Microfluidic labeling of biomolecules with radiometals for use in nuclear medicine

Article abstract of DOI:10.1039/c0lc00162g

Radiometal-based radiopharmaceuticals, used as imaging and therapeutic agents in nuclear medicine, consist of a radiometal that is bound to a targeting biomolecule (BM) using a bifunctional chelator (BFC). Conventional, macroscale radiolabeling methods us

Full text of DOI:10.1039/c0lc00162g

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/loc | Lab on a Chip
Microfluidic labeling of biomolecules with radiometals for use in nuclear
medicine
bc
c
ab
Tobias D. Wheeler,ab Dexing Zeng,c Amit V. Desai,ab Birce Onal, David E. Reichert and Paul J. A. Kenis
€
*
Received 1st July 2010, Accepted 10th September 2010
DOI: 10.1039/c0lc00162g
Radiometal-based radiopharmaceuticals, used as imaging and therapeutic agents in nuclear medicine,
consist of a radiometal that is bound to a targeting biomolecule (BM) using a bifunctional chelator
(BFC). Conventional, macroscale radiolabeling methods use an excess of the BFC–BM conjugate
(ligand) to achieve high radiolabeling yields. Subsequently, to achieve maximal specific activity
(minimal amount of unlabeled ligand), extensive chromatographic purification is required to remove
unlabeled ligand, often resulting in longer synthesis times and loss of imaging sensitivity due to
radioactive decay. Here we describe a microreactor that overcomes the above issues through
integration of efficient mixing and heating strategies while working with small volumes of concentrated
reagents. As a model reaction, we radiolabel 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid
(DOTA) conjugated to the peptide cyclo(Arg-Gly-Asp-DPhe-Lys) with 64Cu2+. We show that the
microreactor (made from polydimethylsiloxane and glass) can withstand 260 mCi of activity over
720 hours and retains only minimal amounts of 64Cu2+ (<5%) upon repeated use. A direct comparison
between the radiolabeling yields obtained using the microreactor and conventional radiolabeling
methods shows that improved mixing and heat transfer in the microreactor leads to higher yields for
identical reaction conditions. Most importantly, by using small volumes (ꢀ10 mL) of concentrated
solutions of reagents (>50 mM), yields of over 90% can be achieved in the microreactor when using
a 1 : 1 stoichiometry of radiometal to BFC–BM. These high yields eliminate the need for use of excess
amounts of often precious BM and obviate the need for a chromatographic purification process to
remove unlabeled ligand. The results reported here demonstrate the potential of microreactor
technology to improve the production of patient-tailored doses of radiometal-based
radiopharmaceuticals in the clinic.
form a complex with the radiometal that is highly stable in vivo,
and possess a functional group that can form a bond with the
biomolecule.
1
Introduction
Nuclear imaging and therapy are vital to several areas of modern
medicine, including oncology, cardiology, hematology, and
studies of the biodistribution of drugs.1,2 These non-invasive
techniques rely on the introduction of radioactive agents
(radiopharmaceuticals) into the body to detect disease via Posi-
tron Emission Tomography (PET) or Single Photon Emission
Computed Tomography (SPECT), or to treat disease with
ionizing radiation. To minimize the systemic exposure of the
body to radiation, and to enhance the specificity and sensitivity
of the PET and SPECT imaging techniques, metallic radionu-
clides (radiometals) are bound to biomolecules (BMs) with bi-
functional chelators (BFCs).3 The BM (e.g., a peptide or an
antibody) is selected to have a high affinity for the tissue of
interest, to deliver and retain radiation only where it is needed.
The BFC is selected to have a high affinity for the radiometal,
Two radionuclides commonly used in nuclear imaging are
the positron emitter 18F (used in PET), and the gamma ray
emitter 99mTc (used in SPECT). These radionuclides have rela-
tively short half-lives (109 minutes and 6 hours, respectively) that
make them favorable for minimizing exposure of the body to
radiation, and have decay characteristics that make them
optimal for their respective imaging modalities. The relatively
short half-life of 18F and its typical labeling conditions (use of
organic solvents) lowers its suitability for use with biomolecules
such as antibodies and peptides. An alternative radionuclide that
has received increasing attention is the positron emitter 64Cu2+.4–6
The decay properties of this radiometal allow it to be used both
as a PET imaging agent4–7 and as a nuclear therapy agent.5,6,8 In
addition, its half-life of 12.7 hours, and the fact that large
quantities with high specific activity (>10 000 mCi molꢁ1) can be
produced at a cyclotron from enriched 64Ni,3 facilitate the
distribution of 64Cu2+ from a central production facility. Further
benefits associated with 64Cu2+ include: (1) its well-documented
coordination chemistry, redox chemistry, and biochemistry and
metabolism in humans; (2) the availability of a variety of aza-
macrocyclic BFCs that can chelate copper in the 2+ oxidation
state with high specificity and stability; and (3) the ability to
perform radiolabeling reactions in protein-friendly, aqueous
aInstitute for Genomic Biology, University of Illinois at Urbana
Champaign, 1206 W. Gregory Dr., Urbana, IL, 61801, USA. E-mail:
kenis@illinois.edu
bDepartment of Chemical & Biomolecular Engineering, University of
Illinois at Urbana Champaign, 600 S. Mathews Ave., Urbana, IL, 61801,
USA
cRadiological Sciences Division, Mallinckrodt Institute of Radiology,
Washington University School of Medicine, Campus Box 8225, 510
South Kingshighway, St Louis, MO, 63110, USA
This journal is ª The Royal Society of Chemistry 2010
Lab Chip, 2010, 10, 3387–3396 | 3387

View Article Online
media (as opposed to the organic solvents required for radio-
labeling with 18F), at pH ꢀ7, and at near-physiological temper-
atures.9
64Cu2+ by 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid
(DOTA) conjugated to the peptide cyclo(Arg-Gly-Asp-DPhe-
Lys)(cyclo(RGDfK)) which is a potential imaging agent in the
detection of cancer via PET.7,12,20–22
Conventional radiolabeling methods for radiometals such as
64Cu2+ typically require the dilution of small quantities of pico-
molar radiometal solution for convenient handling and proper
mixing, resulting in nanomolar concentrations of the radio-
metal.6,10–12 This dilution requires a large excess (ꢀ100-fold
compared to the radiometal concentration) of the potentially
expensive and difficult-to-obtain BFC–BM conjugate to ensure
the desired high percentage of bound radionuclide (>90%) within
a reasonable time (<1 hour). In turn, the use of large excesses of
BFC–BM conjugate necessitates extensive chromatographic
purification to remove unlabeled BFC–BMs and to obtain the
high specific activities that are desirable for application of the
radiopharmaceutical, for example in PET imaging. Chromato-
graphic purification is also potentially required to remove BFC–
BM impurities that may bind more strongly or more quickly to
the radiometal than the desired BFC–BM conjugate. For
instance, if the 100-fold excess of BFC–BM contains 1% impu-
rity, then the molar ratio of impurity to radiometal would be
1 : 1, potentially leading to the synthesis of unwanted radio-
metal–ligand complexes.
