Analyst
View Article Online
COMMUNICATION
A chemical quenching- and physical blocking-based
method to minimize process-mediated aggregation of
Published on 12 August 2013. Downloaded by Osmania University on 07/09/2016 08:13:27.
Cite this: Analyst, 2013, 138, 6277
antibody-crosslinked nanoparticles for imaging
Received 5th July 2013 application†
Accepted 12th August 2013
Chandra K. Dixit,‡*ab Shibsekhar Roy,b Conor Byrne,a Richard O'Kennedyab
DOI: 10.1039/c3an01294h
and Colette McDonaghbc
www.rsc.org/analyst
Critical limitation of nanoparticles (NP) is their aggregation after
functionalisation and antibody cross-linking. We analysed the cause
of this aggregation with respect to functionalities (carboxyls and
amines) on the NP surface. We have devised a low cost novel method
to reduce such aggregations during protein cross-linking and vali-
dated it by probing the platelet surface with platelet surface-specific
anti-CD41 antibody conjugated NPs.
Nanoparticles have enormous potential in the eld of biomed-
ical diagnostics and therapeutics given that there is a wide
variety of them such as metallic, metal oxide, quantum dot,
hybrid silica and polymeric nanoparticles, with applications
extending from inter- and intra-cellular probing to bright labels
for immunoassays to targeted drug-delivery.1–3 Novel silica
nanoparticle synthesis approaches have enabled the doping of
these particles with uorescent dyes, which has opened up a
plethora of applications in uorescence-based detection for
example biosensing, cell imaging using confocal microscopy
and uorescence-assisted cell sorting.4
Dispersion stability and aggregation tendencies of unfunc-
tionalised NPs are the most critical factors for using them in
biological applications. However, unfunctionalized NPs have
been reported to show a certain degree of aggregation under the
inuence of their environments such as solvents due to the
molecular interactions between the silica surface and solvent
molecules. Graing functional groups, for instance carboxyls
and amines, on the NP surface signicantly increases the
number of such interactions in the form of hydrogen bonds, van
der Waals and ionic interactions, according to the physical
a
School of Biotechnology, Dublin City University, Glasnevin, Dublin 9, Ireland
b
National Biophotonics Imaging Platform, Ireland, Dublin City University, Glasnevin,
Dublin 9, Ireland. E-mail: chandra.dixit2@mail.dcu.ie
c
School of Physical Science, Dublin City University, Glasnevin, Dublin 9, Ireland
† Electronic supplementary information (ESI) available: Experimental, Table S1
and Fig. S1–S4. See DOI: 10.1039/c3an01294h
‡ Current address: Biomedical Diagnostics Institute, Dublin City University,
Glasnevin, Dublin 9, Ireland.
This journal is ª The Royal Society of Chemistry 2013
View Journal | View Issue
conditions of the system and may lead to an increase in
aggregation. This kind of aggregation is mainly reversible upon
mild sonication. In addition, functionalised NPs are always
present in a mixed population of bare, partially functionalised
and completely functionalised. Conversely, addition of proteins
to such functionalised nanoparticles only increases the total
number of reactive functional groups. Aggregation in protein-
conjugated NPs can occur due to the covalent linkage between
activated functional groups on the NP/antibody with the func-
tional groups on other NPs/antibodies in addition to non-
covalent interactions. This uncontrolled self/crosslinking
results in aggregates, which may not be separated upon soni-
cation. Therefore, protein-conjugated nanoparticles represent a
complex chemical moiety that can readily undergo aggregation.
Aggregation of chemically functionalised and protein cross-
linked nanoparticles poses a signicant challenge for many
applications5–11 and affects the physico-chemical properties of
nanoparticles and their protein conjugates.12
Efforts have been made to improve nanoparticle dispersion
in biocompatible solvents and buffers. Widely employed
approaches for minimizing aggregation include coating with
functional groups such as carboxyl or amino,13 complex chains
of dendrimers like PAMAM14 and coating with hydrophilic
polymers such as PEG.15 In addition, there are several strategies
such as biotinylation of NPs that promise better dispersion and
have been demonstrated for developing controlled aggregation-
based 3D NP complexes.16 Nanoparticle dispersion was
considerably improved by employing these methods but these
approaches increase NP complexity, the overall functionalisa-
tion time and in some cases resultant size. In addition, these
approaches could not replace chemical cross-linking-mediated
protein conjugation of NPs which is still one of the most
important methods of functionalisation.
