca leak Chan
R
C d C
D C D PD C
dx C
C
P
K
| |
+ + +
|
+
\ .
=
(15)
Inordertoapplythefinitevolumemethodthedomainis
dividedintodiscretecontrolvolumes(Figure1).

Fig.1DomainDiscretization

Taking 20 nodal points in the space between A and B.

Each node is surrounded by a control volume or cell. A
general nodal point is identified by P and its neighbours
inaonedimensionalgeometry,thenodestothewestand
east, are identified by W and E respectively. The west
sides face of the control volume is referred by w and the
eastsidecontrolvolumefacebye.Thedistancesbetween
thenodesWandP,andbetweennodesPandE,areiden
tified by x o . Similarly the distance between face w and
pointPandbetweenPandfaceearedenotedby / 2 x o .
Nodal values to the east and west are available at nodal
values2,3,4............19.Forthesimplicityequation(15)can
bewrittenas
2
1 1 2
0
d C
p C q
dx
+ = (16)
Where
( )( )
( )
max
1 1
0
1
1
1 .
.
R
leak Chan
Ca
leak Chan
Ca
P
p C D P D
D K
C
q D P D
D
| |
= + + +
|
\ .
= +
(17)
The calcium concentration
2
Ca
+
(

is replaced by C for
convenience. Integration of equation (16) over control
volumegives[2],[8]:
2
1 1 2
0
V V V
d C
dV p CdV q dV
dx
A A A
+ =
} } }
(18)
1 1
0
p
e w
dC dC
A A pC A x q A x
dx dx
o o
(
| | | |
+ =
| | (
\ . \ .

(19)
1 1
0
p
e w
dC dC
p C x q x
dx dx
o o
| | | |
+ =
| |
\ . \ .
(20)
1 1
0
P W E P
p
C C C C
p C x q x
x x
o o
o o
| | | |
+ =
| |
\ . \ .
(21)
Thiscanberearrangedas
1 1
1 1 1 1
P W E
p x C C C q x
x x x x
o o
o o o o
(
+ + = + +
(

(22)
The general form for the interior nodal point 2, 3,
4........19isgivenby
P P W W E E u
a C a C a C S = + (23)
1 1
1 1
, , ,
,
W E P W E P
P u
a a a a a S
x x
S p x and S q x
o o
o o
= = = +
= =
(24)
Weapplytheboundaryconditionsatnodepoints1and
20. At node 1 west control volume boundary is kept at
specifiedconcentration
1 1
1
, 0, ,
2 2
,
P W E P W E
P u A
a a a S a a
x
S p x and S C q x
x x
o
o o
o o
= + = =
| | | |
= + = +
| |
\ . \ .
(25)
Similarlyatnode20eastcontrolvolumeboundaryisat
specifiedconcentration.
1 1
1
, , 0,
2 2
,
P W E P W E
P u B
a a a S a a
x
S p x and S C q x
x x
o
o o
o o
= + = =
| | | |
= + = +
| |
\ . \ .
(26)
In equation (23), putting all values of equation (2426)
we have a system of algebraic equations as given below.
Where
A
C and
B
C bethespecifiedboundaryconditions
intermsofcalciumconcentration.
AX B = (27)
Here,
1 2 20
, .................. X c c c = represents the calcium
concentration, A is system matrices and B is the system
vector.

CASEII
When K C <<
Putting K C o = for 0 1 o < <
Then
2
2 2 2
1
1
C
C K o
=
+ +
for 0 1 o < <
Thusequation(12)reducedas
x
x
x x
x
....................................
....................................
1 2
3 4 20
CA CB
JOURNAL OF COMPUTING, VOLUME 3, ISSUE 11, NOVEMBER 2011, ISSN 2151-9617
HTTPS://SITES.GOOGLE.COM/SITE/JOURNALOFCOMPUTING
WWW.JOURNALOFCOMPUTING.ORG 76

( )( )
2
0
1 2
1
max
2
1 . .
1
0
1
Ca leak Chan
R
C d C
D C D PD C
dx C
P
o
| |
+ + +
|
+
\ .
=
+
(28)
For the simplicity equation (28) can be written as
2
2 2 2
0
d C
p C q
dx
+ = (29)
Where
( )( )
( )
2 1
max
2 0 2
1
1 .
1
.
1
leak Chan
Ca
R
leak Chan
Ca
p C D P D
D
P
q C D P D
D o
= + +
| |
= +
|
+
\ .
(30)

Thefinitevolumeschemeisemployedtosolveequations
(29)togetherwith(13)(14).
2
2 2 2
0
V V V
d C
dV p CdV q dV
dx
A A A
+ =
} } }
(31)
2 2
0
p
e w
dC dC
A A p C A x q A x
dx dx
o o
(
| | | |
+ =
| | (
\ . \ .