2
Design and testing
2.1 Design of the microreactor
We present here a PDMS-based, microfluidic reactor for labeling
BFC–BMs with radiometals (Fig. 1). The microreactor consists
of three key elements: (1) a serpentine microchannel for mixing,
in which we defined staggered herringbone grooves19 using soft
lithography; (2) a series of reservoirs for the incubation of the
radiometal–ligand mixture and (3) a thin-film heater for heating
the mixture.
2.1.1 Serpentine mixing channel. The microreactor incorpo-
rates a passive mixer developed by Stroock et al.19 to minimize
diffusive limitations to the overall rate of the radiolabeling reac-
tion. Generally, in the laminar, low Reynolds number flow that
occurs at the microscale, when two streams of reagents are brought
in contact, a depletion zone forms as a result of the consumption
of the reagents at the interface between the two streams. This
depletion zone grows in the transverse direction as the streams flow
axially along the channel. As a result, the reagents must diffuse
across increasingly longer distances in order for the reaction to
proceed. The small scale of microfluidic systems renders them less
susceptible to the growth of large depletion zones than macro-scale
systems (the distance for diffusion, Dr [m], is ultimately limited to
half the width of a microchannel, w/2 z 100 mm). However, the
time scale for diffusion across the channel in microfluidic systems
sD ¼ Dr2/D, where D [m2 sꢁ1] is the diffusivity of the reagent, can
still be significant (on the order of minutes), particularly for large
molecules that have low diffusivities in water (D < 10ꢁ11 m2 sꢁ1),
such as proteins and antibodies, compared to the time scale of the
In this report, we demonstrate the utility of microfluidics to
overcome some of the drawbacks associated with conventional
radiolabeling methods mentioned above. Previous efforts on
microfluidic approaches for the synthesis of radiopharmaceuti-
cals focused primarily on the synthesis of 18F- and 11C-labeled
agents.13–15 For example, Lee et al.16 designed an intricate pol-
y(dimethyl siloxane) (PDMS)-based microreactor for the multi-
step synthesis of 18F-FDG, a radiotracer commonly used to
detect cancer via PET. An improved version of this microreactor
provided yields of 96% after ꢀ15 minutes of total synthesis time,
compared to 75% yield after ꢀ45 minutes using an automated
apparatus for conventional synthesis.17 However, an off-chip
purification step was still required to give a radiochemical purity
of ꢀ99%, and loss of 18F through reaction with or absorption by
the PDMS was noted as a significant issue. In other work, Lu
et al. investigated the effect of infusion rate, i.e., residence time,
on the yields of several 18F and 11C-labeled carboxylic esters using
a T-junction, flow-through glass microreactor.18 Of relevance to
our work, they observed that incomplete diffusive mixing at
lower residence times adversely affected the yield.
reaction, sR ¼1/kC0 , where k [Mꢁn sꢁ1] is the reaction rate
n
constant, C0 [M] is the initial concentration of the reagent, and n is
the order of the reaction. Thus, depletion zones can impose dif-
fusive mass transport limitations on the rate of a reaction that
occurs in a microchannel and should be avoided, particularly for
high-throughput reactions with fast kinetics.
We introduce passive mixing of reagents, generated by stag-
gered herringbone grooves,19 to further reduce, and potentially
eliminate entirely, the limitation to reaction kinetics caused by
depletion zones. The staggered herringbone grooves in the mix-
ing channel induce chaotic stirring in the cross-section of the flow
that stretches and folds the interface between the co-flowing,
laminar streams. This stretching and folding reduces the max-
imum distance, Dr, that the solutes in the initially separate
streams must diffuse in order to form a homogeneous mixture19
and to react, resulting in smaller depletion zones, and thus a
reduced potential for diffusive mass transport limitations on the
rate of reaction. The maximum diffusive distance for chaotic
stirring can be approximated by, Dr ¼ w/2 exp(ꢁDy/l), where w
[m] is the width of the microchannel, Dy [m] is the distance
traveled along the axis of the microchannel, and l [m] is a char-
acteristic length determined by the geometry of the trajectories of
the chaotic stirring.19 Based on this equation and the estimated
Here we report on the design, fabrication, and validation of a
PDMS/glass-based microreactor for radiolabeling biomolecules
with 64Cu2+ or other radiometals that exploits several key attributes
of microfluidics. (1) The ability to manipulate small volumes of
reagents (from mL to pL), which eliminates the need for the dilu-
tion of radionuclides, potentially obviating the need to use excess
ligand and the associated chromatographic purification steps. (2)
The overall small size of the microreactor and peripherals, which
reduces space requirements and radioactivity shielding costs. (3)
The small volumes of reactants and the small dimensions of the
reactor, which sidestep some of the heat and mass transport
limitations encountered in macro-scale radiolabeling. Here we also
use staggered herringbone grooves19 to passively mix reagents and
thereby further reduce mass transport limitations. We successfully
validate the microreactor-based approach by comparison with
conventional procedures for radiolabeling using the chelation of
3388 | Lab Chip, 2010, 10, 3387–3396
This journal is ª The Royal Society of Chemistry 2010

View Article Online
Fig. 1 (a) Photograph and (b) schematic diagram of the microreactor. In (a), the mixing channels and incubation reservoirs are filled with blue dye. In
both (a) and (b), 1 is the radiometal inlet, 2 is the buffer inlet, 3 is the serpentine mixing microchannel, 4 is the BFC–BM inlet, 5 are the incubation
reservoirs, and 6 is the product outlet. In (b), 3a and 3b are the first and second legs of the mixing channel, respectively. The inset in (b) shows an
illustration of the staggered herringbone grooves defined in the PDMS ceiling of the mixing microchannel. The length, width, and height of the PDMS
portion of the microreactor are ꢀ200 ꢃ ꢀ1.500 ꢃ ꢀ0.2500.
value of l z 2 mm for the chaotic stirring induced by the
grooves,19 once the two streams have reached the end of a 3 cm
long, grooved microchannel, Dr will have decreased by roughly
six orders of magnitude, yielding a reduction in sD of 12 orders of
magnitude. This drastic reduction in mixing time eliminates any
mass transport limitations to the reaction rates, as is desired for
the radiometal labeling chemistries pursued here. Furthermore,
the relatively short length of the mixing channel afforded by
using chaotic stirring, as opposed to the longer channel lengths
that would be required when relying solely on diffusion to mix,
enables the use of high flow rates (>500 mL minꢁ1), without
developing prohibitively large pressure drops. This aspect will be
useful for high-throughput radiolabeling for cases in which
chelation rates are fast.
vapor trapped between the liquid and wall. Degassing the
reagents suppresses the formation of these bubbles.
