In vivo behaviour of glyco-NaI@SWCNT ‘nanobottles’

Article abstract of DOI:10.1016/j.ica.2019.05.032

Carbon nanotubes are appealing imaging and therapeutic systems. Their structure allows not only a useful display of molecules on their outer surface but at the same time the protection of encapsulated cargoes. Despite the interest they have provoked in the scientific community, their applications have not yet been fully realised due to the limited knowledge we possess concerning their physiological behaviour. Previously, we have shown that the encapsulation of radionuclide in the inner space of glycan-functionalized single-walled carbon nanotubes (glyco-X@SWCNT) redirected in vivo distribution of radioactivity from the thyroid to the lungs. Here we test the roles played by such glycans attached to carbon nanotubes in controlling sites of accumulation using nanotubes carrying both ‘cold’ and ‘hot’ salt cargoes decorated with two different mammalian carbohydrates, N-acetyl-D-glucosamine (GlcNAc) or galactose (Gal)-capped disaccharide lactose (Gal–Glc). This distinct variation of the terminal glycan displayed between two types of glycan ligands with very different in vivo receptors, coupled with altered sites of administration, suggest that distribution in mammals is likely controlled by physiological mechanisms that may include accumulation in the first capillary bed they encounter and not by glycan-receptor interaction and that the primary role of glycan is in aiding the dispersibility of the CNTs.

Full text of DOI:10.1016/j.ica.2019.05.032

InorganicaChimicaActa495(2019)118933
Contents lists available at ScienceDirect
Inorganica Chimica Acta
journal homepage: www.elsevier.com/locate/ica
Research paper
In vivo behaviour of glyco-NaI@SWCNT ‘nanobottles’
⁎
Sonia De Munaria, Stefania Sandovalb, Elzbieta Pachc, Belén Ballesterosc, Gerard Tobiasb,
,
⁎
⁎
Daniel C. Anthonyd, , Benjamin G. Davisa,
a Chemistry Research Laboratory, Department of Chemistry, University of Oxford, Mansfield Road, Oxford OX1 3TA, UK
b Institut de Ciència de Materials de Barcelona (ICMAB-CSIC), 08193 Bellaterra, Barcelona, Spain
c Catalan Institute of Nanoscience and Nanotechnology (ICN2), CSIC and The Barcelona Institute of Science and Technology, Campus UAB, Bellaterra, 08193, Barcelona,
Spain
d Department of Pharmacology, University of Oxford, Mansfield Road, Oxford OX1 3QT, UK
A R T I C L E I N F O
A B S T R A C T
Keywords:
Carbon nanotubes are appealing imaging and therapeutic systems. Their structure allows not only a useful
display of molecules on their outer surface but at the same time the protection of encapsulated cargoes. Despite
the interest they have provoked in the scientific community, their applications have not yet been fully realised
due to the limited knowledge we possess concerning their physiological behaviour. Previously, we have shown
that the encapsulation of radionuclide in the inner space of glycan-functionalized single-walled carbon nano-
tubes (glyco-X@SWCNT) redirected in vivo distribution of radioactivity from the thyroid to the lungs. Here we
test the roles played by such glycans attached to carbon nanotubes in controlling sites of accumulation using
nanotubes carrying both ‘cold’ and ‘hot’ salt cargoes decorated with two different mammalian carbohydrates, N-
acetyl-D-glucosamine (GlcNAc) or galactose (Gal)-capped disaccharide lactose (Gal–Glc). This distinct variation
of the terminal glycan displayed between two types of glycan ligands with very different in vivo receptors,
coupled with altered sites of administration, suggest that distribution in mammals is likely controlled by phy-
siological mechanisms that may include accumulation in the first capillary bed they encounter and not by
glycan-receptor interaction and that the primary role of glycan is in aiding the dispersibility of the CNTs.
Nanocrystals
Nanocapsules
Glycosylation
Encapsulation
Radioactive
1. Introduction
[20,21]. The salts remain stably confined in the form of ‘nanocrystals’
inside the nanotubes while leaving the outer surface essentially un-
The inner cavities of carbon nanotubes (CNTs) can accommodate a
wide range of guest species [1–4]. Unprecedented structures and
properties compared to those of the same material in the bulk can be
observed when they are confined [5–11]. In the biomedical field,
contrast agents and therapeutic compounds can be either attached to
the external CNT walls or confined within the cavities of the CNTs
[12–18]. The latter is attractive because CNTs can offer striking pro-
tection to chosen payloads, avoiding their interaction with the biolo-
gical milieu [19].
affected, and so ready to be modified by organic molecules.
As-produced, CNTs are insoluble in almost any aqueous solution and
organic solvent, and have been suggested to be toxic to mammalian
cells [22], thereby presenting perceived limitations to their biological
applications [23–27]. Functionalization of CNT side-walls with biolo-
gically- and biotechnologically- relevant molecules (including polymers
[28], peptides [29,30], nucleic acids [31] and carbohydrates [32,33])
allows the generation of potentially stable and biocompatible disper-
sions. For example, non-covalent binding of aromatic molecules by π–π
stacking onto the surface of the nanotubes [34] or covalent modifica-
tion of their polyarenic surface [35] allow loading of multiple mole-
cules along the length of the nanotubes.
Several strategies have been developed for the encapsulation of
materials inside carbon nanotubes. Once filled, unless there is a strong
interaction between the host nanotubes and the guest species, the ends
of the CNTs need to be sealed/closed to allow selective purification
from non-encapsulated materials left external to the CNT. Heating na-
notubes together with inorganic salts at high temperatures allows ca-
pillary permeation of the melted salts inside the nanotubes with the
spontaneous closure of the extremities during the cooling process
We have previously shown that encapsulation of radionuclide into
the inner space of glycan-functionalized single-walled carbon nano-
tubes (glyco-X@SWCNT) may be achieved by molten filling and then
covalent modification, allowing in vivo redirection of the distribution of
the associated radioactivity from the thyroid to the lungs [33]. Here, we
⁎ Corresponding authors.
