Caveolin-Rich Lipid Rafts of The Plasma Membrane of Mature Cerebellar

Cell Calcium 56 (2014) 108123
Contents lists available at ScienceDirect
Cell Calcium
journal homepage: www.elsevier.com/locate/ceca
Caveolin-rich lipid rafts of the plasma membrane of mature cerebellar

granule neurons are microcompartments for calcium/reactive oxygen
and nitrogen species cross-talk signaling
D. Marques-da-Silva, C. Gutierrez-Merino
Department of Biochemistry and Molecular Biology, Faculty of Sciences, University of Extremadura, 06006 Badajoz, Spain
a r t i c l e
i n f o
Article history:
Received 1 December 2013
Received in revised form 28 May 2014
Accepted 7 June 2014
Available online 14 June 2014
Keywords:
Calcium signaling
Calcium microcompartments
Lipid rafts
Caveolin-1
Plasma membrane calcium pump
Sodiumcalcium exchanger
NMDA receptors
L-type voltage-operated calcium channels
Nitric oxide synthase
Cytochrome b5 reductase
Reactive oxygen and nitrogen species
Cerebellar granule neurons
a b s t r a c t
In previous works, we have shown that L-type voltage-operated calcium channels, N-methyl-d-aspartate
receptors (NMDAr), neuronal nitric oxide synthase (nNOS) and cytochrome b5 reductase (Cb5 R) colocalize within the same lipid rafts-associated nanodomains in mature cerebellar granule neurons (CGN).
In this work, we show that the calcium transport systems of the plasma membrane extruding calcium
from the cytosol, plasma membrane calcium pumps (PMCA) and sodiumcalcium exchangers (NCX), are
also associated with these nanodomains. All these proteins were found to co-immunoprecipitate with
caveolin-1 after treatment with 25 mM methyl--cyclodextrin, a lipid rafts solubilizing agent. However, the treatment of CGN with methyl--cyclodextrin largely attenuated the rise of cytosolic calcium
induced by l-glutamate through NMDAr. Fluorescence energy transfer imaging revealed that all of them
are present in sub-microdomains of a size smaller than 200 nm, with a peripheral distribution of the
calcium extrusion systems PMCA and NCX. Fluorescence microscopy images analysis revealed high calcium dynamic sub-microcompartments near the plasma membrane in fura-2-loaded CGN at short times
after addition of l-glutamate. In addition, the close proximity between sources of nitric oxide (nNOS) and
superoxide anion (Cb5 R) suggests that these nanodomains are involved in the fast and efcient cross-talk
between calcium and redox signaling in neurons.
2014 Elsevier Ltd. All rights reserved.
1. Introduction
A highly efcient and rapid functional coupling is particularly relevant for neuronal activity, as neurons have to deliver
fast responses to many repetitive and simultaneous extracellular stimuli coming from different neighbor cells. Protein
Abbreviations: BF, bright eld microscopy images; Cb5 R, cytochrome b5

reductase; CGN, cerebellar granule neurons; DMEM, Dulbeccos-modied Eagles
medium; EC50 , concentration needed to afford 50% activation; EDTA, ethylenediaminetetraacetic acid; EGTA, ethylene glycol-bis(2-aminoethylether)-N,N,N ,N tetraacetic acid; FRET, uorescence or Frster resonance energy transfer; GF, green
uorescence; HEPES, N-[2-hydroxyethyl] piperazine-N -[2-ethanesulfonic acid];
IgG, immunoglobulin G; L-VOCC, L-type voltage-operated calcium channels; MCD,
methyl--cyclodextrin; NCX, sodiumcalcium exchanger; NMDAr, N-methyl-daspartate receptor; nNOS, neuronal nitric oxide synthase; PBS, phosphate-buffered
saline; PMCA, plasma membrane calcium pump; RF, red uorescence; ROI, region of
interest for quantitative microscopy images analysis; ROS, reactive oxygen species;
TBS, tris-buffered saline; TrisHCl, tris(hydroxymethyl)aminomethane hydrochloride; Tween-20, polyoxyethylenesorbitan monolaurate.
Corresponding author. Tel.: +34 924 289419; fax: +34 924 289419.
E-mail address: carlosgm@unex.es (C. Gutierrez-Merino).
http://dx.doi.org/10.1016/j.ceca.2014.06.002
0143-4160/ 2014 Elsevier Ltd. All rights reserved.
compartmentation within microdomains allows for a more

efcient and rapid functional coupling, and studies on calcium
signaling in neurons have played a pioneer role to demonstrate the
outstanding role of subcellular compartmentation in the control
of neuronal activity [13]. In addition, the efcient functional
coupling of several of the major plasma membrane calcium
transport systems allows for a ne control of cytosolic calcium
homeostasis within the narrow range required for neuronal
survival [47]. Deregulation of their functional coupling can lead
to a sustained alteration of intracellular calcium homeostasis in
neurons, a common feature in oxidative stress-mediated neurodegeneration, and plasma membrane calcium transport systems
have been shown to be molecular targets for reactive oxygen
species generated in neurodegenerative insults and diseases,
reviewed in [8,9]. In primary cultures of mature cerebellar granule
neurons (CGN) the entry of calcium through L-type voltageoperated calcium channels (L-VOCC) plays a major role to keep
cytosolic calcium within the optimal 70200 nM concentration
range needed for survival of these neurons in vitro [47]. Indeed,
calcium entry through L-VOCC accounts for more than 75% of
the steady state cytosolic calcium increase in the soma of mature
D. Marques-da-Silva, C. Gutierrez-Merino / Cell Calcium 56 (2014) 108123
CGN in culture after the partial depolarization of the plasma

membrane elicited by an increase of the extracellular K+ concentration from 5 to 25 mM [6], consistent with the fact that low
K+ -induced neuronal apoptosis can be blocked by increasing the
extracellular concentration of potassium [4,5,10]. Calcium entry
through L-VOCC and N-methyl-d-aspartate receptors (NMDAr)
play a major role in the maintenance of cytosolic calcium needed
for CGN survival and excitability [4,6,11,12], and alterations of
their functional response by reactive oxygen species (ROS) or redox
modulation can lead to CGN death in culture either by apoptosis
or by excitotoxicity [6,7,9,13,14]. NMDA receptors are found in
synaptic and in extra-synaptic locations [1517]. As activation of
extra-synaptic NMDA receptors can lead to a less focalized increase
of cytosolic calcium, the extra-synaptic NMDA receptors are likely
to play a role more relevant than synaptic NMDA receptors in the
control of cytosolic calcium homeostasis in the neuronal soma.
Lipid rafts of the plasma membrane are dynamic nanodomains
of a dimension between 10 and 200 nm, characterized by their
heterogeneity and by their high content in cholesterol and esphingolipids [18]. These lipid rafts nanodomains dene cellular
sub-microdomains of the plasma membrane anchoring caveolins,
otillin and also actin microlaments [19], and the presence
of caveolins associated with neuronal plasma membrane in
microdomains without the morphological appearance of caveola invaginations has been well documented during last decade,
revised in [20]. Furthermore, it has been suggested that these
caveolin-rich nanodomains can serve to focalize cell signaling
transduction in neurons [2023].
Experimental evidences supporting the possible implication of
lipid rafts in cytosolic calcium homeostasis and calcium signaling
are merging only during last years, see, e.g. [24]. The most ubiquitous and relevant calcium transport systems involved in the
neuronal cytosolic calcium homeostasis are: NMDAr and L-VOCC
as calcium entry systems, and plasma membrane Ca2+ -pumps
(PMCA) and Na+ /Ca2+ exchanger (NCX) as calcium extrusion systems [1,2,9,25].
L-VOCC have been shown to be associated with lipid rafts in
cardiomyocytes [19,26], and in a previous work we have shown
that in mature CGN in culture L-VOCC is associated with lipid
rafts prepared from these neurons [27]. Since the two major subtypes of L-VOCC present in the brain, namely CaV 1.2 and CaV 1.3,
directly interact with many proteins having the PDZ binding
domain [28,29], proteins that also bind to the NMDA receptor [30],
the association of these receptors with lipid rafts nanodomains
is not an unexpected nding. The presence of NMDA receptors
in isolated lipid rafts has been shown by different investigators
[31,32], and using uorescence resonance energy transfer (FRET)
microscopy imaging their association with L-VOCC-containing lipid
rafts nanodomains in mature primary cultures of cerebellar granule
neurons has been demonstrated in a recent work of our laboratory
[33]. However, the association of the calcium extrusion systems
PMCA and NCX with lipid rafts nanodomains containing L-VOCC
and NMDAr has not been experimentally assessed yet, despite that
PMCA has been shown to be associated with lipid rafts of synaptic
plasma membranes [34] and of primary cortical and hippocampal
rat neurons in culture [35] and NCX binding to lipid rafts has been
demonstrated in coronary artery smooth muscle preparations [36].
Owing to the strong toxicity of sustained and widespread cytosolic
calcium concentrations above 0.4 M, see, e.g. [4,6,7], a close spatial location of these calcium entry and calcium extrusion transport
proteins in the neuronal plasma membrane can provide obvious
advantages for long-term survival of the highly interconnected
brain neurons.
On the other hand, two enzymes that can produce ROS that modulate the activity of these calcium transport systems have also been
shown to associate with the neuronal plasma membrane, namely,
109
nNOS producing nitric oxide [37] and cytochrome b5 reductase

(Cb5 R), which generates a peak of superoxide anion near the plasma
membrane of CGN that plays a critical role in the progress of apoptosis in low K+ medium [38,39]. In previous works, we have shown
that nNOS and Cb5 R are associated with caveolin-rich lipid rafts
nanodomains of the plasma membrane of mature CGN in culture,
close to NMDAr and L-VOCC, respectively [27,33,38]. Indeed, both
nNOS and Cb5 R have been shown to contain protein domains that
can form stable complexes with caveolin-1 [39,40]. The presence of
Cb5 R within L-VOCC and NMDAr-rich nanodomains suggests that
they could be derived from endoplasmic reticulum/plasma membrane junctions which are also known to be associated with lipid
rafts [41]. The possibility of co-localization of nNOS and Cb5 R within
the same nanodomains deserved to be experimentally assessed,
because NO rapidly reacts with superoxide anion to produce peroxynitrite [42], a harmful and strongly neurotoxic agent.
In this work, using preparations of lipid rafts-enriched
membrane fragments from mature CGN in culture, coimmunoprecipitation and FRET imaging of CGN we show that
the calcium extrusion systems (PMCA and NCX) co-localize with
the major calcium entry systems (L-type calcium channels and
NMDA receptors) and with the ROS/RNS enzyme sources (Cb5 R and
nNOS) within lipid rafts-associated sub-microdomains of a size
lower than 200 nm. Our results strongly suggest that in neurons
lipid rafts can be seen as markers of nanodomains of the plasma
membrane of particular relevance in the cross-talk between redox
and calcium signaling.
2. Materials and methods
2.1. Preparation of rat cerebellar granule neurons (CGN)
CGN were obtained from dissociated cerebella of 7 days old
Wistar rats as described in previous works of this laboratory
[6,10,27,38,43]. Briey, cells were plated in Dulbeccos modied
Eagles medium (DMEM) supplemented with 10% heat-inactivated
fetal bovine serum, 5 mM glucose, 19.63 mM KCl, 3.7 ng/mL insulin,
7 M 4-aminobenzoic acid, 50 U/mL penicillin, 25 U/mL streptomycin, 0.91 mM pyruvate and 2 mM glutamine on 35-mm dishes
(Corning, NY, USA) and plates with 24 wells coated with poly-dlysine, at a density of 2.5 106 cells/dish. Cultures were maintained
at 37 C in a humidied atmosphere of 95% air/5% CO2 . Cytosine
arabinofuranoside (10 M) was added to fresh culture medium
48 h after plating to prevent replication of non-neuronal cells.
Seven days after plating, the culture medium was replaced with
the following serum-free DMEM:F12 medium (1:1) supplemented
with 12.5 mM glucose, 20.82 mM KCl, 5 g/mL insulin, 0.1 mg/mL
apo-transferrin, 20 nM progesterone, 50 U/mL penicillin, 25 U/mL
streptomycin, 0.1 mg/mL pyruvate and 2 mM l-glutamine. Cell
viability was experimentally assessed measuring the amount of
colored formazan by the reduction of 3-(4,5-dimethylthiazol-2yl)-2,5-diphenyltetrazolium bromide (MTT) as in previous works
[6,7,10,43].
The composition of the Lockes K25 buffer (pH 7.4 at 37 C)
used in this work is as follows: 4 mM NaHCO3 , 10 mM N[2-hydroxyethyl] piperazine-N -[2-ethanesulfonic acid] (HEPES),
5 mM glucose, 2.3 mM CaCl2 , 1 mM MgCl2 , and 134 mM NaCl and
25 mM KCl.
2.2. Isolation of lipid rafts
Lipid rafts were isolated running sucrose gradients as in a previous work of this laboratory [27,33,39], following a method adapted
from the protocols described in references [44,45]. CGN cultured for
8 days in vitro were washed with Lockes K25 buffer. The lysates
110
were centrifuged for 2 min at 2000 g at 4 C in a refrigerated