The simple features discussed above combine to allow the
microreactor to perform both high-throughput, continuous flow
radiolabeling for fast chelation rates, and semi-batch, incubated
radiolabeling for slow chelation rates. In addition, the simplicity
of the design (i.e., the lack of extensive networks of micro-
channels and of valves and their required ancillary equipment) is
beneficial for its envisioned use in the clinic as an inexpensive,
disposable microreactor for the custom, on-demand synthesis of
radiopharmaceuticals.
2.2 Testing of the microreactor
We validated the microreactor by labeling a BFC–BM conjugate
with 64Cu2+. The BFC–BM conjugate consisted of the bifunctional
chelator 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid
(DOTA), and the peptide cyclo(Arg-Gly-Asp-DPhe-Lys)
(cyclo(RGDfK)). This peptide targets aVb3 integrin, a receptor
that is up-regulated on the surface of cells undergoing angiogen-
esis.20 The high occurrence of angiogenesis in tumors results in a
higher concentration of aVb3 integrin, and therefore a higher
concentration of bound, radiolabeled RGD, in contrast with the
surrounding, healthy tissue. Radiolabeled RGD-based imaging
agents are promising candidates for the imaging of tumors via
PET.7,12,20–22
2.1.2 Reservoirs and heater. For the case when chelation rates
are slow, the microreactor incorporates a series of reservoirs and
a thin film heater. The high degree of homogeneity of the reagent
mixture accomplished through the use of the staggered herring-
bone mixer, allows for the accumulation of clinically relevant
volumes of the mixture (up to 50 mL per run) in the reservoirs,
without provoking mass transport limitations to the rate of the
radiolabeling reaction. If regions of incomplete mixing were
present, these regions, and the depletion zones that develop
within them, would be magnified through the expansion of the
microchannel into the wider reservoir, and diffusive mass
transport limitations would be exacerbated. In the microreactor
presented here, the well-mixed solution of reagents can be incu-
bated for arbitrary periods of time at elevated temperatures to
achieve high yields, and then flushed quickly from the micro-
reactor for use. The microreactor itself can withstand elevated
temperatures, however, bubbles form within the microchannels
The radiolabeling method, summarized in Scheme 1, consists
of three steps: (1) the addition of 64Cu2+ in 0.1 N HCl (the state of
the radiometal as it is received the cyclotron) to 10 mM ammo-
nium acetate (NH4OAc) (pH ¼ 6.8) to form 64Cu(OAc)2 through
ligand exchange; (2) the addition of DOTA-cyclo(RGDfK) in
10 mM NH4OAc (pH ¼ 6.8) to the 64Cu(OAc)2 mixture to form
64Cu-DOTA-cyclo(RGDfK) through chelation; (3) the incuba-
tion of the 64Cu-DOTA-cyclo(RGDfK) mixture at room or
ꢂ
and reservoirs when the system is heated above 65 C, presum-
ably due to the expansion of microscopic pockets of air and water
This journal is ª The Royal Society of Chemistry 2010
Lab Chip, 2010, 10, 3387–3396 | 3389

View Article Online
Scheme 1 Chelation of 64Cu2+ by DOTA-cyclo(RGDfK). (a) The reaction scheme of the chelation of 64Cu2+ by DOTA-cyclo(RGDfK). Of the four
pendant arms of DOTA, two are used in chelation of the radiometal and one is used to conjugate the biomolecule cyclo(RGDfK). (b) The structure of
cyclo(RGDfK) and the position at which DOTA is conjugated to the biomolecule.
elevated temperature to drive the chelation reaction to comple-
tion. The two-step combination of reagents is performed in the
serpentine mixing channel of the microreactor, with radiometal
and buffer solutions mixing in the first leg of the channel (3a in
Fig. 1b) and ligand and radiometal-buffer solutions mixing in the
second leg of the channel (3b in Fig. 1b).
MA) was used to fabricate the SU-8 mold for the microreactor.
Three microlitre flow modular pump components, which included
a syringe pump, a pump driver circuit, and a power supply, were
obtained from Harvard Apparatus (Holliston, MA). The Kapton-
insulated, 200 ꢃ 200, thin film heater, the Omega CN740 tempera-
ture controller, and an Omega SA 1-RTD probe were obtained
from Omega Engineering (Stamford, CT). The Harrick plasma
cleaner was obtained from Harrick Plasma (Ithaca, NY).
Eppendorf tubes were obtained from MIDSCI, Inc. (St Louis,
MO). The BioScan AR-2000 radio-TLC plate reader was
purchased from Bioscan, Inc. (Washington, DC). The Thermo-
mixer was obtained from Eppendorf North America (Hauppauge,
NY). Radio-TLC plates were obtained from Whatman Thin
Layer Chromatography (Piscataway, NJ). Gas-tight, microlitre
syringes were obtained from Hamilton Co. (Reno, NV). The 2695
Waters HPLC system used in the purification of DOTA-
cyclo(RGDfK) was obtained from Waters Corporation (Milford,
MA). The Capintec CRC-712M radioisotope dose calibrator used
to measure activities for retention experiments was obtained from
Capintec, Inc. (Ramsey, NJ). The MS data obtained via LC-MS
were collected with a Micromass ZQ4000 from Waters Corpora-
tion (Milford, MA).
After passing through the mixing channels, the reagents flow
into the reservoirs, where they are incubated at room or elevated
temperatures for a certain period of time. This incubation step is
necessary, as the chelation of 64Cu2+ by DOTA-cyclo(RGDfK)
likely involves the rate-limiting rearrangement of a di- or mono-
protonated intermediate before chelation is complete, as is the
case for the formation of lanthanide complexes (Ln3+) with
DOTA–peptide conjugates.23
3
Materials and methods
3.1 Chemicals
RTV 615 poly(dimethyl siloxane) (PDMS) was obtained from
General Electric Company (Waterford, NY). SU-8 2050 was
obtained from MicroChem Corporation (Newton, MA). (Tride-
cafluoro-1,1,2,2-tetrahydrooctyl)trichlorosilane was obtained from
Gelest, Inc. (Morrisville, PA). Cyclo(RGDfK) was purchased from
Peptides International, Inc. (Louisville, KY). 1,4,7,10-Tetraazacy-
clododecane-1,4,7,10-tetraaceticacidmono(N-hydroxysuccinimide
ester)(DOTA-NHS-ester) was purchased from Macrocyclics
(Dallas, TX). 64Cu2+ in 0.1 M HCl was produced at Washington
University School of Medicine, and obtained through the Radio-
nuclide Resource for Cancer Applications. De-ionized water
(DI-H2O) was produced using a Millipore Milli-Qwater system. All
other chemicals and solvents were purchased from Sigma-Aldrich
(St Louis, MO).