E-mail addresses: gerard.tobias@icmab.es (G. Tobias), daniel.anthony@pharm.ox.ac.uk (D.C. Anthony), ben.davis@chem.ox.ac.uk (B.G. Davis).
https://doi.org/10.1016/j.ica.2019.05.032
Received 3 February 2019; Received in revised form 16 May 2019; Accepted 17 May 2019
Availableonline21May2019
0020-1693/©2019PublishedbyElsevierB.V.

S. De Munari, et al.
InorganicaChimicaActa495(2019)118933
use steam-purified and shortened single-walled carbon nanotubes
(SWCNTs) [36] filled with both ‘cold’ (NaI) and ‘hot’ (Na125I) cargoes
and subsequent functionalization with different carbohydrates to ex-
plore the basis and role of glycan in this redistribution.
hydroxybenzotriazole (HOBt, 743 mg, 5.5 mmol, 2 eq.) and di-iso-
propylethylamine (DIPEA, 960 μL, 5.5 mmol, 2 eq.). Finally, benzyl (2-
(2-(2-aminoethoxy)ethoxy)ethyl)glycinate
(978.8 mg,
3.3 mmol,
1.2 eq.) was added and the mixture stirred at r.t. for 6 h. The reaction
was evaporated to a small volume, recovered with DCM (150 mL) and
washed with H2O (200 mL) and brine (200 mL). The organic layer was
separated, dried over MgSO4, filtered and the filtrate evaporated under
vacuum, to afford a yellow oil (1.846 g, 2.55 mmol, 92%).
2. Experimental
2.1. Purification of SWCNTs
Step 2) To Pd/C (10 mg, 10 wt% loading) under hydrogen, was
added a solution of the product from step 1 (386 mg, 0.53 mmol, 1 eq.)
in MeOH (dry, 3 mL). The reaction mixture was stirred at r.t. for 5 h,
filtered through celite and solvent removed under vacuum. The crude
product was purified by flash column chromatography on silica
(DCM:MeOH, 9:1 to 0:1) to afford the title compound as a colourless oil
(220.4 mg, 0.35 mmol, 66%).
Chemical vapour deposition (CVD) grown SWCNTs, were provided
by Thomas Swan & Co. Ltd (Elicarb®). Steam purification was carried
out in order to remove the amorphous carbon and graphitic shells
formed during the synthesis [36]. Steam treatment was simultaneously
employed to open the SWCNTs ends. For this purpose, 1 g of as-received
SWCNTs were ground with agate mortar and pestle and then placed
inside a tubular furnace. Steam was introduced by bubbling argon
through hot water. The temperature was raised by 20 °C/min till 900 °C,
where the SWCNTs remained for 25 h. After cooling at a rate of 10 °C/
min down to 25 °C, the powder was dispersed in a 6 M HCl solution and
refluxed at 110 °C during 6 h to remove the iron catalyst exposed after
the steam oxidation of the graphitic shells. The mixture was cooled,
filtered using a 0.2 µm polycarbonate membrane, washed with distilled
water until neutral pH was reached and dried overnight at 60 °C. After
this treatment, SWCNTs with a median length of ca. 200 nm are ob-
tained [37].
A suspension of NaI@SWCNTs (6 mg) in DMF (dry, 3 mL) was so-
nicated for 2 min, before a solution of Linker Unit 1.0 (12.7 mg,
0.02 mmol, 1 eq.) and 2,3,5-triiodobenzaldehyde [33] (9.6 mg,
0.02 mmol, 1 eq.) in DMF (dry, 1 mL) was added. The reaction was
refluxed at 130 °C for 4 days, then cooled to r.t. and filtered. The residue
was washed with DMF, MeOH and dried under vacuum to afford di-Boc-
f-NaI@SWCNTs as a black solid (6.8 mg).
A suspension of di-Boc-f-NaI@SWCNTs (12.2 mg) in DCM (dry,
5 mL) was sonicated for 2 min, before TFA (2.5 mL) was added. The
reaction mixture was stirred at r.t. for 24 h, then evaporated, recovered
with MeOH, filtered and washed with MeOH. The solid obtained was
dried under vacuum to afford the title compound as a black solid
(9.46 mg).
2.2. Filling of SWCNTs with NaI by molten phase capillary wetting
Purified and open-ended SWCNTs (100 mg) and NaI (1 g) were
ground together and loaded into a silica ampoule. The system was
evacuated and the ampoule was sealed under vacuum. The sample was
subsequently annealed at 900 °C and slowly cooled down to favour the
crystallization of NaI within the hosting SWCNTs. Afterwards, the
system was air opened and the sample was ground with an agate mortar
and pestle. The sample was next washed in water, to remove the non-
encapsulated NaI, and collected by filtration on top of a polycarbonate
membrane (0.2 μm Whatman).
2.4. Fmoc numbering f-NaI@SWCNTs
A solution of Fmoc chloride (1 mg, 3.86 μmol, 1 eq.) in DCM
(0.5 mL) was added dropwise to a suspension of f-NaI@SWCNTs (1 mg)
in DCM (0.5 mL) at 0 °C. DIPEA (1.5 μL, 8.61 μmol, 2.2 eq.) was added
and the reaction was stirred at r.t. for 16 h. The reaction mixture was
centrifuged, the supernatant discarded and the solid recovered with
MeOH then filtered and washed with MeOH. The solid obtained was
dried under vacuum and treated with DMSO:DMF:DBU (25:24:1 solu-
tion, 1 mL) at r.t. to cleave the Fmoc groups. After 2 h, 20 μL of the
reaction mixture were added to 1000 μL of MeOH. The sample was
centrifuged and the absorbance of the supernatant measured by UV at
295 nm (A = 0.0657, ε = 10,027 cm-1M−1). The functionalization level
of the nanotubes obtained was 0.34 mmol/g.
