This study investigated the use of quantum dots (qdots) to label and visualize two endogenous synaptic proteins, GABAA-1 receptors and glutamate transporters (VGLUT1), in the rat cerebellum. Qdots allowed for the clear visualization of these proteins in very small presynaptic structures like parallel fiber varicosities that are below the diffraction limit of conventional microscopy. Specifically, qdots formed clusters around interneurons and isolated clusters on interneuron dendrites, revealing the presence of GABAA receptors. They also labeled sub-micrometer parallel fibers and 1-2 micrometer presynaptic varicosities containing VGLUT1. While double labeling with two qdot colors was attempted, some receptor sites remained unlabeled

1. RESEARCHARTICLE
Copyright © 2012 American Scientific Publishers
All rights reserved
Printed in the United States of America
Journal of
Nanoscience and Nanotechnology
Vol. 12, 1–5, 2012
Conjugated Quantum Dots Allow for the
Co-Localization of Endogenous Proteins in
Submicroscopic Structures of the Rat Cerebellum
A. El Abed1 ∗
, A. Baudot1
, M. Chat2
, S. Ben Khalifa1
, and G. Louis1
1
Laboratoire de Neuro-Physique Cellulaire, Université Paris Descartes,
Centre Universitaire des Saints-Pères, 75270 Paris Cedex 06, France
2
Laboratoire de Physique Cérébrale, CNRS UMR8118, Université Paris Descartes,
Centre Universitaire des Saints-Pères, 75270 Paris Cedex 06, France
GABA and glutamate are known as the principal inhibitory and excitatory neurotransmitters in the
vertebrate central nervous system, respectively. However, recent electro-physiological and immuno-
gold literature data indicate that GABA may undergo also an excitatory action on presynaptic
varicosities of parallel fibers (PFs) in the molecular layer of the rat cerebellum. PFs are axonal
extensions, with a cross section of about 0.1 m, of the glutamatergic granule cells. Such an unex-
pected excitatory action of GABA indicates clearly the presence of GABA receptors in the PFs of
granule cells. We show in this study that quantum dots may be used specifically and efficiently to
label two endogenous synaptic proteins, namely R-GABAA- 1 receptors (GABAA Rs) and gluta-
mate transporters (VGLUT1) in order to target their localization in very small structures such as the
presynaptic varicosities of the PFs.
Keywords: Quantum Dots, Fluorescence Microscopy, Co-Localization, Interneurons, Parallel
Fibers, Diffraction Limited Resolution.
1. INTRODUCTION
The cerebellum structure of the adult rat consists of four
distinct layers: the white matter, the granule cells layer,
the Purkinje cells somas layer and the molecular layer rat.
The granule layer contains the somas and the dendrites
of the excitatory granule cells. The molecular layer con-
tains two types of inhibitory neurons, called the molecular
layer interneurons (MLIs), the dendrite arborescence of the
Purkinje cells (PCs) and the parallel fibers (PFs). The latter
correspond to axonal extensions of the granule cells, with
a cross section of about 0.1 m, which form presynaptic
varicosities, with a mean diameter of ∼1 m, at an over-
all mean spacing period of about 5.2 m, through which
granule cells excite both PCs and MLIs, see Figure 1 and1
for a review.
Though glutamate and GABA are usually known as
the principal excitatory and inhibitory neurotransmit-
ters in the vertebrate central nervous system, respec-
tively, recent electrophysiological data obtained by Stell
∗
Author to whom correspondence should be addressed.
et al. from presynaptic varicosities of the PFs in the rat
cerebellum molecular layer indicate that GABAA- 1 recep-
tors (GABAA Rs) may undergo also an excitatory action
depending on the experimental conditions.2
More pre-
cisely, Stell et al. showed that activation of GABAA Rs
may induce release of glutamate from PFs onto both MLIs
and PCs. Moreover, immunogold labeling and electronic
microscopy observations2
revealed clearly the presence of
GABAA Rs in presynaptic varicosities of parallel fibers
(PF) whereas immuno-histochemistry experiments realized
by one of the authors (M. Chat) using organic fluorescent
dyes (Rhodamine and Alexa Fluor) have failed to show a
clear co-localization of these receptors in presynaptic vari-
cosities of PFs. Also, one should note also that the activa-
tion of GABAA Rs was observed to be more pronounced
for young rats (post natal day smaller than P13) than for
adult rats (older than P18). This result suggests the pres-
ence of greater number of GABAA receptors in presynaptic
varicosities of PFs for young rats than for adult rats.