Eppendorf microcentrifuge, the cellular pellets were resuspended
and solubilized in 550 l of 0.25% Triton X-100 solubilization buffer
(25 mM TrisHCl, pH 7.4, 150 mM NaCl, 5 mM EDTA, 50 mM NaF,
5 mM NaVO3 and 0.25% Triton X-100, supplemented with the Roche
Biochemicals protease inhibitor cocktail COMPLETER ), and left for
30 min at 4 C. Sucrose concentration of the lysates was adjusted to
41% before they were loaded over two layers of sucrose: 2125 l of
35% sucrose and 500 l of 16% sucrose, prepared in 25 mM TrisHCl,
pH 7.4, 150 mM NaCl, 5 mM EDTA. For lipid rafts isolation, samples
were centrifuged during 5 h at 4 C in a SW60 rotor at 256,000 g
for the average radius, and 10 fractions were collected (from the
top to the bottom, fractions 110). Samples were analyzed by SDSPAGE followed by Western blotting as indicated below.
2.3. SDS-PAGE and Western blotting
SDS-PAGE were run at a concentration of 7.5%, 10.4% or 12.5%
acrylamide depending upon the molecular weights of the target
proteins, using 2 g protein of CGN lysates per lane. Gels were
transferred to nitrocellulose membranes of 0.2 and 0.4 m average pore size (Trans-BloT Transfer Medium, BioRad). Nitrocellulose
membranes were blocked by 1 h incubation at room temperature
with 5% (w/v) non-fat dry milk in PBS supplemented with 0.5% polyoxyethylenesorbitan monolaurate (Tween 20). Then, nitrocellulose
membranes were washed three times with PBS supplemented with
0.5% Tween 20. Immunodetection of proteins was performed with
its specic antibody at a dilution of 1:100 in PBS supplemented with
0.5% Tween 20. After incubation with the rst antibody overnight,
membranes were washed six times with PBS supplemented with
0.5% Tween 20 and incubated for 1 h at room temperature with the
secondary antibody IgG conjugated with horseradish peroxidase
(anti-rabbit, anti-mouse and anti-goat with a dilution of 1:8000,
1:50,000 and 1:500,000, respectively), then washed six times with
PBS supplemented with 0.5% Tween 20, followed by incubation
for 3 min with SuperSignal West Dura Substrate (Pierce). Western
blots were revealed by exposure to an Amersham Hyperlm MP
autoradiography lm (GE Healthcare, UK).
2.4. Immunoprecipitation
Complexes between matrix and antibodies were prepared as
follows. In an Eppendorf tube 5 g of the antibody selected for
immunoprecipitation were mixed with 50 l of the appropriate
matrix (from the same source of the antibody) and 500 l of TBS.
After 1 h incubation in a tube-rotor with continuous shaking, the
matrix was precipitated by 30 s centrifugation at 4 C at 16,100 g
in a refrigerated Eppendorf microcentrifuge 5415R. The supernatant was removed and the precipitated matrix was subjected to
three washes with 500 l of TBS, namely, by 30 s centrifugation at
4 C at 16,100 g in a refrigerated Eppendorf microcentrifuge in
each washing step.
Cell lysates of CGN were prepared as indicated in the isolation
of lipid rafts section and 2 mg of protein lysates were mixed with
40 l of the matrix in an Eppendorf tube for a pre-washing step.
After 30 min incubation in a tube-rotor with continuous shaking,
matrix was precipitated by 30 s centrifugation at 16,100 g and
4 C. The supernatant was carefully withdrawn to avoid removal
of the precipitate, deposited in a separate tube and treated with
25 mM methyl--cyclodextrin and homogenized under mild stirring. Thereafter, the treated supernatant was added to the tube
containing the precipitate of matrix/antibody complex prepared
as indicated above and incubated during 1 h at room temperature
under continuous shaking in a tube-rotor. Then, the matrix was
precipitated by 30 s centrifugation at 16,100 g and 4 C and the
supernatant was carefully removed. The precipitated matrix was
resuspended in TBS and centrifuged again at 16,100 g and 4 C

for a better removal of the remaining supernatant. This washing
step was repeated four times. The washed matrix precipitate was
mixed with standard sample buffer for SDS-electrophoresis supplemented with 125 mM urea and 0.3% saponin, boiled during 5 min
and loaded onto a SDS-PAGE gel. Samples were analyzed by SDSPAGE followed by Western blotting as indicated above.
2.5. Fluorescence microscopy imaging
Fluorescence microscopy images of CGN were acquired with a
Hamamatsu Orca-R2 CCD camera (binning mode 2 2) attached to
a Nikon Diaphot 300 epiuorescence microscope, and quantitative
analysis of the average uorescence intensity of the frames and
of selected neuronal soma was done with the HCImage software
using the region of interest (ROI) tool of this software, as in previous works [27,33,38,39,46]. Only elds devoid of large aggregates
of neurons forming granules or small grain-like structures were
selected for image acquisition, to minimize distortion of images
by the sum of the uorescence contributions of juxtaposed layers of neurons. Images were acquired with an excitation lter of
470 nm, and 510 nm dichroic mirror/520 nm emission lter (donorgreen uorescence) and 580 nm dichroic mirror/590 nm emission
lter (acceptor-red uorescence). Acquired images were exported
as TIFF pseudo-color images for further processing using the Image
J software. Objectives used in this study: NCF Plan ELWD 40 and
CF N Plan LWD 100. Direct-uorescence intensity images are
presented in gray scale (black: very low or no-signal, and white:
saturated signal).
FRET measurements were performed with mature CGN (89
days in vitro) in 24-well plates at a density of 600,000 cells per
well. The wells were washed with Lockes buffer to wash the
phenol red remaining in the plates. Then CGN were xed with
2.5% para-formaldehyde, 3 mM MgCl2 , 2 mM ethylene glycol-bis(2aminoethylether)-N,N,N ,N -tetraacetic acid (EGTA) and 0.32 M
sucrose in PBSs buffer (5 mM sodium phosphate, 137 mM NaCl and
27 mM KCl, pH 7).
After CGN were xed, cells were incubated during 1 h at 37 C
with the rst antibody in PBS supplemented with 0.2% Triton X-100,
and then washed six times with PBS (washing step). Thereafter,
CGN were incubated for 30 min with the appropriate Alexa488labeled secondary antibody in PBS supplemented with 0.2% Triton
X-100, and washed again with PBS before acquisition of uorescence microscopy images stained only with the donor dye. After
nishing the acquisition of images, CGN were incubated during 1 h
at 37 C with the second primary antibody in PBS supplemented
with 0.2% Triton X-100, washed six times with PBS (washing step),
and then incubated for 30 min with the appropriate Cy3-labeled
secondary antibody in PBS supplemented with 0.2% Triton X-100,
and nally washed again with PBS. Thereafter, before obtaining the
uorescence microscopy images of CGN stained with donor and
acceptor dyes the multiplate was placed again in the microscope
and realigned in the same positions using an external ruler to select
observation elds close to those used for the capture of the images
of CGN stained only with the donor dye. Indeed, we avoided to take
the second step images from the very same eld used for rst step
images, to minimize fading by repetitive irradiation of the same
eld, as indicated in previous works [27,38]. Image acquisitions of
CGN double stained with donor and acceptor dyes were routinely
started between 2 and 3 h after nishing the acquisition of images
of CGN stained only with the donor dye, and in all cases both sets
of images were captured within the same day. Before their use for
uorescence microscopy images acquisition, the specicity of primary antibodies for the target protein was routinely assessed by
the presence of a major band at the expected protein molecular
weight accounting for more than 90% bands staining in Western
blotting run with CGN lysates (Supplementary Figure S1), as in

previous works of our laboratory [27,33,38,39]. Antibodies not fullling this requirement were discarded and not used in this work.
The lack of signicant uorescence labeling of CGN treated with the
Alexa488- and Cy3-secondary IgG antibodies used in this study in
the absence of the primary antibodies was assessed before running
FRET experiments, as it yielded red uorescence intensity readings
within the background range of control non-stained CGN, see also
[27,33,38,39]. In addition, to quantify the effects of the treatment
for labeling with the second primary antibody/IgG-Cy3 complex
in the green and red uorescence intensities of CGN stained with
the rst primary antibody plus IgG-Alexa488 we have carried out
control experiments performing these treatments without the second primary antibody. These control experiments showed that on
average the treatment done to label CGN with the red uorescent IgG-Cy3 secondary antibody resulted in 12 3% quenching
of the green uorescence of the IgG-Alexa488 antibody and no
statistically signicant change of the red uorescence intensity
relative to that measured with CGN stained only with the rst primary antibody/IgG-Alexa488 complex. This has been taken into
account in all the calculations of the quenching of green uorescence afforded by CGN double labeling with antibodies shown in
this work. Finally, the exposure times were kept as low as possible to ensure that the contribution of CGN autouorescence was
always lower than 10% of the uorescence intensity readings of
cells stained with the primary and secondary antibodies used in
this work.
2.6. Cytosolic calcium measurements
Intracellular free Ca2+ concentration ([Ca2+ ]i ) was measured
basically as indicated in previous works of our laboratory
[6,7,43,47]. A value of 224 nM has been used for the dissociation constant fura-2/Ca2+ complex to obtain the [Ca2+ ]i values
reported in this work. CGN were loaded with fura-2 by incubation
in DMEM-F12 for 4560 min with 5 M fura-2 acetoxymethyl ester
and 0.025% pluronic-F127 at 37 C. Afterwards, CGN were washed
twice with Lockes K25 buffer and the culture dish was placed in
a thermostatic controlled plate (Warner Instrument Co., Hamden,
CT, USA) of a Nikon Diaphot 300 inverted microscope, equipped
with an epiuorescence attachment and excitation lter wheel. To
measure the [Ca2+ ]i , ratio uorescence images were obtained with
excitation lters of 340 and 380 nm and a dichroic mirror DM510
and absorption lter (emission side) of 510 nm. Digital images were
taken with a Hamamatsu Orca-R2 CCD camera and Lambda 10-2
lter wheel controller and subsequently analyzed with HCImage
software.
2.7. Fluorescence resonance energy transfer (FRET)
For a single donor/acceptor pair with random orientation,
the FRET efciency (E) depends upon their separation (r):
E = r6 /(R0 6 + r6 ), where R0 is the distance at which the efciency of FRET is 50% [48,49]. Alexa488 and Cy3 dyes form a good
donor/acceptor pair for FRET, having a R0 value between 5 and 6 nm
[50]. As the FRET efciency is lower than 0.02 when r = 2 R0 the
maximum distance for experimentally detectable FRET between
a single pair of Alexa488 and Cy3 is lower than 12 nm. However,
as every Cy3 uorescent secondary antibody is labeled with more
than 1 mol of Cy3 per mol of antibody and also due to the presence of more acceptor molecules in the vicinity of the donor in
protein-rich microdomains, FRET between Alexa488-IgG and Cy3IgG secondary antibodies is a case of FRET from one donor to
multiple acceptors. The major advantages of this FRET case are (1)
there is a minimum error in the assumption of random orientation
111
between donor and acceptors and (2) the separation distance for
measurable FRET efciency increases by nearly two-fold [49,51].
The contribution of the wide emission peak of the donor
Alexa488 to the red uorescence of double stained CGN was corrected as indicated in previous works [27,33,39]. Briey, from
red uorescence images of CGN stained only with the primary
and corresponding Alexa488 IgG secondary antibodies we have
determined the signal ratio red/green uorescence intensities
for CGN stained only with Alexa488 tagged IgG antibodies coupled to the primary antibodies (RAlexa488IgG ). This ratio values
have been used for calculation of the red uorescence signal
coming from the Alexa488 dye (RFAlexa488 ) in double stained
CGN: RFAlexa488 = GF RAlexa488IgG , which we have subtracted these
RFAlexa488 values from total RF.
Since both FRET and photobleaching may lead to partial quenching of the donor uorescence (green uorescence), the basic criteria
used to conrm the occurrence of FRET and calculation of FRET
efciency was the simultaneous occurrence of quenching of the
green uorescence (GF) and an increase of the ratio between red
(acceptor uorescence) and green uorescence intensities (ratio
red/green) in CGN stained with Alexa488- and Cy3-secondary
antibodies, as in [27]. Absolute uorescence intensity ratio values were calculated taking into account that within the range of
exposure times used in this work both the average uorescence
intensity and the uorescence intensity per pixel were linearly
dependent with the exposure time, thus, the quenching of green
uorescence (GF) can be quantitatively compared with the parallel
increase of red uorescence (RF) in the same eld of observation.
Under the experimental conditions used in this work the ratio
between the quenching of donor uorescence and the increase
of acceptor uorescence was >0.9, i.e. close to the theoretical
value = 1.
2.8. Materials: chemicals and reagents