Silica NP synthesis and characterization has been reported
previously by our group.4,17,18 The functionalisation and anti-
body cross-linking procedure is described in the ESI.† Prior to
functionalisation, NPs were washed with ethanol and nitrogen
(N2)-dried. Functionalisation of NPs was performed with 2%
Analyst, 2013, 138, 6277–6281 | 6277

Analyst
[(v/v) in ethanol] of amino silane (APTES) and 2% [(v/v) in
deionised water (DIW)] of carboxy-silane for 3 hours with
continuous stirring followed by washing in their respective
solvents. Functionalised NPs were then N2-dried and were
stored in absolute ethanol until further use. Anti-HRP and anti-
CD41 antibodies were then conjugated to the AfNPs and CfNPs
using conventional 1-ethyl-3-(3-dimethylaminopropyl)carbodii-
mide (EDC) chemistry (ESI†).
In this report, we present a facile route to reduce process
induced aggregation of NP–antibody conjugates using a
Published on 12 August 2013. Downloaded by Osmania University on 07/09/2016 08:13:27.
combinatorial surface blocking strategy by employing chemical
quenching with ethanolamine in tandem with physical block-
ing with bovine serum albumin (BSA). We have demonstrated
the improvement of NP distribution homogeneity as a probe by
performing uorescent detection of HRP on the surface. We
have also validated our ndings with HRP by employing platelet
probing using the anti-CD41 antibody conjugated to CfNPs and
AfNPs. The quenching and blocking steps are described in
Scheme 1 for amine (right side) and carboxyl (le side) func-
tionalised NPs. We employed two different surface blocking
schemes. In the rst set, activated groups on the surface of
either the NPs or the antibodies competed with ethanolamine
(at nal concentration of 50 mM) by adding directly from a
stock solution of 1 M at 5 minutes from the start of the conju-
gation reaction between NPs and antibodies. This was per-
formed in order to minimize any conjugation of activated
antibody with other antibodies or with other NPs. In the second
Scheme 1 Gradual centrifugation method for fractionating antibody-conju-
gated NPs as a function of their size. This step was designed to remove larger NP
clusters, if there were any left after the quenching and blocking step, in order to
improve probing homogeneity of the conjugates.
6278 | Analyst, 2013, 138, 6277–6281
View Article Online
Communication
set, we employed a physical blocking step of the NP surface with
1% (v/v) BSA aer quenching the unreacted groups via compe-
tition-based chemical blocking with ethanolamine. Later, we
washed the prepared conjugates by repeated centrifugation
with PBS. We subjected the treated antibody–NP conjugates to
sequential centrifugation as detailed in the ESI† and tested for
their probing efficiencies in the uorescence immunoassay and
confocal microscopy formats.
We have analysed several physical conditions such as pH and
salt strengths of phosphate buffers19 in order to obtain
optimum conditions to minimize any aggregation due to non-
covalent interactions (Table S1†) which could interfere with
covalent-cross-linking-mediated aggregation. In addition, we
present the effects of the chemical and physical blocking on the
aggregation behaviour and probing efficiencies of antibody-
conjugated Af/CfNPs. The hypothesis we developed was found
to corroborate with the uorescence microscopy images pre-
sented in Fig. S1(d) and (e),† which clearly indicate the aggre-
gated nature of antibody conjugated Af and CfNPs on the HRP/
anti-HRP platform where no blocking was performed. We
signicantly reduced this undesirable aggregation due to
uncontrolled cross-linking of NPs and antibodies by applying
the new method reported here (Fig. 2). In addition, we subjected
these antibody-conjugated NP sets to gradual centrifugation as
illustrated in Scheme 1. The rationale is that followed by effi-
cient quenching, there will be either few or no aggregates,
however, by subjecting the conjugates from the quenching
experiments to centrifugation with increasing gravity, the
aggregates should separate out as a function of their cumulative
size. We analysed the sediment of each round for visual
conrmation on a uorescence microscope. In addition, we
have analysed the probing homogeneity of these NP–antibody
conjugates based on their distribution as a function of their
uorescence intensity (Fig. 1) as a three dimensional uores-
cence distribution for the quenched and blocked particles as
probes for HRP. The intensity distribution prole of antibody-
conjugated CfNPs, as shown in Fig. 1a–d, with respect to the
control (untreated CfNP–antibody conjugate) indicates a clear
change of the peak-widths such that they decrease signicantly
from that in the control towards increasing centrifugation
speeds (67, 268, 419 and 603
g), which is accompanied by a
higher degree of sample homogeneity, probe density and target
specicity. In addition, CfNPs were found to be better probes
compared to the amine functionalised NPs (Fig. 1e–h) in terms
of less aggregation and better homogeneity. This evaluation is
supported by the image data in Fig. S2† which clearly demon-
strates better homogeneity and efficiencies for CfNPs. A likely
explanation for the superior performance of CfNPs is that, for
conjugation of antibodies with NPs, it is necessary to activate
carboxyl groups in order to use carbodiimide (EDC) chemistry.
