212Pb@C60 and its water-soluble derivatives: Synthesis, stability, and suitability for radioimmunotherapy

Article abstract of DOI:10.1021/ja068639b

Fullerenes could potentially play a valuable role in radioimmunotherapy by more stably encapsulating radionuclides, especially where conventional chelation chemistry is inadequate due to the physical and/or chemical properties of the radionuclide. One of the therapeutically useful radionuclides that requires improved containment in vivo is ²¹²Pb (τ_1/2 = 10.6 h), the β-emitting parent to α-emitting ²¹²Bi (τ_1/2 = 60.6 min). Myelotoxicity resulting from the accumulation of ²¹²Pb in the bone marrow has limited the use of this radionuclide despite its favorable decay characteristics. In this work, ²¹²Pb@C₆₀ and its malonic ester derivatives were prepared for the first time by allowing the ²¹²Pb to recoil into C ₆₀ following α-decay from its parent, 0.15-s ²¹⁶Po, generated in situ from the decay of ²²⁴Ra (τ_1/2 = 15 days). Repeated washing of the organic phase containing the ²¹²Pb@C₆₀ malonic esters with challenge solutions containing cold Pb²⁺ ions demonstrated that some of the ²¹²Pb could not be exchanged and was apparently inside of the fullerenes. Malonic esters of endohedral α-emitting ²¹³Bi (τ_1/2 = 45 min) fullerenes were prepared by an analogous procedure. Following acidification of the esters, a preliminary biodistribution study in mice was performed with the untargeted water-soluble radiofullerenes. It was found that ²¹²Pb did not accumulate in bone after being administered as an endohedral fullerene, in contrast to results with polyhydroxylated radiofullerenes and conventional polyaminocarboxylate chelators for ²¹²Pb. The results indicate that ²¹²Pb is held more tightly in the fullerene than in other methods and suggest that fullerenes may have an important role in the targeted delivery of ²¹²Pb.

Full text of DOI:10.1021/ja068639b

Published on Web 03/30/2007
212Pb@C60 and Its Water-Soluble Derivatives: Synthesis,
Stability, and Suitability for Radioimmunotherapy
Michael D. Diener,*,† John M. Alford,† Stephen J. Kennel,‡,§ and Saed Mirzadeh‡
Contribution from TDA Research Incorporated, 12345 West 52nd AVenue,
Wheat Ridge, Colorado 80033 and Oak Ridge National Laboratory, 1 Bethel Valley Road,
Oak Ridge, Tennessee 37831-6229
Received December 1, 2006; E-mail: mikee@tda.com
Abstract: Fullerenes could potentially play a valuable role in radioimmunotherapy by more stably
encapsulating radionuclides, especially where conventional chelation chemistry is inadequate due to the
physical and/or chemical properties of the radionuclide. One of the therapeutically useful radionuclides
that requires improved containment in vivo is 212Pb (τ1/2 ) 10.6 h), the â-emitting parent to R-emitting 212Bi
(τ1/2 ) 60.6 min). Myelotoxicity resulting from the accumulation of 212Pb in the bone marrow has limited the
use of this radionuclide despite its favorable decay characteristics. In this work, 212Pb@C60 and its malonic
ester derivatives were prepared for the first time by allowing the 212Pb to recoil into C60 following R-decay
from its parent, 0.15-s 216Po, generated in situ from the decay of 224Ra (τ1/2 ) 15 days). Repeated washing
of the organic phase containing the 212Pb@C60 malonic esters with challenge solutions containing cold
Pb2+ ions demonstrated that some of the 212Pb could not be exchanged and was apparently inside of the
fullerenes. Malonic esters of endohedral R-emitting 213Bi (τ1/2 ) 45 min) fullerenes were prepared by an
analogous procedure. Following acidification of the esters, a preliminary biodistribution study in mice was
performed with the untargeted water-soluble radiofullerenes. It was found that 212Pb did not accumulate in
bone after being administered as an endohedral fullerene, in contrast to results with polyhydroxylated
radiofullerenes and conventional polyaminocarboxylate chelators for 212Pb. The results indicate that 212Pb
is held more tightly in the fullerene than in other methods and suggest that fullerenes may have an important
role in the targeted delivery of 212Pb.
Introduction
antibody fragment, or peptide that is specific for receptors
expressed preferentially on the cancer cells. Following arrival
The empty interior cavity of fullerenes has long been viewed
as an ideal site for containment of radionuclides during in vivo
transport,1 as one component of the effective treatment of cancer
by radioimmunotherapy (RIT).2 Since the chemistry required
to open a hole in a fullerene is complex and exceedingly unlikely
to occur in vivo and the conformational stability of the fullerene
cage is absolute, atoms trapped within fullerenes can only be
released during extremely energetic events. Encapsulating
radionuclides in fullerenes has therefore been proposed as a
strategy to eliminate the undesired toxicity that results from
leakage and catabolism of administered radionuclides that have
been complexed by other chemical methods.
In RIT, metallic radionuclides are generally contained in
polyaminocarboxylate (PAC) chelators wherein three or more
secondary amines and carboxylic acids chelate the radionuclide,
largely sheltering it from the external environment. The
radioisotopes are then delivered to cancer cells by chemically
attaching the radionuclide to a chelator-modified antibody,
of the radionuclide-antibody conjugates at the cancer cell,
energy released from the decay of the radionuclide (mainly in
the form of short-range R and/or â particles as well as X-rays)
can then break strands of the cell’s DNA, thereby killing the
cancerous cell. RIT using PAC chelators and â-emitting
radionuclides has recently become an accepted method for
treatment of B-cell non-Hodgkin’s lymphoma (NHL), and two
radiopharmaceuticals have been approved by the U.S. FDA.3
Effective targeting of the radionuclides remains the primary
challenge in RIT. Current limitations in targeting efficacy
restricts effective use of RIT to attack on cancerous cells that
have preferential exposure to circulating blood (and thus the
circulating antibody conjugate). The hematological malignancies
represent such a tumor system that can be effectively targeted
and have a significant portion of the cancerous cell volume
consisting of single cells or aggregates of a few cells, typically
tens of micrometers in diameter or less. In vivo, â-particles
distribute their energy over millimeters from the radionuclide.
Therefore, â-decays in RIT can damage adjacent healthy tissues
and cells even when they are specifically targeted to cancer cells.
Moreover, a single â-decay seldom breaks more than one strand
of DNA, and single-strand breaks can be readily repaired.
† TDA Research Inc.
‡ Oak Ridge National Laboratory.
§ Current address: University of Tennessee Graduate School of Medicine,
Knoxville, TN.
(1) Edelson, E. Pop. Sci. 1991, Aug, 52.