Primary antibodies: goat anti-NMDAr (sc-1468), rabbit antiL-VOCC (sc-25686), goat anti-L-VOCC (sc-16230), rabbit pananti-PMCA (sc-28765), mouse anti-PMCA1/4 (sc-20028), rabbit
anti-NCX (sc-32881), rabbit anti-NCX1 (sc-30304-R), mouse antinNOS (sc-5302), goat anti-H-Ras (sc-32026), rabbit anti-caveolin-1
(sc-894), rabbit anti-caveolin-2 (sc-7942), goat anti-otillin-1 (sc16640) and rabbit anti-otillin-1 (sc-25506) were supplied by
Santa Cruz Biotechnology (Santa Cruz, CA, USA), and rabbit antiCb5 R (cat. n 10894-1-AP, from ProteinTech Group, Chicago, IL,
USA). Fluorescent-labeled secondary antibodies used to label the
primary antibodies listed above (produced in rabbit, goat or
mouse as indicated in each case): anti-rabbit IgG-Alexa488 (cat.
n A11008), anti-goat IgG-Alexa488 (cat. n A11055) and antimouse IgG-Alexa488 (cat. n A11001) from Invitrogen (Molecular
Probes, Eugene, OR, USA), and anti-rabbit IgG-Cy3 (cat. n C2306),
anti-goat IgG-Cy3 (cat. n C2821) and anti-mouse IgG-Cy3 (cat.
n C2181) from Sigma (St. Louis, MO, USA). All these antibodies
were used in the dilution range recommended in their technical
sheets. Western blots reagents anti-goat IgG horseradish peroxidase, anti-rabbit IgG horseradish peroxidase and SuperSignal West
Dura Extended Duration Substrate were supplied by Pierce (Rockford, IL, USA). Rabbit IgG (Invitrogen Z25305) and goat IgG
(Invitrogen Z25602) were used for the control experiments of
immunoprecipitation.
ST-bodipy dihydropyridine (cat. n S7445) was supplied by
Invitrogen (Molecular Probes, Eugene, OR, USA). Fura-2 acetoxymethyl ester and pluronic F-127 were obtained from Biotium
(Hayward, CA, USA). All other reagents and chemicals were of
analytical grade from SigmaAldrich or RocheMerck (Darmstadt,
Germany).
112
Fig. 1. NMDAr, L-VOCC, PMCA, NCX, nNOS and Cb5 R are associated with lipid rafts isolated from mature CGN in culture. NMDAr, L-VOCC, PMCA, NCX, nNOS and Cb5 R
distribution in membrane fractions were determined by Western blotting using anti-NMDAr (sc-1468), anti-L-VOCC (sc-25686), anti-PMCA (anti-PMCA1/4 sc-20028), antiNCX (anti-NCX1 sc-30304-R), anti-nNOS (sc-5302) and anti-Cb5 R (ProteinTech-10894-1-AP). Fractions distribution of protein raft markers: H-Ras (anti-H-Ras sc-32026),
caveolin-1 (anti-caveolin-1 sc-894), otillin (anti-otillin-1 sc-16640) and caveolin-2 (anti-caveolin-2 sc-7942). Fractions 15 enriched in protein lipid rafts markers are also
enriched in cholesterol as shown in a recent work of our laboratory [39]. NMDAr, L-VOCC, PMCA, NCX, nNOS and Cb5 R are enriched in the fractions 15 where lipid rafts
markers are also enriched. Western blotting images shown are representative of the results obtained in experiments done with at least three different preparations of lipid
rafts. The percentage of each protein in every fraction relative to the total protein in all fractions was calculated from densitometry of the Western blotting images using
Image J, and the cumulative percentages of each protein in the fractions 15 are listed in the right-hand side of the gure. All primary antibodies had a dilution of 1:100 in
PBS with 0.5% Tween 20, the secondary anti-rabbit HRP (Pierce-1858415), the anti-mouse HRP (Pierce-1858413) and the anti-goat HRP (Sigma-A5420) were applied with a
dilution of 1:8000, 1:50,000 and 1:500,000 in PBS with 0.5% Tween 20, respectively. The same amounts of protein were used in each fraction for these experiments.
2.9. Statistical analysis

Results are expressed as mean standard error (S.E.). Statistical
analysis was carried out by MannWhitney non-parametric test.
Signicant difference was accepted at the p <0.05 levels
All the results were conrmed with duplicate measurements of
at least three different CGN preparations.
3. Results
3.1. Lipid rafts-submicrodomains of the plasma membrane are
enriched in the major calcium systems that control cytosolic
calcium in CGN and also in the redox systems nNOS and Cb5 R
In previous works, we have shown by FRET and western blotting that Cb5 R and L-VOCC are proteins associated with lipid rafts
microdomains of the plasma membrane of CGN matured in culture
[27,38,39,52,53] and also nNOS and NMDAr [33].
Membrane fragments prepared from CGN lysates were fractionated in sucrose density gradients as indicated in the Methods.
Western Blot analysis of the fractions showed that typical lipid rafts
markers, H-Ras, otillin, caveolin-1 and caveolin-2 were largely
enriched in fractions 15 (Fig. 1). The more accepted protein markers of lipid rafts, H-Ras and caveolin-1, can only be detected in
fractions 15. Densitometry analysis of the gels led to the percentage of each one of the tested proteins in every fraction (see the
Supplementary Figure S2), and in the right side of this Fig. 1 the
cumulative percentage of each protein in fractions 15 is shown.
The distribution between different fractions of Cb5 R and NMDAr
closely matches that of the typical lipid rafts markers cited above,
and the densitometry analysis yielded in both cases more than
90% of these proteins in fractions 15. L-VOCC and nNOS is also
largely enriched in lipid rafts-enriched fractions (fractions 15),

where more than 70% of each one of these proteins are concentrated, but it can be also detected in other membrane fractions like
otillin (fractions 68) and caveolin-2. Caveolin-2 showed a more
widespread distribution between fractions, in agreement with the
reported presence of caveolin-2 not only in lipid rafts but also in
membranes of the endocytosis pathway [54]. PMCA and NCX were
the proteins less enriched in fractions 15 among all the studied
proteins in this work and they display a more scattered distribution
between subcellular membrane fractions, similarly to caveolin-2
(see above), although it is to be noted that still more than 50% of
the total content of these membrane proteins are in these fractions.
In addition, we have experimentally assessed the co-localization
of PMCA and NCX within FRET distance of protein markers associated with lipid rafts caveolin-1 and otillin by microscopy imaging.
To this end, CGN were treated with anti-PMCA and anti-NCX
primary antibodies and thereafter treated with the secondary
Alexa488-labeled antibodies appropriate for the primary antiPMCA and anti-NCX used, as indicated in Section 2. Under our
experimental conditions, unspecic labeling (i.e. labeling by the
secondary antibodies in the absence of the primary antibodies)
yielded negligible values of background uorescence intensity, e.g.
not signicantly higher that the CGN autouorescence in most of
the cases, and always lower than 10% of the average uorescence
intensity per pixel obtained from images of CGN stained with the
appropriate primary and secondary antibodies. Green and red uorescence microscopy images were acquired as indicated also in
Section 2, after optimizing the exposure time for a negligible level
of autouorescence in non-stained CGN and carefully avoiding uorescence quenching by excessive irradiation of stained CGN with
repetitive light ashes of the xenon lamp to avoid photobleaching
of the dyes (Alexa488 and Cy3). The merge uorescence microscopy
images showed the extensive co-localization in the neuronal soma

of CGN of these proteins with caveolin-1 and otillin stained with
the corresponding primary antibodies/secondary Cy3-labeled antibody, but not in neuronal extensions (Fig. 2). In the case of PMCA,
the uorescence microscopy images indicated co-localization not
only in the periphery of the neuronal soma but also in the neuronal
network of connecting extensions. To demonstrate the occurrence
of FRET using anti-PMCA1/4/IgG-Alexa488 as donor and as acceptors anti-caveolin-1/IgG-Cy3 and anti-otillin/IgG-Cy3 we have
performed a detailed quantitative analysis of the green and red
uorescence images acquired when CGN were labeled only with
anti-PMCA1/4/IgG-Alexa488 and also after CGN double-labeling
with anti-PMCA1/4/IgG-Alexa488 and anti-caveolin-1/IgG-Cy3 or
anti-otillin/IgG-Cy3. First, using the ROI tool of the HCImage software of the microscope we selected the neuronal soma in each
uorescence microscopy image, and then we acquired the uorescence intensity for each pixel within each one of the selected
ROI. The pixels distribution of the green uorescence intensity of
the images revealed that double labeling produced a shift of the
whole Gaussian-shaped peak toward lower values of uorescence
intensity in parallel to the increase of the red uorescence intensity of the images, as illustrated in the Supplementary Figure S3.
Next, we obtained the average uorescence intensity per pixel of
the neuronal soma (average of the pixels within the selected ROI) in
green and red uorescence microscopy images, including controls
of CGN stained with only the uorescence donor (Alexa488-labeled
antibodies) in the absence and presence of the corresponding primary anti-PMCA antibodies. After correction for the uorescence
intensity of unspecic labeling plus autouorescence signal, we
have calculated for each double labeling experiment the following parameters: (i) ratio of the average red uorescence intensity
over the green uorescence intensity, ratio red/green uorescence
(ratio R/G) and (ii) average percentage of quenching of the green
uorescence of the donor. The ratio R/G is a quality criterion for
demonstration of FRET occurrence, as it should increase if the
observed green uorescence quenching is due to FRET, while it
will remain steady for photobleaching-induced green uorescence
quenching. Using the complementary criteria indicated in Section
2, in the experimental conditions used in this work photobleachinginduced uorescence quenching was found to be negligible. Under
these conditions, the average percentage of quenching of the
green uorescence of the donor can be taken as the average FRET
efciency. The average efciencies of FRET were high in double labeling experiments with anti-PMCA1/4/IgG-Alexa488 and
anti-caveolin-1/IgG-Cy3 or anti-otillin/IgG-Cy3, as the average
quenching of donor green uorescence were 30 5% and 46 5%,
respectively. The lower FRET-efciency from anti-PMCA1/4/IgGAlexa488 as donor to anti-caveolin-1/IgG-Cy3 indicated a shorter
distance of close approach for a large sub-population of the
donor/acceptor pair anti-PMCA1/4/otillin than for the pair antiPMCA1/4/caveolin-1, and this is consistent with the results of
distribution of these proteins in the different density fractions
shown in Fig. 1. These results were also conrmed using the
pan-primary antibody anti-PMCA (Supplementary Figure S3). As
both the pan-anti-PMCA and anti-caveolin-1 were rabbit antibodies FRET could not be done in this particular case. CGN staining
with pan-anti-PMCA antibody tagged with IgG-Alexa488 yielded
average FRET efciencies of 45 4% to anti-otillin/IgG-Cy3. Taken
together they conrm that PMCA is extensively associated with
lipid rafts sub-microdomains of the plasma membrane of CGN
matured in culture, in good agreement with the results of studies
reported by other laboratories [34,35], and point out a preferential
partition of this protein closer to otillin than to caveolin-1 in the
cytoskeletal network associated with these sub-microdomains.
Fluorescence microscopy imaging experiments also demonstrated the occurrence of FRET using anti-NCX1/IgG-Alexa488 as
113
donor and anti-otillin/IgG-Cy3 as acceptor, with average FRET efciencies of 29 6%, respectively (Fig. 2 and Supplementary Figure
S3). Anti-caveolin-1/IgG-Cy3 could not be experimentally assessed
because both anti-NCX and anti-caveolin-1 antibodies used in
this work were produced in rabbit. These results were also conrmed using a second primary antibody, anti-NCX tagged with
IgG-Alexa488, which yielded FRET efciencies to anti-otillin/IgGCy3 of 32 6%. However, at difference with PMCA, in the case of
NCX the merge images revealed that NCX co-localization with the
lipid rafts marker otillin is more extensive in the neuronal soma
and only in scattered points of the network of neuronal extensions.
3.2. Co-immunoprecipitation with caveolin-1 revealed the

