1

Stromal cell-Derived Factor-1α directly modulates voltage-dependent

currents of the action potential in mammalian neuronal cells

Authors: A. Guyon, C. Rovère, A. Cervantes, I. Allaeys, & J.L. Nahon

Institut de Pharmacologie Moléculaire et Cellulaire (IPMC)- UMR 6097 CNRS, 660

Route des Lucioles, Sophia Antipolis, 06560 Valbonne, France.

Telephone : 04 93 95 77 50

Fax : 04 93 95 77 08

Correspondence should be addressed to Jean-Louis Nahon (nahonjl@ipmc.cnrs.fr)

List of abbreviations:

AP: action potential

CNS: Central nervous system

DMEM: Dulbecco Modified Eagle’s Medium

FCS: fetal calf serum

PBS: phosphate buffer saline

PCR: polymerase-chain reaction

PFA: paraformaldehyde

RT: reverse transcription

SDF-1α: Stromal derived factor- 1α/CXCL12

TEA: tetra ethyl ammonium

TTX: tetrodotoxin

2

Abstract

Stromal cell -Derived Factor 1α (SDF-1α) is a chemokine whose receptor,

CXCR4, is distributed in specific brain areas including hypothalamus. SDF-1α has

recently been found to play important roles on neurons, although a direct modulation

of voltage-gated ionic channels has never been shown. In order to clarify this issue, we

performed patch-clamp experiments first in fetal mouse hypothalamic neurons in

culture. SDF-1α (10 nM) decreased the peak and rising slope of action potentials and

spike discharge frequency in 22 % of hypothalamic neurons tested. This effect was

blocked by the CXCR4 antagonist AMD 3100 (1 µM) but not by the metabotropic

glutamate receptor antagonist MCPG (500 µM), indicating a direct action of SDF-1α

on its cognate receptor. This effect involved a depression of both inward and outward

voltage-dependent currents of the AP. We confirmed these effects in the human

neuroblastoma cell line SH-SY-5Y, which expresses endogenously CXCR4. Voltage-

clamp experiments revealed that SDF-1α induced a 20% decrease of the peak of

tetrodotoxin-sensitive sodium current and tetraethylammonium-sensitive delayed

rectifier potassium current, respectively. Both effects were concentration-dependant,

and blocked by AMD3100 (200 nM). This dual effect was reduced or blocked by 0.4

mM GTPγS G-protein pre-activation or by G-protein inhibitor Pertussis toxin (200

ng.ml-1) pre-treatement suggesting that it is mediated through activation of a Gi/o-

protein. This study extends the functions of SDF-1α to a direct modulation of voltage-

dependent membrane currents of neuronal cells.

Keywords: Stromal cell -Derived Factor 1α, SH-SY-5Y neuroblastoma cell line,

TTX-sensitive sodium current, delayed rectifier potassium current, hypothalamic

neuron, patch-clamp.

Running title: SDF-1α modulates action potential

3

Introduction

Chemokines are secreted small molecules (8-13 kDa) with chemoattractant

properties whose main accepted role is leukocyte recruitment at inflammatory sites

both at the periphery and in the central nervous system (CNS) (Rossi and Zlotnik

2000; Lazarini et al. 2003). It is well documented that inflammatory processes, either

peripheral or central, profoundly affect the neuroendocrine systems (Besedovsky and

del Rey 1996; Chikanza and Grossman 2000). It might therefore be conceivable that,

among the vast repertoire of chemokines that are mobilized in the course of

inflammation, some participate in the modulation of neuroendocrine activity. Indeed,

in the last few years, emerging evidence accumulated that expand the functional role

of CNS chemokines from biological mediators of inflammatory responses to

molecules playing a developmental and neuromodulatory role (Rossi and Zlotnik

2000; Bajetto et al. 2002).

Among chemokines, Stromal Cell-Derived Factor 1α (SDF-1α) displays many

crucial effects in the brain as migration, proliferation, survival, guidance and

chemoattraction (see review in Lazarini et al. 2003). It has also been suggested that

SDF-1α can have a neuromodulatory role in neurons. It induces intracellular calcium

transients in hippocampal, cortical, cerebellar, human embryonic and dorsal root

ganglion neurons (Bajetto et al. 1999; Limatola et al. 2000; Oh et al. 2002; Liu et al.

2003). In Purkinje cells, SDF-1α induced a slow inward current followed by an

increase of both intracellular Ca2+ level and spontaneous synaptic activity (Limatola et

al. 2000). However, these effects are largely indirect as they appear to be mediated by

SDF-1α-induced glutamate release from astrocytes (Limatola et al. 2000) and, to date,

there is no evidence of an action of CXCR4 stimulation on neuronal voltage-

4

dependent ionic channels. Furthermore, CXCR4 expression has been recently found in

neuronal populations of mouse and rat lateral hypothalamus areas, suggesting that

SDF-1α may modulate the activity of these neurons (Banisadr et al. 2002).

Voltage-dependent sodium and potassium channel are one potential target of

SDF-1α that have wide-ranging effects on activity. We first investigated the putative

effects of SDF-1α on neuronal excitability using mouse hypothalamic neurons in

culture. We then investigated the effects of SDF-1α on a suitable model cell

presenting a high level of expression of the CXCR4 and expressing voltage-gated

currents of the action potential. The human neuroblastoma cell line SH-SY5Y,

established by repetitive subcloning of the SK-N-SH cell line (Biedler et al., 1978),

exhibits the morphological and biochemical features of cells derived from the neural

crest, like sympathetic neurons (Barnes et al., 1981; Ross et al., 1981). These cells

have been previously shown to present spikes or spikelets which are due to the

presence of voltage-activated sodium and potassium currents. SH-SY5Y are thus a

suitable model to study the modulation of excitability of cells of human origin and the

use of this cell line allows to combine molecular biological and electrophysiological

approaches.

Here we show that SDF-1α depresses the action potential discharge of mouse

hypothalamic neurons in culture. Furthermore, we show that SH-SY5Y cells express

the CXCR4 receptor of SDF-1α, and that the stimulation of this receptor by SDF-1α

modulates the excitability of the cell through a depressant action on both sodium and

potassium voltage-activated currents. These results suggest that direct modulation of

action potential may occur in neurons involved in neuroendocrine functions.

5

Materials and Methods

Cell Culture

Human SH-SY5Y cells (kind gift of Dr. E. Bezard, Basal Ganglia Laboratory,

Bordeaux, France) were grown and propagated in DMEM (Invitrogen) containing 10

% heat-inactivated fetal calf serum (FCS, BioWhittaker) and 50 u/ml penicillin and 50

µg/ml streptomycin (Invitrogen). Cells were recorded 2-5 days after plating.