(32)
2 2
0
p
e w
dC dC
p C x q x
dx dx
o o
| | | |
+ =
| |
\ . \ .
(33)
2 2
0
P W E P
p
C C C C
p C x q x
x x
o o
o o
| | | |
+ =
| |
\ . \ .
(34)
This can be rearranged as
2 2
1 1 1 1
P W E
p x C C C q x
x x x x
o o
o o o o
(
+ + = + +
(

(35)
The general form for the interior nodal point 2, 3,
4........19isgivenby
P P W W E E u
a C a C a C S = + (36)

2 2
1 1
, , ,
,
W E P W E P
P u
a a a a a S
x x
S p x and S q x
o o
o o
= = = +
= =
(37)
We apply the boundary conditions at node points 1and 20. At
node 1 west control volume boundary is kept at specified con-
centration
2 2
1
, 0, ,
2 2
,
P W E P W E
P u A
a a a S a a
x
S p x and S C q x
x x
o
o o
o o
= + = =
| | | |
= + = +
| |
\ . \ .
(38)
Similarly at node 20 east control volume boundary is at spe-
cified concentration.
2 2
1
, , 0,
2 2
,
P W E P W E
P u B
a a a S a a
x
S p x and S C q x
x x
o
o o
o o
= + = =
| | | |
= + = +
| |
\ . \ .
(39)
In equation (29), putting all values of equation (36-39) we
have a system of algebraic equations as given below. Where
A
C and
B
C be the specified boundary conditions in terms of
calcium concentration.
AX B = (40)
Here,
1 2 20
, .................. X c c c = represents the calcium
concentration, A is system matrices and B is the system vector.

3. Results and Discussion:
The numerical results for calcium profile against different
biophysical parameters have been obtained using numerical
values of parameter given in table 1 unless stated along with
figures.
Table I
List of physiological parameters used for numerical results

Thenumericalresultsforcalciumprofileagainstdifferent
biophysicalparametershavebeenobtainedusingnumer
ical values of parameter given in table 1 unless stated
alongwithfigures.

Symbol Parameters Values

1
C 0.185
max
R
P

Maximum
2
Ca
+
uptake
(
1
.sec M

)
0.9
leak
D


2
Ca
+
leak flux constant (sec
-
1
)
0.11
Chan
D Channel conductance (sec
-1
) 6
M
R
K

Dissociation constant of
2
Ca
+
to the pump ( ) M
0.1
D Average
2
Ca
+
concentration
( ) M
1.688
JOURNAL OF COMPUTING, VOLUME 3, ISSUE 11, NOVEMBER 2011, ISSN 2151-9617
HTTPS://SITES.GOOGLE.COM/SITE/JOURNALOFCOMPUTING
WWW.JOURNALOFCOMPUTING.ORG 77

Figure2(a)and2(b)spatialvariationofcalciumconcen
tration
Figure 2 (a) and 2 (b) shows the spatial variation of cal
cium In Case I calcium concentration start from 4.5 M
andgoesdownrapidlyup0.5Mat3.5m.InCaseIcal
cium concentration decrease more rapidly than case II
then after become constant as we move far from the
source.

Figure 3 (a) and 3 (b) shows the spatial variation of cal

cium concentration when number of calcium channel be
tweencytosolandER.Sincethecalciumconcentrationis
morehighthanincytosolsoascalciumchannelremain

Figure3(a)and3(b)spatialvariationofcalciumconcen
trationnumberofchannelvaries

closedthecalciumconcentrationremainlittlehigherthan
the channel become open. Due to different number of
channel open calcium profile become low as more num
berofchannelbecomeopened.

Figure4(a)and4(b)spatialvariationofcalciumconcen
trationfordifferentvalueofDiffusioncoefficient
JOURNAL OF COMPUTING, VOLUME 3, ISSUE 11, NOVEMBER 2011, ISSN 2151-9617
HTTPS://SITES.GOOGLE.COM/SITE/JOURNALOFCOMPUTING
WWW.JOURNALOFCOMPUTING.ORG 78

Figure 4 (a) and 4 (b) shows the spatial variation of cal
ciumconcentrationforthreedifferentvaluesofDiffusion
coefficientDCa=250350m
2
/s.Inbothcasesastheval
ueofdiffusioncoefficientincreasesmorenumbersofcal
cium ions get free, hence the calcium concentration in
creases. Calcium concentration decrease rapidly and fi
nally approaches to 0.1 M as we move away from the
source.

Figure5(a)and5(b)spatialvariationofcalciumconcen
trationfordifferentvalueofinfluxatboundary

Figure 5 (a) and 5 (b) shows the spatial variation of cal
ciumconcentrationforfourdifferentvaluesofinflux.The
four different values of influx are
,2 ,3
Ca Ca Ca
o o o
and
4
Ca
o .Inbothcasesasthevalueofinfluxincreasesmore
numbers of calcium ions get free, hence the calcium con
centration increases. Calcium concentration decrease ra
pidlyupto3mandfinallyapproachesto 0.1 M aswe
moveawayfromthesource.

4.Conclusion
It is observed that ER flux has significant effect calcium
concentrationgivesbettercentralregionslittleawayfrom
the source. The Finite Volume Model is developed here
gives us quite interesting results as such models can be
developed to generate information about relationship
amongphysicalandphysiologicalparameterintheprob
lem and give us better insights and understanding of the
chemicalsignalingphenomenainAstrocytes.

