Aminooxy-functionalized DOTA for radiolabeling of oxidized antibodies: Evaluation of site-specific 111In-labeled trastuzumab

Article abstract of DOI:10.1002/jlcr.2950

Antibody-targeted nuclear imaging or radiotherapy rely on chemical modification of the polypeptide backbone that may hamper the antigen binding and hence the targeting ability. In this study, we examine a strategy for covalent modification and radiolabeling of monoclonal antibodies utilizing the oligosaccharide moieties on the molecule: the ¹¹¹In-radiolabeling of the monoclonal antibody trastuzumab (Herceptin, Roche, Basel, Switzerland), which targets the breast cancer-related ErbB2/human epidermal growth factor-2 (HER2) protein. The site-specific radiolabeled antibody was obtained through mild oxidation of the antibody Fc region with periodate and its conjugation with an aminooxy-functionalized tetraazacyclododecanetetraacetic acid (DOTA) chelator. It was further evaluated both in vitro with HER2 expressing cells and in vivo with tumor-bearing mice by single photon emission computed tomography (SPECT), and compared with the DOTA-N-hydroxysuccinimide methodology. Such radiolabeled trastuzumab did not show significant difference in stability when compared with lysine-conjugated antibodies (>95%, 3 days postinjection). They also maintain their immunoreactivity (>90% immunoreactive fraction) and affinity both in vitro (K_d ~ 4-5 nM) and in vivo (12.9 %ID/g in MDA-MB-361 tumors). Our results confirm that the formation of a hydrolytically stable oxime linker between a chelating agent and the oxidized N-glycan residues, away from the variable antigen-binding sites, is an attractive approach for site-specific conjugation of antibodies and accessing radiopharmaceuticals and immunotherapeutics.

Full text of DOI:10.1002/jlcr.2950

Research Article
Received 5 February 2012,
Revised 7 March 2012,
Accepted 7 March 2012
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.2950
Aminooxy-functionalized DOTA for
radiolabeling of oxidized antibodies:
evaluation of site-specific 111In-labeled
trastuzumab
*
Romain Bejot, Julian Goggi, Shebbrin S. Moonshi,
Parasuraman Padmanabhan, and Kishore K. Bhakoo
Antibody-targeted nuclear imaging or radiotherapy rely on chemical modification of the polypeptide backbone that may hamper
the antigen binding and hence the targeting ability. In this study, we examine a strategy for covalent modification and radiolabel-
ing of monoclonal antibodies utilizing the oligosaccharide moieties on the molecule: the 111In-radiolabeling of the monoclonal
antibody trastuzumab (HerceptinW, Roche, Basel, Switzerland), which targets the breast cancer-related ErbB2/human epidermal
growth factor-2 (HER2) protein. The site-specific radiolabeled antibody was obtained through mild oxidation of the anti-
body Fc region with periodate and its conjugation with an aminooxy-functionalized tetraazacyclododecanetetraacetic acid
(DOTA) chelator. It was further evaluated both in vitro with HER2 expressing cells and in vivo with tumor-bearing
mice by single photon emission computed tomography (SPECT), and compared with the DOTA-N-hydroxysuccinimide
methodology.
Such radiolabeled trastuzumab did not show significant difference in stability when compared with lysine-conjugated
antibodies (>95%, 3 days postinjection). They also maintain their immunoreactivity (>90% immunoreactive fraction) and
affinity both in vitro (Kd ~ 4–5 nM) and in vivo (12.9 %ID/g in MDA-MB-361 tumors).
Our results confirm that the formation of a hydrolytically stable oxime linker between a chelating agent and the oxidized
N-glycan residues, away from the variable antigen-binding sites, is an attractive approach for site-specific conjugation of
antibodies and accessing radiopharmaceuticals and immunotherapeutics.
Keywords: nuclear imaging; HER2; trastuzumab; SPECT; In-111; oxime; antibody
amino groups such as lysine residues, which are distributed over
the whole antibody and may thus be present at or near the
variable antigen-binding domains. Hence, covalent modification
near the antigen-binding site may affect the antigen affinity and
specificity. In contrast, antibodies, including the recombinantly pro-
duced proteins, are N-glycosylated on the fragment crystallizable
Introduction
Biotechnology-based target drugs are finding increasing utility in
a number of clinical applications. In the last decade, monoclonal
antibody used as therapeutic agents represented 20% of
the newly developed target drugs.1 Furthermore, as diagnosis
and treatment are becoming increasingly linked, therapeutic
monoclonal antibodies are also have been developed as
region of antibodies. The N-glycans attached to these sites are pre-
dominantly core-a1, 6-fucosylated, and eventually galactosylated.4,5
molecular diagnostic tools. In both cases, immunoconjugates
These sugar residues, present distally from the antigen-binding sites,
are essential: monoclonal antibodies conjugated to cytotoxic
are ideal anchor points for covalent binding of bifunctional chelating
agents.6–8 It is worth noting that some antibodies may also be
drugs, toxins, or radionuclides tend actually to show some
glycosylated at the variable domain, for example, cetuximab.4
increased therapeutic potency, and antibodies radiolabeled with
short-lived isotopes can alternatively provide critical molecular
The functionalization of antibodies through their carbohydrate
information when used with single photon emission computed
groups is typically carried out by mild oxidation of these residues
tomography (SPECT) or positron emission tomography (PET)
with periodates or enzymes to yield aldehydes that present unique
imaging techniques.2
Most methods used for radiolabeling antibodies rely on
bifunctional chelators that are covalently conjugated to the
protein backbone and subsequently form a stable chelate with a
radioactive metal isotope (e.g., In-111, Zr-89, Cu-64, Y-90, Lu-177,
and Re-188). Alternatively, antibodies can be directly radiolabeled
by iodination of available tyrosine residues (e.g., I-123, I-124, and
I-131).3 Bifunctional chelators are typically attached to terminal
Singapore Bioimaging Consortium (A*STAR), Helios, 02-02, 11 Biopolis Way, 138667
Singapore
*Correspondence to: Romain Bejot, Singapore Bioimaging Consortium (A*STAR),
Helios, 02–02, 11 Biopolis Way, 138667 Singapore.