Neuronal cells were prepared from hypothalamus of 15-day-old mouse embryos

(C57/BL6 mice). We used male C57/BL6 adult mice which were bred in the local

animal facilities and maintained on a 12 h dark/light cycle (7 am/7pm) with food and

water ad libitum. All the protocols were carried out in accordance with French

standard ethical guidelines for laboratory animals (Agreement N° 75-178,

05/16/2000). Briefly, cells were mechanically dissociated with a pipette in a

Neurobasal medium (Invitrogen) supplemented with 50 U/ml penicillin and 50 µg/ml

streptomycin. Dissociated cells were then plated at a density of 2 106 cells in 35 mm

(or 1.5 105 cells in 24 mm) plastic tissue-culture dishes precoated for 1-2 hr with

polylysine (10 µg/ml) (Sigma) and grown in a humidified atmosphere of 5% CO2,

95% air. After 30 min incubation, the medium was replaced and B27 supplement

(Invitrogen) was added. Neurons were used for recording 6-8 days after plating.

RNA extraction, reverse-transcription-polymerase chain reaction (RT-PCR)

Total RNA was extracted from the human SH-SY5Y cells or from the hypothalamic

neurons in primary culture (C57/BL6 mice) using RNable method (Eurobio), as

described by manufacturer. The quality of the RNA was monitored by optical density

(260 nm/280 nm ratio) and by ethidium bromide staining of 1% agarose gels after

6

electrophoresis. The reverse transcription (RT) was performed with Superscript II

reverse transcriptase (Invitrogen) with 1 µg RNA of rat hypothalamus or SH-SY5Y

cells and 15-17 mer oligo-d(T) primers. The standard PCR was performed with

HotMaster Taq DNA polymerase (Eppendorf) with human CXCR4 20F, 220R primers

for SH-SY5Y cells or with mouse CXCR4 154F, 302R (see Table 1). The PCR

products were separated through a 1% agarose gel and visualized under U.V. after

staining with ethidium bromide. The PCR fragments of 200 bp and 148 bp (expected

size for the hCXCR4 and mCXCR4 mRNA-derived PCR product, respectively) were

obtained.

Immunohistochemical studies

SH-SY5Y cells or hypothalamic neurons were grown on polylysine coated-coverslips.

Cells were rinsed twice in phosphate buffer saline (PBS) and were fixed for 10 min in

4% paraformaldehyde (PFA) at room temperature. Then, they were permeabilized in

PBS containing 3% horse serum and 0.1% Saponine for 60 min. Cells were incubated

overnight at 4°C with primary antibody diluted in PBS containing 3% horse serum and

0.01% saponine. The CXCR4 antiserum (Santa Cruz Biotechnology), raised in goat,

was used at a dilution of 1:100. Labeling was revealed by incubating cells for 60 min

at room temperature with Alexa Fluor-conjugated rabbit anti goat IgGs (Molecular

Probes). After washes with PBS, coverslips were mounted with Mowiol reagent (25%

Mowiol, 10% glycerol, 0.2M Tris) and dried overnight. Labeled cells were examined

with a Leica TSC SP confocal imaging spectrophotometer equipped with an

argon/krypton laser. Signal was imaged by exciting samples at 488 nm.

7

Whole-cell patch-clamp recordings

Culture dishes were placed under an inverted microscope (Leitz) and superfused at a

flow rate of 1 ml.min-1 with a solution containing (mM) : 150 NaCl, 2,5 KCl, 2 CaCl2,

1 MgCl2, 20 Glucose, 10 Hepes supplemented with 0,05% Bovine Serum Albumine

(BSA), pH 7,4. Recordings were made at room temperature (25 ± 2°C) using a RK-

300 amplifier (Biologic). Patch clamp pipettes were made from hematocrite or

borosilicate glass capillary. They had a resistance of 2-5 MΩ when filled with the

internal solution containing (mM) 120 KCl, 5 MgCl2, 1 CaCl2, 10 EGTA, 4 Na ATP,

0,4 Na GTP, 10 HEPES supplemented with 15 Phosphocreatine and 50 u/ml creatine

phosphokinase (pH adjusted 7.3 with KOH). In one set of experiments, K+ was

replaced by Cs+ in the internal solution and in another set of experiment, EGTA

concentration was lowered to 0.5 mM while KCl was raised to 130 mM. Values of

access resistance ranged from 5 to 10 MΩ at the beginning of the recording to 10-15

MΩ at the end and were compensated at 60-80 %. The value of the access resistance

was checked repeatedly during each experiment. Whole-cell currents were discarded

when the voltage-clamp error (calculated as the uncompensated access resistance x the

recorded current) was higher than 5 mV. Linear components of leak and capacitive

currents were first reduced by analog circuitry and then almost completely cancelled

with the P/N method.

Data analysis

Data were digitized at 5-10 kHz using a Digidata interface coupled to a

microcomputer running pClamp 9 (Axon Instruments). Currents were filtered at 3

KHz. Current/Voltage curves were fitted using a Bolztman equation : I/Imax =

8

1/1+exp ((V-Vo)/k) using SigmaPlot software (Jandel). The slope of the spike

repolarizing phase was estimated as the slope of the asymptote.

Average data are expressed as mean ± standard deviation (s.d.) or mean ±

standard error of mean (s.e.m.) when mentioned (n = number of neurons). Statistical

significance between groups was tested using either the Student's t-test or the ANOVA

followed by Fisher test, and were considered significant at P < 0.05. Statistical

analysis were done using SigmaPlot (Jandel) and Statview softwares.

Drugs

The following drugs were used : TTX (Tocris), TEA and CdCl2 (Sigma), and

(RS)-MCPG (Tocris). The bicyclam AMD 3100 was synthesized by Orga-Link (Gif-

sur-Yvette, France) and kindly provided by P. Kitabgi and W. Rostène (INSERM

E0350 – UPMC, Hôpital Saint-Antoine, Paris Cedex 20, France). Human SDF-

1α (Tebu) was used for experiments with SH-SY5Y cells whereas rat SDF1α (Tebu)

was used for experiments with rat hypothalamus. Frozen stock solutions (in water

except for MCPG, prepared in 1M NaOH) were diluted to their final concentration the

day of the experiment and pH was adjusted. Changes of extracellular solution were

obtained by a fast multi-barrel delivery system positioned close to the cell tested.