2.3. Synthesis of f-NaI@SWCNTs
4-(2-(bis(2-tert-butoxycarbonyl-aminoethyl)amino)ethylamino)-4-ox-
obutanoic acid: To
a solution of tris(2-aminoethyl)amine (6 mL,
41 mmol, 5 eq.) in MeCN (200 mL) at 0 °C was added, dropwise, a so-
lution of succinic anhydride (822 mg, 8.2 mmol, 1 eq.) in MeCN
(100 mL). The reaction mixture was stirred at r.t. for 2 h, before the
supernatant was decanted off and the residue redissolved in MeOH, and
finally dried under vacuum to afford 4-(2-(bis(2-aminoethyl)amino)
ethylamino)-4-oxobutanoic acid as a yellow oil (2.0 g, 8 mmol, 98%).
2.5. Synthesis of Glycosylated-NaI@SWCNTs
2-acetamido-1-thio-(S-2-imido-2-methoxyethyl)-2-deoxy-β-D-glucopyr-
anoside 1.1: To a solution of 2-N-acetamido-3,4,6-tri-O-acetyl-2-deoxy-
1-thio-(S-cyanomethyl)-β-D-glucopyranoside [39] (20 mg, 0.05 mmol,
1 eq.) in MeOH (dry, 1 mL) was added NaOMe (25% in MeOH, 12 μL,
0.05 mmol, 1 eq.). The reaction was stirred at r.t. for 16 h before being
neutralised with DowexH+. The reaction mixture was filtered and
evaporated without heating. The product was obtained as a mixture of
deprotected cyanomethyl (R-SCM) and ‘activated’ sugars (R-IME, 1.1))
in ratio 1:0.5 (determined by 1H NMR (CD3OD)), and used in the next
step without further purification.
To
a solution of 4-(2-(bis(2-aminoethyl)amino)ethylamino)ox-
obutanoic acid (343 mg, 1.39 mmol, 1 eq.) in dioxane (15 mL) was
added di-tert-butyl dicarbonate (611 mg, 2.8 mmol, 2 eq.). The reaction
mixture was stirred for 4 h at r.t., the solvent removed and the residue
purified by flash column chromatography over silica (DCM:MeOH,
10:0.2 to 10:2), to afford the title compound as a white solid (246.8 mg,
0.55 mmol, 40%).
Benzyl (2-(2-(2-aminoethoxy)ethoxy)ethyl)glycinate: To a solution of
benzyl (2-(2-(2-(N-tert-butoxycarbonyl)aminoethoxy)ethoxy)ethyl)gly-
cinate [38] (1.1056 g, 2.79 mmol, 1 eq.) in DCM (5.6 mL) was added
trifluoroacetic acid (TFA, 2.4 mL). The reaction was stirred at r.t. for
1 h, before removing the solvent under vacuum to afford the title
compound as a yellow oil (1.08 g, 2.79 mmol, quant.).
2,3,6-O-tri-O-acetyl-4-O-[2,3,4,6-tetra-O-acetyl-β-D-galactopyranosyl]-
1-thio-(S-cyanomethyl)-β-D-glucopyranoside To lactose (10 g, 29.2 mmol,
1 eq.) were added Ac2O (155 mL, 1636 mmol, 56 eq.) and NaOAc (10 g,
122.6 mmol, 4.2 eq.). The reaction was stirred at 140 °C for 5 h, re-
covered with H2O (200 mL) and extracted with DCM (3 × 200 mL). The
organic layer was washed with NaHCO3 satd. solution (200 mL) and
brine (200 mL), then dried over MgSO4, filtered and the filtrate eva-
porated and co-evaporated with toluene, to afford 1,2,3,6-tetra-O-
acetyl-4-O-[2,3,4,6-tetra-O-acetyl-β-D-galactopyranosyl]-β-D-glucopyr-
anoside [40] as a light yellow foam (19.9 g, 29.2 mmol, quant.).
Linker Unit 1.0
Step 1) To a solution of 4-(2-(bis(2-tert-butoxycarbonyl-aminoethyl)
amino)ethylamino)-4-oxobutanoic acid (1.230 g, 2.75 mmol, 1 eq.) in
DMF (55 mL), were added hexfluorophosphate azabenzotriazole tetra-
methy
uronium
(HATU,
2.091 g,
5.5 mmol,
2 eq.),
2

S. De Munari, et al.
InorganicaChimicaActa495(2019)118933
To
1,2,3,6-tetra-O-acetyl-4-O-[2,3,4,6-tetra-O-acetyl-β-D-galacto-
di-Boc-f-Na125I@SWCNTs as a black solid (1 mg, 9 MBq).
pyranosyl]-β-D-glucopyranoside (14.15 g, 20.85 mmol, 1 eq.) in an ice
bath, were added HBr (33% in AcOH, 23 mL, 129.27 mmol, 6.2 eq.) and
Ac2O (4.2 mL, 43.785 mmol, 2.1 eq.). The reaction was stirred for 3 h,
recovered with DCM (200 mL) and poured into H2O (200 mL). The
mixture was stirred with NaHCO3 satd. solution (300 mL), the organic
layer separated and washed with NaHCO3 satd. solution (3 × 200 mL)
and brine (200 mL). The organic layer was dried over MgSO4, filtered
and the filtrate evaporated to afford 2,3,6-tri-O-acetyl-4-O-[2,3,4,6-
tetra-O-acetyl-β-D-galactopyranosyl]-1-bromo-α-D-glucopyranoside
[41] as a white foam (13.27 g, 19 mmol, 91%).
A suspension of di-Boc-f-Na125I@SWCNTs (1 mg, 9 MBq) in DCM
(dry, 5 mL) was sonicated for 2 min, before TFA (2.5 mL) was added.
The reaction mixture was stirred at r.t. for 24 h, then evaporated, re-
covered with MeOH, filtered and washed with MeOH. The solid ob-
tained was dried under vacuum to afford the title compound f-Na125I@
SWCNTs as a black solid (1 mg, 9 MBq).
GlcNAc-Na125I@SWCNTs: A solution of crude 2-acetamido-1-thio-
(S-2-imido-2-methoxyethyl)-2-deoxy-β-D-glucopyranoside 1.1 (6.5 mg,
20 μmol, 1 eq.) in MeOH:DMSO (2:1, 1.5 mL) was added to a suspension
of f-Na125I@SWCNTs (0.3 mg, 3 MBq) in DCM (0.5 mL). The reaction
was stirred at r.t. for 30 min. The reaction mixture was centrifuged, the
supernatant discarded and the solid recovered with MeOH, filtered and
washed with MeOH. The solid obtained was dried under vacuum to
afford the title compound as a black solid (0.3 mg, 3 MBq).