E-mail: romain_bejot@sbic.a-star.edu.sg
J. Label Compd. Radiopharm 2012
Copyright © 2012 John Wiley & Sons, Ltd.

R. Bejot et al.
functional groups on such molecules. These aldehyde groups can tyrosine kinase found in about 20% of cases of breast cancer.
selectively react with hydrazine-containing or amine-containing The AOD-conjugated trastuzumab is radiolabeled with the
substrates for antibody modification.8 After the antibody conjugates radioactive trivalent metal cation—111In(III)—and evaluated
have been formed, for example via the aldehyde-functionalized in vitro (immunoreactivity and binding affinity against HER2) and
antibody and hydrazine, they can then be stabilized with a suitable in vivo. For comparison, trastuzumab was also functionalized with
reducing agent such as sodium cyanoborohydride.
DOTA by direct coupling to primary amines present on the
Various chelator for conjugation with carbohydrate residues antibody backbone.13–15 In-111 is actually a single g-photon
have been previously investigated, such as the hydrazide-linked emitter (172,247 keV, t1/2 = 2.83 days) that can be used with SPECT
Tc-99 m chelator CYT-395 (Cytogen Corporation, Princeton, NJ, and is compatible with the pharmacokinetics of antibodies in vivo:
USA): Figure 17; or the amine-terminating tetraazacyclododeca- optimal tumor-to-background ratios are typically achieved in
netetraacetic acid (DOTA) and diethylenetriaminepentaacetic 2–4 days. In-111 is also commercially available worldwide and
acid (DTPA) chelators,8,9 such as pendetide, commercially can be replaced in the DOTA complex by alternative radioactive
used as 111In-capromab pendetide (ProstascintW, Cytogen isotopes, such as Cu-64 (PET) or Lu-177 (therapy).
Corporation, Princeton, NJ, USA) or 111In-satumomab pendetide
(OncoScint CR103, Cytogen Corporation, Princeton, NJ, USA).
More recently, there have been reports on the use of O-alkyl
Experimental procedures
hydroxylamines to form stable oxime ethers with oxidized Chemistry
carbohydrate residues of antibodies, hence avoiding the reduc-
All chemicals obtained commercially were purchased from Sigma-Aldrich
tion step.10,11 This oxime linker is actually more hydrolytically
and of analytical grade and used without further purification, unless other-
stable than the isostructural hydrazone.12 Aminooxy-tethered
wise stated. Lucifer Yellow CH was purchased from Biotium (Hayward, CA,
chelates (DOTA and DTPA complexes) have thus been developed
USA). Ammonium acetate and phosphate-buffered saline (PBS) solutions
by PerkinElmer for the introduction of a lanthanide(III)-based
magnetic resonance imaging contrast agent to bioactive mole-
were purified with Chelex 100 (Bio-Rad Laboratories, Hercules, CA, USA)
to remove metal ions. Bovine serum albumin (BSA, Sigma-Aldrich) was fatty
cules.10 A specially designed oxidation-dependent and carbonyl- acid free and low endotoxin grade. DOTA-N-hydroxysuccinimide (NHS)
and penta-t-butyl ((S)-2-(4-aminobenzyl)-1,4,7,10-tetraazacyclododecane-
N,N’,N”,N”’-tetraacetic acid were purchased from Macrocyclics (Dallas, TX, USA).
Ultrafiltration devices were purchased from Millipore Corporation
(Bedford, MA, USA). 111InCl3 was provided in a 0.05 M HCl solution by
PerkinElmer (Boston, MA, USA).
specific mass tag based on an aminooxi-functionalized DOTA
(AOD) has also been developed to covalently tag the residues of
a protein oxidized to aldehyde or keto end products. Such AOD-
tagged peptides can be proteolysed and analyzed by Fourier-
transform ion cyclotron resonance mass spectrometry, allowing
for relative quantitation based on isotopic ratios.11
The previously reported AOD mass tag can also be used to
functionalize antibodies through their carbohydrate residues,
and as a chelator for radioactive metals. Such radiolabeled
Clinical grade trastuzumab (HerceptinW) was purchased from Hoffmann-
La Roche (Basel, Switzerland) and reconstituted to 21 mg/mL in bacterio-
static water for injection, according to the manufacturer’s directions. The
trastuzumab solution was further purified from excipients (histidine, polysor-
bate, and trehalose) by ultrafiltration (Centriprep YM-30, Millipore Corpora-
antibodies, to be used for radioimmunotherapy or molecular tion, 30 kDa cut-off) with PBS and/or acetate buffer.
Thin-layer chromatography was performed on silica gel instant thin-
diagnostic (SPECT, PET), may be obtained with unimpeded
antigen-binding characteristics because the sites of attachment
on oligosaccharides of antibodies are specific and distal to the
antibody-combining site.
Here, we show that AOD can be used as a bifunctional chelator
to access site-specific radiolabeled monoclonal antibodies. AOD
is synthesized and conjugated to the oxidized trastuzumab
(HerceptinW, Roche, Basel, Switzerland). Trastuzumab is a huma-
nized IgG1 monoclonal antibody that targets the ErbB2/human
layer chromatography (ITLC-SG, Biodex Medical Systems, Shirley, NY, USA,
150–771) and analyzed using a radio-thin-layer chromatography strip
scanner fitted with a sodium iodide (NaI)/photomultiplier tube (PMT) detec-
tor (Bioscan Inc., Washington, DC, USA). Size-exclusion chromatography
(SEC) was performed on Superdex 200 10/300 GL column (GE Healthcare
Bio-Sciences AB, Uppsala, Sweden), eluted with 0.5mL/min phosphate buf-
fer (50 mM sodium phosphate, 100 mM NaCl, 1 mM NaN3, pH 7.0) using a
Prominence liquid chromatography system equipped with a diode array
detector (Shimadzu, Kyoto, Japan) and a NaI/PMT-radiodetector (Gamma-
epidermal growth factor-2 (HER2) transmembrane receptor RAM, IN/US Systems Inc., FL, USA).
Figure 1. Structure of various bifunctional chelators for carbohydrate modification. DOTA, tetraazacyclododecanetetraacetic acid; DTPA, diethylenetriaminepentaacetic acid.