(2) (a) Milenic, D. E.; Brady, E. D.; Brechbiel, M. W. Nat. ReV. Drug DiscoVery
2004, 3, 488. (b) Sharkey, R. M.; Goldenberg, D. M. J. Nucl. Med. 2005,
46, 115S.
(3) Srivastava, S.; Dadchova, E. Semin. Nucl. Med. 2001, 27, 330-341.
9
10.1021/ja068639b CCC: $37.00 © 2007 American Chemical Society
J. AM. CHEM. SOC. 2007, 129, 5131-5138
5131

A R T I C L E S
Diener et al.
Figure 1. Decay chains referred to in the text: 224Ra chain (L) and 225Ac chain (R).
Microdosimetry calculations indicate that hundreds of thousands
of â-decays are required to have a 99.99% chance of killing a
cell.4 Conversely, only hundreds of R-decays are required for
the same probability of killing the cell, and the range of an 8
MeV R-particle (from decay of an R-emitting radionuclide) in
vivo is ∼100 µm or ∼10 cell diameters, thereby largely limiting
damage to the cells which are targeted. Although the FDA-
approved NHL therapies use one of two â-emitting radionuclides
(90Y or 131I), it is widely agreed that alpha particles have ideal
characteristics for treatment of the cancers that RIT can most
effectively target.5
obtained from stockpiled 233U.7 Also, generator systems starting
from 225Ac have been developed for its daughter 213Bi (τ1/2
)
45 min)8 (Figure 1). While 213Bi can be satisfactorily chelated
by PAC chelators,9 the radiochemical yield of chelation of 225Ac
is very small.10 Furthermore, the R-emitting daughters of 225Ac
are not retained in the chelator and are released to redistribute
to nontarget organs in vivo. A recent primate study with 225Ac
demonstrated dose-limiting renal toxicity, believed to result from
decay of 213Bi daughters.10 Thus far, none of the approaches to
corral the daughters have been successful,11 and given the high
recoil energy of R-decays, it is also very unlikely that even
fullerenes could be of assistance in retaining the daughters of
225Ac. While some encouraging results have been obtained with
213Bi,9b its short half-life probably requires more efficient
targeting than is currently available.12
Despite the favorable properties of R-particle radiation,
development of R-particle RIT has been limited by the poor
availability and/or physical characteristics of R-emitting radio-
nuclides. Several reviews have debated the relative strengths
and weaknesses of the available choices.5 Currently, most effort
is focused on 211At, 225Ac, and 213Bi. The decay characteristics
of 211At are very attractive (τ1/2 ) 7.2h), but using current state-
of-the-art production methods,6 about 1 day of operation at a
cyclotron capable of accelerating an R-particle to a minimum
energy of 28 MeV produces only about three human doses of
∼8 mCi. 211At is therefore most likely to be prohibitively
expensive for the foreseeable future. Actinium-225 (τ1/2 ) 10
days) is available from the decay of 229Th (τ1/2 ) 7430 yr),
Early RIT studies13 using R-emitting radioisotopes were
performed with 212Bi (τ1/2 ) 60.6 min), in large part because
of the availability of 224Ra (the parent of 212Bi, see Figure 1b).
The short half-life of 212Bi creates the same limitations as for
213Bi, but that problem can potentially be overcome using its
(7) (a) Boll, R. A.; Malkemus, D.; Mirzadeh, S. Appl. Radiat. Isot. 2005, 62,
667-679. (b) Mirzadeh, S. Appl. Radiat. Isot. 1998, 49, 345-349. (c) Boll,
R. A.; Garland, G.; Mirzadeh, S. In International Conference on Nuclear
Data for Science and Technology; Haight, R. C., Talou, P., Kawano, T.,
Chadwick, M. B., Eds.; AIP Conference Proceedings 769; American
Institute of Physics: New York, 2005; pp 1674-1675.
(8) (a) Boll, R. A.; Mirzadeh, S.; Kennel, S. J. Radiochim. Acta 1997, 79,
145-149. (b) Ma, D.; McDevitt, M. R.; Finn, R. D.; Scheinberg, D. A.
Appl. Radiat. Isot. 2001, 55, 667-678.
(9) (a) McDevitt, M. R.; Ma, D.; Lai, L. T.; Simon, J.; Borchardt, P.; Frank,
R. K.; Wu, K.; Pellegrini, V.; Curico, M. J.; Miederer, M.; Bander, N. H.;
Scheinberg, D. A. Science 2001, 294, 1537-1540. (b) McDevitt, M. R.;
Sgouros, G.; Finn, R. D.; Humm, J. L.; Jurcic, J. G.; Larson, S. M.;
Scheinberg, D. A. Eur. J. Nucl. Med. 1998, 25, 1341-1351.
(10) Miederer, M.; McDevitt, M. R.; Sgouros, G.; Kramer, K.; Cheung, N.-K.
V.; Scheinberg, D. A. J. Nucl. Med. 2004, 45, 129-137.
(4) (a) Humm, J. L.; Cobb, L. M. J. Nucl. Med. 1990, 31, 75-83. (b) Macklis,
R. M.; Yin, J. Y.; Beresford, B.; Atcher, R. W.; Hines, J. J.; Humm, J. L.
Radiat. Res. 1992, 130, 220-226. (c) O’Donoghue, J. A.; Bardies, M.;
Wheldon, T. E. J. Nucl. Med. 1195, 36, 1902-9.
(5) (a) Couturier, O.; Supiot, S.; Degraf-Mougin, M.; Faivre-Chauvet, A.;
Carlier, T.; Chatal, J.-F.; Davodeau, F.; Cherel, M. Eur. J. Nucl. Med. Mol.
Imaging 2005, 32, 601-614. (b) Mulford, D. A.; Scheinberg, D. A.; Jurcic,
J. G. J. Nucl. Med. 2005, 46, 199S-204S. (c) Zalutsky, M. R.; Pozzi, O.
R. Quantum J. Nucl. Med. Mol. Imaging 2004, 48, 289-296. (d) Mirzadeh,
S. Appl. Radiat. Isot. 1998, 49, 345-349.
(11) Jaggi, J. S.; Kappel, B. J.; McDevitt, M. R.; Sgouros, G.; Flombaum, C.
D.; Cabassa, C.; Scheinberg, D. A. Cancer Res. 2005, 65, 4888-4895.
(12) Adams, G. P.; Shaller, C. C.; Chappell, L. L.; Wu, C.; Horak, E. M.;
Simmons, H. H.; Litwin, S.; Marks, J. D.; Weiner, L. M.; Brechbiel, M.
W. Nucl. Med. Biol. 2000, 27, 339-346.
(13) Macklis, R. M.; Kinsey, B. M.; Kassis, A. I.; Ferrara, J. L. M.; Atcher, R.