3.3 Synthesis and purification of DOTA-cyclo(RGDfK)
DOTA-cyclo(RGDfK) was prepared by reacting a solution of
6.03 mg cyclo(RGDfK) and 14.0 mL triethylamine (10 eq.) in
1.0 mL dimethylformamide (DMF) with 8.38 mg DOTA-NHS
(1.1 eq.). After stirring for 3 hours at room temperature, 3 mL of
DI-H2O were added and stirred for another 30 minutes to
hydrolyze the excess DOTA-NHS ester. The crude DOTA-
cyclo(RGDfK) was then purified on a Waters HPLC system
using an Alltech Econosil C18 semi-preparative column (10 mm,
4.6 mm ꢃ 250 mm) with a mobile phase of 15 vol% acetonitrile
and 85 vol% de-ionized water with 0.1 vol% trifluoroacetic acid
(TFA), at a flow rate of 3.0 mL minuteꢁ1. Under these condi-
tions, the DOTA-cyclo(RGDfK) eluted at 13.5 minutes. The
identity and purity of the purified DOTA-cyclo(RGDfK) peptide
was confirmed by LC-MS using an Phenomenex Luna C18(2)
3.2 Equipment
Three sets of female 1/4–28 to female Luer lock adaptors with
ferrules and nuts, PEEK tubing of 1/1600 OD and 0.0100 ID, and
a NanoTight kit with 1/1600 ꢃ 0.02700 FEP sleeves, obtained from
Upchurch Scientific (Oak Harbor, WA) and Microbore PTFE
tubing (0.01200 ID ꢃ 0.03000 OD) obtained from Cole-Parmer
(Vernon Hills, IL) were assembled to connect syringes to the
microreactor. The 75 mm ꢃ 50 mm ꢃ 1 mm glass slide, used to
assemble the microreactor, and the Fisher Vortex Genie 2 were
obtained from Fisher Scientific (Pittsburgh, PA). A test grade, 300
silicon wafer, obtained from University Wafer (South Boston,
˚
column (5 mm, 100 A, 4.6 mm ꢃ 150mm). The analysis
was performed using the following gradient of 0.1 vol% TFA in
de-ionized water (A) and 0.1 vol% TFA in acetonitrile (B) with
a flow rate of 1.0 mL minuteꢁ1: 0–3 minutes: 90% A, 10% B; 13
minutes: 85% A, 15% B; 20–24.5 minutes: 50% A, 50% B; 25–30
minutes: 90% A, 10% B. A UV chromatogram and a TIC
3390 | Lab Chip, 2010, 10, 3387–3396
This journal is ª The Royal Society of Chemistry 2010

View Article Online
chromatogram are shown in Fig. 2a and b, respectively. The
mass spectrum of the peak at ꢀ12 minutes is shown in Fig. 2c,
and demonstrates that the peak corresponds to pure DOTA-
cyclo(RGDfK), with M1+ ¼ 990.9 and M2+ ¼ 496.4.
induced by the geometry of the grooves.19 Once the buffer and
radiometal solutions reached the end of the first mixing channel,
the ligand (BFC–BM) solution was pumped into the micro-
reactor through inlet 4 and mixed with the radiometal–buffer
mixture in a second mixing microchannel with staggered
herringbone grooves defined in the ceiling. The mixed reagents
then filled a series of five hexagonal reservoirs (5 in Fig. 1), after
which the flow was stopped and the mixture was heated to the
desired temperature and allowed to react for a specified incu-
bation time. After this incubation time, buffer solution was
pumped into the microreactor to flush out the product, through
the outlet (6 in Fig. 1), for collection. A schematic diagram of the
system used to operate the microreactor is shown in Fig. 3. The
flow of each reagent stream, radiometal, buffer, and ligand, into
the microreactor was driven by syringe pumps that were
controlled by a LabVIEW virtual instrument (VI). Communi-
cation between the VI and the syringe pump drivers was facili-
tated by a serial connection. The VI was used to set the individual
flow rates and regulate the timing of the pumps, with inputs for
controlling the incubation time, the total flow rate, the desired
volume of product (the maximum being the total volume of
the reservoirs, ꢀ50 mL), and the stoichiometric molar ratios
of buffer-to-radiometal and ligand-to-radiometal. For radio-
labeling at elevated temperatures, a thin-film heater, placed in
3.4 Fabrication of the microreactor
The microreactor was fabricated using SU-8-based soft-lithog-
raphy techniques for PDMS.24 The dimensions of the features
shown in Fig. 1 are as follows: serpentine microchannel (100 mm
high, 200 mm wide, 10.7 cm long), five hexagonal reservoirs
(100 mm high, 5 mm wide, 3 cm long), microchannels between
reservoirs (100 mm high, 200 mm wide, 7.2 mm long). The stag-
gered herringbone grooves in the mixing channels (see expanded
view in Fig. 1b) were designed to have the same geometry as
those developed by Stroock et al.19 An SU-8 mold of the features
was fabricated as follows: (1) negative images of the microchannels,
inlets, outlets, and reservoirs were printed on one transparency film
and those of the grooves were printed on a second transparency
film, using a 5080 dpi printer; (2) the pattern of the microchannels,
inlets, outlets, and reservoirs was transferred from the first trans-
parency to a 100 mm thick layer of SU-8 2050 spun onto a silicon
wafer, via photolithography; (3) the pattern of the grooves was
transferred from the second transparency to a second, 50 mm thick
layer of SU-8 2050 that was spun over the first layer, following an
alignment step so as to define the grooves directly above the
mixing channel; (4) unexposed SU-8 was dissolved using
propylene glycol methyl ether acetate (PGMEA), and the exposed
silicon surface was passivated via vapor deposition of (trideca-
fluoro-1,1,2,2-tetrahydrooctyl) trichlorosilane, under vacuum.
PDMS (10 : 1 ratio of precursor to curing agent) was then poured
over the SU-8 mold and cured at 60 ꢂC for 2+ hours. The cured
PDMS was removed from the mold, inlet and outlet holes were
punched, and the PDMS was bonded to a 75 mm ꢃ 50 mm ꢃ 1
mm glass slide, after 1 minute plasma treatment of both compo-
nents with a Harrick Plasma Cleaner (RF level: Hi, Pressure: 500–
1000 mTorr), followed by baking at 60 ꢂC for 12 hours.
3.5 Operation of the microreactor
Fig. 3 Schematic diagram of the radiolabeling system. Three LabVIEW-
controlled syringe pumps drive the flow of the BFC–BM (ligand), buffer,
radiometal solutions into the microreactor. A thin-film heater, controlled
using a temperature controller and a resistance temperature detector
(RTD) probe affixed to the heater, sets the incubation temperature. Lead
shielding surround the microreactor and radiometal pump when oper-
ating with radioactive reagents.
The microreactor was operated in semi-batch mode. Upon
entering the microreactor through inlets 1 and 2, respectively,
a radiometal solution and a buffer solution flowed through the
mixing channel (3 in Fig. 1). As the two solutions flowed along
the microchannel, they were mixed by the chaotic advection
Fig. 2 Purity of DOTA-cyclo(RGDfK), as indicated by (a) UV and (b) total ion current (TIC) chromatograms. The mass spectrum of the material
corresponding to the peak with an elution time of 12.6 minutes in (b) is shown in (c).