2,3,6-tri-O-acetyl-4-O-[2,3,4,6-tetra-O-acetyl-β-D-galactopyr-
anosyl]-1-bromo-α-D-glucopyranoside (14.58 g, 20.85 mmol, 1 eq.) was
mixed with thiourea (3.5 g, 45.87 mmol, 2.2 eq.) and dissolved in
acetone (50 mL). The reaction was stirred for 3 h at 80 °C, then eva-
porated. The solid was recovered with acetone (60 mL), and Na2S2O5
(10.3 g, 54.21 mmol, 2.6 eq.), K2CO3 (4 g, 29.19 mmol, 1.4 eq.) and
chloroacetonitrile (20 mL, 316.92 mmol, 15.2 eq.) were added. The
reaction mixture was stirred at r.t. for 3 h, evaporated and purified by
flash column chromatography on silica (EtOAc:PE, 1:1) to afford the
title compound as a white foam (8.7621 g, 12.6 mmol, 63%).
4-O-[β-D-galactopyranosyl]-1-thio-(S-2-imido-2-methoxyethyl)-β-D-glu-
copyranoside 1.2: To a solution of 2,3,6-tri-O-acetyl-4-O-[2,3,4,6-tetra-
O-acetyl-β-D-galactopyranosyl]-1-thio-(S-cyanomethyl)-β-D-glucopyr-
anoside (100 mg, 0.14 mmol, 1 eq.) in MeOH (dry, 1 mL) was added
NaOMe (25% in MeOH, 33 μL, 0.14 mmol, 1 eq.). The reaction was
stirred at r.t. for 16 h before being neutralised with DowexH+. The
reaction mixture was filtered and evaporated without heating. The
product was obtained as a mixture of the deprotected cyanomethyl (R-
SCM) and ‘activated’ sugars (R-IME, 1.2) in ratio 1:0.7 (determined by
1H NMR (CD3OD)), and used in the next step without further pur-
ification.
Lac-Na125I@SWCNTs:
A
solution of crude 4-O-[β-D-galactopyr-
1.2
anosyl]-1-thio-(S-2-imido-2-methoxyethyl)-β-D-glucopyranoside
(6.0 mg, 20 μmol, 1 eq.) in MeOH:DMSO (2:1, 1.5 mL) was added to a
suspension of Na125I@SWCNTs (0.3 mg, 3 MBq) in DCM (0.5 mL). The
reaction was stirred at r.t. for 30 min. The reaction mixture was cen-
trifuged, the supernatant discarded and the solid recovered with MeOH,
filtered and washed with MeOH. The solid obtained was dried under
vacuum to afford the title compound as a black solid (0.3 mg, 3 MBq).
For additional schemes, structures and characterization details,
namely, TGA, HRTEM, STEM, TLC, NMR mass spectrometry and in-
frared spectroscopy see the Supporting Information.
3. Results and discussion
3.1. Preparation of ‘cold’ glyco-NaI@SWCNTs
To prepare the nanotubes for filling they were first treated with
steam, followed by an HCl (aq) wash, in order to remove graphitic
nanoparticles, amorphous carbon and metal catalysts that remain as
impurities from their generation [36]. This method also results in si-
multaneous shortening of the nanotubes via a process believed to in-
volve oxidation and decarboxylation of more reactive carbon sites
present at their tips [36]. TEM images of both as-received and steam-
purified SWCNTs are shown in Fig. S1; these revealed some residual
iron-derived nanoparticles (from the preparative catalyst) after steam
treatment. Sodium iodide (hot or cold) was filled into these carbon
nanotubes to generate NaI@SWCNTs by adaptation of the protocol
developed by Green et al. for the creation of KI ‘nanocrystals’ in
SWCNTs [5]. Thus, a mixture of steam purified SWCNTs and the metal
halide was annealed above the melting point of the inorganic salt
(m.p.NaI = 661 °C) inside an evacuated silica ampoule. Heating at
900 °C not only drove encapsulation of salt inside the nanotubes, but
also induced the closing of their tips [20]. In this way, internal NaI
crystals were isolated from the outer environment by the formation of
carbon ‘nanocapsules’ [42] (or ‘nanobottles’ [30]). As a result, any re-
sidual, external NaI present after synthesis was easily removed simply
by washing the sample in refluxing water.
GlcNAc-NaI@SWCNTs: A solution of crude 2-acetamido-1-thio-(S-2-
imido-2-methoxyethyl)-2-deoxy-β-D-glucopyranoside
1.1
(6.5 mg,
20 μmol, 1 eq.) in MeOH:DMSO (2:1, 1.5 mL) was added to a suspension
of f-NaI@SWCNTs (1 mg) in DCM (0.5 mL). The reaction was stirred at
r.t. for 30 min. The reaction mixture was centrifuged, the supernatant
discarded and the solid recovered with MeOH, filtered and washed with
MeOH. The solid obtained was dried under vacuum to afford the title
compound as a black solid (0.88 mg). Elemental analysis: (C, H, N)
67.34, 1.08, 2.98.
Lac-NaI@SWCNTs: A solution of crude 4-O-[β-D-galactopyranosyl]-
1-thio-(S-2-imido-2-methoxyethyl)-β-D-glucopyranoside 1.2 (6.5 mg,
20 μmol, 1 eq.) in MeOH:DMSO (2:1, 1.5 mL) was added to a suspension
of f-NaI@SWCNTs (1 mg) in DCM (0.5 mL). The reaction was stirred at
r.t. for 30 min. The reaction mixture was centrifuged, the supernatant
discarded and the solid recovered with MeOH, filtered and washed with
MeOH. The solid obtained was dried under vacuum to afford the title
compound as a black solid (0.8 mg). Elemental analysis: (C, H, N)
67.32, 1.11, 3.01.