formation of stable protein complexes of calcium transport and
redox systems within these sub-microdomains
Caveolins form a polymeric network associated with stable
lipid rafts sub-microdomains within the cells [19,20]. As indicated
above, caveolin-1 is more specically focalized within these submicrodomains than caveolin-2. In addition, caveolin-1 has been
shown to form complexes with nNOS [40], with Cb5 R [39] and
with other cytoskeleton proteins associated with lipid rafts [19,20].
Therefore, caveolin-1 is the isoform of choice as a biomarker
for the lipid rafts isolated from CGN. The high FRET efciency
observed between labeled caveolin-1 and calcium transport and
redox systems enriched in these submicrodomains pointed out
a close proximity between these proteins and caveolin-1. On
these grounds, we experimentally conrmed that immunoprecipitation of lipid rafts fractions by incubation with matrix-conjugated
rabbit anti-caveolin-1 led to co-immunoprecipitation of these proteins, while no signicant co-immunoprecipitation was observed
in control experiments when lipid rafts fractions were incubated
with matrix-conjugated rabbit IgG (Supplementary Figure S4), as
expected on the basis of the Western blotting results shown in Fig. 1.
Since lipid rafts can be solubilized with methyl--cyclodextrin
(MCD), the possibility that some of them can form stable proteinprotein complexes with caveolin-1 after solubilization of lipid rafts
fractions deserved to be studied. To this end we have experimentally assessed the co-immunoprecipitation with matrix-conjugated
with rabbit anti-caveolin-1 of the above-mentioned calcium transport and redox systems after solubilization of lipid rafts with
25 mM MCD, under experimental conditions were immunoprecipitation of these proteins with matrix-conjugated with rabbit
IgG is negligible (control experiments shown in the Supplementary Figure S4). Fig. 3 shows that NMDAr is almost completely
co-precipitated with anti-caveolin-1, pointing out a tight association between both proteins. In addition, nNOS, Cb5 R and L-VOCC,
and to a lower extent PMCA and NCX, are also immunoprecipitated with anti-caveolin-1, pointing out that a signicant fraction
of these proteins remain also associated with remnant lipid/protein
complexes containing caveolin-1 after solubilization of lipid rafts
with MCD, although in the latter cases most of these proteins
remain in the supernatant. Solubilization of lipid rafts with MCD
should be expected to have a strong effect on the formation of stable
complexes between caveolin-1 and non-integral membrane proteins, such as nNOS and Cb5 R, and with other cytoskeleton proteins
associated with lipid rafts. Indeed, the extent of nNOS and Cb5 R
co-immunoprecipitation with non-solubilized lipid rafts is higher
than 75% (data not shown), and it drops to 27.8% and 11.4% with
MCD-solubilized lipid rafts, respectively (Fig. 3B).
These results were conrmed using matrix-conjugated with
goat anti-NMDAr as the immunoprecipitating agent and matrixconjugated with goat IgG for the control experiments. nNOS, Cb5 R
and L-VOCC were found to co-immunoprecipitate with the remnant
114
Fig. 2. Extensive FRET from PMCA and NCX tagged with IgG-Alexa488 (as uorescence donor) to characteristic lipid rafts proteins (caveolin-1 and otillin) tagged with
IgG-Cy3 antibodies forming an appropriate donor/acceptor pair. (Panel A) Representative quantitative uorescence microscopy images of CGN stained with anti-PMCA1/4
(sc-20028)/IgG-Alexa488 (PMCA1/4*A488, AD) or with anti-PMCA1/4/IgG-Alexa488 and anti-caveolin-1 (sc-894)/IgG-Cy3 (PMCA1/4*A488/Cav1*Cy3, EH). Merged
images of green (GF) and red uorescence (RF) are pasted over a semi-transparent bright eld (BF) mask. Green and red areas display the donor and acceptor uorescence,
respectively, and the orange-yellow areas point out the higher intensity FRET regions (D and H). The exposure time for green uorescence images was 243.1 ms and for
red uorescence images was 261.8 ms. (Panel B) Representative quantitative uorescence microscopy images of CGN stained with anti-PMCA1/4 (sc-20028)/IgG-Alexa488
(PMCA1/4*A488, AD) or with anti-PMCA1/4/IgG-Alexa488 and anti-otillin-1 (sc-16640)/IgG-Cy3 (PMCA1/4*A488/Flot*Cy3, EH). Other experimental details as in Panel A.
(Panel C) Representative quantitative uorescence microscopy images of CGN stained with anti-NCX1 (sc-30304-R)/IgG-Alexa488 (NCX1*A488, AD) or with anti-NCX1/IgGAlexa488 and anti-otillin-1 (sc-16640)/IgG-Cy3 (NCX1*A488/Flot*Cy3, EH). Other experimental details as in Panel A. (Panel D) Ratio of red/green uorescence obtained
115
Fig. 3. Co-immunoprecipitation with caveolin-1 after solubilization of CGNs lipid rafts with 25 mM MCD. The experimental protocol is indicated in detail in Section 2.
(Panel A) Western blotting of the proteins precipitated with matrix/anti-caveolin-1 complexes showed the co-immunoprecipitation with caveolin-1 of L-VOCC, NMDAr,
PMCA, NCX, nNOS and Cb5 R. The presence of these proteins in the matrix/anti-caveolin-1 precipitates (M) and in the supernatants (S) was revealed using the antibodies
listed in the legend for Fig. 1. Supernatants were rst concentrated by centrifugation using Amicon lters with a cutoff molecular weight of 10 kDa and an aliquot of the
supernatants was loaded in the gels. The approximate molecular weights of the immunoreactive bands against each one of these antibodies are indicated in the right-hand
side of the gure. (Panel B) Partition of the proteins co-immunoprecipitated with caveolin-1 between precipitated matrix and supernatant. The results are expressed as
percentage of the total amount of each protein found in the total volumes of precipitated matrix (IP) and supernatants.
protein matrix solubilized with MCD precipitated with antiNMDAr (Supplementary Figure S5).
3.3. l-Glutamate-induced increase of cytosolic calcium revealed
transient calcium microcompartments near the plasma
membrane of CGN
It is well known that the application of 50100 M l-glutamate
to mature CGN in culture raises cytosolic calcium, largely through
activation of NMDAr as this increase is nearly completely blocked
by NMDAr inhibitors [7,47]. Fig. 4A is a representative image to
show the occurrence of population heterogeneity in the kinetics
of ratio 340/380 increase in the neuronal soma in CGN neurons
in culture. The presence of subpopulations of mature CGN neurons at 89 DIV with a different rate of calcium rise in response to
l-glutamate is an expected result, because the amplitude and kinetics of miniature NMDA-excitatory postsynaptic currents become
larger and faster several days later in CGN cultures, i.e. up to
14 DIV, in parallel with the shift from NR2B to NR2A-containing
NMDAr complexes [55,56]. This fact also leads to a relatively slow
kinetics of the rise of the ratio 340/380 when averaging the data
obtained for >400 neuronal soma, data shown in Fig. 4B. A close
inspection of the images of ratio 340/380 increase at short times

after addition of l-glutamate revealed the appearance of a scattered
distribution of pixels with ratio 340/380 values higher than the
average concentrated near the plasma membrane of the neuronal
soma (Fig. 5). To further highlight this fact we performed image
analysis with the HCI software of the uorescence microscope by
sequential rise of the threshold intensity while maintaining the
amplication gain of the ratio signal, e.g. images displaying only
the pixels of the acquired image with an intensity higher than the
pre-xed threshold value, and representative images of selected
neuronal soma are shown in the Supplementary Figure S6 and
Fig. 5C. As shown in the Supplementary Figure S6 the increase
of the threshold value helps to visually localize the occurrence
and regional distribution within the neuronal soma of the pixels
displaying higher intensity of uorescence with 340 nm excitation lter, without affecting their distribution pattern. As expected,
100 M l-glutamate elicited a large and time-dependent increase
in the total number of pixels within the neuronal somas with high
intensity of uorescence (red colored pixels). The observation of
images captured with a 4 s interval revealed a steadily high frequency of the higher ratio 340/380 and uorescence intensity pixels
scattered near the periphery of the neuronal soma, i.e. close to the
from the analysis of uorescence intensity data for CGN stained with anti-PMCA1/4/IgG-alexa488 only or with anti-NCX1/IgG-Cy3 only (First Antibody*A488, control
experiments), and double stained with anti-PMCA1/4/IgG-Alexa488//anti-caveolin-1/IgG-Cy3 (PMCA1/4*A488/Cav1*Cy3), anti-PMCA1/4/IgG-Alexa488//anti-otillin/IgGCy3 (PMCA1/4*A488/Flot*Cy3) and anti-NCX1/IgG-Alexa488//anti-otillin/IgG-Cy3 (NCX1*A488/Flot*Cy3). The results shown in this panel D are the mean S.E. evaluated
from the uorescence intensity readings of the following number of neuronal soma of at least three different CGN preparations in each case: 4305 (PMCA1/4*A488), 1813
(NCX1*A488), 2113 (PMCA1/4*A488/Cav1*Cy3), 2192 (PMCA1/4*A488/Flot*Cy3) and 1953 (NCX1*A488/Flot*Cy3). *p < 0.05, i.e. statistically signicant, with respect to the
control (CGN labeled with the donor only).
116
Fig. 4. Kinetics of cytosolic calcium increase in the neuronal soma after l-glutamate application. Mature CGN in culture (910 DIV) were loaded with fura-2 as indicated in
Section 2 and then changed to Lockes K25 buffer (37 C) and treated with 100 M l-glutamate. (Panel A) Representative ratio images showing populational heterogeneity
in the kinetics of cytosolic calcium increase in the neuronal soma after l-glutamate in CGN cultures. (Panel B) Population average of the l-glutamate-induced kinetics of
cytosolic calcium increase in the neuronal soma. Analysis of CGN response to l-glutamate was done in the presence (open circles) or absence (closed squares) of 10 M
MK-801, a selective non-competitive NMDAr antagonist. Data acquisition and analysis was done after selection of the neuronal soma using the region of interest (ROI) tool of
this software as indicated in Section 2, with exposure times lower than 0.4 s. The ratio 340/380 values shown are the average S.E. of experiments done with three different
preparations of CGN (n > 400 neuronal soma of elds taken from 6 plates).
Fig. 5. l-Glutamate evoked dot-like cytosolic calcium concentration peaks near the plasma membrane of CGN. (Panel A) Representative ratio 340/380 images of fura-2-loaded
CGN captured at the times post-application of l-glutamate indicated in the left side (in seconds). (Panel B) Processing of the ratio 340/380 image with the enhance contrast
tool of the Image J software suggested focalized uctuations of cytosolic calcium concentration within individual neuronal somas after application of l-glutamate. (Panel C)
Pseudocolor images obtained with the 340 nm excitation lter (pseudocolor 340 nm/510 nm uorescence microscopy images) processed using the HCImage software of the
uorescence microscope raising the threshold intensity level to 600 to view only the pixels of higher intensity, i.e. pixels with an intensity >70% of the maximum intensity
reading (850), better highlighted the occurrence of focal points within the neuronal soma displaying higher calcium levels. (Panel D) Results of pseudocolor 340 nm/510 nm
uorescence microscopy image averaging. (Panel D1) Image showing the result of the average of the rst 48 pseudocolor images captured at intervals of 4 s after the addition
of l-glutamate with the 340 nm excitation lter and processed with a threshold intensity level of 600. (Panel D2) Surface plots of the uorescence intensity of red plus green
pixels (A) and solely of red pixels (B) generated by Image J software of the neuronal soma marked with arrows in the average image shown in panel D1.
117
Fig. 6. Blockade of the response of NMDAr to l-glutamate by MCD. (Panel A) Representative ratio images (340/380) showing that 10 mM MCD elicits a blockade of the
increase of ratio (340/380) after addition of 100 M of l-glutamate as potent as that induced by 10 M MK-801. (Panel B) Population average of the effect of millimolar
concentrations of MCD on l-glutamate-induced kinetics of cytosolic calcium increase in the neuronal soma. The results obtained with MK-801 (see Fig. 4) are also included
in this gure for a direct comparison with those obtained with MCD. The ratio 340/380 values shown are the average S.E. of experiments done by triplicate with at least
two different preparations of CGN (n > 400 neuronal soma of elds taken from at least 6 plates). See Fig. 4 for further experimental details. (Panel C) The cell viability of
CGN is not affected upon incubation during 30 min with up to 20 mM MCD. Cell viability was measured as indicated in Section 2. The data shown are the average S.E. of
experiments done by triplicate with two different preparations of CGN. *p < 0.05, i.e. statistically signicant with respect to the control.
plasma membrane, particularly at early times after the addition of