This journal is ª The Royal Society of Chemistry 2010
Lab Chip, 2010, 10, 3387–3396 | 3391

View Article Online
contact with the glass surface of the microreactor, was used to
heat the microreactor, and a temperature controller and resis-
tance temperature detector (RTD) probe were used to maintain
the temperature within ꢄ1 ꢂC of the desired temperature set-
point. The radiometal syringe pump, heater, and microreactor
were shielded on four sides and below with 200 of lead shielding.
The radiometal syringe was fitted with a 7.5 cm length of PEEK
tubing connected to an Upchurch Luer lock fitting. The PDMS
microreactor was fit with a 10 cm length of PTFE tubing con-
nected to an Upchurch NanoTite fitting. The syringe and PEEK
‘‘needle’’ were inserted into the fitting and tightened.
(pH ¼ 6.8) to give final concentrations of 200, 100, 40, and 2 mM
DOTA-cyclo(RGDfK).
3.7.1 Conventional radiolabeling procedures. Conventional
radiolabeling was performed using two methods: (1) mixing,
incubation and heating of a total volume of 10.5 mL of solution
with a Thermomixer, and (2) mixing of a total volume of 105 mL
of solution using a vortexer, followed by continued mixing,
incubation and heating with a Thermomixer. For method (1),
3.0 mL of 100 mM DOTA-cyclo(RGDfK) solution were first
added to 1.5 mL of 10 mM NH4OAc (pH ¼ 6.8) in a 1.6 mL
Eppendorf tube. Then, 6.0 mL of 50 mM 64Cu2+ solution were
added to the mixture, and the Eppendorf tube was placed in the
Thermomixer and incubated for 12 minutes at a temperature of
23, 37 or 47 ꢂC. For method (2), 30 mL of 100 mM DOTA-
cyclo(RGDfK) solution were first added to 15 mL of 10 mM
NH4OAc (pH ¼ 6.8) in a 1.6 mL Eppendorf tube. Then, 60 mL of
50 mM 64Cu2+ solution were added to the mixture, the mixture
was vortexed for 10 seconds, and the Eppendorf tube was placed
in the Thermomixer and incubated for 12 minutes at a tempera-
ture of 23, 37 or 47 ꢂC. The final concentration of 64Cu2+ and
DOTA-cyclo(RGDfK) was 28.6 mM in both methods. Following
the incubation period, the extent of reaction was determined as
described in Section 3.7.3.
3.6 Measurement of the retention of 64Cu2+
A 3 cm long, 200 mm wide, 100 mm tall microchannel with
staggered herringbone grooves, fabricated from PDMS and glass
and with a volume of ꢀ0.6 mL, was used for these experiments.
Prior to use, the microchannel was cleaned with 50 mL of 1 N
nitric acid to remove trace metals, and flushed with 5 mL of
10 mM NH4OAc buffer solution.
3.6.1 Pre-treatment with Cu2+/Na+. 50 mL of a 10 mM CuCl2/
NaCl solution were injected into the microchannel and allowed
to sit for 30 minutes. The microchannel was then flushed with
5 mL of 10 mM NH4OAc buffer solution (pH ¼ 6.8).
3.7.2 Microreactor radiolabeling procedure. The general
mode of operation of the microreactor is described in Section 3.5.
For the radiolabeling experiments summarized in Fig. 6a (effect
of residence time on extent of reaction) and Fig. 6b (effect of
mixing method and temperature on extent of reaction), the
following solutions were used: (1) 50 mM carrier-added 64Cu2+
solution, (2) 100 mM DOTA-cyclo(RGDfK) stock solution, and
(3) 10 mM NH4OAc (pH ¼ 6.8) buffer. Using the LabVIEW
interface, the pumps were programmed to combine the radio-
metal (1), ligand (2), and buffer (3) solutions with the same
volumetric ratios as those used for the conventional radio-
labeling procedure, giving a ligand-to-metal molar ratio of 1 : 1
and a final concentration of 64Cu2+ and DOTA-cyclo(RGDfK) of
28.6 mM. The total flow rate was set to 50 mL minuteꢁ1, the total
volume was set to 20 mL (an additional 30 mL of buffer was
pumped into the microreactor to fill all the reservoirs). For the
experiments in Fig. 6a, the temperature was set to 37 ꢂC, and the
incubation time was set to 5, 10, and 20 minutes (resulting in total
residence times, tRES ¼ 7, 12, and 22 minutes: e.g., 1 minute for
filling + 10 minutes for incubation + 1 minute for flushing). For
the experiments in Fig. 6b, the incubation time was set to 10
minutes (tRES ¼ 12 minutes), and the temperature was set to 23,
37 and 47 ꢂC. After the collection of product, the extent of
reaction was determined as described in Section 3.7.3.
3.6.2 Retention measurement. 30 mL of no carrier-added (only
radioactive) 64Cu2+ in 10 mM NH4OAc buffer solution (pH ¼
6.80) were injected into the microchannel. After 10 minutes, the
64Cu2+ solution was displaced from the microchannel by 200 mL
of air, via syringe. The microchannel was then flushed with 200
mL of 10 mM NH4OAc buffer. The percent of injected activity
retained in the microchannel was calculated from the known
activity of the injected solution, and the difference between the
activity left behind in the microchannel after flushing and the
initial activity of the microchannel measured before the 64Cu2+
solution was injected, all measured with the Capintec radioiso-
tope dose calibrator.
3.7 Radiolabeling of DOTA-cyclo(RGDfK)
Stock solutions of carrier-added 64Cu2+ were prepared by first
diluting the solution obtained from the cyclotron (20 mCi 64Cu2+
in 0.1 M HCl) with 10 mM NH4OAc (pH ¼ 6.8) to give a specific
activity of 1 mCi mLꢁ1. The various copper solutions were then
prepared from this original stock solution. In order to prepare
a 200 mM carrier-added 64Cu2+ solution, 20 mL of the 1 mCi mLꢁ1
64Cu2+ solution were mixed with 20 mL of 10 mM non-radioactive
copper solution and 960 mL of 10 mM NH4OAc (pH ¼ 6.8). The
different concentrations of carrier-added 64Cu2+ solutions listed
in Sections 3.7.1 and 3.7.2 were obtained by the dilution of the
above 200 mM carrier-added 64Cu2+ solution with 10 mM
NH4OAc.