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012

R. Bejot et al.
(20 mM HIO4, 0.15 M NaCl, 10 mM NaOAc, pH 4.5) at 4 ꢀC. The solution
was incubated for 1 h at 4 ꢀC in the dark. The reaction was stopped by
adding 50 mL ethylene glycol and incubated for 5 min at 4 ꢀC.16 Purifica-
tion of the oxidized antibody was performed by ultrafiltration (Centriprep,
30 kDa) with acetate buffer (0.15 M NaCl, 10 mM NaOAc, pH 4.5,
4 ꢁ 15 mL) at 1500 ꢁg. The volume was adjusted to 5 mL, and 35 mg
AOD (60 mmol) in 1 mL water was added. The solution was incubated for
2 h at room temperature. Purification of the functionalized antibody was
performed by ultrafiltration (Centriprep, 30 kDa) with metal-free PBS
(4 ꢁ 15 mL) and 0.25 M metal-free ammonium acetate (2 ꢁ 15 mL) at
1500 ꢁg. The protein concentration was measured at 280 nm and
adjusted to a final concentration of 4.0 mg/mL with 0.25 M metal-free
ammonium acetate. The solution of DOTA-oxime-trastuzumab was
aliquoted into 0.5 mL portions and stored at ꢂ80 ꢀC.
Synthesis of aminooxy-functionalized DOTA (AOD)
((S)-2-(4-(2-aminooxy)-acetamido)-benzyl)-1,4,7,10-tetraazacyclododecane-
N,N’,N”,N”’-tetraacetic acid (AOD) was synthesized according to the
procedure reported by S. Lee et al.: Scheme 1.11 A 260mg of N-Boc-
aminooxyacetic acid (1.35mmol) was dissolved in 0.5mL anhydrous
N,N-dimethylformamide and further diluted with 4.5mL anhydrous dichlor-
omethane. The solution was cooled to 0 ꢀC, and 104 mL N,N’-diisopropylcar-
bodiimide (0.66 mmol) was added. The solution was stirred for 30 min at
0 ꢀC prior to the evaporation of dichloromethane under vacuum. To the
residue was added a solution of 100 mg penta-t-butyl ((S)-2-(4-aminoben-
zyl)-1,4,7,10-tetraazacyclododecane-N,N’,N”,N”’-tetraacetic acid (0.14mmol)
and 50mL diisopropylethylamine (0.3mmol) in 2 mLN,N-dimethylforma-
mide. The solution was stirred for 2 h at room temperature, before the addi-
tion of 20mL 1% aqueous triethylamine. The aqueous phase was extracted
with ethyl acetate (6ꢁ 5mL). The combined organic extracts were dried
with MgSO4 and concentrated under vacuum. The crude product as a yel-
low oil was stirred overnight at room temperature in 2mL trifluoroacetic
acid and further concentrated under vacuum. The residue was triturated
in diethyl ether and filtered to give 74 mg AOD as a white solid (91% yield,
582.6 g/mol). 1 H NMR spectrum (400 MHz, dimethyl sulfoxide,d6) was
consistent with the literature11: d (ppm) 1.0–4.0 (m) 4.13 (s), 7.19–7.22 (m)
7.56–7.62 (m), 9.66 (s).
Direct DOTA-conjugation of antibody
The anti-HER2 monoclonal antibody trastuzumab was modified with
DOTA by direct coupling of one of the four DOTA carboxylate groups to
primary amines present in the antibody structure. A 1 mL of reconstituted
trastuzumab solution (21 mg/mL) was purified from other excipients by
ultrafiltration (Centriprep, 30 kDa) with PBS (4 ꢁ 15 mL) at 1500 ꢁg. The
protein concentration was measured at 280 nm (e =1.4(mg/mL)ꢂ1.cmꢂ1):
3.45 mg/mL. DOTA-NHS ester of 8.3 mg (10 mmol) was dissolved in 1 mL
PBS. The DOTA-NHS ester solution of 114 mL (10 mM) were added to 4.9 mL
trastuzumab (3.45 mg/mL, 114 nmol) at 4 ꢀC, and the reaction mixture
incubated overnight at 4 ꢀC. Purification of the functionalized antibody was
performed by ultrafiltration (Centriprep, 30 kDa) with metal-free PBS
(2 ꢁ 15 mL) and 0.25 M metal-free ammonium acetate (2 ꢁ 15 mL) at
1500 ꢁg. The protein concentration was measured at 280 nm and adjusted
to a final concentration of 4.0 mg/mL with 0.25 M metal-free ammonium
acetate. The solution of DOTA-amide-trastuzumab was aliquoted into
0.5 mL portions and stored at ꢂ80 ꢀC.
Oxidation conditions
The oxidation conditions were investigated at various pH (4.0, 4.5, 5.0,
and 5.5) and temperatures (4, 20, and 37 ꢀC), with two periodate concen-
trations (2 and 10 mM) and for a number of durations (15, 30, 60, and
120 min). The extent of oxidation was evaluated by conjugation with
the carbohydrazide (CH)-containing fluorescent dye, Lucifer Yellow CH.
The reconstituted trastuzumab solution (1 mL, 21 mg) was purified
from excipients by ultrafiltration (Centriprep YM-30, 30 kDa cut-off) with
PBS (2 ꢁ 15 mL) and acetate buffer (0.15 M NaCl, 10 mM NaOAc,
2 ꢁ 15 mL) at 1500 ꢁg. The protein concentration was measured at
280 nm (e = 1.4 (mg/mL)ꢂ1.cmꢂ1) and adjusted to a final concentration
of 1 mg/mL with acetate buffer (0.15 M NaCl, 10 mM NaOAc). Periodic acid
was dissolved in acetate buffer, and the pH adjusted. To the antibody
solution (200 mL, 1 mg/mL) was added the periodate solution (200 mL),
and the mixture was incubated in the dark. The reaction was stopped
by adding 50 mL ethylene glycol and incubated for 5 min. Purification of
the oxidized antibody was performed by ultrafiltration (Amicon Ultra-4,
30 kDa cut-off) with acetate buffer (2 ꢁ 4 mL), PBS pH 6.0 (3 ꢁ 4 mL) at
4000 ꢁg and used immediately after. To the solution of oxidized antibody
(200 mL) was added Lucifer Yellow CH dipotassium salt (50 mL, 8 mg/mL in
PBS, pH 6.0), and the solution was incubated for 2 h at room temperature.