l-glutamate (see Supplementary Video 1). Next, using the Image J
software we performed accumulation and averaging of sequentially
captured images (Fig. 5, panel D and data not shown), a standard
procedure to increase the signal to noise ratio for weak signals. To
obtain images with a better spatial resolution the pseudocolor RGB
images captured with excitation lter 340/emission lter 510 nm
(pseudocolor 340 nm/510 nm images), i.e. experimental conditions
corresponding to the uorescence peak of the fura-2:Ca2+ complex, instead of the ratio images were used. Averaging and analysis
of the sequential images captured after addition of l-glutamate
using the Image J software pointed out that: (i) background intensity maxima were eliminated from the averages of >32 images
using a tolerance level 5%, indicating that this was the threshold level of noise in the pixels intensity, (ii) the ratio between
the number of intensity maxima reported by the Image J software
increases after addition of l-glutamate and, more relevant in statistical terms, the increase is stronger for the number of intensity
maxima obtained with a tolerance level 15%, and (iii) the number
of intensity maxima with a tolerance level 15% and 20% remained
constant when 32, 48 and 64 sequential images were averaged (data
not shown). The surface plots of uorescence intensity within representative CGN generated with Image J software from averages
of 48 pseudocolor (340 nm/510 nm) images sequentially captured
with a 4 s interval are shown in Fig. 5, panel D. These surface plots
monitored the presence of dynamic high-calcium compartments
of an average size of only a few pixels localized near the plasma

membrane of the neuronal soma. As each pixel has a dimension of
0.2 m 0.2 m, the surface plots highlighted the occurrence of a
large network of dynamic high calcium microcompartments localized at less than 1 m from the plasma membrane, with the ability
to generate a steady high calcium ring forming a crown-like contour
and focalized peaks of calcium of a width of only 23 pixels, i.e. of
a width lower than 0.6 m. A simple hypothesis to account for this
result is that the major calcium extrusion systems of the neuronal
plasma membrane are located near the major calcium entry points
(NMDAr), such that they can account for localized high calcium
transients and attenuate the spreading of calcium ions through
the cytosol. Therefore, we experimentally addressed this point by
FRET-imaging using specic antibodies for the major extrusion systems of the neuronal plasma membrane, PMCA and NCX, which are
expressed in mature CGN in culture [57,58].
The treatment of CGN with the lipid rafts solubilizing agent
MCD strongly impairs the cytosolic calcium rise after addition of
100 M l-glutamate before a signicant loss of cell viability could
be noticed, as shown in Fig. 6. The results pointed out that incubation of CGN during 10 min with millimolar concentrations of MCD,
from 2.5 to 10 mM, elicits an almost complete blockade of the
ratio increase after addition of l-glutamate, close to the blockade
observed in the presence of the NMDAr inhibitor MK-801. Moreover, the density and intensity of pixels displaying ratio 340/380
values well above the average in the periphery of the neuronal
118
Fig. 7. Extensive FRET between NMDAr and PMCA and between NMDAr and NCX tagged with antibodies forming an appropriate donor/acceptor pair. (Panel A) Representative quantitative uorescence microscopy images of CGN stained with anti-NMDAr (sc-1468)/IgG-Alexa488 (NMDAr*A488, AD) or with anti-NMDAr/IgG-Alexa488
and pan-anti-PMCA (sc-28765)/IgG-Cy3 (NMDAr*A488/PMCA*Cy3, EH). Merged images of green (GF) and red uorescence (RF) are pasted over a semi-transparent bright
eld (BF) mask. Green and red areas display the donor and acceptor uorescence, respectively, and the orange-yellow areas point out the higher intensity FRET regions (D
and H). The exposure time for green uorescence images was 243.1 ms and for red uorescence images was 261.8 ms. (Panel B) Representative quantitative uorescence
microscopy images of CGN stained with anti-NMDAr (sc-1468)/IgG-Alexa488 (NMDAr*A488, AD) or with anti-NMDAr/IgG-Alexa488 and anti-NCX (sc-32881)/IgG-Cy3
(NMDAr*A488/NCX*Cy3, EH). Other experimental details as in Panel A. (Panel C) Distribution of the uorescence intensity of pixels of 1229 neuronal somas selected from
uorescence microscopy images like those shown in the Panel A obtained with three different CGN preparations. Distribution of pixels versus uorescence intensity in
green uorescence images and red uorescence images. (Panel D) Ratio of red/green uorescence obtained from the analysis of uorescence intensity data for CGN stained with
somas, i.e. near the plasma membrane, are largely lowered in CGN
incubated with MCD with respect to their values in non-treated
CGN (control experiment).
3.4. FRET revealed that the major calcium extrusion systems of
the plasma membrane of CGN, PMCA and NCX, are located near
NMDAr
Representative uorescence microscopy images of CGN double
stained with anti-NMDAr/IgG-Alexa488 (anti-NMDAr*A488) and
anti-PMCA/IgG-Cy3 (anti-PMCA*Cy3) or anti-NCX/IgG-Cy3 (antiNCX*Cy3) are presented in Fig. 7. Both direct inspection of merge
images and pixels-intensity analysis within neuronal soma demonstrated the occurrence of extensive FRET in the neuronal soma, but
not in the network of neuronal extensions, between Alexa488- and
Cy3-tagged antibodies (see also Supplementary Figure S7). As indicated before, the detailed analysis of the uorescence microscopy
images was performed within the neuronal somas, after their selection with the ROI tool of the HCImage software of the microscope.
Therefore, the conclusions derived from this analysis are applicable
in this case only to the neuronal soma.
The widening and in some cases also the distortion of the
Gaussian shaped distribution of the intensity of uorescence
pixels observed in experiments of double-labeling with antibodies, pointed out the occurrence of a diffuse and scattered
distribution of PMCA and NCX around NMDAr. The average
efciencies of FRET for these pairs estimated from the quenching of donor green uorescence are also high, 38 11% for the
anti-NMDAr*A488/anti-PMCA*Cy3 pair and 39 6% for the antiNMDAr*A488/anti-NCX*Cy3 pair, pointing out that they are within
the range of FRET-distance using labeled secondary antibodies. The
results were further conrmed using a second primary antibody for
PMCA (anti-PMCA1/4) and NCX (anti-NCX1), which yielded average FRET-efciencies of 32 8% for anti-PMCA1/4 and 19 6% for
anti-NCX1. For the case of anti-NMDAr/anti-PMCA1/4 pair it was
also observed a large distortion of the Gaussian-shaped distribution of the intensity of uorescence pixels. However, these FRET
efciencies are only slightly lower than those obtained for the
FRET pairs used to highlight the proximity between NMDAr and
L-VOCC, 4964% [33], and this indicated that both PMCA and NCX
are separated from NMDAr only a few nanometers more than LVOCC, although the distribution of PMCA and NCX around NMDAr
appears to be more heterogeneous, with only a fractional part of
these proteins in the proximity of the NMDAr. Noteworthy, in
these cases the efciency of FRET from anti-PMCA/IgG-Alexa488
and from anti-NCX/IgG-Alexa488 to the uorescent ligand of LVOCC ST-bodipy-dyhydropyridine are very weak to be considered
signicant, e.g. 5% (data not shown). Thus, PMCA and NCX are separated by a distance 50 nm from L-VOCC, as this is the maximum
distance for FRET detection using this latter approach.
On the other hand, as no signicant FRET was found between
anti-PMCA/IgG-Alexa488 and anti-NCX or anti-NCX1/IgG-Cy3
(data not shown), we can conclude that PMCA and NCX are not vicinal proteins within the lipid rafts-associated sub-microdomains,
as they must be separated by more than 80 nm.
3.5. FRET also revealed that nNOS is located near Cb5 R
Earlier, we have demonstrated that Cb5 R is in the vicinity of L-VOCC following FRET-imaging approaches like the ones
119
used in this work [27,53]. Moreover, Cb5 R associated with lipid

rafts microdomains of mature CGN in culture can account for a
large fraction of superoxide anion production near this subcellular compartment [38], which is an early event playing a major
role in CGN apoptosis induced by extracellular K+ deprivation
[10,39,43].
Because of the implications of the proximity to nNOS of a
superoxide producing system for the focalized generation of peroxynitrite, a short-lived and highly neurotoxic cellular strong
oxidant reported to mediate neurodegeneration in ischemiareperfusion, inammation and some neurodegenerative diseases
[5964], we have studied whether nNOS and Cb5 R are within
the FRET-distance range explored by a combination of primary
and uorescent secondary antibodies (Fig. 8). Both direct inspection of merge images and pixels-intensity analysis demonstrated
the occurrence of extensive FRET between Alexa488- and Cy3tagged IgG antibodies in the neuronal soma, but not in the network
of neuronal extensions. The average efciency of FRET for antinNOS/IgG-Alexa488//anti-Cb5 R/IgG-Cy3, i.e. 32 10%, was lower
than that obtained for the FRET pairs used to highlight the proximity between NMDAr and L-VOCC, but close to that found for
the pair NMDAr/PMCA labeled with Alexa488- and Cy3-tagged IgG
antibodies, see above. In addition, green uorescence quenching
did not lead to signicant distortions of the Gaussian shape of the
intensity distribution of green uorescence pixels in the images
analyzed, and the intensity distribution of pixels of red uorescence images analyzed were also single Gaussians. Thus, our results
support that sub-microdomains associated with lipid rafts of the
plasma membrane can generate focalized peroxynitrite oxidative stress upon stimulation of superoxide anion production by
Cb5 R.
4. Discussion
In previous studies we have shown that major calcium transport systems channeling calcium entry from the extracellular
medium into the cytosol of CGN, like NMDAr and L-VOCC, and
enzyme generating ROS/RNS, like Cb5 R and nNOS, are associated
with caveolin-rich lipid rafts sub-microdomains of the plasma
membrane [27,33,38,39,53]. The analysis of isolated lipid raftsenriched membrane fractions done in this work demonstrated
that also the major calcium extrusion systems of the neuronal
plasma membrane (PMCA and NCX) are associated with these submicrodomains in primary cultures of mature CGN. Our results are
consistent with previous reports showing that PMCA is associated
with lipid rafts of synaptic plasma membranes [34] and of primary cortical and hippocampal rat neurons in culture [35], and also
with NCX binding to lipid rafts in coronary artery smooth muscle
preparations [36].
Moreover, the results of the immunoprecipitation experiments
showed that all these calcium transport and redox systems were
strongly anchored to the caveolin-1 protein network, as all of them
were found in the precipitates of matrix/anti-caveolin-1 complexes
after solubilization of lipid rafts with MCD. The quantitative analysis of the co-immunoprecipitation data pointed out that all NMDAr
present in CGN lysates was co-precipitated with caveolin-1, as well
as a large fraction of total nNOS. Noteworthy, several of these proteins have been shown to form complexes with caveolin-1, e.g.
nNOS [40], the L-VOCC subunit 22 expressed in the cerebellum [65] and Cb5 R [39], and NCX from vascular endothelial cells,
anti-NMDAr/IgG-alexa488
only
(NMDAr*A488,
control
experiments),
and
double
stained
with
anti-NMDAr/IgG-Alexa488//pan-anti-PMCA/IgG-Cy3
(NMDAr*A488/PMCA*Cy3) and anti-NMDAr/IgG-Alexa488//anti-NCX/IgG-Cy3 (NMDAr*A488/NCX*Cy3). The results shown in this panel D are the mean S.E. evaluated from the uorescence intensity readings of the following number of neuronal soma of at least three different CGN preparations in each case: 3955 (NMDAr*A488), 1229
(NMDAr*A488/PMCA*Cy3) and 2726 (NMDAr*A488/NCX*Cy3). *p < 0.05, i.e. statistically signicant with respect to the control (CGN labeled with the donor only).
120
Fig. 8. FRET between nNOS and Cb5 R tagged with antibodies forming an appropriate donor/acceptor pair. (Panel A) Representative quantitative uorescence microscopy
images of CGN stained with anti-nNOS (sc-5302)/IgG-Alexa488 (nNOS*A488, AD) or with anti-nNOS/IgG-Alexa488 and anti-Cb5 R (Protein Tech 10894-1-AP)/IgG-Cy3
(nNOS*A488/Cb5 R*Cy3, EH). Merged images of green (GF) and red uorescence (RF) are pasted over a semi-transparent bright eld (BF) mask. Green and red areas display the
donor and acceptor uorescence, respectively, and the orange-yellow areas point out the higher intensity FRET regions (D and H). The exposure time for green uorescence
images was 205.7 ms and for red uorescence images was 243.1 ms. (Panel B) Distribution of the uorescence intensity of pixels of 1314 neuronal somas selected from
uorescence microscopy images like those shown in the Panel A obtained with three different CGN preparations. Distribution of pixels versus uorescence intensity in green
uorescence images and red uorescence images. (Panel C) Ratio of red/green uorescence obtained from the analysis of uorescence intensity data. The results shown in
panels C and D are the mean S.E. evaluated from the uorescence intensity readings of at least 20 different eld frames. *p < 0.05, i.e. statistically signicant with respect
to the control (CGN labeled with the donor only).
which was detected in caveolin-rich membrane fractions containing caveolin-1 [66], has been shown to interact with caveolin-3
[67]. Therefore, our results led to the conclusion that the treatment
with MCD elicited a partial dissociation of these complexes, but
it was not able to dissociate NMDAr from the caveolin-1 protein
network. Thus, NMDAr is anchored to lipid rafts-associated submicrodomains by stronger protein-protein interactions than nNOS,
L-VOCC, Cb5 R and NCX. Moreover, the co-precipitation of L-VOCC
and nNOS with matrix/anti-NMDAr complexes lend additional support to a major conclusion attained in our previous FRET study,
namely, that NMDAr, L-VOCC and nNOS are vicinal proteins within
lipid rafts-associated nanodomains in mature CGN [33].
On these grounds, using staining with uorescent antibodies
forming appropriate FRET donoracceptor pairs we have analyzed
whether the major systems for calcium extrusion from the neuronal
cytosol, PMCA and NCX, also co-localized within caveolin-1-rich
nanodomains containing the major systems for calcium entry from
the extracellular medium in mature CGN, NMDAr and L-VOCC.
Noteworthy, our results pointed out an extensive co-localization
of PMCA and NCX with caveolin-1 in the neuronal soma, but
not in neuronal extensions. Thus, our results suggest that these