W.; Hines, J. J.; Coleman, C. N.; Adelstein, S. J.; Burakoff, S. J. Science
1988, 249, 1024-6.
(6) Zalutsky, M. R.; Zhao, X. G.; Alston, K. L.; Bigner, D. J. Nucl. Med.
2001, 42, 1508-1515.
9
5132 J. AM. CHEM. SOC. VOL. 129, NO. 16, 2007

212Pb@C60 and Its Water-Soluble Derivatives
A R T I C L E S
440.4 keV (26.1%), respectively. To quantitate 212Bi, the intensity of
the 583.1 keV (32.5%) γ-ray of its 3.07-min 208Tl daughter was
measured at transient equilibrium with the parent. Consequently, at least
10 min were allowed to elapse between chemical processing and activity
measurements. Since the intensity of the γ-ray from 225Ac at 188.1
keV is only 0.47%, 225Ac was quantitated by measurement of intensities
of 218.0 keV (11.58%) from its 4.9-min 221Fr daughter at transient
equilibrium. All relevant nuclear data are taken from the 1986 Table
of Radioisotopes.17 Gross radioactivity was measured in an ionization
chamber (CRC-7, Capintec Inc., NJ). An automated γ-ray scintillation
counter (Wallac Wizard) consisting of a well-type NaI(Tl) detector was
used for biodistribution studies of 212Pb.
Synthesis by Recoil. The 224/225Ra sources were prepared by an
electrodeposition technique. The electrolysis apparatus was of conven-
tional design consisting of a 20-mL Pyrex vial, a Pt working electrode
(1.5 mm o.d. Pt rod bent into a circle and spot welded to a ∼1 cm
diameter circle of Pt mesh), and a counter Pt electrode (1.5 mm o.d.
rod). The distance between the two electrodes was ∼5 mm. The
electrolyte, 10 mL of 0.01 M HNO3, was purged with a gentle flow of
N2. In a typical experiment, the purified 224Ra (as dry nitrate) was
dissolved in 200 µL of 1.2 M HNO3, and the mixture was transferred
to the electrolysis cell. Electrodeposition was conducted under a constant
potential of 12 V. Under our experimental setup, the initial current
was ∼0.5 mA, which dropped to ∼0.1 mA over a 2-h period. When
assay of the electrolyte indicated removal of 224Ra from solution, the
working electrode was carefully removed from the electrolysis cell
without opening the circuit. The working electrode was then left to
dry in air for 2 h and assayed for radioactivity.
parent, 212Pb (τ1/2 ) 10.6 h), as a vehicle for transporting 212Bi
to the target tissue. However, while the PAC chelator known
as DOTA (1,4,7,10-tetraazacyclododecane-N,N′,N′′,N′′′-tetraace-
tic acid) can chelate 212Pb satisfactorily in saline solutions, in
vivo tests resulted in severe myelotoxicity,14 presumably the
result of bone-targeting 212Pb2+ lost or transchelated out of the
DOTA. Furthermore, 36% internal conversion of the γ-ray
which follows the â-decay of 212Pb to 212Bi also caused release
of 36% of 212Bi from DOTA.15 Since we began this work, two
reports of new chelators for 212Pb have appeared,16 but the results
appear to indicate that further development or a new approach
is still required. It was thought that use of a fullerene instead
of a chelator could potentially solve both problems with 212Pb,
thereby making the 212Pb/212Bi system attractive for RIT. If the
fullerene was successful, the only significant remaining draw-
back to using 212Pb in RIT is the high-energy γ-ray emitted by
the 208Tl daughter. Compared to the outstanding problems with
225Ac and 213Bi, solutions for shielding health-care personnel
from high-energy γ-rays are well known and relatively inex-
pensive compared to the costs of 211At production. Thus,
development of endohedral 212Pb fullerenes could make R-par-
ticle RIT feasible, greatly impacting the way hematological
malignancies and micrometastes are treated.
In this paper, we report the synthesis of 212Pb@C60 by recoil
following R-decay of its short-lived parent, formation of water-
soluble 212Pb@C60 malonic acids, the stability of the radio-
fullerene during â--decay of 212Pb to 212Bi, and a preliminary
biodistribution study of the untargeted water-soluble radio-
fullerene in mice. To our knowledge, this is the first report of
an endohedral fullerene formed by recoil from R-decay, the first
endohedral lead fullerene, the first fullerene encapsulating an
R-emitting radionuclide with a therapeutically useful half-life,
and (despite some limitations) the first biodistribution study of
radiofullerene malonic acids.
A coating of C60 was applied by spraying a saturated toluene solution
of C60 over the mesh with an artist’s airbrush. The toluene rapidly
evaporated, leaving a thick film of C60 over the radionuclide-plated Pt
mesh. The coated mesh was then left to stand for 24-36 h, except as
noted. At the end of this contact time, the coated Pt mesh was assayed
for radioactivity.
Esterification. Because M@C60 are generally insoluble,18 the
fullerenes (both C60 and R@C60) were removed from the Pt mesh by
derivatizing them to their malonic ester derivatives,19 which are then
soluble in tetrahydrofuran (THF) and ethyl acetate (EA). After assaying
for radioactivity, the mesh was placed in a two-neck 10-mL pear-shaped
flask which was purged with argon. Four milliliters of THF containing
3 mg of suspended and dispersed NaH was added. Diethylbromoma-
lonate (0.02 mL) was also added, and the flask was vortexed vigorously
for 10 min. The THF slurry was transferred by syringe from the flask
into a 15-mL polypropylene conical bottom centrifuge tube and
centrifuged for 2 min at ∼1000g to sediment the sodium salts. The
supernatant was transferred by syringe into a fresh centrifuge tube and
evaporated under argon. A 0.75 mL amount of EA was added, vortexed
briefly to dissolve the fullerene esters, and then transferred to a 1.5
mL conical-bottomed centrifuge tube. Alternately, the THF could be
evaporated in the flask, 0.75 mL of EA added, vortexed, removed to a
1.5 mL centrifuge tube, and centrifuged for 1 min at 10 000g to
sediment the NaH. The EA supernatant was then transferred to a fresh
1.5 mL conical-bottomed centrifuge tube. The mesh was assayed again
for radioactivity at a later convenient time.
Experimental Section
Radioisotopes. As needed, 224Ra and 225Ra were extracted from a
stock of Th containing both 228Th and 229Th. Chemical separation of
Ra and Ac from Th was achieved using the macroporous anion-
exchange resin MP1 in 8 M HNO3 media, and separation of Ra from
Ac was accomplished on a low-cross-linking cation-exchange resin
AG50-X4 using 1.2 M HNO3 as eluent. A detailed procedure for these
separations is given by Boll et al.7a At the time of separation, the activity
ratio of 224Ra to 225Ra was 1:2.5.