Purification of the Lucifer Yellow-conjugated antibody was performed by
ultrafiltration (Amicon Ultra-4, 30 kDa) with PBS pH 7.4 (5 ꢁ 4 mL) at
4000 ꢁg and final filtration through a 0.22 mm filter. The amount of
recovered antibody and the antibody concentration was determined with a
bicinchoninic acid protein assay (Thermo Scientific, Waltham, MA, USA). The
number of oxidation sites were calculated from the extent of conjugation with
Lucifer Yellow, determined by fluorescence (lex = 425 nm, lem = 528 nm).
DOTA-functionalized antibodies characterization
The DOTA-oxime-trastuzumab and DOTA-amide-trastuzumab conjugates
were assayed for the number of DOTA moiety per antibody molecule.
The conjugates (50 mL, 4.0 mg/mL) were mixed with 50 mL of nonradioac-
tive indium chloride (200–500 mM in 0.25 M ammonium acetate, pH 7)
spiked with 111InCl3 (0.66 MBq/mL) and incubated overnight at 37 ꢀC.
An excess of 50 mL aqueous DTPA (10 mM, pH 7) was added and
incubated at 37 ꢀC for 6 h to chelate any unbound indium. The mixtures
were analyzed by ITLC-SG developed in aqueous DTPA (10 mM, pH 5.0).
The proportion of 111In-labeled trastuzumab (Rf = 0.0) or 111In-DTPA
(Rf = 0.5–1.0) was measured using an automatic gamma counter (2470
Wizard2, Perkin Elmer, Waltham, MA, USA). The proportion of 111In-labeled
trastuzumab was multiplied by the molar ratio of InCl3/antibody to
calculate the DOTA-conjugation level.
Radiochemistry
111In-Radiolabeling
111InCl3 (1850 MBq/mL in 0.05 M HCl, PerkinElmer, Boston, MA, USA)
was diluted 1:2 with 0.25 M metal-free ammonium acetate (pH 7), and
50–100 MBq were added to the DOTA-oxime-trastuzumab (25 mL,
4.0 mg/mL) and DOTA-amide-trastuzumab conjugates. The reaction
mixture was incubated at 37 ꢀC for 45 min, then 50 mL aqueous DTPA
Oxidation and DOTA-conjugation of antibody
Chemical oxidation of trastuzumab oligosaccharides was performed by
mixing 7 mL purified trastuzumab (4.2 mg/mL, 200 nmol) in acetate buffer
(0.15 M NaCl, 10 mM NaOAc, pH 4.5) and 7 mL periodic acid solution
Scheme 1. Aminooxy-DOTA (AOD) synthesis. DOTA, tetraazacyclododecanetetraacetic acid; DMF, N,N-dimethylformamide; DCM, dichloromethane; TFA, trifluoroacetic acid;
DIPEA, diisopropylethylamine; DIPCI, N,N0-diisopropylcarbodiimide; AOD, aminooxy-functionalized DOTA.
J. Label Compd. Radiopharm 2012
Copyright © 2012 John Wiley & Sons, Ltd.
www.jlcr.org

R. Bejot et al.
(10 mM, pH 7) was added, and the mixture was incubating for another radioactivity was plotted versus the radioligand concentration to obtain
5 min. The level of In-111 incorporation was determined by ITLC-SG the saturation receptor-binding curve. The equilibrium dissociation
developed in aqueous DTPA (10 mM, pH 5.0). Purification of the radiola- constant Kd and the density of receptors Bmax were obtained by nonlinear
beled antibodies was performed by ultrafiltration (Amicon Ultra-0.5, regression analysis of the specific binding [total–nonspecific binding] for
30 kDa cut-off, pre-coated with 0.5% BSA in PBS) with PBS (3 ꢁ 400 mL) a single binding site (Prism 5.0c, GraphPad Software).
at 14,000 ꢁg. Purified 111In-labeled antibodies were recovered in 200 mL
PBS (0.2% BSA) by centrifugation at 1000 ꢁg and diluted with 0.9% sal-
ine. The specific activity was in the range of 500–1000 MBq/mg antibody In vivo biodistribution and targeting of
111In-DOTA-conjugated trastuzumab
at the end of synthesis.
Quality control of 111In-labeled trastuzumab
Animals, housing, and environment
Outbred female nude CD-1W mice were purchased from Charles River
(Wilmington, MA, USA). Animal experiments were performed in accor-
dance with the Institutional Animal Care and Use Committee (IACUC)
guidelines under ethics number IACUC 090437.
After purification, the radiochemical purity was determined by ITLC-SG
and SEC. For thin-layer chromatography, 5 mL 111In-labeled trastuzumab
were mixed with 10 mL aqueous DTPA (10 mM, pH 7.0) for 10 min at room
temperature and further analyzed by ITLC-SG developed in aqueous
DTPA (10 mM, pH 5.0). SEC was performed by injecting 20 mL of 1:40
diluted 111In-labeled trastuzumab on Superdex 200 10/300 GL column.
Breast cancer cells
Radioactivity was measured by a NaI/PMT flow detector.
Human breast cancer cells MDA-MB-231 (human ductal carcinoma) and
In vitro assays
MDA-MB-361 (human ductal carcinoma) were purchased from the ATCC.
Cells were cultured in Dulbecco’s modified Eagle medium (Invitrogen,
Grand Island, NY, USA), supplemented with 10% fetal bovine serum,
100 U/mL penicillin, and 100 mg/mL streptomycin, at 37 ꢀC in a humidified
atmosphere, with 5% of CO2.
Human epidermal growth factor receptor-2 overexpressing cells, SK-BR-3
(human ductal breast carcinoma), were purchased from the American
Type Culture Collection (ATCC, Manassas, VA, USA). Cells were cultured
in RPMI 1640, supplemented with 10% fetal bovine serum, 100 U/mL
penicillin, and 100 mg/mL streptomycin, at 37 ꢀC in a humidified atmo-
sphere with 5% of CO2.