For the radiolabeling experiments summarized in Fig. 7 (effect
of concentration on extent of reaction), combinations of stock
solutions and buffer-to-metal molar ratio inputs were used to
give the following final concentrations of radiometal and ligand
(1 : 1 molar ratio): 1 mM (2 mM stock solutions, 1 : 1 buffer-to-
radiometal ratio), 10 mM (40 mM stock solutions, 500 : 1
buffer-to-radiometal ratio), 20 mM (40 mM stock solutions, 1 : 1
buffer-to-radiometal ratio), 30 mM (100 mM stock solutions,
130 : 1 buffer-to-radiometal ratio), 50 mM (100 mM stock solu-
tions, 1 : 1 buffer-to-radiometal ratio), and 90 mM (200 mM stock
Stock solutions of 10 mM DOTA-cyclo(RGDfK) were
prepared by first dissolving 10.0 mg of purified DOTA-
cyclo(RGDfK) (Section 3.3) in 100 mL of a mixture of 10 mM
NH4OAc (pH ¼ 6.8) and acetonitrile (MeCN) (50 : 50 vol%),
and then diluting these solutions with 10 mM NH4OAc
3392 | Lab Chip, 2010, 10, 3387–3396
This journal is ª The Royal Society of Chemistry 2010

View Article Online
solutions, 11 : 1 buffer-to-radiometal ratio). Experiments were
performed with the same settings as listed above, and with
tRES ¼ 12 minutes and at a temperature of 37 ꢂC. After the
collection of product, the extent of reaction was determined as
described in Section 3.7.3.
Throughout the course of radiolabeling experiments, three
trial runs were performed before data were recorded to ensure
proper functioning of the microreactor after syringe pump or
heater settings were changed and after depleted syringes were
replenished.
3.7.3 Measurement of radiolabeling yield. At the end of each
reaction run, a 1 mL aliquot of product was spotted onto a radio-
TLC plate and developed using a mobile phase of methanol/10%
ammonium formate buffer (2 : 1 volume ratio). After developing,
the radio-TLC plates were then counted using the Bioscan radio-
TLC scanner to quantify free and bound 64Cu2+. The radiolabeling
yields were calculated by dividing the area of the radioactivity
peak obtained for the bound 64Cu2+ by the total area of the
radioactivity peaks obtained for both bound and free 64Cu2+.
Fig. 4 Retention of 64Cu2+ in a PDMS/glass microchannel with stag-
gered herringbone grooves. The retention measured without blocking the
surface of the microchannel using non-radioactive Cu2+ (white bars) is
compared to the retention measured after blocking the surface of the
microchannel using non-radioactive Cu2+ (grey bars). Error bars repre-
sent the standard deviations of the averages of multiple experiments
(n ¼ 1 for no Cu2+ pre-treatment, n ¼ 2 for Cu2+ pre-treatment).
4
Results and discussion
4.1 Compatibility of PDMS/glass with 64Cu2+
We performed the experiments discussed in this section to
determine the compatibility of 64Cu2+with the materials
comprising the microreactor. For microfluidic radiolabeling with
18F, the absorption of the radionuclide by PDMS is a significant
problem;14,16,17 up to 95% of an amount of injected activity can be
retained.17 To determine whether similar issues will affect radio-
labeling in the current microreactor, we examined the retention
of 64Cu2+ and the ability of PDMS and glass to withstand the
emissions of this radionuclide.
responsible for the retention observed in Fig. 4, we used a Na+
solution to block the surface of the microchannel. Following this
pre-treatment, we observed the same decrease in the percentage
of injected activity retained as for the case when we pre-treated
the microchannel with the Cu2+ solution (data not shown). From
this and the preceding results, we conclude that the retention of
64Cu2+ on the glass surface is primarily due to electrostatic
interactions and not due to absorption of the ions into the
PDMS. The retention of 64Cu2+ in the microchannel can be
minimized either by changing the charge of the glass surface of
the microchannel to positive, by functionalizing the glass with
a positively charged silane, for example,25 or by fabricating the
microreactor completely in PDMS. Note: in the radiolabeling
experiments discussed in later sections, we ignore data for the
first three reactions performed in the microreactor. This proce-
dure ensures that the influence of 64Cu2+ retention on the
measured extent of reaction is minimal, by saturating the surface
of the microreactor with copper ions.
InFig. 4, weplot thepercent of injected activity of 64Cu2+ retained
in a single microchannel (length z 3 cm, volume z 0.6 mL) with
staggered herringbone grooves defined in the ceiling, for a series of
injections of activity. The white bars represent data for a micro-
channel that has been washed with nitric acid, and the grey bars
represent data for a microchannel that has been washed with nitric
acid and then pre-treated with a non-radioactive Cu2+ solution
(Section 3.6). The microchannel that was not pre-treated with Cu2+
solution retained a substantial portion of the activity of the first
injection (ꢀ70%), but retained only ꢀ5% of the activity of subse-
quent injections. The microchannel that was pre-treated with Cu2+
solution retained only ꢀ15% of the activity of the first injection and
also retained only ꢀ5% of the activity of subsequent injections. The
differencebetweentheretentionoftheactivityinthefirstinjectionin
the pre-treated and non-pre-treated microchannels, and the
decreaseinretentionofactivity insubsequentinjections suggest that
64Cu2+ adheres to the walls of the microchannel, though once the
surface is saturated, no further adhesion occurs. The adhesion of the
64Cu2+ is presumably due to non-specific, electrostatic interactions
between the positively charged copper ions and the negatively
charged glass surface; polymer coatings with similar chemical
groups to those of PDMS are used to minimize electrostatic inter-
actions between positively charged solutes and the surfaces of
capillary electrophoresis equipment made of glass.25
Over the course of the experiments performed in this work, we
used a single microreactor. In total, the device was exposed to 260
mCi of radiation over 720 hours. Only after this exposure did we
observe signs of damage—the microreactor began leaking, indi-
cating that the seal between the PDMS and glass had decayed.
This leaking was the only instance of mechanical failure that we
observed, and occurred after the completion of all experiments
reported here (after approximately six months of use). Thus, the
device is sufficiently robust for its envisioned role as a single- or
multiple-use (<20), disposable microreactor for radiolabeling.
4.2 Microfluidic radiolabeling
In this section, we examine the radiolabeling of DOTA-
cyclo(RGDfK) with 64Cu2+ using a 1 : 1 stoichiometric ratio of
the two reagents. For the purpose of evaluating the performance
To confirm the hypothesis that non-specific electrostatic
interactions between the microchannel surface and the 64Cu2+ are
This journal is ª The Royal Society of Chemistry 2010
Lab Chip, 2010, 10, 3387–3396 | 3393

View Article Online
of the microreactor, all experiments were performed using
carrier-added 64Cu2+, so that the concentration of Cu2+ was
known, and to maximize the number of experiments that could
be performed with one batch of the radiometal. Thus, although
we refer to 64Cu2+solutions, over 99% of the copper present in the
solutions is not radioactive. The chemical equivalence of radio-
active and non-radioactive Cu-DOTA-cyclo(RGDfK) is
demonstrated by the identical chromatographic elution times of
the two molecules (see Fig. 5). For the purpose of operating the
microreactor in the clinic, in the production of PET imaging
agents, for example, undiluted radiometal solution obtained
directly from the cyclotron would be used instead.
temperatures. ‘Conventional, 10 mL’ and ‘Conventional, 100 mL’
represent data obtained when the reaction was performed using
conventional radiolabeling procedures with small volumes (ꢀ10
mL) and large volumes (ꢀ100 mL), respectively (Section 3.7.1).