In the case of CfNPs, the carboxyl groups of NPs are activated
and the antibodies are further reacted with them. However, in
the case of AfNPs, the carboxyl groups of the antibodies are
activated, which increases the probability of intermolecular
cross-linking involving many antibodies (as in Fig. S1(c)†) that
could result in large aggregates. Conversely, CfNPs have the
least probability of inter-NP cross-linking, as is evident from
This journal is ª The Royal Society of Chemistry 2013

Communication
Fig. 1 Fluorescence imaging of the horseradish peroxidase (HRP) probing
homogeneity and efficiency of the anti-HRP antibody-conjugated NPs. Functional
group quenching and physical blocking of NPs were employed to reduce NP
aggregates in carboxy-functionalised NPs (a–d). The probing homogeneity along
with density significantly improved from ‘a’ to ‘d’ as is evident from their density
Published on 12 August 2013. Downloaded by Osmania University on 07/09/2016 08:13:27.
profiles. Antibody-conjugated amine-functionalised NPs (e–h) were found to have
relatively low probing density in comparison to carboxy-functionalised NPs. The
images a–d and e–h correspond to the supernatants of sequential centrifugation
of the chemically quenched samples at speeds described in Scheme 1.
Fig. 2 (a) Platelet probing using optimized anti-CD41 antibody-conjugated
carboxy-functionalised NPs at 480 and 630 nm laser excitation for measuring
auto-fluorescence and signal, respectively. Platelet probing efficiency was high in
comparison to the control along with high specificity as is clear from the super-
imposed image on the right. AF is auto-fluorescence. (b) ImageJ-based areal
intensity distribution analysis of 480 nm AF and 630 nm Ab–NP frames from
figure (a). Panel b ‘1’ corresponds to 630 nm laser-based imaging of anti-CD41
antibody–CfNP conjugate prepared by the reported novel method; ‘2’ is 480 nm
laser-based auto-fluorescence measurement of platelets; 3 and 4 are respective
630 and 480 nm laser measurements for double negative control of 1 and 2
whereas, 5 and 6 corresponds to the single negative control (details on controls in
results section). CfNP is carboxy-functionalised NPs. (c)–(f) represents the intensity
distribution profiles of antibody-conjugated CfNPs on platelets captured over the
fibrinogen-coated surface in order to qualitatively analyze aggregation. (c) and
(d) are two different views of the distribution profile for the 630 nm channel
which is the centre image in (a). Comparison of (d) with the controls (e) and (f)
demonstrate the effectiveness of the aggregation reduction protocol.
Fig. S1(e).† However, without the blocking, there are still large
aggregates in both the NP systems. This is clear from Fig. S3†
which is complementary to Fig. 2 and to Fig. S1.† Here, the data
are treated using ImageJ-analysis where the aggregates are
This journal is ª The Royal Society of Chemistry 2013
View Article Online
Analyst
represented in the X, Y and Z directions as explained in the
caption and in the ESI.† The three Z-plane depths represent
small (0%), intermediate (6%) and large (12%) aggregates.
Panels ‘a–c’ and ‘g–i’ show images for untreated AfNPs and
CfNPs respectively which indicates that, although aggregates
are formed, the aggregate size is less for CfNPs (g–i). Panels d–f
and j–l are images corresponding to quenched and blocked
AfNPs and CfNPs respectively which illustrate the much
reduced aggregate size and is consistent with Fig. 1(d) and (f)
corresponding to the conjugate centrifugation at 603
g. In
particular, the reduction in aggregation is evident from the
comparison of Fig. S3(g1) (control) and (j1) (treated CfNPs).† It
is clear from these data that competitive chemical quenching
and blocking, in conjunction with the multi-step centrifugation,
has signicantly reduced the aggregation in both sets because
maximum probe density is obtained in the last round of
centrifugation which is evident from the 3D intensity distribu-
tion proles (Fig. 1, S2 and S3†). In addition, based on the
widths of the peaks in 3D intensity distribution proles the size
of NP–antibody conjugates could be predicted qualitatively
(Fig. S3†).
Finally, in order to validate our ndings we performed a cell
surface analysis experiment where we probed platelets with
anti-CD41 antibody conjugated to NPs that were treated
according to the optimized conditions of aggregation reduction.