GTPγS is a non-hydrolysable GTP analogue that irreversibly activates G-

proteins. It has been demonstrated that inclusion of GTPγS in the recording pipette

solution occludes G-protein-mediated inhibition of different voltage-dependent

currents by various neurotransmitters (Catterall 1997; Dascal 2001). In our

experiment, 0.4 mM GTPγS (Sigma) was used to replace the 0.4 mM GTP in the

pipette solution. Once the whole-cell configuration was established, GTPγS diffused

into the recorded cell and activated G-proteins, leading to an inhibition of voltage-

9

dependent potassium and sodium current within 10 min (not shown). It has been

demonstrated that Pertussis toxin (PTX) blocks Gi/o-protein mediated inhibition of

voltage-dependent currents. To determine whether the SDF-1α-mediated effects on

potassium currents were sensitive to PTX, we first treated the cultures with 200 ng.ml-

1 PTX (Calbiochem) for 24 h, and then tested the effects of 10 nM SDF-1α on

potassium and sodium currents.

10

Results

SDF-1α modulates the firing of mouse hypothalamic cultured neurons

We used a culture of mouse hypothalamic neurons to study the effects of SDF-

1α (Fig. 1a). RT-PCR experiments showed that CXCR4 mRNAs are expressed in cells

of this preparation (Fig 1b). Immunocytochemistry experiments showed that a high

proportion of neurons expresses CXCR4, mainly at their plasma membrane (Fig 1c).

Hypothalamic neurons in culture were recorded in the whole cell configuration

of the patch-clamp technique. They had a membrane resistance and a capacitance of

respectively 579 ± 630 MΩ and 13.9 ± 7,7 pF (n = 46). Their mean membrane

potential was –39.7 ± 15.0 mV (n = 36). Continuous current was injected when

necessary to maintain the potential of the neurons around –60 mV before applying step

depolarizing current. In 22% of the neurons tested (n = 8 out of 36), SDF-1α (10 nM)

modified the pattern of discharge of the neurons recorded in current clamp, without

any significant change in their resting membrane potential (Fig 1d) whereas it had no

effect in the remaining tested neurons (not shown). SDF-1α effect was a 10.2 ± 5 %

slowing-down of the mean frequency of action potential (AP) discharge in response to

a depolarizing current pulse (from 48.5 ± 8.7 Hz to 43,7 ± 9.1 Hz, paired t test, p <

0.001, n = 8) as illustrated in Figure 1d-e. This effect was accompanied with a slight

but significant decrease of the maximum rising slope of the APs evoked during the

depolarizing pulse (of respectively 6.83 ± 5.65 %, p < 0.02; 10.18 ± 7.56 %, p < 0.05

and 15.79 ± 9.13, p < 0.02 for 1st, 2nd and 3rd AP, paired t test, n = 8) and of the peak

amplitude of the APs (of respectively 3.08 ± 3.04 %, p< 0,05; 7,29 ± 4.19 %, p< 0.01,

and 7.17 ± 4.04, p<0.01, for the 1st, 2nd and 3rd AP, paired t test, n = 8). In addition,

11

SDF-1a induced a decrease in the rate of decay time of the AP (of respectively 4.69 ±

2.73 % and 14.6 ± 9.42.%, for the 1st and 2nd AP, paired t test p<0.01, n = 8).

These effects reversed after washout of SDF-1α and were subjected to

desensitization when repeated applications of SDF-1α were performed (not shown). In

addition, these effects were blocked by the pre-application of the specific CXCR4

antagonist AMD 3100 (1µM) (Fig 2a, n = 3).

It has been shown previously (Limatola et al. 2000) that following CXCR4

stimulation by SDF-1α astrocytes can release glutamate which can then stimulate

metabotropic Gi/o coupled receptors present at the neuronal plasma membrane.

Astrocytes are present in our preparation of hypothalamic neurons in culture. In order

to determine whether the effect of SDF-1α occurs via the stimulation of CXCR4

present on neurons or indirectly through astrocyte glutamate release via neuronal

mGluR, we performed experiments with the mGluR antagonist MCPG. In 500 µM

MCPG, SDF-1α induced a significant slowing down of the spike frequency from 56 ±

16 Hz to 47 ± 13 Hz, paired t test, p < 0.01, n = 8) as illustrated in Figure 2b. This

represents a 15.8 ± 9.44 % decrease, which was not significantly different from what

was observed in control under SDF-1α treatment only. Therefore, MCPG did not

prevent the effect of SDF-1α on the firing pattern of the hypothalamic neurons, ruling

out the hypothesis of an effect through astrocyte glutamate release on neuronal

mGluR.

The effect of SDF-1α observed in current clamp could involve intracellular

calcium elevations as it has been described in other cells (Bajetto et al. 1999; Limatola

et al. 2000; Oh et al. 2002; Liu et al. 2003) and the activation of a calcium-induced

potassium current. However, this was unlikely because we used 10 mM of the calcium

buffer EGTA in the internal solution. Moreover, in experiments performed with an

12

internal solution containing only 0.5 mM EGTA, 10 nM SDF1α induced a decrease in

frequency of 13.4 ± 3.29 % (n = 3), not significantly different from what was observed

in 10 mM EGTA (Fig 2c). Alternatively, the effect of SDF-1α observed in current

clamp could involve a depression of the sodium current (which would explain the

depression of the rate of rise time and peak of the AP) and a depression of the delayed

rectifier potassium current (which would explain the slowing in the frequency of AP

discharge and the decreased rate of decay time of the AP), although reducing the

sodium current could also reduce the potassium current due to decreased activation. In

order to test these hypothesis we recorded hypothalamic neurons in voltage clamp. At

a holding potential of -60 mV, application of 10 nM SDF-1α did not modified the

basal current and did not induce any change in the membrane conductance (n = 24, not

shown). In 23% of neurons tested, SDF-1α decreased by 7.8 ± 7.3% (paired t test P <

0.05, n = 6) the peak of the total outward currents induced by a depolarizing pulse

from -60 to + 50 mV (Fig 2d). Although the effects on sodium currents were more

difficult to study because of space-clamp problems, SDF-1α depressed the peak of the

inward sodium current by – 15.0 ± 8.7% (n = 3, fig 2e) in 11.5% of neurons.

The human neuroblastoma cell line SH-SY5Y is a good model to study the

direct effects of SDF-1α on voltage-gated currents of the AP

The human neuroblastoma cell line SH-SY5Y, established by repetitive

subcloning of the SK-N-SH cell line (Biedler et al. 1978), exhibits the morphological

and biochemical features of cells derived from the neural crest, like sympathetic

neurons (Barnes et al. 1981; Ross et al. 1981). SH-SY5Y cells appeared to be a good

model to study the effects of SDF-1α on voltage-gated currents of neuronal cells of

human origin, for the following reasons: (1) they present a few short processes (Fig

13

3a), thus they can be easily voltage-clamped, (2) they present a high level of CXCR4

expression as shown by RT-PCR (Fig 3b) and immunocytochemistry experiments (Fig

3c), (3) they present a delayed rectifier potassium current whose properties fulfilled

those of the Kv3.1 (Tosetti et al. 1998; Friederich et al. 2001; Friederich et al. 2003)

(Fig 4) and voltage dependent sodium currents, of as yet unknown molecular type

(Toselli et al. 1996) (Fig 5).