The yields obtained using the microreactor at 37 and 47 ꢂC were
significantly higher than those obtained using the 10 mL
conventional procedure. We speculate that the superior mixing
of small volumes of reagents achieved by the microreactor
resulted in this improved performance. To support this hypoth-
esis, we compared the yields obtained using the microreactor to
those obtained using conventional procedures with a larger
volume (Conventional, 100 mL), in which more efficient mixing is
possible through vortexing. For both methods, the yield is the
same at 23 and 47 ꢂC, within experimental error, indicating that
the mixing achieved by the microreactor and by macro-scale
vortexing is sufficient to overcome the diffusive mass transfer
limitations to the radiolabeling reaction that hinder the perfor-
mance of the ‘Conventional, 10 mL’ method. For the micro-
reactor, the yield obtained at 37 ꢂC is the same as that obtained at
47 ꢂC, within experimental error. This result suggests that the
microreactor achieves the maximum yield possible for reagent
concentrations of ꢀ30 mM and tRES ¼ 12 minutes at a lower
temperature than the conventional method, and may also suggest
improved performance through enhanced heat transfer in the
microreactor.
In Fig. 6a, we plot the radiolabeling yield (i.e., the percent of
64Cu2+ chelated by DOTA-cyclo(RGDfK)) obtained by mixing
50 mM solutions of 64Cu2+ and DOTA-cyclo(RGDfK) at a 1 : 1
stoichiometric ratio, and incubating the mixture at 37 ꢂC for
residence times, tRES, of 7, 12 and 22 minutes. The data indicate
that, at this concentration of reagents, over 80% of the radiometal
is chelated after 12 minutes. To test reproducibility, the radio-
labeling yield for this residence time was measured for solutions
made from three separate productions of 64Cu2+. The standard
deviation for this average is larger than those for experiments that
were performed with the same batch of 64Cu2+ (7 and 22 minutes),
indicating the variability of the specific activity of the 64Cu2+
between productions. This variability results in changing amounts
of trace metal contaminants that can compete with Cu2+ for
chelation with DOTA (e.g., Ni, Fe, and Zn).
In addition to enhancing reaction rates through efficient mass
and heat transfer, the ability of microfluidic systems to manip-
ulate small volumes of concentrated reagents can potentially lead
to high radiolabeling yields. In Fig. 7, we plot radiolabeling yield
as a function of the final concentration of 64Cu2+, MF, (i.e., the
concentration after all solutions are mixed together) for a stoi-
chiometric ratio of 64Cu2+ to DOTA-cyclo(RGDfK) of 1 : 1,
tRES ¼ 12 minutes, and an incubation temperature of 37 ꢂC. The
data show that, by increasing the final concentration of the
reagents to >50 mM, yields approaching >90% are obtainable
with these reaction conditions. Due to the 1 : 1 stoichiometric
ratio, the need for the separation of unlabeled BFC–BMs is
eliminated once yields reach >90%. The concentration of radi-
ometal obtained from the cyclotron is typically 1–2 mCi z 4
picomoles in ꢀ10 mL, or ꢀ0.4 mM. Final concentrations $50 mM
could be obtained for the clinical production of radiopharma-
ceuticals by (1) minimizing the dilution of the radiometal solu-
tion by using highly concentrated solutions of buffer and ligand,
and (2) implementing a microfluidic means of increasing the
concentration of the radiometal, such as through the evaporation
of water from the radiometal solution.
To contextualize the improvement in radiolabeling provided
by the microreactor, we make a direct comparison between the
radiolabeling yield obtained by the microreactor and those
obtained through conventional radiolabeling methods, again
using a 1 : 1 stoichiometric ratio of 64Cu2+ to DOTA-cyclo
(RGDfK). To make this comparison, we chose a residence time
of 12 minutes and a radiometal/ligand concentration of ꢀ30 mM,
such that the radiolabeling yield is high enough to be measured
accurately, but does not reach 100% in the microreactor, based
on the results in Fig. 6a. Fig. 6b shows the radiolabeling yield for
the chelation of 64Cu2+ by DOTA-cyclo(RGDfK) obtained using
three different radiolabeling methods at three different
5
Conclusions
We present a simple, inexpensive microreactor made from
PDMS and glass for radiolabeling biomolecule-bifunctional
chelator conjugates (BFC–BMs) with radiometals for applica-
tion as imaging or therapeutic agents in nuclear medicine. The
microreactor is able to efficiently mix small volumes of reagents
(ꢀ10 mL or less) by using chaotic advection that is induced by
staggered herringbone grooves19 defined in a mixing micro-
channel in which reagents are combined, and to incubate the
reaction mixture at elevated temperatures after it fills several
micro-reservoirs (total volume ¼ 50 mL). We tested the
Fig. 5 UV and radio chromatograms for carrier-added 64Cu-DOTA-
cyclo(RGDfK). Chromatograms of a 50 mM sample synthesized using the
conventional method, at T ¼ 37 ꢂC and tRES ¼ 12 minutes (see Section
3.7.1), were obtained using the LC-MS method described in Section 3.3.
3394 | Lab Chip, 2010, 10, 3387–3396
This journal is ª The Royal Society of Chemistry 2010

View Article Online
Fig. 6 Radiolabeling yield for the production of 64Cu-DOTA-cyclo(RGDfK) as a function of (a) residence time, tRES, and (b) radiolabeling method at
various temperatures, T. In (a), error bars represent standard deviation in the extent of reaction for experiments using radiometal solutions made from
the same batch of 64Cu2+ for tRES ¼ 7 and 22 minutes (n ¼ 3) and between experiments using radiometal solutions made from three different batches of
64Cu2+ for tRES ¼ 12 minutes (3 repetitions for each batch, n ¼ 9). Experiments were performed at T ¼ 37 ꢂC. Lines between points are guides for the eye.
In (b), for the ‘Microreactor’ method, error bars represent the standard deviation in extent of reaction between three experiments using radiometal
solutions made from two (T ¼ 23 and 47 ꢂC, n ¼ 6) and three (T ¼ 37 ꢂC, n ¼ 9) different batches of 64Cu2+. For the other two methods, error bars
represent the standard deviation in extent of reaction between three experiments using radiometal solutions made from the same batch of 64Cu2+ (n ¼ 3).