References:

[1] Agulhon C, Fiacco TA, McCarthyKD (2010) Hippocampal
short and longterm plasticity are not modulated by astrocyte
Ca2signaling.Science327:12501254.
[2] B.K. Jha, N. Adlakha, M.N. Mehta , Finite Volume Model to
Study the Effect of Buffer on Cytosolic Ca
2+
Advection Diffu
sionInternationalJournalofEngineeringandNaturalSciences,
WASET,4(3):1601632010
[3] CornellBell AH, Thomas PG, Smith SJ (1990a) The excitatory
neurotransmitter glutamate causes filopodia formation in cul
turedhippocampalastrocytes.Glia3:322334.
[4] CharlesA(1998)Intercellularcalciumwavesinglia.Glia24:
3949.
[5] De Young, G. W., and Keizer, J., A singlepool inositol 1,4,5
triphosphaatereceptorbased model for agoniststimulated os
cillations in Ca
2+
concentration, Proc. Natl, Acad. Sci. USA 89,
(1992),98959899.
[6] FiaccoTA,AgulhonC,TavesSR,PetraviczJ,CasperKB,Dong
X, Chen J, McCarthy KD (2007) Selective stimulation of astro
cytecalciuminsitudoesnotaffectneuronalexcitatorysynaptic
activity.Neuron54:611626.
[7] HajimeHirase,LifenQian,PeterBartho,GyorgyBuzsaki,Cal
cium Dynamics of Cortical Astrocytic Networks In Vivo, PLoS
Biology2004Vol.2(4)494499
[8] H. K. Versteeg, W. Malalasekera, An introduction to computa
tional fluid dynamics the finite volume method, Longman Sci
entific&Technical,1995
[9] J. Kevin Foskett, Carl White, KingHo Cheung, and DonOn
Daniel Mak , Inositol Trisphosphate Receptor Ca2_ Release
Channels,PhysiolRev87:593658,2007
[10] J.W. Deitmer, A.J. Verkhratsky, C. Lohr, Calcium signalling in
glialcells,CdCalcium(1998)24(5/6),405416
[11] Nahoko Kuga,Takuya Sasaki,Yuji Takahara, Norio Matsuki,1
and Yuji Ikegaya LargeScale Calcium Waves Traveling
through Astrocytic Networks In Vivo The Journal of Neurosci
ence,February16,201131(7):26072614
[12] Neher E. Vesicle pools and Ca2_ microdomains: new tools for
understanding their roles in neurotransmitter release. Neuron
20:389399,1998.
[13] Othmer,H.,andTang,Y.,OscillationsandWavesinaModelof
InsP3Controlled Calcium Dynamics. In Experimental and
Theoretical Advances in Biological Pattern formation, Edition,
NewYork:PlenumPress,(1993)pp.277313.
[14] Pauline C. C., Mathematical modeling of calcium dynamics in
astrocytes,thesis,RiceUniversity.2004

Brajesh Kumar Jha is a research scholar in department of Applied

Mathematics and Humanities from Sardar Vallabha Bhai National
Institute of Technology, Surat. He has completed his M.Sc. in Ap
plied Mathematics from SOMAAS, Jiwaji University, Gwalior. He
has number of publication in international journals. His current
JOURNAL OF COMPUTING, VOLUME 3, ISSUE 11, NOVEMBER 2011, ISSN 2151-9617
HTTPS://SITES.GOOGLE.COM/SITE/JOURNALOFCOMPUTING
WWW.JOURNALOFCOMPUTING.ORG 79

research interests are Mathematical Modelling and computational
biology.

Dr.NeeruAdlakhaisworkingasAssociatProfessorinDepartment
ofAppliedMathematicsandHumanitiesfromSardarVallabhaBhai
National Institute of Technology, Surat. She has done Ph.D in
mathematics in area of computational biology from jiwaji univer
sity,Gwalior.Shehaswonnumberofawardsforherresearchwork.
SheisceretaryofspecialgroupofintrestofBioinformatics(CSI).She
isalsomemberofworkgrouponBioinformaticsandTC5ofIFIP.

Dr. M. N. Mehta is Professor in Mathematics, Deptt. Of Applied

Mathematics and Humanities, SVNIT, Surat, Gujarat (INDIA). For
mer Head of Applied Science and Humanity Deptt. SVNIT, Surat.
HedidB.Sc.,M.Sc.,andPh.D.fromSouthGujaratUniversity,Surat.
Hehaspublished70Researchpapers,organizedmanyNationaland
International Conferences, Seminars and workshops. Teaching
UG/PG courses since 36 years and wrote some books for UG stu
dents. He supervised 14 Ph. D. and some M.Phill candidate. His
researchinterestisFluidflowthroughPorousMedia.

JOURNAL OF COMPUTING, VOLUME 3, ISSUE 11, NOVEMBER 2011, ISSN 2151-9617
HTTPS://SITES.GOOGLE.COM/SITE/JOURNALOFCOMPUTING
WWW.JOURNALOFCOMPUTING.ORG 80