Thus, SH-SY5Y cells present the properties required to study the effects of

activation of CXCR4 by its ligand SDF-1α on voltage-gated currents of the AP.

SDF-1α modulates the AP in SH-SY5Y cells

We first performed current-clamp experiments in order to test the effects of

SDF-1α in SH-SY5Y cells. When using the potassium internal solution, their

membrane potential was -51,3 ± 21,0 mV (n = 8). In response to a stimulation by a

rectangular depolarizing current pulse, SH-SY5Y cells generated spikes or spikelets

as previously described (Toselli et al. 1996; Bajetto et al. 1999).

In every cell tested presenting an AP (n = 8), SDF-1α induced a slight but

significant decrease of the slope of the repolarizing phase (Fig 3d and h ) as would be

expected by a decrease in potassium delayed rectifier current. As shown in Fig 3e, a

moderate block of the potassium current by TEA induces a slowing of the slope of the

spike decay, and induces an increase in the peak of the spike. SDF-1α induces a slight

but significant decrease in the maximum rising slope (Fig 3g) and in the peak of the

spike (Fig 3f) as expected by the simultaneous depressant effect on the voltage-

dependent sodium current.

The effect of SDF-1α observed in current clamp suggests that both a

depression of the delayed rectifier potassium current (which would explain the

14

slowing in the frequency of AP discharge) and a depression of the sodium current

(which would explain the depression of the peak and maximum rising slope of the AP)

could be involved. In order to test this hypothesis we recorded the SH-SY5Y cells in

voltage clamp.

SDF-1α decreases potassium currents in SH-SY5Y cells

We first investigated the effects of SDF-1α on the voltage-activated potassium

delayed rectifier current recorded in the presence of TTX (3 µM). In 93% of the cells

tested (n = 14), 10 nM SDF-1α reversibly depressed this current (Fig 4a and b)

without affecting the activation/inactivation properties of the current (Fig 4c and d).

The magnitude of this effect was 20.6 ± 3,5 % (mean ± s.e.m, paired t test P < 0.001, n

= 14, Fig 4e). As illustrated in Fig 4e, the effect of SDF-1α was concentration-

dependent. At concentrations of respectively 0.1 nM and 1 nM, SDF-1α induced a

response in 29% and 44 % of the cells tested (n = 7 and 25, respectively) and the

current decrease was reduced to respectively (mean ± s.e.m) 0.14 ± 2.3 % (n = 7) and

7.2 ± 2.1% (n = 25). The inhibitory effect induced by 10 nM SDF-1α was significantly

blocked by the bicyclam AMD 3100 (200 nM, n = 9 cells, Fig 4e), which had no effect

per se (not shown), suggesting that it was indeed mediated through the activation of

CXCR4. Conversely, SDF-1α did not affect the current recorded with the internal Cs+

solution (Fig 4f, n.s., n = 21).

SDF-1α decreases sodium currents in SH-SY5Y cells

In 11 out of 17 cell tested (65%), 10 nM SDF-1α reversibly depressed the

sodium current recorded in TEA (Fig 5a and b) without affecting its

activation/inactivation properties (Fig 5c). The magnitude of this effect was 21.9 ± 2,3

15

% (mean ± s.e.m, paired t test P < 0.001, n = 11). The effect was concentration-

dependent (Fig 5d): in 0.1 nM and 1 nM SDF1α, the percentage decrease of the

current was reduced to respectively (mean ± s.e.m) 12.4 ± 2.9% (n = 6) and 18.9 ±

2.0%, (n = 6). The effect induced by 10 nM SDF-1α was significantly blocked by 200

nM AMD (n = 7), suggesting that it is mediated through CXCR4 stimulation (Fig 5d).

Involvement of G-proteins in SDF-1α-induced inhibition

We speculated that the depressive action of SDF-1α on potassium and sodium

currents in SH-SY5Y cells was mediated by Gi/o-proteins as found in many cellular

systems (Tanabe et al. 1997; Bajetto et al. 1999; Klein et al. 1999; Zheng et al. 1999;

Lazarini et al. 2000). First, we irreversibly activated G-proteins with GTPγS (Catterall

1997; Dascal 2001) and then tested whether SDF-1α could induce further inhibition of

potassium and sodium currents. We used also an inhibitor of Gi/o-proteins, PTX

(Tanabe et al. 1997; Bajetto et al. 1999; Klein et al. 1999; Reisine 1990; Rosenthal et

al. 1990), to test the effects of SDF-1α.

In the presence of GTPγS, SDF-1α (10 nM) did decrease the potassium or

sodium current (respectively n = 9 and n = 6, paired t test n.s., Fig 4g and 5e),

whereas in control cells recorded in the same preparation, SDF-1α induced a

significant depression of both the potassium (8.1 ± 1.7 %, mean ± s.e.m, n = 11, paired

t test P < 0.001) and the sodium currents (25.4 ± 6.9 %, mean ± s.e.m , n = 9, paired t

test P < 0.01). Therefore, there was a significant block by GTPγS of the SDF-1α-

induced inhibition of both the potassium and the sodium current (P < 0.001 and P <

0.01, respectively). In cells pretreated with PTX, SDF-1α (10 nM) had no significant

effect on the potassium (n = 10, Fig 4g) and sodium currents (n = 5, Fig 5e), in

contrast to control cells. This represents a significant block by the PTX treatment of

16

the SDF-1α−induced depression of both the sodium and potassium currents (P < 0.02

and P < 0.05, respectively).

17

Discussion

SDF-1α controls the activation and directed migration of leucocytes, and

participates in the inflammatory processes in host defense (Lazarini et al., 2003). In

the brain SDF-1α activates the CXCR4 receptor expressed in a variety of neural cells

and this signaling results in diverse biological response. However, this effect has been

shown to be indirect through a glutamatergic mechanism, being inhibited by an

antagonist of metabotropic glutamate receptor (Limatola et al. 2000). Here we provide

the first electrophysiological characterization of the effects of SDF-1α on voltage-

operated currents in the human neuroblastoma SH-SY5Y cell line and mouse

hypothalamic neurons, adding another potent way for this chemokine to exert a

neuromodulatory role in the brain.