In vivo imaging studies
Preliminary studies were carried out in naive CD1 nude mice to determine
the acute distribution of 111In-DOTA-oxime-trastuzumab. CD1 nude mice
(n = 2) were injected with a solution of 111In-DOTA-oxime-trastuzumab
(~26 MBq in 0.2 mL) via the lateral tail vein and the animals imaged at
10, 90, and 180 min, and 24 and 48 h postinjection (p.i.), using a
NanoSPECT/computed tomography (CT) camera (Bioscan Inc., Washington,
DC, USA).
Immunoreactive fraction
The immunoreactive fraction (IRF) of both 111In-radiolabeled antibodies
was determined by extrapolating the cell-binding at infinite antigen
excess.17,18 SK-BR-3 cells were harvested by centrifugation and resus-
pended at 8 ꢁ 106 cells/mL in Dulbecco PBS. Increasing cell concentra-
tions (0.125–4 ꢁ 106 cells in 0.5 mL PBS, 0.5% BSA) were prepared by
two parallel serial 1:2 dilutions in 1.5 mL polypropylene centrifuge tubes,
pre-coated with 1% BSA. A fixed amount of 111In-radiolabeled antibodies
(50 mL, 5.5–6.0 ng, about 2 kBq) in PBS (1% BSA) was added to the cell sus-
pension (duplicates). After incubation for 1 h at 4 ꢀC, total radioactivity in
each tube was measured using an automatic gamma counter, and cell
suspensions were centrifuged (8 min, 200 ꢁg) and washed twice with
PBS (0.2% BSA). Relative binding was calculated as the ratio of cell-bound
radioactivity (radioactivity of pellet obtained after the last centrifugation
step) to total radioactivity. Nonspecific binding was determined with an
excess of unlabeled trastuzumab (1.3 mM). Bound/total radioactivity was
plotted versus the cell concentration. The IRF was obtained by nonlinear
regression analysis of the binding ratio (Prism 5.0c, GraphPad Software,
La Jolla, San Diego, CA, USA).
Studies were subsequently carried out in female CD1 nude mice
bearing MDA-MB-231 and MDA-MB-361 tumors to determine the longi-
tudinal distribution of both 111In-radiolabeled antibodies. Mice (n = 10)
were inoculated with the HER2 expressing MDA-MB-361 cells laterally
into the upper left thigh and with low-HER2 expressing MDA-MB-231
cells laterally into the upper right thigh. In both cases, the cells were
implanted subcutaneously at a volume of 100 mL (2 ꢁ 107 cells/mL),
mixed with 100 mL of Matrigel (BD Biosciences, Franklin Lakes, NJ, USA).
To ensure growth of the estrogen-receptor-positive MDA-MB-361 cells,
all the animals were implanted subcutaneously with 0.72 mg, 60-day
sustained release 17b-estradiol pellets (Innovative Research of America,
Sarasota, FL, USA) 24 h prior to tumor inoculation. The tumors were
allowed to grow for approximately 30 days.
Once the tumors had reached approximately 300 mm3, the tumors
were measured with calipers, and an F-test was performed (no significant
Saturation receptor-binding assay
The dissociation constants (Kd) for both preparations of 111In-radiolabeled variance). Subsequently, the animals were injected with the radiolabeled
antibodies were measured by radioligand saturation receptor-binding antibodies (10–30 MBq in 0.2 mL via the lateral tail vein) and imaged at 24,
assays. The initial 111In-radiolabeled antibody concentrations were 48, 72, and 144 h p.i., using a NanoSPECT/CT camera. At the correct time
measured using the Easy-Titer Human IgG (H + L) assay kit (Pierce/Thermo point, the animals were anesthetized using inhalational isoflurane at 2%
Fischer Scientific Inc., Rockford, IL, USA), and various concentration were alveolar concentration and placed onto the imaging bed: CT images
prepared by 1:2 serial dilution (1.56–100 nM) in PBS (0.5% BSA), using (40 kV, 500 mA; 4 ꢁ 4 binning, 200 mm resolution) were acquired to obtain
1.5 mL polypropylene tubes pre-coated with 1% BSA. HER2 overexpres- anatomical information, and SPECT images were acquired consecutively
sing SK-BR-3 cells were harvested by centrifugation and resuspended at for 40 min. Images were reconstructed using the image reconstruction,
2 ꢁ 106 cells/mL in Dulbecco PBS. For nonspecific binding measurement, visualization, and analysis program supplied by the manufacturer.
half of the cell suspension was supplemented with 200 nM nonlabeled
Single photon emission computed tomography and CT data were
Herceptin. In a 96-well plate, pre-coated with 1% BSA, were incubated analyzed by using Amide software (Sourceforge 10.1, http://amide.
200,000 SK-BR-3 cells (100 mL, in the presence [nonspecific binding] or sourceforge.net). The SPECT and CT images were co-registered to
absence [total binding] of an excess unlabeled trastuzumab) with increas-
confirm anatomical location of the tumors. Uptake of radioactivity in the
ing concentration of radiolabeled 111In-DOTA-trastuzumab (0.78–50 nM; tumor was determined by placement of a Region of Interest (ROI) around
100 mL, triplicates). After incubation for 1 h at 4 ꢀC, cell-bound and free the tumor border delineated using the CT images. An ROI was also placed
radioactivity was separated by filtration (Cell harvester, GF/C filter, around a region of the muscle of the lower left hind limb, as well as a
PerkinElmer, Waltham, MA, USA). The cell suspensions were washed region of the liver and hind limb knee joint to provide reference tissue
six times with PBS (0.2% BSA). GF/C filters were separated, and the values. All imaging data are expressed as percentage of the injected dose
radioactivity was measured using an automatic gamma counter. Bound per gram of tissue (%ID/g).
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012

R. Bejot et al.
Figure 2. Size-exclusion chromatograms of 111In-DOTA-conjugated trastuzumab (left). Size-exclusion chromatogram analysis of radioactive metabolites in plasma 3 days
postinjection (right). DOTA, tetraazacyclododecanetetraacetic acid; CPM, counts per minute.