2.6. Synthesis of glycosylated-Na125I@SWCNTs
Characterization by high-angle annular dark-field scanning trans-
mission electron microscopy (HAADF-STEM) allowed encapsulated salt
to be clearly discerned from the walls of the nanotubes (Fig. 1). Visual
inspection allowed filled SWCNTs containing heavy atoms (Fig. S2,
white arrowed) to be readily distinguished from empty (red arrowed).
This also confirmed that washing with water after filling successfully
removed all residual salt from the outer surfaces of the CNTs in the
sample. Successful encapsulation of NaI was also confirmed using high-
resolution transmission electron microscopy (HRTEM, Fig. 1b), al-
lowing observation of even the crystalline lattice of encapsulated NaI.
Finally, analysis of the sample with energy dispersive X-ray spectro-
scopy (EDX, Fig. 1c) further confirmed the presence of Na and I within
the encapsulated cargo.
Filling of SWCNTs with Na125I: Steam-purified SWCNTs (1 mg) were
loaded in a silica ampoule with Na125I solution (10-5 M, 100 μL,
20 MBq), dried and sealed under high vacuum with an oxygen-propane
flame. The ampoule was heated in a furnace up to 900 °C using the same
ramp described for the cold samples. The sample was recovered from
the ampoule with H2O (1 mL), sonicated and filtered. Final
activity = 9 MBq (1 mg, RCY = 45%).
f-Na125I@SWCNTs: A suspension of Na125I@SWCNTs (1 mg, 9 MBq)
in DMF (dry, 3 mL) was sonicated for 2 min, before a solution of Linker
Unit 1.0 (12.7 mg, 0.02 mmol, 1 eq.) and 2,3,5-triiodobenzaldehyde
(9.6 mg, 0.02 mmol, 1 eq.) in DMF (dry, 1 mL) was added. The reaction
was refluxed at 130 °C for 4 days, then cooled to r.t. and filtered. The
solid was washed with DMF, MeOH and dried under vacuum to afford
Following successful filling to form NaI@SWCNTs, covalent
3

S. De Munari, et al.
InorganicaChimicaActa495(2019)118933
Fig. 1. NaI-filled SWCNTs (NaI@SWCNTs). (a) HAADF-STEM; since the intensity of the signal scales up to ca the square of the atomic number, the heavier atoms of
the guest (Na and I) appear as brighter lines along the inner cavities of the nanotubes compared to C from their walls; (b) HRTEM image; (c) EDX spectroscopy
confirmed the presence of Na and I. Fe and Cu signals correspond to those from SWCNT-generation catalyst residuals and the copper grid used for supporting the
sample, respectively. Si signal arises from the EDX detector.
Scheme 1. Functionalization of NaI@SWCNTs with biantennary linker reagent 1.0 and subsequent glycosylation. Reagents and conditions: a) DMF, 130 °C, 4 days; b)
DCM, TFA, r.t., 24 h; c) 1.1 or 1.2, MeOH:DMSO (2:1), r.t., 30 min.
functionalization of the sidewalls was achieved under mild conditions
via 1,3-dipolar cycloaddition using appropriately functionalized azo-
methine ylids [14,43,44] (Scheme 1). The choice of ylid also simulta-
neously allowed introduction of a 2,3,5-triiodophenyl motif as a
‘marker’ or ‘tagging’ motif,[45] suitable for the ready detection of
successful functionalization using electron microscopy. To allow more
ready diversification of glycan, a prior strategy devised by Hong et al.
[33] was altered through first introducing a Boc-protected bis-amine
branched linker moiety 1.0 to the CNT (Scheme 1), followed by de-
protection using trifluoroacetic acid (TFA) to form a functionalized
SWCNT as a divergent intermediate, f-NaI@SWCNT. Fmoc-numbering
[33] was employed to quantify the loading of the primary amine groups
thus introduced; the degree of functionalization at 0.34 mmol/g.
Intermediate f-NaI@SWCNT was then subsequently glycoconju-
gated with appropriate amine-reactive glycan-IME imidate reagents
[46,47], GlcNAc-IME 1.1 or Lac-IME 1.2 to introduce GlcNAc or lactose
(Galβ1,4–Glc) to the linkers attached to surface of the NaI@SWCNT,
respectively, forming [GlcNAc]2-NaI@SWCNT or [Lac]2-NaI@SWCNT
(Scheme 1).
HAADF-STEM imaging of the samples after GlcNAc functionaliza-
tion Fig. 2 revealed a higher intensity in the bundle area when com-
pared to the as-filled NaI@SWCNTs, due to the presence of the 2,3,5-
triiodophenyl ‘marker’ combined into the linker moiety, thereby con-
firming successful covalent functionalization. The introduction of the
linker to the surface of NaI@SWCNTs in f-NaI@SWCNTs, [GlcNAc]2-
NaI@SWCNT and [Lac]2-NaI@SWCNT was also confirmed by TGA
(Fig. 3). Unlike the NaI@SWCNTs whose combustion starts ∼400 °C,
weight loss at lower temperatures even down to ∼200 °C was observed
that can be attributed to the combustion of the organic fraction at-
tached to the nanotubes. GlcNAc and Lac functionalized SWCNTs
showed higher thermal stabilities than the linker-functionalized
SWCNT f-NaI@SWCNTs. During functionalization inorganic materials
are employed (e.g. MgSO4) that appear to remain in the functionalized
samples and that contribute to an unexpected increase in the amount of
inorganic residue collected after the TGA for those samples that have
been through these processes (SI Fig. S4). Successful glycoconjugation
4

S. De Munari, et al.
InorganicaChimicaActa495(2019)118933
Fig. 2. HAADF-STEM images of (a) as-filled sample and (b) after GlcNAc functionalization to form [GlcNAc]2-NaI@SWCNT. The respective HAADF intensity profiles
along the white dashed boxes demonstrate that the intensity in the bundle area is higher for the [GlcNAc]2-NaI@SWCNT, in agreement with the presence of iodide
(Z = 53) in surface functional groups that have a substantial contribution to the signal.
glycosylated nanotubes in vivo we created ‘hot nanobottle’ isotopologue
variants in which we replaced ‘cold’ NaI with ‘hot’, radioactive Na125I.