Radioactivity Measurement. γ-ray spectrometry was used for
measurement of all radioisotopes used in this study. The γ-ray
spectrometer consisted of a calibrated intrinsic Ge detector (crystal
active volume ≈ 100 cm3) and a PC-based multichannel analyzer
(MCA) (Canberra Industries, Meriden, CT). The detector has a
resolution of 0.8 keV at 5.9 keV, 1.0 keV at 123 keV, and 1.9 keV at
1332 keV. Energy and efficiency calibrations were determined with
γ-ray sources traceable to the National Institute of Standard and
Technology (NIST). The γ-ray energies and their absolute intensities
in parentheses used for determination of 212Pb, 224Ra, 225Ra, and 213Bi
radioactivity were 238.6 (43.6%), 240.8 (3.90%), 40.1 (30.0%), and
Removal of Unencapsulated Radionuclides. Free or otherwise
unencapsulated radionuclides were removed by repeated washings of
the EA ester solution with an aqueous 0.012 M HNO3 solution that
was also 5 µN in each of cold Pb2+, Bi3+, and Tl3+ ions. One-half
milliliter of the washing solution was added to the EA and vortexed
for 10 s, followed by centrifugation at 2000 rpm for 30 s. The upper
organic layer was carefully removed by a syringe with a fine-gauge
(14) Ruble, G.; Wu, C.; Squire, R. A.; Gansow, O.; Strand, M. Int. J. Radiat.
Oncol. Biol. Phys. 1996, 34, 609-616. (b) Horak, E.; Hartmann, F.;
Garmestani, K.; Wu, C.; Brechbiel, M. W.; Gansow, O.; Landolfi, N. F.;
Waldmann, T. A. J. Nucl. Med. 1997, 38, 1944-1950.
(15) Mirzadeh, S.; Kumar, K.; Gansow, O. A. Radiochim. Acta 1993, 60, 1-10.
(16) A) Chappell, L. L.; Dadachova, E.; Milenic, D. E.; Garmestani, K.; Wu,
C.; Brechbiel, M. W. Nucl. Med. Biol. 2000, 27, 93-100. (b) Chong, H.
S.; Milenic, D. E.; Garmestani, K.; Brady, E. D.; Arora, H.; Pfiester, C.;
Brechbiel, M. W. Nucl. Med. Biol. 2006, 33, 459-467.
(17) Browne, E.; Firestone, R. B. In Table of Radioisotopes; Shirley, V. S.,
Ed.: Wiley-Interscience: New York, 1986.
(18) Diener, M. D.; Alford, J. M. Nature 1998, 393, 668-670.
(19) Bolskar, R. D.; Benedetto, A. F.; Husebo, L. O.; Price, R. E.; Jackson, E.
F.; Wallace, S.; Wilson, L. J.; Alford, J. M. J. Am. Chem. Soc. 2003, 125,
5471-8.
9
J. AM. CHEM. SOC. VOL. 129, NO. 16, 2007 5133

A R T I C L E S
Diener et al.
needle and placed in a new centrifuge tube. The organic and aqueous
phases were then sequentially assayed for radioactivity. The washing
sequence was then repeated by contacting the EA solution with fresh
aqueous acid until the aqueous wash contained less than 1 nCi of
212Pb. Note that the vast majority of fullerene esters in solution after
the washing are empty fullerenes.
Hydrolysis of the Esters.19 The EA containing the ester was
transferred to a 10 mL round-bottom flask and purged with N2,
evaporating the EA. A slurry of 1-2 mg of NaH in toluene was then
added by syringe to the purged flask. This flask, still under N2, was
immersed in a water bath, stirred, and heated to ∼80 °C for 1 h. It was
then removed from the bath, and slow dropwise addition of 0.2 mL of
methanol was begun immediately from a fine-gauge syringe. The
reaction of the MeOH with the NaH slurry causes frothing, which could
be minimized by limiting the rate of addition. Methanol addition was
discontinued when a red precipitate formed and the toluene became
clear. The toluene was then carefully removed with a syringe and fine-
gauge needle. One milliliter of deionized water was added to the flask,
which immediately dissolved the red precipitated fullerene acid. The
solution was transferred to a centrifuge tube, and the solids were spun
down at 10 000g for 5 min. All three fractions were then assayed for
radioactivity.
Figure 2. Radioactivity of 212Pb and 212Bi during aqueous washing
following esterification of C60. 212Pb and 212Bi points are offset slightly for
clarity. Wash #0 is the initial radioactivity in the ethyl acetate, prior to any
washing. Error bars represent the uncertainty associated with the radioactiv-
ity count rate. Logarithmic scale is used on the ordinate to show both the
organic and aqueous washes on the same figure.
At this point, the aqueous phase containing the 212Pb@C60 malonic
acids was extremely basic, pH ≈ 14. In the first batch, the pH was
neutralized by dropwise addition of 6 M HCl. The process was
improved upon in subsequent batches by adding 6 M HCl dropwise to
the basic water, which fully protonates and precipitates the fullerene
acid. The product was separated by centrifugation (10 000g for 5 min),
and the acid supernatant was removed. 2-(N-Morpholino)ethanesulfonic
acid (MES) buffer was then added to redissolve the fullerene acid. One
molar NaOH was added to adjust the pH to 7. The radiofullerene acids
were recovered quantitatively.
Preliminary Biodistribution Study. Female mice were obtained
from Taconic, Inc., and the biodistribution studies were carried out in
an AAALAC-approved animal care facility at ORNL according to
approved protocols #0256 and #0267. Two different batches of
212Pb@C60 acids were used on different dates for the in vivo testing.
The first batch used 212Pb@C60 acid that had not been reprecipitated
and therefore had a high amount of salt present. To minimize the
deleterious effects of the salt on the animals, two injections of
radiofullerene acids were made very slowly into the tail vein, separated
by 2 min. The first two mice, female severe combined immunodefi-
ciency (SCID) mice, received 200 µL in each injection, for a total
administered dose of 29.2 ( 0.1 and 28.9 ( 0.1 nCi, respectively. The
third mouse was a female Balb/c mouse, and it received two 125 µL
injections totaling 17.9 ( 0.8 nCi. The mice were sacrificed 1 h after
injection. The following tissues were collected from each mouse and
placed in previously weighed plastic scintillation vials: skin, leg muscle,
femur and sternum (referred to as “bone” henceforth), stomach and
intestines (which were emptied first), ovaries and uterus, liver, spleen,
kidneys, heart, and lungs. Blood was also collected.
electroplated within 2 h. Following the contact time with C60
and subsequent esterification of the fullerenes, it was found that
96.1 ( 13.2% of the electroplated 224Ra remained on the
platinum mesh. Conversely, no visible residue of the fullerene
coating remained on the mesh after the esterification process.