Our study shows that CXCR4 stimulation by SDF-1α negatively modulates the

action potential discharge in both preparations. Our experiments were performed in the

presence of 10 mM of the calcium buffer EGTA in the internal solution. Additional

experiments performed in the presence of 0.5 mM EGTA in the intra-pipette recording

solution did not reveal further difference. This suggest a weak involvement of

intracellular calcium elevations and activation of calcium dependant potassium

currents in the SDF-1α effect observed in current-clamp experiments. Conversely,

SDF-1α depressed both outward and inward voltage-dependent currents in both

preparations. These effects are likely to be involved in the negative modulation of

action potential discharge observed in current clamp.

In SH-SY5Y cells, the outward current is carried by potassium ions, on account

of its block by internal cesium and external TEA. Its features are those of the "delayed

rectifier" potassium current already described in these cells and shown to be

18

responsible for the repolarizing phase of spikes (Tosetti et al. 1998). A decrease of the

amplitude of this current induces a slowing of repolarization of the spike, which was

indeed observed in current clamp experiments. Fast inward currents were fully

inactivating and TTX-sensitive, suggesting they were voltage-dependent sodium

currents, already described in SH-SY5Y cells (Toselli et al. 1996). These currents

were also depressed by SDF-1α explaining why the peak of the spike is decreased by

SDF-1α instead of increased as would be expected by the sole depression of potassium

currents. Their inactivation curve could not be well fitted by a single Boltzman curve,

suggesting that the sodium currents were heterogeneous in SH-SY5Y cells. An

heterogeneity of sodium currents has already been reported in these cells between

undifferentiated cells and cells differentiated by retinoic acid in vitro (Toselli et al.

1996). In our preparation, although we did not induce differentiation, it appears that

SH-SY5Y cells spontaneously start to differentiate after 2 days plating, as suggested

by their morphology and the faster rising phase of their spikes. Therefore, this

heterogeneity of sodium currents could be correlated to the presence of the two types

of sodium channels in these cells during differentiation.

In SH-SY5Y cells, modulatory effects on both sodium and potassium currents

occurred through the direct stimulation of CXCR4 as they were blocked by the

selective antagonist of this receptor, the bicyclam AMD 3100. In culture neurons,

where astrocytes are present which could release glutamate in response to SDF-

1α (Limatola et al. 2000), the solution containing SDF-1α was puffed close to the cell

body. Moreover, the kinetics of the effect was similar to the one observed in SH-

SY5Y cells suggesting that the effects observed in this preparation were also mediated

through a direct stimulation of CXCR4. Indeed, the effects were blocked by the

selective CXCR4 antagonist AMD 31000 and persisted in the presence of an

19

antagonist of mGluR receptors (MCPG, 500 µM). Therefore, the effects of SDF-1α

which we observe on hypothalamic neurons are likely to occur through the direct

stimulation of the CXCR4 present at the neuronal plasma membrane.

In term of intracellular signaling, the modulatory effects of SDF-1α in SH-

SY5Y cells involved a G protein of the Gi/o family as demonstrated by their

sensitivity to PTX. CXCR4 stimulation has already been shown to induce a PTX-

sensitive G protein activation in neurons (Bajetto et al. 1999; Zheng et al. 1999).

Moreover, potassium delayed rectifier currents have been shown to be modulated

through the activation of a G-protein (ffrench-Mullen et al. 1994; Xu et al. 2000; Huan

et al. 2001). Concerning the voltage-gated sodium channel, although early studies led

to the idea that they were not subject to modulation, a G-protein-dependent

modulation of brain sodium voltage-gated channels has now been described (Ma et al.

1994; Carr et al. 2002). The intracellular regulatory cascades activated by SDF-1α in

neuronal cells are presently unknown. To date, activation of CXCR4 has been shown

to induce inhibition of adenylate cyclase (Zheng et al. 1999), intracellular Ca2+

transients (see review in Ragozzino 2002) and activation of the mitogen-activated

protein Kinase, ERK1/2 (Lazarini et al. 2000). It will be of interest to determine which

pathways are involved in the modulation of the two currents by the stimulation of

CXCR4 receptor, taking advantage of the use of the neuroblastoma cell line

preparation.

In heterologous expression systems, it has been shown that CXCR4 activation

can decrease voltage-activated calcium currents (Oh et al. 2002). Here we show in two

neuronal models expressing endogeneously both voltage-dependent currents and

CXCR4, that SDF1α simultaneously modulates the two currents which are mainly

involved in the action potential. Since these currents are particularly implicated in the

20

control of input-output relationships between neurons, it is tempting to suggest that

SDF-1α may be involved in neuromodulation. There is growing evidence that voltage-

dependent sodium channels regulate not only somatic spike generation but also

dendritic integration of synaptic input (Schwindt and Crill 1995; Stuart and Sakmann

1995; Gonzalez-Burgos and Barrionuevo 2001). Therefore, SDF-1α inhibition of

neuronal excitability could result in a decrease of neurotransmitter release and a

reduction of back-propagating spikes in dendrites (Magee and Johnston 1997).

In conclusion, the similarity between the CXCR4-mediated effects of SDF-1α

in SH-SY5Y cells and in hypothalamic neurons, both on the shape of the AP and on

the kinetics of the response to the chemokine, indicates that the SH-SY5Y cell line

represents a relevant model to study CXCR4 -mediated effects in neurons. It further

suggests that, as in neuroblastoma cells, SDF-1α acted on hypothalamic neurons via

CXCR4 to decrease potassium and sodium currents of the AP. It will be of interest

now to determine which hypothalamic neurons are modulated by SDF-1α.

21

Acknowledgements

We are grateful to Drs. Stéphane Melik-Parsadanian, Patrick Kitabgi and Dr. William

Rostène for proving CXCR4 antagonist AMD 3100 and for discussion of our data. We

warmly thank Drs. Jean Chemin, Fabrice Duprat, Eric Honoré and Georges Romey

from Prof. Michel Lazdunski's laboratory (IPMC, Valbonne, France) for technical

support and helpful advices. We thank also Drs. Bruno Cardinaud for helpful

discussions, Christine Ortola and Natacha Grand for excellent assistance in cell

culture, Franck Aguila and Jacqueline Kervella for contribution in artwork and

manuscript assistance. We are grateful to the Zeiss Society for their support in taking

the photomicrographs. This work was supported by the « association pour la

Recherche sur le Cancer » (ARC, subvention n° 3375-2003-2004), and in part by the

Centre National de la Recherche Scientifique (CNRS) (programme OHLL, 2003). A.

Guyon is supported by a “délégation” position from the CNRS/french Education

Minister (University of Pierre et Marie Curie, Paris).