Experiments were performed with tRES ¼ 12 minutes. All experiments were performed using a 1 : 1 stoichiometric ratio of 64Cu2+ to DOTA-
cyclo(RGDfK).
microreactor is blocked with positively charged ions (ꢀ5%
retention of injected activity). Furthermore, from a comparison
between the radiolabeling yields of the microreactor and of two
conventional radiolabeling methods at various temperatures, we
conclude that the higher radiolabeling yield observed for the
microreactor was achieved through efficient mixing with chaotic
advection and through enhanced heat transfer. Finally, we
demonstrate that, by using small volumes of concentrated radi-
ometal (ꢀ50 mM), it is possible to achieve high radiolabeling
yields (>90%) without resorting to using an excess of BFC–BM
to accelerate the rate of reaction, as is necessary in macro-scale,
conventional radiolabeling procedures that require dilution of
the radiometal. High yields with a 1 : 1 stoichiometric ratio of
radiometal to BFC–BM eliminate the need for chromatographic
purification of the product to remove unlabeled BFC–BMs,
resulting in shorter synthesis times and therefore a higher specific
activity of the resulting imaging or therapeutic agent.
The results we report here suggest that this microreactor-based
Fig. 7 Radiolabeling yield as a function of the final concentration of
64Cu2+, MF, using a 1 : 1 stoichiometric ratio of 64Cu2+ to DOTA-
cyclo(RGDfK). Error bars represent standard deviation in the extent of
reaction for three experiments (tRES ¼ 12 minutes, T ¼ 37 ꢂC), using
radiometal solutions made from the same batch of 64Cu2+. Lines between
points are guides for the eye.
approach has potential for improving the preparation of radio-
pharmaceuticals in the clinic. We plan to further optimize
chelation conditions for radiolabeling with other radiometals,
such as 99mTc and 68Ga, and to integrate pre-concentration and
purification elements in the microreactor to ensure that high
radiolabeling yields are obtained reproducibly, and that the
radiometal solution is sufficiently pure and concentrated for the
eventual clinical application of resulting radiopharmaceuticals in
humans. Furthermore, we plan to develop a complementary
microreactor for the conjugation of bifunctional chelators to
disease-specific biomolecules. Such a system would provide great
versatility for the production of patient-tailored doses of imaging
and therapeutic agents for nuclear medicine.
performance of the microreactor by radiolabeling DOTA-
cyclo(RGDfK) with 64Cu2+. We show that the materials from
which the microreactor is made can withstand substantial doses
of radiation (260 mCi), and interact minimally with the radio-
metal, once the negatively charged glass surface of the
This journal is ª The Royal Society of Chemistry 2010
Lab Chip, 2010, 10, 3387–3396 | 3395

View Article Online
12 Y. Wu, X. Zhang, Z. Xiong, Z. Cheng, D. R. Fisher, S. Liu,
S. S. Gambhir and X. Chen, J. Nucl. Med., 2005, 46, 1707–
1718.
Acknowledgements
We acknowledge financial support from the Department of
Energy Office of Biological and Environmental Research, grant
number DE-FG02-08ER64682. The production of 64Cu2+ was
supported by the National Institutes of Health, grant number
R24 CA86307.
13 H. Audrain, Angew. Chem., Int. Ed., 2007, 46, 1772–1775.
14 A. M. Elizarov, Lab Chip, 2009, 9, 1326–1333.
15 P. W. Miller, J. Chem. Technol. Biotechnol., 2009, 84, 309–315.
16 C.-C. Lee, G. Sui, A. Elizarov, C. J. Shu, Y.-S. Shin, A. N. Dooley,
J. Huang, A. Daridon, P. Wyatt, D. Stout, H. C. Kolb,
O. N. Witte, N. Satyamurthy, J. R. Heath, M. E. Phelps,
S. R. Quake and H.-R. Tseng, Science, 2005, 310, 1793–1796.
17 A. M. Elizarov, M. R. van Dam, Y. S. Shin, H. C. Kolb,
H. C. Padgett, D. Stout, J. Shu, J. Huang, A. Daridon and
J. R. Heat, J. Nucl. Med., 2010, 51, 282–287.
References
1 T. Leitha, Curr. Pharm. Des., 2009, 15, 173–187.
2 W. J. G. Oyen, L. Bodei, F. Giammarile, H. R. Maecke, J. Tennvall,
M. Luster and B. Brans, Ann. Oncol., 2007, 18, 1782–1792.
3 S. Liu, Adv. Drug Delivery Rev., 2008, 60, 1347–1370.
4 C. J. Anderson and R. Ferdani, Cancer Biother. Radiopharm., 2009,
24, 379–393.
18 S. Y. Lu, P. Watts, F. T. Chin, J. Hong, J. L. Musachio, E. Briard and
V. W. Pike, Lab Chip, 2004, 4, 523–525.
19 A. D. Stroock, S. K. Dertinger, A. Ajdari, I. Mezic, H. A. Stone and
G. M. Whitesides, Science, 2002, 295, 647–651.
20 I. Dijkgraff and O. C. Boerman, Cancer Biother. Radiopharm., 2009,
24, 637–347.
21 G. Z. Ferl, R. A. Dumont, I. J. Hildebrandt, A. Armijo, R. Haubner,
G. Reischl, H. Su, W. A. Weber and S.-C. Huang, J. Nucl. Med., 2009,
50, 250–258.
5 M. Shokeen and C. J. Anderson, Acc. Chem. Res., 2009, 42, 832–841.
6 S. V. Smith, J. Inorg. Biochem., 2004, 98, 1874–1901.
7 X. Chen, E. Sievers, Y. Hou, R. Park, M. Tohme, R. Bart,
R. Bremner, J. R. Bading and P. S. Conti, Neoplasia (Ann Arbor,
MI, U. S.), 2005, 7, 271–279.
22 Z. Liu, Y. Yan, S. Liu, F. Wang and X. Chen, Bioconjugate Chem.,
2009, 20, 1016–1025.
23 L. M. De Leon-Rodriguez and Z. Kovacs, Bioconjugate Chem., 2008,
19, 391–402.
24 Y. N. Xia and G. M. Whitesides, Angew. Chem., Int. Ed., 1998, 37,
551–575.
25 C. Huhn, R. Ramautar, M. Wuhrer and G. W. Somsen, Anal.
Bioanal. Chem., 2010, 396, 297–314.
8 J. S. Lewis, R. Laforest, T. L. Buettner, S.-K. Song, Y. Fujibayashi,
J. M. Connett and M. J. Welch, Proc. Natl. Acad. Sci. U. S. A.,
2001, 98, 1206–1211.
9 T. J. Wadas, E. H. Wong, G. R. Weisman and C. J. Anderson, Curr.
Pharm. Des., 2007, 13, 3–16.
10 D. L. Kukis, S. J. DeNardo, G. L. DeNardo, R. T. O’Donnell and
C. F. Meares, J. Nucl. Med., 1998, 39, 2105–2110.
11 T. J. Wadas and C. J. Anderson, Nat. Protocols, 2006, 1, 3062–3068.
3396 | Lab Chip, 2010, 10, 3387–3396
This journal is ª The Royal Society of Chemistry 2010