This result demonstrated that it is possible to reuse 224Ra-
electroplated meshes for many sequential esterifications.
Behavior of the 224Ra Decay Products. A typical washing
sequence is shown in Figure 2 for the daughters of 224Ra. The
212Pb ions outside of the fullerene were expected to exchange
and be extracted into the aqueous phase while the endohedral
212Pb atoms remained in the ethyl acetate. For the sample shown
in Figure 2, 220 nCi of 212Pb@C60 esters (0.43% yield based
on 212Pb, vide infra) were recovered in the ethyl acetate after
exhaustive washing. Derivatization and washing for this sample
were performed over 6.5 h, although the technique was later
refined to be complete in less than 1 h and could occur in about
15 min if the intermediate washes did not have to be analyzed.
Unlike 212Pb, the 208Tl signal in the aqueous wash stopped
tailing off and attained a relatively steady value after the fourth
wash (Figure 2). All of the 208Tl should break out of the fullerene
as it is created by R-decay from 212Bi. With a half-life of 3
min, 208Tl rapidly attains secular equilibrium with 212Bi. After
the washing was completed, the fractions were separated and
after 5-10 min the 208Tl signal in the ethyl acetate fraction was
primarily from 212Bi present in the ethyl acetate fraction, from
which it decayed after the washing and before the counting. In
Figure 2, 21 ( 1 min passed between washings, during which
time the 208Tl had attained over 99% of equilibrium to the 212Bi
present. By the same reasoning, the 208Tl signal in the aqueous
fraction indicated the presence of 212Bi in the aqueous fraction,
since any 208Tl created before the fractions were separated would
have decayed prior to counting. Thus, the values for 208Tl in
Figure 2 are an accurate measurement of 212Bi.
The second batch of 212Pb@C60 acids had previously been acidified,
precipitated, and redissolved as described above to minimize the
extraneous salts. Six female Balb/c mice each received a single tail
vein injection of 200 µL containing 10.3 ( 0.8 nCi of 212Pb@C60 acid
in pH 7 buffer. Three mice were sacrificed at 3 h postinjection and the
remaining three at 8 h postinjection. The same tissues were collected
as before and counted for 10 min each in the automatic γ-ray detector.
Standards consisting of 10 µL of the injected solutions were also
counted immediately before and after the tissues as were blank vials
to provide a background. Raw data were corrected for the decay of
212Pb, and % ID/g values for each sample were calculated.
To verify this, the last aqueous wash was counted at various
intervals up to 2000 min after contact with the ethyl acetate
(Figure 3). The observed decay for the 583 keV γ-ray could be
fit with two components: an initially large component with a
slope of 60.6 min (i.e., 212Bi initially present in the aqueous
Results
Electrodeposition and Adherence of Radium. Under our
experimental conditions, typically 60-80% of the 224Ra present
9
5134 J. AM. CHEM. SOC. VOL. 129, NO. 16, 2007

212Pb@C60 and Its Water-Soluble Derivatives
A R T I C L E S
Figure 4. Activities of 225Ra daughters during aqueous washing following
esterification of C60. Errors calculated as for Figure 1. “Calculated 213Bi”
is the activity of 213Bi that was present in ethyl acetate at the time of that
particular wash, back calculated from the activity of 213Bi present at the
time of Wash #7.
Figure 3. Counting of a single aqueous wash of R@C60 esters and fitting
(green line) with two radionuclide decays with half-lives corresponding to
212Pb (blue line) and 212Bi (red line).
phase) and a very small contribution with a slope of 10.6 h
(i.e., 212Bi created from the decay of a minute amount of 212Pb
present in the aqueous, presumably from incomplete separation
of the aqueous and organic layers following vortex mixing).
The only possible conclusion is that a significant amount of
212Bi becomes accessible to the aqueous solution during the time
between each wash. The fraction of 212Bi activities found in
the last three aqueous washes shown in Figure 2 were 37%,
48%, and 39%, respectively, of the 212Bi produced during the
time elapsed since the previous wash.
Encapsulation Efficiency. The fraction of 212Pb that recoiled
into C60, the encapsulation efficiency, was calculated from the
washing data. For the sample shown in Figure 2, 150 µCi of
224Ra was electroplated onto a Pt mesh. The amount of 212Pb
created during the 18 h of contact between the fullerene and
the electroplated Ra was calculated to be 78 µCi using the well-
known solution to the first-order differential equation governing
the decay and growth of radioactivity (see for example ref 15),
Figure 5. 212Pb activity during aqueous washing following esterification
of C60 in a sample where the C60 contacted the electroplated radium for
only 25 min.
0
i.e., A2 ) [(λ2 - λ1/λ2)]/A1 [exp(-λ1t) - exp(-λ2t)], where
0
A2 denotes the activity of 212Pb at the end of contact time, A1
while the 213Bi persisted in the ethyl acetate, and after seven
washes, essentially no 213Bi was removed from the organic
phase. Thus, 213Bi behaved similarly to 212Pb, while 225Ac did
not. Encapsulation efficiencies for 213Bi were comparable to
those observed for 212Pb.
For one experiment, the esterification process was performed
very soon after the C60 coating was dry, resulting in a contact
time between C60 and the electroplated radionuclides of only
25 min, versus the more typical 18-36 h. The activities present
in the washings following esterification are shown in Figure 5.
The encapsulation efficiencies of 212Pb and 213Bi were 0.31%
and 0.59%, respectively.
Conversion of Esters to Acids. Radioactivity assay of the
aqueous portion after hydrolysis and separation from the toluene
and the solids revealed yields between 55% and 75% for the
212Pb fullerenes in the aqueous phase with ∼7% of the 212Pb
remaining in each of the toluene and the insoluble material.
Preliminary Biodistribution Study. The results of the
biodistribution study are shown in Figure 6 and Table 2. The
data show that there was significant uptake in the liver and
spleen. As the livers of young Balb/c mice typically weigh a
little less than 1 g, around 50% of the injected dose was found
is the 224Ra activity at the beginning of contact (150 µCi), t is
the contact time (decay period), and λ1 and λ2 are the decay
constants (ln 2/τ1/2) of 224Ra and 212Pb, respectively. The activity
of 212Pb at the end of the washing period was simply back
extrapolated to the end of the contact time using the relationship
0
A2 ) A2 [exp(-λ2t)], where here A20 is the 212Pb activity at the
end of the contact period (78 µCi), A2 the 212Pb activity after
washing (220 nCi), and t the time period between the end of
contact time and the end of washing (6.5 h). The decay-corrected
activity of 212Pb after washing was 336 nCi; therefore, the
encapsulation efficiency calculated by this method was 0.336/
78 ) 0.43%. Using the same method, we obtained a range of
values between 0.1% and 0.6% for both 212Pb and 213Bi,
presumably varying according to the C60 film quality and purity
of the electroplated material.