Gamma emission from 125I allows sensitive quantification and hence
whole body distribution of constructs through ‘gamma counting’ of
samples. Empty SWCNTs underwent an adapted molten phase filling
process: SWCNTs and an aqueous solution of the radioactive salt were
placed in a silica ampoule and heated to remove water. The ampoule
was then sealed under vacuum and annealed following the protocol
employed for ‘cold’ NaI. The functionalization and glycoconjugation
process was carried out in an essentially analogous manner to that used
for non-radioactive samples to generate ‘hot’ f-Na125I@SWCNTs
(0.3 mg, 3 MBq), ‘hot’ [GlcNAc]2-Na125I@SWCNTs (0.3 mg, 3 MBq) and
‘hot’ [Lac]2-Na125I@SWCNTs (0.3 mg, 3 MBq).
3.3. Biodistribution analyses of ‘hot’ and ‘cold’ glyco-NaI@SWCNTs.
Previous studies performed on the robustness of radionuclide-filled
Fig. 3. TGA analyses of NaI filled SWCNTs (NaI@SWCNTs), and after being
nanotubes in biological environments have shown a strong correlation
externally functionalized (f-NaI@SWCNTs) and glycosylated ([GlcNAc]2-NaI@
between radioactivity and nanotube distribution, suggesting negligible
SWCNTs and [Lac]2-NaI@SWCNTs).
leakage of radionuclide salts [33,48]. This is particularly the case when
using radio-iodide as ‘cargo’; free iodide is readily taken up by the
was further confirmed by elemental combustion analyses (see Supple-
mentary Information).
thyroid leading to sensitive, easily detected observation – no such free
iodide was detected in these prior studies. We used this correlation to
quantify the distribution of nanotubes via gamma emission from the
encapsulated Na125I cargo. ‘Hot’ glyco-Na125I@SWCNTs were injected
in CD-1 male mice (0.1 mg/100 µL) and biodistribution after 1 h de-
termined from gamma radiation levels in different organs (lungs, liver,
spleen, brain, gut, kidneys, muscle and heart) ex vivo (Fig. 4).
3.2. Preparation of ‘hot’ glyco-Na125I@SWCNTs.
One mode in which glycosylated-‘nanocapsules/bottle’ might be
exploited is in the highly controlled and localized delivery of radio-
activity. This also potentially allows quantification of biodistribution
and even direct imaging of this delivery. To test the potential of our
Two variants of mammalian glycans were chosen to test influence
upon biodistribution. Both have putative endogenous receptors in
5

S. De Munari, et al.
InorganicaChimicaActa495(2019)118933
4. Conclusions
Functionalized and glycosylated nanotubes can act as ‘nano-
capsules’ or ‘nanobottles’ to redirect the accumulation of inorganic
radionuclide ‘cargo’ concealed in their inner void (here iodide) from a
natural physiological target (here thyroid) [50] to alternative targets.
At the functionalization loadings used here, different glycosylation
patterns on these glyco-‘nanobottle’ constructs did not modulate in vivo
distribution profiles. Instead, these appear to only aid dispersibility;
these dispersed carbon nanotubes then appear to accumulate in a
manner that is primarily determined by their physical properties within
physiology (e.g. recruitment at the first capillary bed that they en-
counter) in a manner that is more likely to be determined by admin-
istration site than by any specific ligand-receptor targeting interaction
at these lower loading of surface ligands on filled SWCNTs. It should be
noted that overly high rates of injection can affect distribution. We
injected 100 μL of suspension over a 10-second period in which the
heart would normally pump ∼2.5–3.0 mL of blood. The mouse has a
blood volume of ∼1.5 mL. Thus our injection was less than 4% volume
at a rate well below suggested typical maximum of, e.g. 50 μL/s [51].
These ‘nanocapsules/bottles’ may represent a useful, sealed ‘source
of radiation’. If this could be combined with the benefits of a delivery
system, then this could create a potentially suitable form of ‘nano-
brachytherapy’. Given the seemingly dominant control of physical
properties rather than biochemical properties upon distribution ob-
served here, we speculate that future manipulation of not only surface
functionalization (e.g. at higher levels or with different types) but even
via changes to carbon nanostructure at a SWCNT level might usefully
influence their in vivo capabilities.
Fig. 4. Biodistribution of amine-functionalized and glycosylated Na125I@
SWCNTs. Organs of CD-1 mice were harvested 1 h after tail-vein injection of
amine-functionalized (f-Na125I@SWCNTs) and glycosylated nanotubes
(GlcNAc-Na125I@SWCNTs; Lac-Na125I@SWCNTs). The distribution of the
samples in the tissues was determined by ɣ-counting and expressed as per-
centage of the injected dose per g of tissue (%ID/g).
mammals with different in vivo distribution. GlcNAc was used in our
prior study in which functionalized NaI@SWCNTs containing radio-
active iodide was successfully delivered into the lungs [33] and was
again used here in [GlcNAc]2-Na125I@SWCNT. Galactosyl-terminated,
Lac, is sometimes considered to be a ‘liver-targeting’ agent due to its
interaction as a ligand with the asialoglycoprotein receptors expressed
on hepatocytes [49] and was selected in this study as the second glycan
and used in [Lac]2-Na125I@SWCNT. The distribution profile de-
termined from both [GlcNAc]2-Na125I@SWCNT and [Lac]2-Na125I@
SWCNT (Fig. 4), however, showed no significant differences in biodis-
tribution with essentially complete accumulation of the radioactivity in
the lungs for both samples after tail-vein injection. This negligible in-
fluence of glycan was further confirmed by the identical in vivo beha-
viour of non-glycosylated precursor f-Na125I@SWCNT (Fig. 4).