nanodomains are more enriched in extrasynaptic NMDAr than in
synaptic NMDAr.
We have set 10% FRET efciency as the minimum level for statistically signicant FRET, on the basis of the size of the error
bars obtained in our measurements. Having in mind the volume
and geometry of IgG [68], the value of the distance for 50% FRET
efciency of the donor/acceptor pairs used herein (R0 values),
between 5 and 6 nm [50], and the fact that FRET efciency is lower
than 10% for a separation distance of 2 R0 in the case of a single
donor/acceptor pair [48,49] and also in the case of a highly packed
and close to hexagonal-centered geometry of multiple acceptors
per donor [49,51], we can conclude that the absence of statistically signicant FRET implies a separation distance higher than
80 nm when using double staining with antibodies to generate the
donor/acceptor pair and higher than 50 nm when only one of the
proteins of interest is stained using the primary and labeled secondary antibody complex. Taking into account the molecular sizes
of the plasma membrane calcium transport systems studied herein,
e.g. NMDAr and L-VOCC (entry systems) and PMCA and NCX (extrusion systems), their location within domains of a size lower than
100 nm in a rigid lipid environment (lipid rafts) strongly suggest
that these nanodomains operate as a calcium microchip-like structure of the plasma membrane, likely devised for a tight control of
neuronal excitability by calcium signaling.
Because of the high diffusion coefcient of calcium ions in the
cytosol, 300 m2 s1 [3], the location of the major calcium extrusion systems of the neuronal plasma membrane within 100 nm
distance from the points of calcium entry is not needed on purely
theoretical basis for a ne control of the steady-state concentration of bulk cytosolic calcium. On the contrary, this close location
between calcium entry and calcium extrusion systems is a need to
generate local calcium gradients, i.e. cytosolic calcium microcompartments, near the plasma membrane. Indeed, the occurrence of
these microcompartments has been highlighted and demonstrated
in this work by ratiometric uorescence microscopy images of CGN
loaded with fura-2 taking at short times after addition of l-Glu.
On these grounds, we wish to note that our calculated distance
for NMDAr and L-VOCC separation within these nanodomains is
20 nm [33]. As shown in Parekh [3], at this short distance L-VOCC
activation can elicit transients of micromolar local Ca2+ concentration nearby NMDAr. The strong attenuation by MCD of the
cytosolic calcium increase induced by l-glutamate through NMDAr
pointed out that the association of NMDAr with these lipid rafts
nanodomains bears relevant functional consequences. Let us recall
here that calmodulin-dependent protein kinase II (CaMKII) has
been shown to interact with L-VOCC subunit 2a and with NMDA
receptors subunit NR2B [69], implying that this protein is present
in neuronal nanodomains associated with these lipid rafts (a point
that we have conrmed experimentally, data not shown), and the
synaptic clustering of -amino-3-hydroxy-5-methylisoxazole-4propionic acid (AMPA) receptors upon phosphorylation by CaMKII
has been shown to potentiate NMDAr activation in the induction of long-term post-synaptic potentiation [70]. On the other
hand, Cav-1/ knockout mice show premature neuronal aging
and degeneration [71] and an impaired induction of mGluRdependent long-term depression at CA3-CA1 synapses [72]. In
addition, caveolin-1 knockdown by siRNA has led to the conclusion
that caveolin-1 expression is essential for NMDAr-mediated Src and
ERK-1/2 activation [21]. Thus, the vicinal location of L-VOCC and
NMDAr within lipid rafts-associated nanodomains plays a facilitating role for lowering the threshold for neuronal excitability by l-Glu
and for triggering excitotoxic cascades induced by l-Glu. Indeed,
synergism between activation of L-VOCC and NMDAr has been
reported in short-term nociceptive plasticity in the dorsal horn of
spinal cord [73] and in frog retina [74].
121
As illustrated in [3], activation of a typical 2.6 pS conductance

channel can generate a calcium microdomain with calcium concentrations higher than 1 M extending up to 100 nm with low
calcium buffering capacity, as the mean path length increases from
several nm up to 82 nm in the presence of 0.1 mM of the calcium chelator fura-2. Because the conductance of NMDAr in CGN
is >30 pS [75], the theoretical approach developed by Neher [76]
allows to conclude that upon activation of the NMDAr this gradient will extend up to more than 200 nm distance from the calcium
pore of this receptor in the case of micromolar calcium buffering
capacity in these microdomains. Noteworthy, the maximum size
accepted for lipid rafts is 200 nm [18]. Our FRET imaging results
led to the conclusion that PMCA and NCX must be separated more
than 50 nm away from L-VOCC, as non-signicant FRET efciency
was observed between the uorescent L-VOCC ligand ST-bodipy
dihydropyridine and IgG-labeled PMCA or NCX. In addition, NCX
is at least 80 nm apart from L-VOCC and PMCA, because the FRET
efciency between these proteins labeled with antibodies forming
a donor/acceptor pair was not found to be statistically signicant.
Taken together, all FRET results obtained in this and previous works
of our laboratory lead to the conclusion that PMCA and NCX must
be located in the periphery of lipid rafts-associated nanodomains.
As peripheral proteins, they should be expected to be less tightly
bound to these nanodomains and also more rapidly exchangeable
with the rest of the plasma membrane. Thus, the low percentage
of PMCA and NCX immunoprecipitated with matrix/anti-caveolin1 complexes and the distorted Gaussian distributions observed
for FRET pairs labeling NMDAr and PMCA or NCX provide additional experimental support to the peripheral location of PMCA
and NCX in these nanodomains. The location of PMCA and NCX
close to the limits of these sub-microdomains is equivalent to introduce an immobile calcium buffering capacity near the periphery of
the volume element dening a microcompartment in the diffusion
equations shaping the spatial prole of the Ca2+ signal in the theoretical approach derived by Neher [76]. Therefore, experimental
conditions yielding to a steady-state level of activation of the major
calcium entry systems of the lipid-rafts associated nanodomains of
CGN, e.g. NMDAr activation by l-glutamate and L-VOCC activation
by plasma membrane depolarization, are the most suited for monitoring these dynamic high calcium sub-microcompartments. The
image analysis done in this work with fura-2 loaded CGN revealed
that under these experimental conditions it can be observed a
large heterogeneity of cytosolic calcium concentration in the neuronal soma, with the highest steady-state calcium concentrations
reached in discrete volume elements of a size lower than 0.6 m
forming a dot-like ring close to the plasma membrane.
Our results also pointed out the presence within these submicrodomains of redox systems capable of generating ROS, namely
nNOS and Cb5 R. As noted in a previous work, the presence of Cb5 R in
these sub-microdomains suggests that they can be, at least in part,
derived from endoplasmic reticulum/plasma membrane junctions
[39]. Owing to the dependence on free Ca2+ concentration of NO
production by nNOS, EC50 0.20.4 M [77], a vicinal sub-cellular
location is the simplest way to allow for a fast and more potent
activation of this enzyme upon NMDAr and/or L-VOCC activation,
as noted in [33]. Yet, the results pointed out that nNOS and Cb5 R are
within the FRET distance limit when using primary antibodies and
IgG-labeled with dyes forming a donor/acceptor pair as secondary
antibodies. Therefore, both proteins are separated on average less
than 80 nm. Co-immunoprecipitation of nNOS with matrix/antiCb5 R complexes after treatment with methyl--cyclodextrin gave
additional experimental support to the strong anchorage of both
proteins within the same protein aggregates associated with lipid
rafts of the plasma membrane.
Deregulation of Cb5 R associated with the plasma membrane
of CGN can lead to superoxide anion in the early stage of
122
induction of neuronal apoptosis [38,39]. As NO rapidly reacts with