22

References

Bajetto A., Bonavia R., Barbero S. and Schettini G. (2002) Characterization of

chemokines and their receptors in the central nervous system: physiopathological

implications. J Neurochem 82, 1311-1329.

Bajetto A., Bonavia R., Barbero S., Piccioli P., Costa A., Florio T. and Schettini

G. (1999) Glial and neuronal cells express functional chemokine receptor CXCR4 and

its natural ligand stromal cell-derived factor 1. J Neurochem 73, 2348-2357.

Banisadr G., Fontanges P., Haour F., Kitabgi P., Rostene W. and Melik

Parsadaniantz S. (2002) Neuroanatomical distribution of CXCR4 in adult rat brain and

its localization in cholinergic and dopaminergic neurons. Eur J Neurosci 16, 1661-

1671.

Barnes E. N., Biedler J. L., Spengler B. A. and Lyser K. M. (1981) The fine

structure of continuous human neuroblastoma lines SK-N-SH, SK-N-BE(2), and SK-

N-MC. In Vitro 17, 619-631.

Besedovsky H. O. and del Rey A. (1996) Immune-neuro-endocrine interactions:

facts and hypotheses. Endocr Rev 17, 64-102.

Biedler J. L., Roffler-Tarlov S., Schachner M. and Freedman L. S. (1978)

Multiple neurotransmitter synthesis by human neuroblastoma cell lines and clones.

Cancer Res 38, 3751-3757.

Carr D. B., Cooper D. C., Ulrich S. L., Spruston N. and Surmeier D. J. (2002)

Serotonin receptor activation inhibits sodium current and dendritic excitability in

prefrontal cortex via a protein kinase C-dependent mechanism. J Neurosci 22, 6846-

6855.

Catterall W. A. (1997) Modulation of sodium and calcium channels by protein

phosphorylation and G proteins. Adv Second Messenger Phosphoprotein Res 31, 159-

181.

Chikanza I. C. and Grossman A. B. (2000) Reciprocal interactions between the

neuroendocrine and immune systems during inflammation. Rheum Dis Clin North Am

26, 693-711.

Dascal N. (2001) Ion-channel regulation by G proteins. Trends Endocrinol Metab

12, 391-398.

23

ffrench-Mullen J. M., Plata-Salaman C. R., Buckley N. J. and Danks P. (1994)

Muscarine modulation by a G-protein alpha-subunit of delayed rectifier K+ current in

rat ventromedial hypothalamic neurones. J Physiol 474, 21-26.

Friederich P., Benzenberg D., Trellakis S. and Urban B. W. (2001) Interaction of

volatile anesthetics with human Kv channels in relation to clinical concentrations.

Anesthesiology 95, 954-958.

Friederich P., Dilger J. P., Isbrandt D., Sauter K., Pongs O. and Urban B. W.

(2003) Biophysical properties of Kv3.1 channels in SH-SY5Y human neuroblastoma

cells. Receptors Channels 9, 387-396.

Gonzalez-Burgos G. and Barrionuevo G. (2001) Voltage-gated sodium channels

shape subthreshold EPSPs in layer 5 pyramidal neurons from rat prefrontal cortex. J

Neurophysiol 86, 1671-1684.

Huan C., Zhou M., Wu M., Zhang Z. and Mei Y. (2001) Activation of melatonin

receptor increases a delayed rectifier K+ current in rat cerebellar granule cells. Brain

Res 917, 182-190.

Klein R. S., Williams K. C., Alvarez-Hernandez X., Westmoreland S., Force T.,

Lackner A. A. and Luster A. D. (1999) Chemokine receptor expression and signaling

in macaque and human fetal neurons and astrocytes: implications for the

neuropathogenesis of AIDS. J Immunol 163, 1636-1646.

Lazarini F., Tham T. N., Casanova P., Arenzana-Seisdedos F. and Dubois-Dalcq

M. (2003) Role of the alpha-chemokine stromal cell-derived factor (SDF-1) in the

developing and mature central nervous system. Glia 42, 139-148.

Lazarini F., Casanova P., Tham T. N., De Clercq E., Arenzana-Seisdedos F.,

Baleux F. and Dubois-Dalcq M. (2000) Differential signalling of the chemokine

receptor CXCR4 by stromal cell-derived factor 1 and the HIV glycoprotein in rat

neurons and astrocytes. Eur J Neurosci 12, 117-125.

Limatola C., Giovannelli A., Maggi L., Ragozzino D., Castellani L., Ciotti M. T.,

Vacca F., Mercanti D., Santoni A. and Eusebi F. (2000) SDF-1alpha-mediated

modulation of synaptic transmission in rat cerebellum. Eur J Neurosci 12, 2497-2504.

Liu Z., Geng L., Li R., He X., Zheng J. Q. and Xie Z. (2003) Frequency

modulation of synchronized Ca2+ spikes in cultured hippocampal networks through

G-protein-coupled receptors. J Neurosci 23, 4156-4163.

24

Ma J. Y., Li M., Catterall W. A. and Scheuer T. (1994) Modulation of brain Na+

channels by a G-protein-coupled pathway. Proc Natl Acad Sci U S A 91, 12351-

12355.

Magee J. C. and Johnston D. (1997) A synaptically controlled, associative signal

for Hebbian plasticity in hippocampal neurons. Science 275, 209-213.

Oh S. B., Endoh T., Simen A. A., Ren D. and Miller R. J. (2002) Regulation of

calcium currents by chemokines and their receptors. J Neuroimmunol 123, 66-75.

Ragozzino D. (2002) CXC chemokine receptors in the central nervous system:

Role in cerebellar neuromodulation and development. J Neurovirol 8, 559-572.

Reisine T. (1990) Pertussis toxin in the analysis of receptor mechanisms. Biochem

Pharmacol 39, 1499-1504.

Rosenthal W., Hescheler J., Eckert R., Offermanns S., Schmidt A., Hinsch K. D.,

Spicher K., Trautwein W. and Schultz G. (1990) Pertussis toxin-sensitive G-proteins:

participation in the modulation of voltage-dependent Ca2+ channels by hormones and

neurotransmitters. Adv Second Messenger Phosphoprotein Res 24, 89-94.

Ross R. A., Biedler J. L., Spengler B. A. and Reis D. J. (1981) Neurotransmitter-

synthesizing enzymes in 14 human neuroblastoma cell lines. Cell Mol Neurobiol 1,

301-311.

Rossi D. and Zlotnik A. (2000) The biology of chemokines and their receptors.

Annu Rev Immunol 18, 217-242.

Schwindt P. C. and Crill W. E. (1995) Amplification of synaptic current by

persistent sodium conductance in apical dendrite of neocortical neurons. J

Neurophysiol 74, 2220-2224.