Next, in order to test whether biodistribution could be attributed to
dominant, inherent physiological properties of the nanotubes, we at-
tempted to assess the effect of an alternative injection site via intra-
aortic injection that necessitated preferential use of ‘cold’ glyco-
SWCNTs and quantification instead by ICP-MS of tissue-derived sam-
ples. These analyses calibrated by internal standard revealed a lower
method detection limit for iodide in tissue under these conditions
of ∼25 ng/g.tissue; iodide was also successfully observed in samples
from animals treated with glyco-NaI@SWCNT at levels of up to 126 ng/
g.tissue. However, although iodide levels were found in certain organs
(liver, lung and spleen) of up to 60 ng/g.tissue above those found in
control, untreated animals, these should be treated as only suggestive
qualitative indications – statistical analyses do not allow support of any
quantitative significance. Thus, whilst these may be indicative of
broader organ distribution caused by an alternative injection site, the
use of ‘cold iodide’ tracking of filled, functionalized NaI@SWCN in vivo
by ICP-MS analysis of tissues appears to lack the sensitivity required for
unambiguous biodistribution analysis.
Acknowledgements
This works was financially supported by EU FP7-ITN Marie-Curie
Network programme RADDEL [grant number 290023] and the EU FP7-
Integrated Infrastructure Initiative–I3 programme ESTEEM2 [grant
number 312483]. We also acknowledge financial support from Spanish
Ministry of Economy and Competitiveness through the “Severo Ochoa”
Programme for Centres of Excellence in R&D [grant numbers SEV-2015-
0496, ICMAB; SEV-2017-0706, ICN2]. The ICN2 is funded by the
CERCA programme. We would like to thank Thomas Swan & Co. Ltd.
for supplying Elicarb® SWNT.
Finally and above all, this work on probing the potential of such
‘nanobottles’ in physiology has been inspired by many stimulating and
fruitful conversations with Prof Malcolm Green – his insight and vision,
as on many other occasions, has proven invaluable.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.ica.2019.05.032.
References
[1] J. Sloan, A.I. Kirkland, J.L. Hutchison, M.L.H. Green, Chem. Commun. (2002)
1319–1332.
[2] M. Monthioux, E. Flahaut, Mater. Sci. Eng. C 27 (2007) 1096–1101.
[3] U.K. Gautam, P.M.F.J. Costa, Y. Bando, X. Fang, L. Li, M. Imura, D. Golberg, Sci.
Technol. Adv. Mater. 11 (2010) 054501.
[4] A.N. Khlobystov, ACS Nano 5 (2011) 9306–9312.
Together these data suggested that, at the limited loadings of sur-
face functionalization used here (< 0.7 mmol of glycan / g of CNT),
glycosylation serves only to aid dispersibility rather than providing any
form of ‘targeting’. Instead, once dispersed, glyco-SWCNTs may deliver
their ‘cargo’ via various physiological mechanisms that may include
recruitment at the first capillary bed that they encounter after i.v. ad-
ministration, e.g. tail → lung.
[5] R.R. Meyer, J. Sloan, R.E. Dunin-Borkowski, A.I. Kirkland, M.C. Novotny,
S.R. Bailey, J.L. Hutchison, M.L.H. Green, Science 289 (2000) 1324–1326.
[6] E. Philp, J. Sloan, A.I. Kirkland, R.R. Meyer, S. Friedrichs, J.L. Hutchison,
M.L.H. Green, Nat. Mater. 2 (2003) 788.
[7] D.G. Calatayud, H. Ge, N. Kuganathan, V. Mirabello, R.M.J. Jacobs, N.H. Rees,
C.T. Stoppiello, A.N. Khlobystov, R.M. Tyrrell, E.D. Como, S.I. Pascu,
ChemistryOpen 7 (2018) 144–158.
[8] A. Vasylenko, S. Marks, J.M. Wynn, P.V.C. Medeiros, Q.M. Ramasse, A.J. Morris,
J. Sloan, D. Quigley, ACS Nano 12 (2018) 6023–6031.
[9] L. Cabana, B. Ballesteros, E. Batista, C. Magén, R. Arenal, J. Oró-Solé, R. Rurali,
6

S. De Munari, et al.
InorganicaChimicaActa495(2019)118933
G. Tobias, Adv. Mater. 26 (2014) 2016–2021.
Actuator B-Chem. 122 (2007) 672–682.
[10] M. Hart, J. Chen, A. Michaelides, A. Sella, M.S.P. Shaffer, C.G. Salzmann, Angew.
Chem. Int. Edit. 57 (2018) 11649–11653.
[32] B.K. Gorityala, J. Ma, X. Wang, P. Chen, X.-W. Liu, Chem. Soc. Rev. 39 (2010)
2925–2934.
[11] S. Sandoval, D. Kepić, Á. Pérez del Pino, E. György, A. Gómez, M. Pfannmoeller,
G.V. Tendeloo, B. Ballesteros, G. Tobias, ACS Nano 12 (2018) 6648–6656.
[12] K.B. Hartman, D.K. Hamlin, D.S. Wilbur, L.J. Wilson, Small 3 (2007) 1496–1499.
[13] Z. Liu, W. Cai, L. He, N. Nakayama, K. Chen, X. Sun, X. Chen, H. Dai, Nat.
Nanotechnol. 2 (2006) 47.
[33] S.Y. Hong, G. Tobias, K.T. Al-Jamal, B. Ballesteros, H. Ali-Boucetta, S. Lozano-Perez,
P.D. Nellist, R.B. Sim, C. Finucane, S.J. Mather, M.L.H. Green, K. Kostarelos,
B.G. Davis, Nat. Mater. 9 (2010) 485.
[34] Z. Liu, X. Sun, N. Nakayama-Ratchford, H. Dai, ACS Nano 1 (2007) 50–56.
[35] C.R. Martin, P. Kohli, Nat. Rev. Drug Discov. 2 (2003) 29.
[36] B. Ballesteros, G. Tobias, L. Shao, E. Pellicer, J. Nogués, E. Mendoza, M.L.H. Green,
Small 4 (2008) 1501–1506.