Radiolabeled antibodies metabolism
in excellent radiochemical purity >99% (assessed by both thin-
layer chromatography and SEC) and with a high specific activity
of 500–1000 MBq/mg antibody at the end of synthesis. Radio-
chemical identification was performed by SEC and revealed that
the main radioactive peak for both radiolabeled antibodies
Blood samples (~50 mL) were collected from the tail vein of mice at 63–65 h
p.i., into 0.5 mL heparinized polypropylene centrifuge tubes. Blood samples
were centrifuged at 20,000 ꢁg to collect serum; blood cells were washed
with 50 mL PBS (0.1% BSA) and centrifuged at 20,000 ꢁg. The combined
supernatant fractions were fractionated on Superdex 200 10/300 GL column corresponds to a protein of 150–180 kDa (tr = 24.5 min) as
(GE Healthcare Life Sciences), eluted with 0.5 mL/min phosphate buffer
and analyzed with both an online NaI/PMT radioactivity detector (IN/US)
and an offline automatic gamma counter (2470 Wizard2, Perkin Elmer).
expected (Figure 2, left). As it turns out, SEC also indicated
the formation of 5 ꢃ 1% impurity product (about 350–400 kDa,
tr = 21.1 min) along with 111In-DOTA-oxime-trastuzumab: it was
supposed that a dimerization occurred during the oxidation or
DOTA-conjugation of the antibody.
Results and discussion
DOTA-functionalization and 111In-radiolabeling of
trastuzumab
111In-labeled conjugates immunoreactivity and
binding affinity
The aminooxy-functionalized chelator AOD was readily obtained in
91% yield from the protected (S)-2-(4-aminobenzyl)-DOTA tetra-
t-butyl ester and subsequently conjugated to trastuzumab.
Trastuzumab is a recombinant humanized anti-HER2 monoclonal
antibody that contains carbohydrate residues (GnGnF6) attached
to the CH2 domain of the heavy chains.4,19,20 Mild oxidation of
the antibody was investigated using the methodology reported
by D. Hage16: various parameters, such as periodate concentration,
reaction time, pH, and temperature, were assayed, and the extent
of antibody oxidation was determined by conjugation of the
oxidized antibody with the carbohydrazide-containing fluorescent
dye Lucifer Yellow CH (data not shown). With the use of optimized
conditions (~3–6 aldehydes/antibody), the antibody was oxidized
for 1 h at 4 ꢀC with periodic acid (10 mM) in acetate buffer (pH
4.5) and further functionalized with DOTA in the presence of a
300-fold molar excess of AOD to yield the DOTA-oxime-
trastuzumab conjugate. Trastuzumab was also functionalized with
DOTA through its primary amine residues, by direct conjugation
with 10-fold molar excess of the activated N-hydroxysuccinimidyl
DOTA ester to yield the DOTA-amide-trastuzumab conjugate.
The DOTA-oxime-trastuzumab and DOTA-amide-trastuzumab
conjugates were assayed for the number of DOTA moiety per
antibody molecule. The molar ratios of DOTA/antibody were
determined to be 5.1 ꢃ 0.7 and 5.6 ꢃ 0.9, respectively.
The IRF of 111In-DOTA-oxime-trastuzumab and 111In-DOTA-amide-
trastuzumab were determined to be 0.99 ꢃ 0.03 and 0.93 ꢃ 0.08,
respectively (meanꢃ standard deviation, Figure 3). The nonspe-
cific binding, determined by adding an excess of unlabeled
trastuzumab with 1.6ꢁ 106 cells, was less than 0.3% and was not
taken into account. The immunoreactive fractions of both radiola-
beled antibodies were highly preserved and were not statistically
different. The conjugation of DOTA to the carbohydrate residues
of trastuzumab did not significantly improve the immunoreactivity
of the radiolabeled antibody. It did not prove detrimental either,
despite the presence of the radiolabeled protein dimer that may
not bind HER2 to the same extent.
The antibodies were further radiolabeled for in vitro evalua-
tion and nuclear imaging. 111In-Radiolabeling of DOTA-oxime-
trastuzumab and DOTA-amide-trastuzumab conjugates was
performed with [111In]indium(III) chloride, at 37 ꢀC for 45 min:
incorporation of 111InCl3 was 93% and 94%, respectively
(assessed by thin-layer chromatography). Purification was
Figure 3. Immunoreactive fractions. Each point represents meanꢃ standard
performed by ultrafiltration, and both conjugates were obtained deviation. DOTA, tetraazacyclododecanetetraacetic acid; IRF, immunoreactive fraction.
J. Label Compd. Radiopharm 2012
Copyright © 2012 John Wiley & Sons, Ltd.
www.jlcr.org

R. Bejot et al.
Both 111In-radiolabeled antibodies demonstrated saturable In vivo biodistribution and targeting of 111In-DOTA-
binding to HER2-overexpressing SK-BR-3 cells, which was conjugated trastuzumab
displaceable by an excess of unlabeled trastuzumab. The disso-
ciation constants (Kd) were determined using an equilibrium The uptake of both 111In-labeled antibodies in MDA-MB-361 and
saturation-binding assay and was 4.9 ꢃ 1.0 nM and 4.5 ꢃ 0.9 nM, MDA-MB-231 tumor-bearing mice is shown in Figures 5 and 6.
respectively (Figure 4). As expected, both methods provide The percentage-injected dose per gram of tissue (%ID/g) was mea-
radiolabeled antibodies with similar binding affinity and within sured using ROI analysis in the high (MDAMB361) and low
the range reported in the literature for radiolabeled trastuzumab: (MDAMB231) HER2-expressing tumors; muscle and liver uptake
1.7–8.2 nM.21–23 The maximal number of binding sites Bmax was were measured as reference regions. Over the time, course studied
determined to be 410,000–500,000 receptors/cell.
the reference tissues (muscle and liver) displayed decreasing
Figure 4. Saturation receptor-binding assay. Each point represents mean ꢃ standard deviation. DOTA, tetraazacyclododecanetetraacetic acid.
Figure 5. Time-activity curves of Region of Interests in naive and tumor-bearing mice for 111In-DOTA-oxime-trastuzumab (left) and 111In-DOTA-amino-trastuzumab (right).
Each point represents mean ꢃ standard deviation. DOTA, tetraazacyclododecanetetraacetic acid.
Figure 6. Transaxial single photon emission computed tomography images of 111In-DOTA-oxime-trastuzumab (left) and 111In-DOTA-amino-trastuzumab (right) in tumor-
bearing mice 3 days postinjection. The central areas with radioactivity are the kidneys and flow through to the bladder. DOTA, tetraazacyclododecanetetraacetic acid.