We show in this study that functionalized quantum dots
(qdots) may be used specifically and efficiently to label
two endogenous synaptic proteins, namely GABAA- 1
J. Nanosci. Nanotechnol. 2012, Vol. 12, No. xx 1533-4880/2012/12/001/005 doi:10.1166/jnn.2012.4922 1

2. RESEARCHARTICLE
Conjugated Quantum Dots Allow for the Co-Localization of Endogenous Proteins El Abed et al.
Fig. 1. Simplified representation of the rat cerebellum transverse
section; inset: detail of presynaptic varicosities of the parallel fibers (PFs)
with Purkinje cells (A) and with Molecular Layer Interneurons (B).
receptors and glutamate transporters (VGLUT1), in order to
target their localization in submicroscopic and microscopic
structures of the rat cerebellum. Such structures consist of
the PFs of the granule cells and their presynaptic varicosi-
ties with the molecular layer interneurons, respectively.
Thanks to their unique optical properties, qdots opened
a new area in cellular imaging and single molecule detec-
tions, see for a review references.3 4
For example, in the
field of neuroscience, Dahan et al.5
was the first to use
conjugated qdots to track individual glycine receptors and
analyzed their lateral dynamics in the synaptic region of
living cultured neurons during periods of time ranging
from milliseconds to few minutes. However, because of the
ability of neurons to establish complex and wide networks
in the central nervous system, results obtained generally
from isolated cultured neurons do not meet very much
attention among the neurobiologists community, for whom
brain slice, cut either from living or fixed tissues, repre-
sent a more suitable material model. However, the relative
big size of functionalized qdots by regards to organic flu-
orescents dyes size, i.e., 10 nm versus 1 nm, the heteroge-
neous structure of tissue slices and their auto-fluorescence
require different fixation and permeabilization conditions
for optimum immuno-staining of slices.
2. EXPERIMENTAL DETAILS
Typically, tissue labeling process begins with the fixation
of the whole animal, a chemical treatment that crosslinks
quickly proteins in order to freeze the structure of the
whole cell and its environment in a given sate. In our
study, rats were anaesthetized by an intraperitoneal injec-
tion of 100 to 150 l Pentobarbital (Sanofi) diluted 5×
in a 0.9% NaCl solution. Then, animals were transcar-
dially perfused with a cold (∼5 C) 0.9% NaCl solution,
followed by a cold fixative solution prepared just before
use. The best results were obtained with fixative solu-
tion which consists of 4% paraformaldehyde (PFA) alone
Fig. 2. P21 rat cerebellum slice as imaged by optical and fluorescence
microscopy which shows aggregates of goat F(ab’)2 anti-rabbit IgG Qdot
605 conjugates (AR-Qds conjugates). AR-Qds aggregates (red domains),
distribute randomly on the surface of the cerebellum slices but more
particularly on their borders and the cerebellum white matter (smooth
central area); scale bar = 50 m.
(without glutaraldehyde or picric acid) in phosphate buffer
(0.15 M, pH = 7 4). After 20 minutes, the cerebellum ver-
mis was removed and post fixed overnight in the same
fixative solution at 4 C. Then, 40 m thick sagittal cere-
bellar slices were cut in a cold phosphate buffer 0.15 M,
pH = 7 4 with a Vibratome (VT 1000S Leica). All incuba-
tions were performed under continuous agitation at room
temperature in 24-well culture plates. The sections were
thoroughly washed in phosphate buffer saline, PBS 0.15 M,
then incubated 2 hours in bovine serum albumin (BSA from
Amersham), 2% in Tris buffer saline (TBS-HCl 0.05 M,
pH = 7 6) with 0.9% NaCl and incubated overnight with
specific monoclonal primary antibodies which consisted
of either rabbit anti-GABAA- 1 (R-GABAA 1), a kind
gift from Prof. Sieghart (Vienna), directed against the
1 subunit of GABAA receptors, or mouse anti-VGLUT1
(M-VGLUT1) in TBS containing 0.05% Triton for perme-
abilization and 2% of BSA (TBST). Two types of conju-
gated quantum dots were purchased from Invitrogen and
used in this study. The first type consisted of Qd®
strepta-
vidin conjugates which are made from a nanosized CdSe-
ZnS semiconductor core–shell, which is further coated
with a polymer shell and coupled to streptavidin. We used
two Qdot®
streptavidin conjugates with two narrow sym-
metric emission maximum near 525 nm (Q10141MP) and
605 nm (Q10101MP), named in this study sterp-Qd525
and strep-Qd605, respectively. The second type of the used
conjugated quantum dots consisted of a Qdot®
605 sec-
ondary antibody conjugate for which the polymer shell is
coupled to a secondary antibody (goat F(ab’)2 anti-rabbit
IgG antibody, Q11402MP), named in this study AR-Qds.