superoxide anion to produce peroxynitrite, a diffusion-controlled
reaction having a bimolecular rate constant of 6.7 109 M1 s1
[42], the co-localization of nNOS and Cb5 R within these submicrodomains allows to envisage them as focal points of cellular
production of the harmful oxidant peroxynitrite. Indeed, it has
been shown by many studies performed along last decade that
peroxynitrite is a strong neurotoxic oxidant producing cell death
in neuronal cultures in vitro [6,13,14] and also as a major
agent in oxidative stress-induced neurodegeneration in brain
ischemia/reperfusion [59], inammation [60,62] and neurodegenerative diseases [61,63,64,78,79]. Noteworthy, all the major
calcium transport systems of the neuronal plasma membrane studied in this work (NMDAr, L-VOCC, PMCA and NCX) have been shown
to be molecular targets for ROS overproduced during the progress
of these neurodegenerative processes in brain and peripheral neuronal structures, reviewed in [8,9]. Moreover, a sustained alteration
of neuronal cytosolic calcium homeostasis is also observed since the
early stages of neuronal apoptosis [4,5] and neurodegenerative disorders [8,9,80,81]. As a sustained and widespread rise of cytosolic
calcium is very harmful for neurons, a focalized three dimensionalstructure in the neuronal plasma membrane for redox regulation
of the calcium transport systems more relevant for the control of
cytosolic calcium homeostasis afford many advantages for the rapid
ne and delicate tuning between neuronal excitability and survival. Because of the well recognized and relevant role of respiring
mitochondria in ROS production in neurons and other mammalian
cells [60,82,83], neuronal life is strongly threatened by metabolic
hyperactivity, as exemplied by glutamate excitotoxicity. A focalized redox/calcium integrative structure at the points initiating the
calcium signaling pathway is probably the more efcient strategy to guarantee a faster and modulated response for cells like
neurons that have to deliver integrative responses to many possible simultaneous incoming signals from neighboring cells, without
dying by metabolic hyperactivity. However, as actin polymerization and actin laments are distorted by exposure to peroxynitrite
[47,84,85] cytoskeleton-linked lipid rafts structures are likely to be
highly sensitive to RNS-mediated oxidative stress. Whether these
redox/calcium molecular microchips of the plasma membrane are
altered in neurons more prone to undergo degeneration is a challenging possibility that deserve to be further explored.
Acknowledgements
This work has been funded by Grant BFU2011-30178 of the
Spanish Ministerio de Ciencia e Innovacin and GR10092 of the
Gobierno de Extremadura, with FEDER co-nancing. DMS has
been a Predoctoral Fellow of the Spanish Ministerio de Ciencia
e Innovacin. Maria I. Piedehierro, our laboratory technician, is
acknowledged for her help with CGN preparations and western
blotting experiments.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.ceca.2014.06.002.
References
[1] M.J. Berridge, Neuronal calcium signaling, Neuron 21 (1998) 1326.
[2] M.J. Berridge, P. Lipp, M.D. Bootman, The versatility and universality of calcium
signalling, Nat. Rev. Mol. Cell Biol. 1 (2000) 1121.
[3] A.B. Parekh, Ca2+ microdomains near plasma membrane Ca2+ channels: impact
on cell function, J. Physiol. 586 (2008) 30433054.
[4] J.L. Franklin, E.M. Johnson Jr., Suppression of programmed neuronal death
by sustained elevation of cytoplasmic calcium, Trends Neurosci. 15 (1992)
501508.
[5] J.L. Franklin, E.M. Johnson Jr., Block of neuronal apoptosis by a sustained
increase of steady-state free Ca2+ concentration, Philos. Trans. R. Soc. Lond.
B: Biol. Sci. 345 (1994) 251256.
[6] Y. Gutierrez-Martin, F.J. Martin-Romero, F. Henao, C. Gutierrez-Merino, Alteration of cytosolic free calcium homeostasis by SIN-1: high sensitivity of L-type
Ca2+ channels to extracellular oxidative/nitrosative stress in cerebellar granule
cells, J. Neurochem. 92 (2005) 973989.
[7] M.A. Garcia-Bereguiain, A. Samhan-Arias, F.J. Martin-Romero, C. GutierrezMerino, Hydrogen sulde raises cytosolic calcium in neurons through
activation of L-type Ca2+ channels, Antioxid. Redox Signal. 10 (2008) 3142.
[8] C. Hidalgo, P. Donoso, Crosstalk between calcium and redox signalling: from
molecular mechanisms to health implications, Antioxid. Redox Signal. 10
(2008) 12751312.
[9] C. Gutierrez-Merino, Redox modulation of neuronal calcium homeostasis
and its deregulation by reactive oxygen species, in: C. Gutierrez-Merino, C.
Leeuwenburgh (Eds.), Free Radicals in Biology and Medicine, 2008, Research
Signpost, Kerala, 2008, pp. 67101.
[10] F.J. Martin-Romero, E. Garcia-Martin, C. Gutierrez-Merino, Inhibition of the
oxidative stress produced by plasma membrane NADH oxidase delays lowpotassium induced apoptosis of cerebellar granule cells, J. Neurochem. 82
(2002) 705715.
[11] R. Balazs, O.S. Jorgensen, N. Hack, N-methyl-d-aspartate promotes the survival
of cerebellar granule cells in culture, Neuroscience 27 (1988) 437451.
[12] E.C. Toescu, Activity of voltage-operated calcium channels in rat cerebellar
granule neurons and neuronal survival, Neuroscience 94 (1999) 561570.
[13] H. Ischiropoulos, D. Duran, J. Horwitz, Peroxynitrite-mediated inhibition of
DOPA synthesis in PC12 cells, J. Neurochem. 65 (1995) 23662372.
[14] J.L. Trackey, T.F. Uliasz, S.J. Hewett, SIN-1-induced cytotoxicity in mixed cortical
cell culture: peroxynitrite-dependent and -independent induction of excitotoxic cell death, J. Neurochem. 79 (2001) 445455.
[15] G. Rumbaugh, S. Vicini, Distinct synaptic and extrasynaptic NMDA receptors in
developing cerebellar granule neurons, J. Neurosci. 19 (1999) 1060310610.
[16] R.S. Petralia, Y.X. Wang, F. Hua, Z. Yi, A. Zhou, L. Ge, F.A. Stephenson, R.J.
Wenthold, Organization of NMDA receptors at extrasynaptic locations, Neuroscience 167 (2010) 6887.
[17] S.T. Li, J.G. Ju, Functional roles of synaptic and extrasynaptic NMDA receptors in
physiological and pathological neuronal activities, Curr. Drug Targets 13 (2012)
207221.
[18] L.J. Pike, Rafts dened: a report on the keystone symposium on lipid rafts and
cell function, J. Lipid Res. 47 (2006) 15971598.
[19] K.M.M. O Connell, J.R. Martens, M.M. Tamkun, Localization of ion channels to
lipid raft domains within the cardiovascular system, Trends Cardiovasc. Med.
14 (2004) 3742.
[20] B.P. Head, P.A. Insel, Do caveolins regulate cells by actions outside of caveolae?
Trends Cell Biol. 17 (2007) 5157.
[21] B.P. Head, H.H. Patel, Y.M. Tsutsumi, Y. Hu, T. Mejia, R.C. Mora, P.A. Insel,
D.M. Roth, J.C. Drummond, P.M. Patel, Caveolin-1 expression is essential for Nmethyl-d-aspartate receptor-mediated Src and extracellular signal-regulated
kinase 1/2 activation and protection of primary neurons from ischemic cell
death, FASEB J. 22 (2008) 828840.
[22] B.P. Head, Y. Hu, J.C. Finley, M.D. Saldana, J.A. Bonds, A. Miyanohara, I.R.
Niesman, S.S. Ali, F. Murray, P.A. Insel, D.M. Roth, H.H. Patel, P.M. Patel, Neurontargeted caveolin-1 protein enhances signaling and promotes arborization of
primary neurons, J. Biol. Chem. 286 (2011) 3331033321.
[23] M. Toselli, G. Biella, V. Taglietti, E. Cazzaniga, M. Parenti, Caveolin-1 expression
and membrane cholesterol content modulate N-type calcium channel activity
in NG108-15 cells, Biophys. J. 89 (2005) 24432457.
[24] T. Murata, M.I. Lin, R.V. Stan, P.M. Bauer, J. Yu, W.C. Sessa, Genetic evidence
supporting caveolae microdomain regulation of calcium entry in endothelial
cells, J. Biol. Chem. 282 (2007) 1663116643.
[25] M. Brini, E. Carafoli, Calcium signalling: a historical account, recent developments and future perspectives, Cell. Mol. Life Sci. 57 (2000) 354370.
[26] R.C. Balijepalli, J.D. Foell, D.D. Hall, J.W. Hell, T.J. Kamp, Localization of cardiac L-type Ca2+ channels to a caveolar macromolecular signaling complex is
required for 2 -adrenergic regulation, Proc. Natl. Acad. Sci. U. S. A. 103 (2006)
75007505.
[27] D. Marques-da-Silva, A.K. Samhan-Arias, T. Tiago, C. Gutierrez-Merino, L-type
calcium channels and cytochrome b5 reductase are components of protein
complexes tightly associated with lipid rafts microdomains of the neuronal
plasma membrane, J. Proteomics 73 (2010) 15021510.
[28] J.P. Weick, R.D. Groth, A.L. Isaksen, P.G. Mermelstein, Interactions with PDZ
proteins are required for L-type calcium channels to activate cAMP response
element-binding protein-dependent gene expression, J. Neurosci. 23 (2003)
34463456.
[29] H. Zhang, A. Maximov, Y. Fu, F. Xu, T.S. Tang, T. Tkatch, D.J. Surmeier, I. Bezprozvanny, Association of CaV1.3 L-type calcium channels with Shank, J. Neurosci.
25 (2005) 10371049.
[30] K. Prybylowski, K. Chang, N. Sans, L. Kan, S. Vicini, R.J. Wenthold, The synaptic
localization of NR2B-containing NMDA receptors is controlled by interactions
with PDZ proteins and AP-2, Neuron 47 (2005) 845857.
[31] S. Besshoh, S. Chen, I.R. Brown, J.W. Gurd, Developmental changes in the
association of NMDA receptors with lipid rafts, J. Neurosci. Res. 85 (2007)
18761883.
[32] I. Delnte-Ramirez, P. Salcedo-Tello, F. Bermudez-Rattoni, Spatial memory formation induces recruitment of NMDA receptor and PSD-95 to synaptic lipid
rafts, J. Neurochem. 106 (2008) 16581668.

[33] D. Marques-da-Silva, C. Gutierrez-Merino, L-type voltage-operated calcium
channels, N-methyl-d-aspartate receptors and neuronal nitric-oxide synthase
form a calcium/redox nano-transducer within lipid rafts, Biochem. Biophys.
Res. Commun. 420 (2012) 257262.
[34] M.R. Sepulveda, M. Berrocal-Carrillo, M. Gasset, A.M. Mata, The plasma membrane Ca2+ -ATPase isoform 4 is localized in lipid rafts of cerebellum synaptic
plasma membranes, J. Biol. Chem. 281 (2006) 447453.
[35] L. Jiang, D. Fernandes, N. Mehta, J.L. Bean, M.L. Michaelis, A. Zaidi, Partitioning of
the plasma membrane Ca2+ -ATPase into lipid rafts in primary neurons: effects
of cholesterol depletion, J. Neurochem. 102 (2007) 378388.
[36] I. Kuszczak, S.E. Samson, J. Pande, D.Q. Shen, A.K. Grover, Sodiumcalcium
exchanger and lipid rafts in pig coronary artery smooth muscle, Biochim. Biophys. Acta 1808 (2011) 589596.
[37] J.E. Brenman, D.S. Bredt, Synaptic signalling by nitric oxide, Curr. Opin. Neurobiol. 7 (1997) 374378.
[38] A.K. Samhan-Arias, M.A. Garcia-Bereguiain, F.J. Martin-Romero, C. GutierrezMerino, Clustering of plasma membrane-bound cytochrome b5 reductase
within lipid rafts microdomains of the neuronal plasma membrane, Mol. Cell.
Neurosci. 40 (2009) 1426.
[39] A.K. Samhan-Arias, D. Marques-da-Silva, N. Yanamala, C. Gutierrez-Merino,
Stimulation and clustering of cytochrome b(5) reductase in caveolin-rich lipid
microdomains is an early event in oxidative stress-mediated apoptosis of cerebellar granule neurons, J. Proteomics 75 (2012) 29342949.
[40] Y. Sato, I. Sagami, T. Shimizu, Identication of caveolin-1-interacting sites in
neuronal nitric-oxide synthase. Molecular mechanism for inhibition of NO formation, J. Biol. Chem. 279 (2004) 88278836.
[41] B. Pani, H.L. Ong, X. Liu, K. Rauser, I.S. Ambudkar, B.B. Singh, Lipid rafts
determine clustering of STIM1 in endoplasmic reticulum-plasma membrane
junctions and regulation of store-operated Ca2+ entry, J. Biol. Chem. 283 (2008)
1733317340.
[42] R.E. Huie, S. Padmaja, The reaction of NO with superoxide, Free Radic. Res.
Commun. 18 (1993) 195199.
[43] A.K. Samhan-Arias, F.J. Martin-Romero, C. Gutierrez-Merino, Kaempferol blocks
oxidative stress in cerebellar granule cells and reveals a key role for the plasma
membrane NADH oxidase activity in the commitment of apoptosis, Free Radic.
Biol. Med. 37 (2004) 4861.
[44] J. Herreros, T. Ng, G. Schiavo, Lipid rafts act as specialized domains for tetanus
toxin binding and internalization into neurons, Mol. Biol. Cell 12 (2001)
29472960.
[45] J.L. MacDonald, L.J. Pike, A simplied method for the preparation of detergentfree lipid rafts, J. Lipid Res. 46 (2005) 10611067.
[46] A.K. Samhan-Arias, M.A. Garcia-Bereguiain, F.J. Martin-Romero, C. GutierrezMerino, Regionalization of plasma membrane-bound avoproteins of cerebellar granule neurons in culture by uorescence energy transfer imaging, J.
Fluoresc. 16 (2006) 393401.
[47] T. Tiago, D. Marques-da-Silva, A.K. Samhan-Arias, M. Aureliano, C. GutierrezMerino, Early disruption of the actin cytoskeleton in cultured cerebellar granule
neurons exposed to 3-morpholinosydnonimine-oxidative stress is linked to
alterations of the cytosolic calcium concentration, Cell Calcium 49 (2011)
174183.
[48] L. Stryer, Fluorescence energy transfer as a spectroscopic ruler, Annu. Rev.
Biochem 47 (1978) 819846.
[49] C. Gutierrez-Merino, F. Centeno, E. Garcia-Martin, J.M. Merino, Fluorescence
energy transfer as a tool to locate functional sites in membrane proteins,
Biochem. Soc. Trans. 22 (1994) 784788.
[50] R.P. Haugland, The Handbook: A Guide to Fluorescent Probes and Labeling
Techniques, 10th ed., Invitrogen Corp., Carlsbad, 2005.
[51] C. Gutierrez-Merino, Quantitation of the Forster energy transfer for bidimensional systems: II. Protein distribution and aggregation state in biological
membranes, Biophys. Chem. 14 (1981) 259266.
[52] A.K. Samhan-Arias, C. Gutierrez-Merino, Plasma membrane-bound cytochrome
b5 reductase is associated with lipid rafts in cerebellar granule neurons in
culture, in: T. Grune (Ed.), Proceedings of the SFFR-Europe 2008, Bologna, Medimond International Proceedings, 2008, pp. 7578.
[53] D. Marques-da-Silva, A.K. Samhan-Arias, M.A. Garcia-Bereguiain, T. Tiago,
C. Gutierrez-Merino, Major plasma membrane-associated redox centres colocalize with L-type calcium channels in neuronal lipid rafts microdomains,
in: D. Caporosi, F. Pigozzi, S. Sabatini (Eds.), Free Radicals, Health and Lifestyle,
Bologna, Medimond International Proceedings, 2009, pp. 127130.
[54] M. Shmuel, E. Nodel-Berner, T. Hyman, A. Rouvinski, Y. Altschuler, Caveolin-2
regulates endocytosis and trafcking of the M1 muscarinic receptor in MDCK
epithelial cells, Mol. Biol. Cell 18 (2007) 15701585.
[55] Z. Fu, S.M. Logan, S. Vicini, Deletion of the NR2A subunit prevents developmental changes of NMDA-mEPSCs in cultured mouse cerebellar granule neurones,
J. Physiol. 563 (2005) 867881.
[56] C. Lu, Z. Fu, I. Karavanov, R.P. Yasuda, B.B. Wolfe, A. Buonanno, S. Vicini, NMDA
receptor subtypes at autaptic synapses of cerebellar granule neurons, J. Neurophysiol. 96 (2006) 22822294.
[57] D. Guerini, E. Garca-Martn, A. Gerber, C. Volbracht, M. Leist, C. GutierrezMerino, E. Carafoli, The expression of PMCa pump isoforms in cerebellar
granule neurons is dependent on cytosolic free Ca2+ , J. Biol. Chem. 274 (1999)
16671676.
123
[58] M.V. Pulina, R. Rizzuto, M. Brini, E. Carafoli, Inhibitory interaction of the plasma
membrane Na+ /Ca2+ exchangers with the 14-3-3 proteins, J. Biol. Chem. 281
(2006) 1964519654.
[59] C. Iadecola, Bright and dark sides of nitric oxide in ischemic brain injury, Trends
Neurosci. 20 (1997) 132139.
[60] R.C. van der Veen, D.R. Hinton, F. Incardonna, F.M. Hofman, Extensive peroxynitrite activity during progressive stages of central nervous system inammation,
J. Neuroimmunol. 77 (1997) 17.
[61] M.F. Beal, Energetics in the pathogenesis of neurodegenerative diseases, Trends
Neurosci. 23 (2000) 298304.
[62] F. Bao, D. Liu, Peroxynitrite generated in the rat spinal cord induces neuron
death and neurological decits, Neuroscience 115 (2002) 839849.
[63] P. van Damme, M. Dewil, W. Robberecht, L. van den Bosch, Excitotoxicity and
amyotrophic lateral sclerosis, Neurodegener. Dis. 2 (2005) 147159.
[64] A.M. Duenas,