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012

R. Bejot et al.
radioactivity retention for both of the radiolabelled analogs consis- (10 mM, pH 4.5) to an extent that is compatible with nuclear
tent with nonspecific binding and clearance. The tumor uptake of imaging or radioimmunotherapy. Such oxidized antibodies can
the two analogs, however, reached a peak at 72 h p.i. suggesting be readily conjugated to AOD and purified using well-established
specific binding and retention in the tumors. For both the methods, similar to the lysine conjugation method.
111In-DOTA-oxime and 111In-DOTA-amide labeled analogs, the
The AOD methodology allows for the synthesis of hydrolytically
radioactivity retention at 72 h p.i. was higher in the high-HER2- stable oxime-linked and site-specific labeled immunoconjugates,
expressing (MDA-MB-361) xenograft tumors (12.9ꢃ 1.2 %ID/g, which can be used to target antigen-exposed cells both in vitro
p < 0.05, and 17.7 ꢃ 0.7%ID/g, respectively) than in low-HER2- and in vivo. The method was exemplified by the 111In-radiolabeled
expressing (MDA-MB-231) tumors (6.7ꢃ 1.2 %ID/g, p < 0.05, DOTA-oxime antibody conjugate, which demonstrated specific
and 11.8 ꢃ 3.0%ID/g), indicating that the site-specific labeled anti- binding with high-HER2-expressing cells in vitro and in high-
body (111In-DOTA-oxime-trastuzumab) and 111In-DOTA-amide- HER2-expressing tumors in vivo. The results did not show a
trastuzumab differentially bind to the tumors on the basis of significant advantage for site-specific labeling of trastuzumab,
their HER2-expression levels in vivo.
but the method could be readily applied to various glycoproteins
Overall, the retention of both 111In-labeled trastuzumab of interest, especially antibodies whose immunoreactive fraction
analogs measured in the high-HER2-expressing xenografts were is impeded by lysine modification. Furthermore, engineered
comparable with that reported in the literature;23,24 however, antibody fragments such as minibodies, which exhibit a rapid
111In-DOTA-oxime-trastuzumab displayed significantly lower blood clearance, could be site-specific labeled with AOD: antibody
retention than 111In-DOTA-amide-trastuzumab.
fragments with a glycosylation site (e.g., asparagine) in the variable
In agreement with previously published studies, both the domain can be glycosylated and further radiolabeled via conjuga-
labeled analogs displayed low retention in the muscle reference tion to the oxidized carbohydrate. As reported by S. V. Govindan
tissue (<1.55 %ID/g from 24 h p.i. onward) and high liver et al., glycosylation at the k light chain variable domain (VK) does
excretion (111In-DOTA-oxime-trastuzumab; 9.7 ꢃ 0.7 %ID/g and not interfere with immunoreactivity,9 hence may be superior to
111In-DOTA-amide-trastuzumab; 12.0 ꢃ 1.1 %ID/g at 24–72 h p.i). the lysine conjugation methods.
Interestingly, hotspots of activity were observed in the synovium
for both antibodies suggesting a specific accumulation of the
Conflicts of Interest
radiolabelled antibodies in the rat synovium (24.7ꢃ 4.6 %ID/g
and 27.2 ꢃ 8.6%ID/g, respectively, at 72 h p.i). High ErbB2 expres-
The authors declare that they have no conflict of interest.
sion has been shown in human synovium,25 which may account
for the high accumulation of the 111In-DOTA-oxime-trastuzumab
and 111In-DOTA-amide-trastuzumab at these sites. Overall, the
References
pharmacokinetic analysis of 111In-DOTA-oxime-trastuzumab in
[1] M. Rask-Andersen, M. S. Almén, H. B. Schiöth. Trends in the
mice obtained by SPECT imaging demonstrated the in vivo
exploitation of novel drug targets. Nat. Rev. Drug Discov. 2011, 10,
579–90.
targeting potential of site specific 111In-labeled trastuzumab.
[2] A. M. Wu, P. D. Senter. Arming antibodies: prospects and challenges
Stability of 111In-labeled antibodies
for immunoconjugates. Nat. Biotechnol. 2005, 23, 1137–46.
[3] R. M. Reilly, The Radiochemistry of Monoclonal Antibodies and Peptides
Monoclonal Antibody and Peptide-targeted Radiotherapy of Cancer,
The stability of the radiolabeled antibodies in vivo was investigated
Wiley, Hoboken, NJ, 2010.
by SEC analysis of mouse plasma, 3 days postinjection (Figure 2,
[4] J. Stadlmann, M. Pabst, D. Kolarich, R. Kunert, F. Altmann. Analysis of
right). Both radiolabelled antibodies eluted mostly as a single peak
of about 150–180 kDa corresponding to the parent labeled
antibody and showed no significant sign of aggregation or
transchelation to transferrin and only a small fraction of low
molecular weight component (1–4%). The small fraction of free
radioactive metabolites in mice serum, 3 days p.i., indicates that
the free radioactive signal detected in SPECT is essentially due to
the radiolabeled antibodies. However, this is not evidence
that the antibodies have not degraded as demetallation may
release free radioactive 111In(III) that is more prone to elimination
than the radiolabeled glycoprotein. Some of the imaging data
suggests that the 111In-DOTA-amide-trastuzumab may be degrading
after 72 h as a uniform decrease in retention is observed in both
tumors and the synovium from 72–144 h p.i. (Figure 5).
immunoglobulin glycosylation by LC-ESI-MS of glycopeptides and
oligosaccharides. Proteomics 2008, 8, 2858–71.
[5] T. Shantha R. Terminal sugars of Fc glycans influence antibody
effector functions of IgGs. Curr. Opin. Immunol. 2008, 20, 471–8.
[6] J. N. Tinianow, H. S. Gill, A. Ogasawara, J. E. Flores, A. N. Vanderbilt,
E. Luis, et al. Site-specifically 89Zr-labeled monoclonal antibodies
for ImmunoPET. Nucl. Med. Biol. 2010, 37, 289–97.
[7] M. A. Stalteri, S. J. Mather, B. A. Belinka, D. J. Coughlin, V. U. Chengazi,
K. E. Britton. Site-specific conjugation and labelling of prostate
antibody 7E11C5.3 (CYT-351) with technetium-99m. Eur. J. Nucl.