Behavior of the 225Ra Decay Products. As our Ra source
was a mixture of 224Ra and 225Ra, the decay products of 225Ra
could also be traced through the washing process following
esterification. The distribution of 225Ac (measured by its
daughter 221Fr) and 213Bi is shown in Figure 4. No observable
225Ac remained in the ethyl acetate following the second wash,
9
J. AM. CHEM. SOC. VOL. 129, NO. 16, 2007 5135

A R T I C L E S
Diener et al.
incorporated into fullerenes during the gas-phase synthesis of
the fullerene, encapsulation of most other elements requires
forcing the element through the fullerene cage by one means
or another.20 Thus, most of the known chemistry of endohedral
fullerenes has concerned lanthanides and alkaline-earth metals,
with a few reports of actinide encapsulation.21 While there are
several desirable â--emitting radioisotopes among the lan-
thanides, the elements that have posed difficulties for PAC
chelators are located in other parts of the periodic table. One of
the more intriguing uses of fullerenes in nuclear medicine would
be encapsulation of gaseous isotopes, which are unsuited to
traditional methods for in vivo transport. Significant progress
has been made toward that goal by neutron activating stable
(gas) atoms that are mixed with fullerenes and allowing the
recoil following prompt γ-ray emission to incorporate the
product nuclide into the fullerene.22 Besides lanthanides and
gases, neutron activation has also been used to create endohedral
fullerenes with antimony, arsenic, germanium, and selenium.23
However, the isotopes created in these experiments do not have
obvious uses in nuclear medicine. Since there are no simple
neutron activation pathways that lead to R-emitting radioisotopes
of interest, a new approach was required.
Figure 6. Average (n ) 3) biodistribution of 212Pb at 1, 3, and 8 h
postinjection of 212Pb@C60(C[CO2H]2)x in female mice, showing standard
deviation. GI ) stomach + intenstines, ut+ov ) uterus + ovaries; %ID/g
) percent of injected dose per gram of tissue. See text for caveats regarding
preparation of the radiofullerene, administered dose and mouse strain.
Table 1. Activities of 224Ra (in µCi) on the Platinum Mesh Before
and After Esterification of the Fullerenea
224Ra radiaoctivity (
At (theor)
µCi)
expt
no.
elapsed
time (h)
Part of the appeal of this method for encapsulating radioiso-
topes in fullerenes is that it can be performed on a very small
scale without vacuum apparatus and without aerosolizing
radioisotopes as would occur by the standard electric arc or laser
ablation processes for generating endohedral metallofullerenes.
It was not a priori apparent that endohedral metallofullerenes
could be formed from recoil of the daughter nucleus following
R-decay. Recoil energies associated with R-decay are much
higher than those from prompt γ-ray emission following neutron
capture. Rare-earth and noble-gas nuclides have kinetic energies
between 100 and 500 eV due to the recoil imparted by the
emission of a γ-ray following neutron capture. The kinetic
energy of the newly formed 212Pb nucleus (after decay of 216Po)
is ∼33 keV, almost 2 orders of magnitude larger than the recoil
energy of any previously implanted nuclide. Conversely,
â--decays typically impart 10-20 eV of recoil energy to the
daughter nucleus (except for very low Z nuclei) and are generally
too small to implant or remove a nucleus from C60 (ref 15).
However, molecular dynamics simulations show that a nucleus
with as little as 5 eV kinetic energy can penetrate the fullerene
wall in the extremely low probability case that it is aimed exactly
at the middle of a hexagon in the fullerene.23 Thrash et al.
convincingly established that recoil energies less than 200 eV
are sufficient to remove a 166Ho nucleus from the inside of C60.24
Due to the rapid recoil of the nucleus relative to the kinetic
energy of its electrons, some outer electrons will be stripped
off of the 212Pb radioisotope at the time of its creation. As the
A0
At (expl)
R (%)
1
2
3
18.2
34.0
53.9
51.9 ( 3.1
53.8 ( 2.8
33.1 ( 1.8
45.2 ( 2.7
41.0 ( 2.2
21.7 ( 1.2
42.5 ( 2.3
41.6 ( 2.2
20.0 ( 1.2
average
94.0 ( 7.6
102 ( 8
92.2 ( 7.6
96.1 ( 13.2
a Elapsed time is the time between application of C60 on the 224Ra-
plated Pt mesh and dissolution of the C60 layer from Pt mesh. A0 signifies
the measured 224Ra radioactivity on the Pt mesh at time zero, just before
application of C60. At(theor) is the decay-corrected 224Ra radioactivity on
the Pt mesh at elapsed time t, from column 2. At(expl) denotes the measured
224Ra radioactivity on the Pt mesh post dissolution of the C60 layer at time
t (also from column 2). R is the 224Ra adherence to the Pt mesh, At(expl) ×
100/At(theor).
Table 2. Average (n ) 3) Biodistribution of 212Pb at 1, 3, and 8 h
Postinjection of 212Pb@C60(C[CO2H]2)x in Female Micea
percent injected dose per gram of tissue, % ID/g
tissue
1 h
3 h
8 h
liver
31.6 ( 4.3
9.9 ( 1.9
28.7 ( 3.7
5.2 ( 1.5
6.9 ( 0.5
2.5 ( 0.6
1.7 ( 0.3
2.5 ( 0.6
2.3 ( 0.3
0.4 ( 0.1
0.5 ( 0.1
63.1 ( 5.5
25.9 ( 1.7
9.2 ( 1.0
3.8 ( 0.9
3.1 ( 0.2
2.6 ( 0.6
2.6 ( 0.5
1.9 ( 0.2
2.0 ( 0.1
0.8 ( 0.2
0.7 ( 0.2
53.7 ( 7.1
27.2 ( 3.2
6.8 ( 0.9
2.9 ( 1.3
1.4 ( 1.0
4.0 ( 2.3
2.3 ( 0.5
1.3 ( 0.3
1.7 ( 0.3
0.6 ( 0.1
0.5 ( 0.1
spleen
kidney
lungs
blood
heart
uterus + ovaries
stomach + intestines
bone
muscle
skin
a See text for caveats regarding preparation of the radiofullerene,
administered dose, and mouse strain. Values are ( the standard deviation.
(20) Endofullerenes: A New Family of Carbon Clusters; Akasaka, T., Nagase,
S., Eds.; Kluwer Academic Publishers: Doordrecht, The Netherlands, 2002.