2 J. Nanosci. Nanotechnol. 12, 1–5, 2012

3. RESEARCHARTICLE
El Abed et al. Conjugated Quantum Dots Allow for the Co-Localization of Endogenous Proteins
The overall mean size of the used Qdot®
conjugates is
about 15–20 nm.
Fixed slices were imaged by confocal microscopy,
at room temperature in an open chamber, with a Zeiss
LSM 510 confocal microscope, using a 63X oil-immersion
objective with a numerical aperture of N O = 1 4, equipped
with an Argon laser ( = 488 nm) or Hg Lamp. The
pinhole aperture and the laser power were ∼96 m and
2.5 mW, respectively. The controls consisted of incu-
bations without the primary antibodies and with the
secondary antibodies and strepatvidin coated qdots (strep-
Qds). Other controls consisted of incubated slices with
strep-Qds alone. To test the fixation of the used primary
and secondary antibodies, we used also controls with con-
ventional dyes such as Rhodamine Red-X ( abs = 570 nm,
em = 590 nm) and Alexa Fluor 488 X ( abs = 495 nm,
em = 519 nm).
To achieve a reasonable compromise between specific
and non specific labeling of conjugated qdots in cerebel-
lum slices, we varied different parameters such as fix-
ation strength, permeabilization and antibody incubation
strength and time.
3. RESULTS AND DISCUSSION
3.1. Single Labeling
Two approaches were investigated in order to achieve for
a labeling of GABAA Rs and VGLUT1 transporters with
qdots. The first approach consisted of using a Qdot®
605
goat F(ab’)2 anti-rabbit IgG conjugate (AR-Qds). Unfortu-
nately, despite many attempts, we did not succeed to label
neither GABAA receptors nor VGLUT1 transporters with
such markers. We observed many aggregates of conjugated
qdots, distributed randomly on the surface of the slices and
more particularly on their borders. We observed also that
conjugated qdots stain more particularly the white matter.
The second approach consisted of incubating first cere-
bellum slices with the AR-biotin secondary GABAA and
Fig. 3. Fluorescence confocal microscopy images of two interneurons located in the molecular layer of two P21 rat cerebellum slices where GABA
receptors were labeled with Qd605 (red, left) and where VGLUT1 transporters were labeled with Qd525 (green, right).
VGLUT1 antibodies, then with strep-qdots. In this case,
staining was significantly enhanced. Slices were mounted
on glass slides in Prolong Anti-fade kit mounting medium
(Molecular Probes P 7481).
Figure 3(A) shows a region of interest (ROI) located
in the molecular layer of a P21 rat cerebellum slice
where GABAA receptors were labeled with strep-Qd605.
As expected, qdots form numerous clusters around the
soma of the gabaergic interneurons, but form also isolated
clusters located away from the somas, revealing hence the
presence of GABAA receptors on interneurons dendrites
and also in presynaptic varicosities of the PFs, as we show
in this study. One should remark also that because of the
used scanning confocal technique, the blinking feature of
individual qdots could not be observed.
Figures 3(B and C) shows images of a P21 rat cere-
bellum slices where VGLUT1 transporters were labeled
with strep-Qd525, following the same procedure as for
the labeling of GABAA receptors. One observes the exis-
tence of two types of structures: i) sub-micrometric clus-
ters which should correspond to PFs (Fig. 3(C)) and ii)
microscopic structures (Fig. 3(B)), with a typical size of
∼1 to 2 m. Such structures should correspond to presy-
naptic varicosities of PFs.
As one may observe, despite the PFs cross section size
being of the order of or smaller than the diffraction limited
resolution allowed by optical microscopy, i.e., ∼0.2 m,
the fluorescence of qdots enables the observation of such
structures with a spatial resolution of about 0.02 m,
as shown in Figure 3(C). Such high spatial resolution is
mainly due to the well known high sensitivity and signal-
to-noise ratio (SNR) of qdots. This feature leads to a typi-
cal SNR ∼30 for qdots which may allow for the obtaining
of a spatial resolution of ∼1 nm.6
3.2. Double Labeling
Following the classical method used generally for double
labeling, one incubates first slices with both AR-GABA
J. Nanosci. Nanotechnol. 12, 1–5, 2012 3

4. RESEARCHARTICLE
Conjugated Quantum Dots Allow for the Co-Localization of Endogenous Proteins El Abed et al.