Stuart G. and Sakmann B. (1995) Amplification of EPSPs by axosomatic sodium

channels in neocortical pyramidal neurons. Neuron 15, 1065-1076.

Tanabe S., Heesen M., Yoshizawa I., Berman M. A., Luo Y., Bleul C. C.,

Springer T. A., Okuda K., Gerard N. and Dorf M. E. (1997) Functional expression of

the CXC-chemokine receptor-4/fusin on mouse microglial cells and astrocytes. J

Immunol 159, 905-911.

Toselli M., Tosetti P. and Taglietti V. (1996) Functional changes in sodium

conductances in the human neuroblastoma cell line SH-SY5Y during in vitro

differentiation. J Neurophysiol 76, 3920-3927.

25

Tosetti P., Taglietti V. and Toselli M. (1998) Functional changes in potassium

conductances of the human neuroblastoma cell line SH-SY5Y during in vitro

differentiation. J Neurophysiol 79, 648-658.

Xu R., Clarke I. J., Chen S. and Chen C. (2000) Growth hormone-releasing

hormone decreases voltage-gated potassium currents in GH4C1 cells. J

Neuroendocrinol 12, 147-157.

Zheng J., Thylin M. R., Ghorpade A., Xiong H., Persidsky Y., Cotter R., Niemann

D., Che M., Zeng Y. C., Gelbard H. A., Shepard R. B., Swartz J. M. and Gendelman

H. E. (1999) Intracellular CXCR4 signaling, neuronal apoptosis and neuropathogenic

mechanisms of HIV-1-associated dementia. J Neuroimmunol 98, 185-200.

26

Table 1 Primer sequences and PCR conditions used to determine neuropeptide

gene expression (number of PCR cycles performed: 35)

Primers F : Forward R: Reverse Sequences 5’-3’ Gene bank number

mouse CXCR4 F154

R 302

human CXCR4 F20

R220

TTGACTGGCATAGTCGGCAATG

ATGGCATCAACTGCCCAGAAG

AGGTAGCAAAGTGACGCCG

GATGGTGGGCAGGAAGATT

Gi 31560813

Gi 34189293

27

Figure Legends

Figure 1 SDF-1α effects on hypothalamic neurons in culture

a- Photomicograph of the mouse hypothalamic neurons in culture. Note the long

processes of the neurons.

b- Expression of CXCR4 mRNA in hypothalamic neurons in primary culture. Total

RNA (2 µg) were submitted to RT-PCR analysis as described under « Methods ». A

negative control without reverse transcriptase (NRT) was performed.

c- Immunocytochemical localization of the CXCR4 protein. Immunostaining of

hypothalamic neurons in primary culture with an anti-CXCR4 antibody at a 1/100

dilution.

d- Dual effect of SDF-1α on the AP discharge induced by a +20 pA rectangular

current step in one hypothalamic cultured neuron.

e- Histogram of the mean frequency (± s.e.m, n = 8) of AP discharge measured in

trains of action potentials obtained in response to currents steps of the same amplitude

in control conditions and in the presence of 10 nM SDF-1α (dots represent individual

values).

Figure 2 Characterization of SDF-1α effects on hypothalamic neurons in

culture

a- The effect of SDF-1α is blocked by the pre-application of the CXCR4 antagonist

AMD 3100 (1 µM, AMD).

b- The effect of SDF-1α is still observed in the continuous presence of the

metabotropic glutamate receptor antagonist MCPG (500 µM).

28

c- The effect of SDF-1α is not modified with 0.5 mM EGTAin the pipette recording

solution.

d- SDF-1α (10 nM) depresses the potassium outward currents recorded in TTX (3

µM). Currents were obtained in response to a voltage step from –60 mV to + 50 mV

and were leak-subtracted and filtered at 1 kHz.

e- SDF-1α (10 nM) depresses the sodium inward current recorded in the presence of

20 mM TEA. Currents were obtained in response to a voltage step from –60 mV to -10

mV and were leak-subtracted and filtered at 1 kHz.

Figure 3 SH-SY5Y is a good model to study the effects of SDF-1α on voltage-

dependent currents

a- Photomicograph of the SH-SY5Y cell line preparation. Note the absence of long

processes.

b- Expression of CXCR4 mRNA in SH-SY5Y cells. Total RNA (2 µg) were submitted

to RT-PCR analysis as described in Material and Methods. A negative control without

reverse transcriptase (NRT) was performed.

c- Immunocytochemical localization of the CXCR4 protein. Immunostaining of SH-

SY5Y cells with an anti-CXCR4 antibody at a 1/100 dilution.

d- When recorded in the whole-cell configuration of the patch-clamp technique, SH-

SY5Y cells presented a capacitance an a membrane resistance of respectively 14,6 ±

8,02 pF and 782,4 ± 538,3 MΩ (n = 50). Their resting membrane potential was -51 ±

21 mV (n = 8). CXCR4 stimulation by SDF-1α induced a decrease in the peak and in

the repolarization slope of the AP of SH-SY5Y cells. Spike recorded in response to a

rectangular 200 pA current pulse applied to the cell in the absence (Control) or

29

presence of 10 nM SDF-1α. The inset is a detail of the peak of the spike with a higher

magnification.

e- Spike recorded in another cell in response to a rectangular 200 pA current pulse in

the absence (Control) or at the beginning of the application of 20 mM TEA (TEA).

f, g, h- The rising slope (f), the peak (g) and the descending slope (g) of the AP were

plotted for 8 cells in control condition, in the presence of SDF-1α and after SDF-1α

has been washed-out. Mean values ± SD (n = 8) are given by the bars. ** P < 0.01.

Figure 4 CXCR4 stimulation by SDF-1α decreases the delayed rectifier

potassium current in SH-SY5Y cells

a- Whole-cell voltage-clamp experiments performed in the presence of TTX (3 µM)

revealed that 68% of SH-SY5Y cells present an outward delayed rectifier current (with

a maximum amplitude of 704 ± 632 pA, n = 34). Current traces were obtained in

response to a depolarizing pulse at +50 mV from a holding potential of -60 mV before

(Control) during and after (Wash) the application of 10 nM SDF-1α (n = 14). Notice

that the activation kinetics of the current was unchanged.

b- Time course of the SDF-1α effect. Depolarizing pulses of + 50 mV were applied

every 10 s and peak (•) and baseline (o) currents were plotted against time.