(21) (a) Guo, T.; Diener, M. D.; Chai, Y.; Alford, J. M.; Haufler, R. E.; McClure,
S. M.; Ohno, T.; Weaver, J. H.; Scuseria, G. E.; Smalley, R. E. Science
1992, 257, 1661-3. (b) Akiyama, K.; Zhao, Y.; Sueki, K.; Tsukada, K.;
Haba, H.; Nagame, Y; Kodama, T.; Suzuki, S.; Ohtsuki, T.; Sakaguchi,
M.; Kikuchi, K.; Katada, M.; Nakahara, H. J. Am. Chem. Soc. 2001, 123,
181-2. (c) Diener, M. D.; Smith, C. A.; Veirs, D. K. Chem. Mater. 1997,
9, 1773.
in the liver between 3 and 8 h postinjection. Accumulation in
the spleen was also observed, reaching about one-half the
specific activity as the liver. In contrast, accumulation in bone
was modest (∼2% ID/g), and, most importantly, did not increase
over time. None of the other tissues sampled showed high
accumulation of 212Pb within the limitations of the study.
(22) (a) Gadd, G. E.; Evans, P. J.; Hurwood, D. J.; Wood, J.; Morica, S.;
Blackford, M.; Elcombe, M.; Kennedy, S.; James, M. Chem. Phys. Lett.
1996, 261, 221-7. (b) Jimenez-Vasquez, H. A.; Cross, R. J.; Saunders,
R.; Poreda, R. J. Chem. Phys. Lett. 1994, 229, 111-4.
Discussion
(23) Ohtsuki, T.; Ohno, K.; Shiga, K.; Kawazoe, Y.; Maruyama, Y.; Masumoto,
K. J. Chem. Phys. 2000, 112, 2834.
Encapsulation of Radionuclides by Recoil Following
R-Particle Decay. While most Group 2 and 3 elements can be
(24) Thrash, T. P.; Cagle, W. D.; Alford, J. M.; Wright, K.; Ehrhardt, G. J.;
Mirzadeh, S.; Wilson, L. J. Chem. Phys. Lett. 1999, 308, 329.
9
5136 J. AM. CHEM. SOC. VOL. 129, NO. 16, 2007

212Pb@C60 and Its Water-Soluble Derivatives
A R T I C L E S
velocity of the recoiling 212Pb decreases, it will regain electrons
from its surroundings. Theoretically, the range of 212Pb2+ ions
with kinetic energy of 32 keV in graphite is ∼2 µm, corre-
sponding to roughly 1500 C60 molecules. Therefore, the 212Pb
ions likely smash through over a thousand fullerenes before
losing enough energy to stay encapsulated. This may account
for the relatively low encapsulation efficiency.
the chemical procedures carried out were identical regardless
of contact time, hypotheses regarding some chemical derivati-
zation of Pb and Bi (but not Ac) to impart organic solubility
can be ruled out. Additionally, such hypotheses would require
derivatization to occur in some unknown way where the
radionuclide is not subject to exchange during washing.
Fate of 212Bi Following âsDecay of 212Pb. Previous work
showed that ∼36% of the 212Bi formed from a DOTA-chelated
212Pb was lost from the chelator.15 This is significant because
36% of the time the 238-keV γ-ray which follows â--decay of
212Pb undergoes internal conversion (IC).15 In the IC process,
the electromagnetic interaction between the excited nucleus and
the electronic shell results in ejection of electrons (mainly from
the K shell). The ejected electrons would have a kinetic energy
equal to the nucleus excitation energy minus the electron binding
energy. From the point of view of an external observer, however,
the process appears as an internal photoelectric effect, i.e., a
γ-ray collides with an electron, ejecting the electron out of its
orbit. For an atom, the consequence of an IC process is similar
to that of a K-capture (KC) process. The positive charge
generated in the sea of electrons is immediately filled by the
electrons from upper shells, generating corresponding X-rays
which, in turn, may under go internal photoelectric effect,
thereby knocking out electrons from upper shells. The cascade
of the event through the electronic shells always results in loss
of a number of electrons from the binding shells, leaving the
atoms highly charged. If a radioactive atom undergoing an IC
or KC process is part of a molecule, then the daughter atom is
almost always ejected from the molecule. This is due to the
fact that the electron transfers within the molecule, needed to
quench the excitation generated by IC or KC, are very slow
relative to nuclear events.
It was found that little of the parent radioisotope 224Ra was
removed from the source mesh during removal of the fullerenes
by esterification (Table 1). Potentially, many films of fullerenes
could therefore be applied and removed from a single source
of electroplated 224Ra. Removal of the esterified 212Pb@C60
daughters from the 224Ra electroplated and immobilized on the
mesh is akin to a conventional radioisotope generator in that
the daughters can be extracted every so often without disturbing
the parent radionuclides. Even with 1% encapsulation efficiency,
four human doses (estimated as 5 mCi) of 212Pb@C60 could be
produced from 1 Ci of 224Ra in a 5-day week. Theoretically,
3.8 Ci of 224Ra can be extracted every week from a 5 Ci
stockpile of 228Th. With a half-life of slightly under 2 years, 5
Ci of 228Th can produce roughly 750 doses of 212Pb@C60 over
∼3 years, which would be suitable for a Phase II clinical trial.
Furthermore, irradiating 1 g of 226Ra in the ORNL High Flux
Isotope Reactor can produce ∼400 Ci of 228Th7cssufficient for
production of 60 000 doses of 212Pb@C60 over a 3-year period.
An encapsulation efficiency of 1% is therefore not prohibitive
to a therapy based on 212Pb@C60.
Radiolytic Purification of 212Pb@C60 Malonic Esters. The
principle behind washing the esterified solutions is that radio-
active metal ions outside of the fullerene exchange rapidly with
the cold metal ions present in the aqueous washing solution.
The exchange occurs because there are several orders of
magnitude more cold metal ions than radioactive ones. Although
the enthalpies of the exchange reactions are zero, the free
energies are negative due to the increase in entropy of the
systems. However, if the radioactive metal ion is inside of a
fullerene ester (which is not soluble in water), it cannot exchange
through the fullerene cage and remains in the organic phase.
The esters can be dissolved in toluene (instead of ethyl acetate)
and then washed, but it requires more washes to remove the
exohedral radionuclides from toluene solution than when ethyl
acetate is used as the organic solvent.
The net result is that in the 36% of the time where IC occurs,
212Bi is created in a highly ionized state (+5 or even higher).
Experimentally, the portion of the 212Bi created during the time
between washings that was extracted into the aqueous phase
on successive washes (39%, 48%, and 37%) correlates reason-
ably well with the percent of 212Bi that undergoes internal
conversion (36%). It therefore seems most likely that the internal
conversion process ejected 212Bi ions from the fullerene ester
or otherwise modified the fullerene ester so as to make it water
soluble.