Fig. 4. Fluorescence microscopy images of P21 (A, B, C) and P11 (D, E, F) rats cerebellum slices where GABAA Rs and VGLUT1 were labeled
respectively with Qd605 (red) and Alexa 488 (green); scale bar = 2 m.
(anti rabbit) and AM-VGLUT1 (anti mouse). To label
GABAA Rs first, one incubates slices with AR-biotin, fol-
lowed by several washes and a further incubation with
strep-Qd605 (followed by several washes). After the label-
ing of GABAA receptors with strep-Qd605, one labels
VGLUT1 transporters by incubating slices with AM-biotin
(anti-mouse) which binds selectively to AM-VGLUT1,
followed by several washes. Finally, one incubates slices
with strep-Qd525, followed by several washes.
Unfortunately, we noticed that despite the use of many
different concentrations of AR-biotin, GABAA Rs were
labeled with both Qd605 and Qd525. This result means that
during the first labeling process with strep-Qd525, some of
the AR-biotin sites remained free as confirmed by a control
where a single labeling of GABAA Rs with strep-Qd605,
was followed by a last incubation with strep-rhodamine.
We observed in this control that some of GABAA Rs were
also labeled with rhodamine. For this reason, we decided
to label VGLUT1 transporters with AM-Alexa488 while
GABAA receptors were labeled with strep-Qd605.
Figures 4(A–C) shows a ROI of an adult rat (P21) cere-
bellum slice where VGLUT1 transporters were labeled
with AM-Alexa 488 (green) and GABAA receptors were
labeled with strep-Qd605 (red). This image shows that
GABAA receptors are located around the soma of an
interneuron and also nearby parallel fibers (green). One
can notice that no co-localization of GABAA Rs and
VGLUT1 transporters is observed in this case.
Figures 4(D–F) shows an image of cerebellum slice
of young rat (P11) where GABAA Rs were labeled
with strep-Qd605 and where VGLUT1 were labeled with
AM-Alexa488. One observes that some clusters of GABAA
Rs are localized individually away from the interneurons
somas, i.e., on dendrites. More interesting, one notices that
some of the GABAA Rs are co-localized with VGLUT1
transporters, leading to yellow clusters (indicated by arrows
in Fig. 4(F)) located around circular structures which
should correspond to presynaptic varicosities of PFs.
4. CONCLUSIONS
We have shown in this preliminary study that quantum
dots may be used specifically and efficiently to label
two endogenous synaptic proteins, namely R-GABAA − 1
receptors (GABAA Rs) and glutamate transporters
(VGLUT1) in order to target the localization of such
receptors in microscopic and sub-microscopic cerebellum
structures which consist of parallel fibers of the granule
cells, presynaptic parallel fibers varicosities and molecu-
lar layer interneurons. We observed also, for the first time
by fluorescence microscopy, the co-localization of GABAA
Rs and VGLUT1 transporters in presynaptic parallel fibers
in the case of young rats, in agreement with the recent
literature data. This work should be continued in order to
optimize the labeling of GABAA Rs and VGLUT1 trans-
porters with secondary antibodies conjugated qdots.
Acknowledgments: We would like to thank Prof.
W. Sieghart for the gift of R-GABAA- 1 antibody and
Dr A. Marty for his kind help and fruitful discussions.
4 J. Nanosci. Nanotechnol. 12, 1–5, 2012

5. RESEARCHARTICLE
El Abed et al. Conjugated Quantum Dots Allow for the Co-Localization of Endogenous Proteins
References and Notes
1. G. M. G. Shepherd and M. Raastad, The Cerebellum 2, 110 (2003).
2. B. M. Stell, P. Rostaing, A. Triller, and A. Marty, J. Neuroscience
Vol. 27, Pages 9022 (2007).
3. Z.-B. Li, W. Cai, and X. Chen, J. Nanosci. Nanotechnol. 7, 2567
(2007).
4. X. Michalet, F. F. Pinaud, L. A. Bentolila, J. M. Tsay, S. Doose, J. J.
Li, G. Sundaresan, A. M. Wu, S. S. Gambhir, and S. Weiss, Science
307, 538 (2005).
5. M. Dahan, S. Levi, C. Luccardini, P. Rostaing, B. Riveau, and
A. Triller, Science 302, 442 (2003).
6. M. Dahan, Nanoparticles in Biomedical Imaging, edited by J. W. M.
Bulte and M. M. J. Modo, Fundamental Biomedical Technologies,
Springer (2007), Vol. 3, pp. 427–441.
Received: 1 December 2010. Accepted: 1 May 2011.
J. Nanosci. Nanotechnol. 12, 1–5, 2012 5