R. Goold, P. Giunti, Molecular pathogenesis of spinocerebellar
ataxias, Brain 129 (2006) 13571370.
[65] A. Davies, L. Douglas, J. Hendrich, J. Wratten, A. Tran van Minh, I. Foucault, D.
Koch, W.S. Pratt, H.R. Saibil, A.C. Dolphin, The calcium channel 22 subunit
partitions with CaV2.1 into lipid rafts in cerebellum: implications for localization and function, J. Neurosci. 26 (2006) 87488757.
[66] M. Teubl, K. Groschner, S.D. Kohlwein, B. Mayer, K. Schmidt, Na+ /Ca2+ exchange
facilitates Ca2+ -dependent activation of endothelial nitric-oxide synthase, J.
Biol. Chem. 274 (1999) 2952929535.
[67] J. Bossuyt, B.E. Taylor, M. James-Kracke, C.C. Hale, Evidence for cardiac
sodiumcalcium exchanger association with caveolin-3, FEBS Lett. 511 (2002)
113117.
[68] E.O. Saphire, P.W.H.I. Parren, C.F. Barbas 3rd, D.R. Burtona, I.A. Wilson,
Crystallization and preliminary structure determination of an intact human
immunoglobulin, b12: an antibody that broadly neutralizes primary isolates of
HIV-1, Acta Crystallogr. D57 (2001) 168171.
[69] C.E. Grueter, S.A. Abiria, I. Dzhura, Y. Wu, A.J.L. Ham, P.J. Mohler, M.E. Anderson,
R.J. Colbran, L-type Ca2+ channel facilitation mediated by phosphorylation of
the subunit by CaMKII, Mol. Cell 23 (2006) 641650.
[70] Y. Hayashi, S.H. Shi, J.A. Esteban, A. Piccini, J.C. Poncer, R. Malinow, Driving
AMPA receptors into synapses by LTP and CaMKII: requirement for GluR1 and
PDZ domain interaction, Science 287 (2000) 22622267.
[71] B.P. Head, J.N. Peart, M. Panneerselvam, T. Yokoyama, M.L. Pearn, I.R. Niesman,
J.A. Bonds, J.M. Schilling, A. Miyanohara, J. Headrick, S.S. Ali, D.M. Roth, P.M.
Patel, H.H. Patel, Loss of caveolin-1 accelerates neurodegeneration and aging,
PLoS ONE 5 (2010) e15697.
[72] Y. Takayasu, K. Takeuchi, R. Kumari, M.V. Bennett, R.S. Zukin, A. Francesconi,
Caveolin-1 knockout mice exhibit impaired induction of mGluR-dependent
long-term depression at CA3-CA1 synapses, Proc. Natl. Acad. Sci. U. S. A. 107
(2010) 2177821783.
[73] P. Fossat, I. Sibon, G. Le Masson, M. Landry, F. Nagy, L-type calcium channels
and NMDA receptors: a determinant duo for short-term nociceptive plasticity,
Eur. J. Neurosci. 25 (2007) 127135.
[74] A. Baginskas, A. Kuras, L-type Ca2+ current in frog tectal recurrent neurons
determines the NMDA receptor activation on efferent neuron, Exp. Brain Res.
193 (2009) 509517.
[75] S.G. Cull-Candy, M.M. Usowicz, Multiple-conductance channels activated
by excitatory amino acids in cerebellar neurons, Nature 325 (1987)
525528.
[76] E. Neher, Usefulness and limitations of linear approximations to the understanding of Ca2+ signals, Cell Calcium 24 (1998) 345357.
[77] D.S. Bredt, S.H. Snyder, Nitric oxide: a physiologic messenger molecule, Annu.
Rev. Biochem 63 (1994) 175195.
[78] R. Lagoa, C. Lopez-Sanchez, A.K. Samhan-Arias, C.M. Ganan,

V. Garcia-Martinez,
C. Gutierrez-Merino, Kaempferol protects against rat striatal degeneration
induced by 3-nitropropionic acid, J. Neurochem. 111 (2009) 473487.
[79] C. Gutierrez-Merino, C. Lopez-Sanchez, R. Lagoa, A.K. Samhan-Arias, C. Bueno,
V. Garcia-Martinez, Neuroprotective actions of avonoids, Curr. Med. Chem. 18
(2011) 11951212.
[80] D.W. Choi, Calcium: still center-stage in hypoxia-ischemic neuronal death,
Trends Neurosci. 18 (1995) 5860.
[81] M.P. Mattson, S.L. Chan, Dysregulation of cellular calcium homeostasis in
Alzheimers disease: bad genes and bad habits, J. Mol. Neurosci. 17 (2001)
205224.
[82] M.T. Lin, M.F. Beal, Mitochondrial dysfunction and oxidative stress in neurodegenerative diseases, Nature 443 (2006) 787795.
[83] R.B. Hamanaka, N.S. Chandel, Mitochondrial reactive oxygen species regulate
cellular signaling and dictate biological outcomes, Trends Biochem. Sci. 35
(2010) 505513.
[84] T. Tiago, S. Ramos, M. Aureliano, C. Gutierrez-Merino, Peroxynitrite induces Factin depolymerization and inhibition of myosin ATPase stimulation by actin,
Biochem. Biophys. Res. Commun. 342 (2006) 4449.
[85] T. Tiago, A.K. Samhan-Arias, C. Gutierrez-Merino, Cytoskeleton dynamics
modulates the regionalization of plasma membrane-bound cytochrome b5
reductase in cerebellar granule neurons in culture, in: T. Grune (Ed.),
Proceedings of the SFFR-Europe 2008, Bologna, Medimond International
Proceedings, 2008, pp. 9194.

Caveolin-Rich Lipid Rafts of The Plasma Membrane of Mature Cerebellar

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Caveolin-Rich Lipid Rafts of The Plasma Membrane of Mature Cerebellar

Uploaded by

Copyright:

Available Formats

Cell Calcium 56 (2014) 108123

Contents lists available at ScienceDirect

Caveolin-rich lipid rafts of the plasma membrane of mature cerebellar

Abbreviations: BF, bright eld microscopy images; Cb5 R, cytochrome b5

compartmentation within microdomains allows for a more

D. Marques-da-Silva, C. Gutierrez-Merino / Cell Calcium 56 (2014) 108123

CGN in culture after the partial depolarization of the plasma

nNOS producing nitric oxide [37] and cytochrome b5 reductase

D. Marques-da-Silva, C. Gutierrez-Merino / Cell Calcium 56 (2014) 108123

were centrifuged for 2 min at 2000 g at 4 C in a refrigerated

resuspended in TBS and centrifuged again at 16,100 g and 4 C

D. Marques-da-Silva, C. Gutierrez-Merino / Cell Calcium 56 (2014) 108123

blotting run with CGN lysates (Supplementary Figure S1), as in

2.8. Materials: chemicals and reagents

D. Marques-da-Silva, C. Gutierrez-Merino / Cell Calcium 56 (2014) 108123

2.9. Statistical analysis

largely enriched in lipid rafts-enriched fractions (fractions 15),

D. Marques-da-Silva, C. Gutierrez-Merino / Cell Calcium 56 (2014) 108123

images showed the extensive co-localization in the neuronal soma

3.2. Co-immunoprecipitation with caveolin-1 revealed the

D. Marques-da-Silva, C. Gutierrez-Merino / Cell Calcium 56 (2014) 108123

D. Marques-da-Silva, C. Gutierrez-Merino / Cell Calcium 56 (2014) 108123

inspection of the images of ratio 340/380 increase at short times

D. Marques-da-Silva, C. Gutierrez-Merino / Cell Calcium 56 (2014) 108123

D. Marques-da-Silva, C. Gutierrez-Merino / Cell Calcium 56 (2014) 108123

plasma membrane, particularly at early times after the addition of

of an average size of only a few pixels localized near the plasma

D. Marques-da-Silva, C. Gutierrez-Merino / Cell Calcium 56 (2014) 108123

D. Marques-da-Silva, C. Gutierrez-Merino / Cell Calcium 56 (2014) 108123

used in this work [27,53]. Moreover, Cb5 R associated with lipid

D. Marques-da-Silva, C. Gutierrez-Merino / Cell Calcium 56 (2014) 108123

D. Marques-da-Silva, C. Gutierrez-Merino / Cell Calcium 56 (2014) 108123

not in neuronal extensions. Thus, our results suggest that these

As illustrated in [3], activation of a typical 2.6 pS conductance

D. Marques-da-Silva, C. Gutierrez-Merino / Cell Calcium 56 (2014) 108123

induction of neuronal apoptosis [38,39]. As NO rapidly reacts with

D. Marques-da-Silva, C. Gutierrez-Merino / Cell Calcium 56 (2014) 108123

[64] A.M. Duenas,

[78] R. Lagoa, C. Lopez-Sanchez, A.K. Samhan-Arias, C.M. Ganan,

You might also like