Med. 1997, 24, 651–4.
[8] J. D. Rodwell, V. L. Alvarez, C. Lee, A. D. Lopes, J. W. Goers, H. D. King,
et al. Site-specific covalent modification of monoclonal antibodies:
in vitro and in vivo evaluations. Proc. Natl. Acad. Sci. U.S.A. 1986,
83, 2632–6.
[9] S. V. Govindan, D. M. Goldenberg, G. L. Griffiths, S. O. Leung,
M. J. Losman, H. J. Hansen. Site-specific modifications of light chain
glycosylated antilymphoma (LL2) and anti-carcinoembryonic antigen
(hImmu-14-N) antibody divalent fragments. Cancer Res. 1995, 55,
5721s–5s.
[10] J. Hovinen. Synthesis of aminooxy-functionalized lanthanide(III)
chelates for carbonyl-group conjugation. Chem. Biodivers. 2006,
3, 296–303.
[11] S. Lee, N. L. Young, P. A. Whetstone, S. M. Cheal, W. H. Benner,
C. B. Lebrilla, et al. Method to site-specifically identify and quantitate
carbonyl end products of protein oxidation using oxidation-
dependent element coded affinity tags (O-ECAT) and nanoliquid
chromatography Fourier transform mass spectrometry. J. Proteome
Res. 2006, 5, 539–47.
Conclusion
The site-specific conjugation of chelators to trastuzumab though
the carbohydrate residues on the fragment crystallizable region
can be readily performed using the aminooxy-functionalized
DOTA (AOD). The oxidation conditions of the oligosaccharide
may require substrate-specific optimizations, but it is expected
that at low temperature (4ꢀC) for a maximum of 1 h in the dark,
a N-glycosylated antibody can be oxidized by a periodate solution
J. Label Compd. Radiopharm 2012
Copyright © 2012 John Wiley & Sons, Ltd.
www.jlcr.org

R. Bejot et al.
[12] J. Kalia, R. T. Raines. Hydrolytic stability of hydrazones and oximes.
Angew. Chem. Int. Ed. 2008, 47, 7523–6.
[13] P. Paudyal, B. Paudyal, H. Hanaoka, N. Oriuchi, Y. Iida, H. Yoshioka,
et al. Imaging and biodistribution of Her2/neu expression in non-
small cell lung cancer xenografts with Cu-labeled trastuzumab PET.
Cancer Sci. 2010, 101, 1045–50.
to human IgG. PDB ID: 3D6G, RCSB protein data bank, available from:
http://www.pdb.org [23 July 2012].
[20] E. Boeggeman, B. Ramakrishnan, M. Pasek, M. Manzoni, A. Puri,
K. H. Loomis, et al. Site specific conjugation of fluoroprobes to
the remodeled Fc N-glycans of monoclonal antibodies using mutant
glycosyltransferases: application for cell surface antigen detection.
Bioconjug. Chem. 2009, 20, 1228–36.
[14] S. Rasaneh, H. Rajabi, M. H. Babaei, F. J. Daha, M. Salouti. Radiolabeling
of trastuzumab with 177Lu via DOTA, a new radiopharmaceutical for [21] K. McLarty, B. Cornelissen, D. A. Scollard, S. J. Done, K. Chun, R. M. Reilly.
radioimmunotherapy of breast cancer. Nucl. Med. Biol. 2009, 36, 363–9.
[15] J. Y. Wong, A. Raubitschek, D. Yamauchi, L. E. Williams, A. M. Wu,
P. Yazaki, et al. A pretherapy biodistribution and dosimetry study
of indium-111-radiolabeled trastuzumab in patients with human
Associations between the uptake of 111In-DTPA-trastuzumab, HER2
density and response to trastuzumab (Herceptin) in athymic mice
bearing subcutaneous human tumour xenografts. Eur. J. Nucl. Med.
Mol. Imaging 2009, 36, 81–93.
epidermal growth factor receptor 2-overexpressing breast cancer. [22] Y. Tang, J. Wang, D. A. Scollard, H. Mondal, C. Holloway, H. J. Kahn,
Cancer Biother. Radiopharm. 2010, 25, 387–94.
[16] D. S. Hage. Periodate oxidation of antibodies for site-selective
immobilization in immunoaffinity chromatography. Methods Mol.
Biol. 2000, 147, 69–82.
[17] T. Lindmo, P. Bunn Jr. Determination of the true immunoreactive
fraction of monoclonal antibodies after radiolabeling. Methods
Enzymol. 1986, 121, 678.
et al. Imaging of HER2/neu-positive BT-474 human breast cancer
xenografts in athymic mice using 111In-trastuzumab (Herceptin)
Fab fragments. Nucl. Med. Biol. 2005, 32, 51–8.
[23] D. L. Costantini, C. Chan, Z. Cai, K. A. Vallis, R. M. Reilly. 111In-labeled
trastuzumab (Herceptin) modified with nuclear localization sequences
(NLS): an Auger electron-emitting radiotherapeutic agent for HER2/
neu-amplified breast cancer. J. Nucl. Med. 2007, 48, 1357–68.
[18] S. Konishi, K. Hamacher, S. Vallabhajosula, P. Kothari, D Bastidas, [24] M. N. Lub-de Hooge, J. G. Kosterink, P. J. Perik, H. Nijnuis, L. Tran,
N. Bander, et al. Determination of immunoreactive fraction of
radiolabeled monoclonal antibodies: what is an appropriate
method? Cancer Biother. Radiopharm. 2004, 19, 706–15.
J. Bart, et al. Preclinical characterisation of 111In-DTPA-trastuzumab.
Br. J. Pharmacol. 2004, 143, 99–106.
[25] K. Satoh, S. Kikuchi, M. Sekimata, Y. Kabuyama, M. K. Homma,
Y. Homma. Involvement of ErbB-2 in rheumatoid synovial cell
growth. Arthritis Rheum. 2001, 44, 260–5.
[19] D. Moiani, M. Salvalaglio, C. Cavalotti, A. D. Bujacz, I. Redzynia,
G. D. Bujacz, et al. A protein-a mimetic peptide dendrimer bound
www.jlcr.org
Copyright © 2012 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2012