Corresponding traces 1, 2, 3 are the traces presented in part a of the figure (1 = control,

2 = 10 nM SDF-1α, 3 = wash).

c- Voltage dependence of the SDF-1α effect. The traces are current recorded in

response to ramps of potential from -100 to 80 mV in 1500 ms every 10 s before

(Control) during and after (Wash) the application of 10 nM SDF-1α. Notice that there

was few voltage-dependence of the effect of SDF-1α.

30

d- SDF-1α (10 nM) does not affect the activation/inactivation properties of the

current.

Activation (•,o ) and inactivation (,∇) curves in the absence (filled symbols) or the

presence (open symbols) of SDF-1α were realized by averaging the normalized

activation and inactivation curves obtained in 5 cells by measuring the maximum

outward current induced by depolarizing steps. Activation protocol consisted in

successive depolarizing pulse of 100 ms from a holding potential of -60 mV by

increments of 10 mV every 10 s (example of traces in the inset on the top right).

Inactivation protocol consisted in 1 s depolarizing pre-pulses from a holding potential

of -60 mV with 10 mV increments followed by a depolarizing pulse at +50 mV every

10 s (see inset on the bottom right). Activation and inactivation curves were fitted with

Boltzman curves (control = solid line, SDF-1α = dash line) with following parameters:

(n = 5) V1/2 (mV) k (mV)

Control activation

Activation in SDF-1α

Control inactivation

Inactivation in SDF-1α

10.9 ± 0.9

9.2 ± 0.7

2.7 ± 1.9

3.2 ± 3.9

12.2 ± 0.8

12.1 ± 0.6

-17.2 ± 1.5

-14.4 ± 3.2

e- Concentration- dependence of the effect and block by the antagonist AMD 3100

(AMD). Bars represent the mean percentage of decrease of IK measured in all the cells

tested (n in parenthesis) at the various concentrations indicated in the absence (-) or

presence of the antagonist (*** P < 0.01, Fisher test after ANOVA).

f- Lack of effect of SDF-1α on the residual current recorded with an internal cesium

solution (in which all the potassium was substituted with cesium). The inset on top

illustrates the block of the delayed rectifier current by 20 mM TEA (currents are

31

induced by a pulse from –60 to 50 mV, and leak subtracted). When using Cs+ instead

of K+ in the internal solution, there remained an outward current with a mean

amplitude of 102 ± 8 pA at -50 mV (n = 21) which did not inactivate. The traces are

current recorded in response to ramps of potential from -100 to 80 mV of 1500 ms

every 10 s. The residual current was unaffected by 10 nM SDF-1α but was blocked by

20 mM TEA. This current only represents 14,5% of the outward current recorded at +

50 mV with the K+ internal solution.

g- Bar graph show the effects of SDF-1α (10 nM) on potassium currents in control

cells recorded with the 0.4 mM GTP internal solution (Control), in cells recorded with

the 0.4 mM GTPγS internal solution (GTPγS) and in PTX pre-treated cells (PTX). Pre-

activation of G-protein by GTPγS and pre-treatment with PTX block the SDF-1α-

evoked reduction of potassium current.

Bars are mean (± s.e.m, n in parenthesis) of the current decrease relative to control

current before SDF-1α application, ** P < 0.01, *** P < 0.001. The experiments

were performed 2-3 days after plating but in g were performed 4-5 days after plating,

when cells express both sodium and potassium currents.

Figure 5 CXCR4 stimulation by SDF-1α induces a depression of sodium current

in SH-SY5Y cells

a- In the presence of 20 mM TEA, voltage depolarizing steps induced a transient

inward current in 46 % of SH-SY5Y cells. The maximum amplitude of this current

ranged between 70 and 600 pA with a mean of 258,7 ± 141,6 pA (n = 23). Current

traces were obtained in response to a depolarizing pulse at +10 mV from a holding

potential of -60 mV before (Control) and during the application of 10 nM SDF-1α.

Notice that the kinetics of the current was unchanged.

32

b- Time course of the SDF-1α effect. Depolarizing pulses of + 10 mV were applied

every 10 s and peak (•) and baseline (°) currents were plotted against time.

Corresponding traces 1 and 2 are illustrated in part a of the figure.

c- Lack of effect of SDF-1α 10 nM on the activation/inactivation properties of the

sodium current. Activation (•,o ) and inactivation (,∇) curves in the absence (filled

symbols) or the presence (open symbols) of SDF-1α were realized by averaging the

normalized activation and inactivation curves obtained in 3 cells by measuring the

maximum inward current induced by depolarizing steps. Activation protocol consisted

in successive depolarizing pulse of 50 ms from a holding potential of -60 mV by

increments of 5 mV every 10 s (example of traces corresponding to depolarization to -

40, -30, -20, -10, 0 +10 and +20 mV are given in the inset on the top right).

Inactivation protocol consisted in 500 ms depolarizing pre-pulses from a holding

potential of -60 mV with 5 mV increments followed by a depolarizing pulse at -10

mV. Inset on the bottom right illustrates an example of traces obtained with this

protocol. The inset on top illustrates the full block of the current by 3 µM TTX

(currents were leak-subtracted).

Parameters of the Boltzman fit (control = solid line, SDF-1α = dash line) are the

following:

(n = 3) V1/2 (mV) k (mV)

Control activation

Activation in SDF-1α

Control inactivation

Inactivation in SDF-1α

-17.1 ± 1.1

-16.9 ± 1.3

-58.7 ± 1.7

-58.6 ± 1.8

5.4 ± 1.0

6.7 ± 1.1

-20.9 ± 1.7

-20.1= ± 1.7

It is worth noting that the expression of this current progressively increased from 2 to

5 days plating and that inactivation curves could not be well fitted by a single

33

Boltzman curve, suggesting that the sodium currents were heterogeneous in SH-SY5Y

cells.

d- Concentration-dependence of the effect and block by the antagonist AMD 3100

(AMD). Bars represent the mean percentage of decrease of the sodium current (INa)

measured in all the cells which responded to SDF-1α (mean ± s.e.m, n in parenthesis)

at the various concentrations indicated in the absence (-) or presence of 200 µM of the

antagonist (* P < 0.05, *** P < 0.01, Fisher test after ANOVA).

e- Pre-activation of G-protein by GTPγS and pre-treatment with PTX block the SDF-

1α-evoked reduction of sodium current.

Bar graph showing the effects of SDF-1α (10 nM) on sodium currents in control cells

recorded with the 0.4 mM GTP internal solution (Control), in cells recorded with the

0.4 mM GTPγS internal solution (GTPγS) and in PTX pre-treated cells (PTX). See

Methods. Bars are mean (± s.e.m, n in parenthesis) of the current decrease relative to

control current before SDF-1α application, * p < 0.05 ** p<0.01 . Cells were

recorded 4-5 days after plating when sodium current expression increases.