[14] R. Singh, D. Pantarotto, L. Lacerda, G. Pastorin, C. Klumpp, M. Prato, A. Bianco,
K. Kostarelos, Proc. Natl. Acad. Sci. USA 103 (2006) 3357.
[15] M.R. McDevitt, D. Chattopadhyay, B.J. Kappel, J.S. Jaggi, S.R. Schiffman,
C. Antczak, J.T. Njardarson, R. Brentjens, D.A. Scheinberg, J. Nucl. Med. 48 (2007)
1180–1189.
[37] M. Kierkowicz, E. Pach, A. Santidrián, S. Sandoval, G. Gonçalves, E. Tobías-Rossell,
M. Kalbáč, B. Ballesteros, G. Tobias, Carbon 139 (2018) 922–932.
[38] K. Kordatos, T. Da Ros, S. Bosi, E. Vázquez, M. Bergamin, C. Cusan, F. Pellarini,
V. Tomberli, B. Baiti, D. Pantarotto, V. Georgakilas, G. Spalluto, M. Prato, J. Org.
Chem. 66 (2001) 4915–4920.
[16] M. Prato, K. Kostarelos, A. Bianco, Accounts Chem. Res. 41 (2008) 60–68.
[17] L. Cabana, M. Bourgognon, J.T.-W. Wang, A. Protti, R. Klippstein, R.T.M. de
Rosales, A.M. Shah, J. Fontcuberta, E. Tobías-Rossell, J.K. Sosabowski, K.T. Al-
Jamal, G. Tobias, Small 12 (2016) 2893–2905.
[39] S. De Munari, T. Schiffner, B.G. Davis, Isr. J. Chem. 55 (2015) 387–391.
[40] B.S. Patil, V.V.S. Babu, Indian J. Chem. 43B (2004) 1288–1291.
[41] M. Bergmann, Justus Liebigs Ann. Chem. 434 (1923) 79–110.
[42] G. Tobias, B. Ballesteros, M.L.H. Green, Phys. Status Solidi C 7 (2010) 2739–2742.
[43] W. Wu, S. Wieckowski, G. Pastorin, M. Benincasa, C. Klumpp, J.-P. Briand,
R. Gennaro, M. Prato, A. Bianco, Angew. Chem. Int. Edit. 44 (2005) 6358–6362.
[44] M. Maggini, G. Scorrano, M. Prato, J. Am. Chem. Soc. 115 (1993) 9798–9799.
[45] S.Y. Hong, G. Tobias, B. Ballesteros, F. El Oualid, J.C. Errey, K.J. Doores,
A.I. Kirkland, P.D. Nellist, M.L.H. Green, B.G. Davis, J. Am. Chem. Soc. 129 (2007)
10966–10967.
[18] S.S. Wong, E. Joselevich, A.T. Woolley, C.L. Cheung, C.M. Lieber, Nature 394
(1998) 52.
[19] C.J. Serpell, K. Kostarelos, B.G. Davis, ACS Central Sci. 2 (2016) 190–200.
[20] L. Shao, G. Tobias, Y. Huh, M.L.H. Green, Carbon 44 (2006) 2855–2858.
[21] M. Martincic, S. Vranic, E. Pach, S. Sandoval, B. Ballesteros, K. Kostarelos,
G. Tobias, Carbon 141 (2019) 782–793.
[22] H.-F. Cui, S.K. Vashist, K. Al-Rubeaan, J.H.T. Luong, F.-S. Sheu, Chem. Res. Toxicol.
23 (2010) 1131–1147.
[23] V.L. Colvin, Nat. Biotechnol. 21 (2003) 1166.
[46] P. Garnier, X.-T. Wang, M.A. Robinson, S. van Kasteren, A.C. Perkins, M. Frier,
A.J. Fairbanks, B.G. Davis, J. Drug Target. 18 (2010) 794–802.
[47] Y.C. Lee, C.P. Stowell, M.J. Krantz, Biochemistry 15 (1976) 3956–3963.
[48] H. Ge, P.J. Riss, V. Mirabello, D.G. Calatayud, S.E. Flower, R.L. Arrowsmith,
T.D. Fryer, Y. Hong, S. Sawiak, R.M.J. Jacobs, S.W. Botchway, R.M. Tyrrell,
T.D. James, J.S. Fossey, J.R. Dilworth, F.I. Aigbirhio, S.I. Pascu, Chem 3 (2017)
437–460.
[24] D. Cui, F. Tian, C.S. Ozkan, M. Wang, H. Gao, Toxicol. Lett. 155 (2005) 73–85.
[25] G. Jia, H. Wang, L. Yan, X. Wang, R. Pei, T. Yan, Y. Zhao, X. Guo, Environ. Sci.
Technol. 39 (2005) 1378–1383.
[26] L.P. Zanello, B. Zhao, H. Hu, R.C. Haddon, Nano Lett. 6 (2006) 562–567.
[27] K. Kostarelos, Nat. Biotechnol. 26 (2008) 774.
[28] P. Liu, Eur. Polym. J. 41 (2005) 2693–2703.
[29] V. Georgakilas, N. Tagmatarchis, D. Pantarotto, A. Bianco, J.-P. Briand, M. Prato,
Chem. Commun. (2002) 3050–3051.
[49] R.J. Stockert, A.G. Morell, I.H. Scheinberg, Science 186 (1974) 365.
[50] D.K. Marsee, A. Venkateswaran, H. Tao, D. Vadysirisack, Z. Zhang, D.D. Vandre,
S.M. Jhiang, J. Biol. Chem. 279 (2004) 43990–43997.
[30] C.J. Serpell, R.N. Rutte, K. Geraki, E. Pach, M. Martincic, M. Kierkowicz, S. De
Munari, K. Wals, R. Raj, B. Ballesteros, G. Tobias, D.C. Anthony, B.G. Davis, Nat.
Commun. 7 (2016) 13118.
[51] K.H. Diehl, R. Hull, D. Morton, R. Pfister, Y. Rabemampianina, D. Smith, J.M. Vidal,
C. van de Vorstenbosch, J. Appl. Toxicol. 21 (2001) 15–23.
[31] S. Daniel, T.P. Rao, K.S. Rao, S.U. Rani, G.R.K. Naidu, H.-Y. Lee, T. Kawai, Sens.
7