We employed a two channel optical system corresponding to
480 and 630 nm laser excitation to study co-localization of
antibody–NP conjugates over platelets for their probing effi-
ciency. Each pixel in the images here corresponds to 300 nm
which is equivalent to the focal width (pin hole) employed in
this study. We have made an important assumption here that
each pixel corresponds to a group of three NPs (diameter of NP
¼ 100 nm) which is justied by the analysis represented in
Fig. 3. The dark circular structures in Fig. 2 are brinogen
stamps which are used to facilitate platelet binding.20 We used
the 480 nm excitation laser to measure any molecular uores-
cence of the immobilized platelets (green channel, Fig. 2a le
Fig. 3 Analysis of the pixelation density over the platelet surface. A platelet
cluster (b) was taken from the original image (a), which was further zoomed in to
visualize pixel distribution. The focal width (pin hole) was 300 nm that corre-
sponds to the size of a pixel. This indicates that each pixel could accommodate at
the maximum of three NPs. The distribution of pixels and hence NPs in (c) indi-
cates a homogeneously distributed NP population with minimal aggregation.
Analyst, 2013, 138, 6277–6281 | 6279

Analyst
image) whereas, a 630 nm laser to measure the uorescence of
the NIR664 dye in the NPs on the platelet surface (centre image).
Fig. 2a (right) clearly shows the co-localisation of the platelet
and dye uorescence indicating efficient probing efficiency of
the treated CfNPs in comparison to the untreated NP–antibody
conjugates (Fig. S4†).
For Fig. 2b, an areal intensity distribution from the 1 cm2
coverslip, was obtained for frames corresponding to both
channels (Fig. 2a) using the region of interest (ROI) analysis
function of ImageJ, which allows comparison of the same area
Published on 12 August 2013. Downloaded by Osmania University on 07/09/2016 08:13:27.
on both of the individual image frames. In Fig. 2b, 1 and 2
correspond to the 630 nm (red dye; specic signal obtained with
dye) and 480 nm (non-specic signal obtained with auto-
uorescence) channels respectively. Two controls were
employed in this study (Fig. S4†), a double-negative mouse IgG
control which is not specic to the CD41 marker, conjugated to
NPs that were not treated according to the reported novel
method (‘3’ corresponds to the measurements performed with
630 nm and ‘4’ with 480 nm) while the second control was a
single-negative anti-CD41 antibody-conjugated NPs (specic
antibody but no treatment to reduce aggregation) 5 (at 630 nm)
and 6 (480 nm). The increased intensity in blocks 1 and 5
demonstrates the high efficiency of probing of platelets using
the antibody–NP conjugates [against appropriate controls
(Fig. S4†)] prepared following the protocol described in this
paper. However, a high signal was obtained for block 5 which
was approximately equivalent to the block 1. It represents that
aggregated nanoparticles may also have high signals as
compared to the treated sample given that they possess the
specic probing element (anti-platelet antibody in this case).
CfNP–antibody conjugate images were also subjected to inten-
sity distribution proling in order to assess the homogeneity of
probes across the platelet surface along with the appropriate
controls and these data are shown in Fig. 2c–f. The width of the
peak along XY- (width of the aggregate) and Z-plane (height of
the aggregate) allowed the qualitative analysis of the aggregate
size as a function of intensity distribution. Panels 2e and f
correspond to the controls for this analysis such that 2e is the
double negative control and 2f is single negative. In 2e only a few
peaks were observed that may correspond to some non-speci-
city toward the surface while in 2f large aggregates could be
observed as a function of intensity distribution in the XY-plane
(width of the aggregate) and Z-plane (height of the aggregate).
Panels 2c and d represent the intensity distribution in the red
channel (Fig. 2a centre) for the treated CfNP-conjugated to HRP
antibodies for two different views. In particular, it is clear from
Fig. 2d that the intensity distribution is more homogeneous for
the particles treated using the newly developed protocol in
comparison to the controls 2e and f. In addition, NP distribu-
tion over a cluster of seven platelets is shown as a factor of
pixelation of the intensity (Fig. 3). A clearly homogeneous pix-
elation was found which indicates effective reduction in the
aggregation of NPs that has signicantly improved the NP
distribution over the platelet surface. It is important to mention
here that platelets are highly sticky in nature and their capture
distribution over a surface is uncontrolled and therefore, they
tend to form clusters.