While it was originally anticipated that the fullerene would
retain all of the 212Bi after â--decay from 212Pb, the loss of
36% of 212Bi should be put in perspective relative to 225Ac-
based strategies. Consider that all of the 221Fr created by â-decay
from 225Ac (as well as its two daughters) will be separated from
any chelator (or fullerene) and left to deposit their energy
according to the biochemistry of the daughter ions. In the 225Ac
case, ca. 300% of the administered dose is lost, compared to
36% loss for endohedral 212Pb.
Biodistribution of the Untargeted 212Pb@C60 Malonic
Acids. The preliminary biodistribution results showed slow
clearance of 212Pb@C60 malonic acids relative to PAC chelators,
which are largely expelled within an hour. This was also the
case with the only other biodistribution study of radio-
fullerenes.25 That work studied the biodistribution of
166Hox@C82(OH)y (x ) 1,2; y ) 16-24), and found 20-25%
Note that this procedure will readily exchange radioactive
metal atoms from PAC chelators as well as other chemically
bonded systems. Failure to exchange the 212Pb out of the ethyl
acetate solution provides definitive evidence that the 212Pb is
located inside of the fullerene. Furthermore, the encapsulated
212Pb is exceedingly unlikely to be transchelated in vivo as there
is no evidence to suggest that a hole can be opened in the
fullerene by any enzymatic action in a time frame comparable
the 10.6-h half-life of 212Pb.
This theory was corroborated by the finding that 225Ac was
not retained in the organic phase. Since 225Ac is produced by
â--decay from 225Ra, it does not have enough recoil energy to
penetrate a fullerene wall. On the other hand, daughters by
R-decay of both the 224Ra and the 225Ra decay chains are found
in the C60 with similar encapsulation efficiencies. Reducing the
contact time between the fullerenes and the radionuclides greatly
reduced the amount of radiofullerenes produced, but the
encapsulation efficiencies remained essentially constant. Since
(25) Cagle, D. W.; Kennel, S. J.; Mirzadeh, S.; Alford, J. M.; Wilson, L. J.
Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 5182.
9
J. AM. CHEM. SOC. VOL. 129, NO. 16, 2007 5137

A R T I C L E S
Diener et al.
ID/g in the liver between 1 and 4 h, with it clearing slowly
afterward. We hoped that derivatization by malonic acids would
improve the clearance of the radiofullerene relative to polyhy-
droxylate derivatives, but that is clearly not the case. However,
slow clearance of untargeted radiofullerenes should be irrelevant
following proper conjugation to an antibody or other targeting
agent.
fullerene malonates, perhaps this is still the case, even after 8
h. A longer biodistribution study, with higher administered
activities, will be required to determine the cause of the observed
212Pb activity in the bone as well as establish the clearance rate
of the fullerene malonic acid. Application to RIT will ultimately
require conjugation of the fullerene malonates to antibodies (or
other targeting agents), which has recently been accomplished
for the first time.27
One crucial aspect of the biodistribution which did greatly
improve was the accumulation in bone. Over 8% ID/g of the
polyhydroxylate was present in the bone at 1 h, and that figure
increased to over 11% ID/g by 24 h.25 Conversely, the malonate
investigated herein accumulated only ∼2% ID/g in the bone. It
is possible that some 212Pb@C60 malonate localizes in the bone
due to formation of ionic and hydrogen bonds between
mineralized bone and the partially protonated form of the
malonic acid groups used to induce water solubility in the
radiofullerene.26 If this were true, it would explain why 2% ID/g
was found in the bone when none was expected and imply that
a different water-solubilizing group could potentially further
reduce radiofullerene accumulation in bone. In any case, the
observed ∼2% ID/g in the bone is comparable to the best value
obtained from a new set of PAC chelators.16b In that work, the
% ID/g found in the femur for 203Pb was much higher than for
the other radionuclides tested in the PAC chelators (177Lu, 86Y,
and 205/6Bi), indicating the continuing need for improved
methods for in vivo transport of lead and other bone-seeking
radioisotopes.
Conclusions
For the first time, endohedral metallofullerenes were prepared
containing R-emitting radionuclides suitable for use in radio-
immunotherapy, specifically 212Pb and 213Bi. A new preparation
route for radiofullerenes, using the recoil following R-decay of
a parent radionuclide, was also demonstrated. Malonic esters
were then added to the radiofullerenes, and repeated washing
of ethyl acetate solutions of the radiofullerene esters with
aqueous challenge solutions was found to be a rapid technique
for removal of exohedral radionuclides. While 212Pb@C60
appeared stable in solution, about 36% of the 212Bi formed
within the C60 by â-decay from the 212Pb@C60 was apparently
released from the fullerene or otherwise damaged the fullerene.
In a preliminary biodistribution study, encapsulation of 212Pb
in C60 malonic acid derivatives prevented the 212Pb from
accumulating in bone but demonstrated rather slow clearance.
This work demonstrates the potential of fullerenes for solving
problems that have not been solved by PAC chelators but also
illustrates some limitations to the use of fullerenes.
In contrast to chelation, the fullerene should be able to
maintain containment of 212Pb under any in vivo condition, as
the ability of the fullerene to retain encapsulation of an atom is
unaffected by pH, intermolecular complexation, or charge state
of the internal atom. The presence of any lead in the bone after
encapsulation within a fullerene is therefore surprising. Initially,
perhaps some activity could be attributed to blood and/or
adventitial tissues. Given the generally slow clearance of the
Acknowledgment. This work was supported by the National
Cancer Institute’s SBIR grant no.R43 CA96356 (M.D.D.). Work
at ORNL was also supported in part by the U.S. Department of
Energy, Office of Biological and Environmental Research and
Office of Nuclear Energy, Isotope Program. ORNL is managed
under contract DE-AC05-00OR22725 by UT-Battelle, LLC for
the U.S. Department of Energy. We acknowledge Dr. John
Gibson, Dr. Russ Knapp, and John Wright for their critical
review of the manuscript.
(26) As pointed out by a reviewer: Wilson, T. M.; Charifson, P. S.; Baxter, A.
D.; Geddie, N. G. Bioorg. Med. Chem. Lett. 1996, 6, 1043.
(27) Ashcroft, J. M.; Tsyboulski, D. A.; Hartman, K. B.; Zakharian, T. Y.; Marks,
J. W.; Weisman, R. B.; Rosenblum, M. G.; Wilson, L. J. Chem. Commun.
2006, 3004-6.
JA068639B
9
5138 J. AM. CHEM. SOC. VOL. 129, NO. 16, 2007