6280 | Analyst, 2013, 138, 6277–6281
View Article Online
Communication
Conclusions
We established from this study that chemical quenchers, in
conjunction with physical blocking using blocker proteins,
could efficiently be employed for blocking the activated func-
tional groups on the surface of a uorescent silica nanoparticle,
which contribute to undesired aggregation effects and can
subsequently have a negative impact on applications, such as
uorescence-based assays and cell probing. We also found that
carboxy-functionalised NPs have less tendency to aggregate
both before and aer undergoing the chemical and physical
blocking protocol. The aggregation behaviour could further be
minimized by using a sequential centrifugation procedure.
These optimized NPs were successfully employed in a model
HRP assay and for probing the CD41 surface receptor on the
platelets which validated the benets of the new protocol.
Acknowledgements
This work was supported through the National Biophotonics
and Imaging Platform, Ireland, and funded by the Irish
Government's Programme for Research in Third Level Institu-
tions, Cycle 4, National Development Plan 2007–2013.
Notes and references
1 A. C. Sabuncu, J. Grubbs, S. Qian, T. M. Abdel-Fattah,
M. W. Stacey and A. Bwskok, Colloids Surf., B, 2012, 95, 96–
102.
2 A. Albanese, P. S. Tang and W. C. W. Chan, Annu. Rev.
Biomed. Eng., 2012, 14, 1–16.
3 H. Xing, N. Y. Wong, Y. Xiang and Y. Lu, Curr. Opin. Chem.
Biol., 2012, 16(3–4), 429–435.
4 S. Roy, R. Woolley, B. D. MacCraith and C. McDonagh,
Langmuir, 2010, 26(17), 13741–13746.
5 D. F. Moyano and V. M. Rotello, Langmuir, 2011, 27(17),
10376–10385.
6 A. G. Roc, D. Carmona, N. Miguel-Sancho, O. Bomat-Miguel,
F. Balas, C. Piquer and J. Santamaria, Nanotechnology, 2012,
23, 155603–155612.
7 T. Kim, C.-. H. Lee, S.-W. Joo and K. Lee, J. Colloid Interface
Sci., 2008, 318(2), 238–243.
8 J. Fresnais, C. Lavelle and J. F. Berret, J. Phys. Chem. C, 2009,
113(37), 16371–16379.
9 S. Basu, S. K. Ghosh, S. Kundu, S. Panigrahi, S. Praharaj,
S. Pande, S. Jana and T. Pal, J. Colloid Interface Sci., 2007,
313(2), 724–734.
10 D. Zhang, O. Neumann, H. Wang, V. M. Yuwono,
A. Barhoumi, M. Perham, J. D. Hartgerink, P. Wittung-
Stafshede and N. J. Halas, Nano Lett., 2009, 9(2), 666–671.
11 M. S. Bakshi, J. Phys. Chem. C, 2011, 115(29), 13947–13960.
12 B. Gilbert, R. K. Ono, K. A. Ching and C. S. Kim, J. Colloid
Interface Sci., 2009, 339(2), 285–295.
13 R. P. Bagwe, L. R. Hilliard and W. Tan, Langmuir, 2006, 22(9),
44357–44362.
14 T. Tao, Y. Yang, S. Liu, Y. Zheng, J. Fu and J. Chen, Acta
Biomater., 2013, 9(5), 6431–6438.
This journal is ª The Royal Society of Chemistry 2013

Communication
15 Y. Cho, R. Shi, A. Ivanisevic and R. B. Borgens, J. Neurosci.
Res., 2010, 88(7), 1433–1444.
16 E. Katz, A. N. Shipway and I. Willner, in Nanoparticles: From
Theory to Application, ed. G. Schmid, WILEY-VCH Verlag
GmbH & Co. KGaA, Weinheim, Germany, 2004, pp. 368–421.
17 S. Roy, R. Woolley, B. D. MacCraith and C. McDonagh,
Langmuir, 2010, 26(17), 13741–13746.
Published on 12 August 2013. Downloaded by Osmania University on 07/09/2016 08:13:27.
This journal is ª The Royal Society of Chemistry 2013
View Article Online
Analyst
18 S. Roy, C. K. Dixit, R. Woolley, B. D. MacCraith, R. O'Kennedy
and C. McDonagh, Langmuir, 2010, 26(23), 18125–18134.
19 J. Jiang, G. Oberdoster and P. Biswas, J. Nanopart. Res., 2009,
11, 77–89.
20 L. Basabe-Desmonts, S. Ramstrom, G. Meade, S. O'Neill,
A. Riaz, L. P. Lee, A. J. Ricco and D. Kenny, Langmuir, 2010,
26(18), 14700–14706.
Analyst, 2013, 138, 6277–6281 | 6281