HEHEHE-Tagged Affibody Molecule May Be Purified by IMAC, Is

Bioconjugate Chem. 2010, 21, 2013–2022

2013

HEHEHE-Tagged Affibody Molecule May Be Purified by IMAC, Is Conveniently Labeled with [99mTc(CO)3]+, and Shows Improved Biodistribution with Reduced Hepatic Radioactivity Accumulation Vladimir Tolmachev,*,†,‡ Camilla Hofstro¨m,§ Jennie Malmberg,† Sara Ahlgren,‡ Seyed Jalal Hosseinimehr,†,| Mattias Sandstro¨m,⊥ Lars Abrahmse´n,# Anna Orlova,† and Torbjo¨rn Gra¨slund§ Division of Biomedical Radiation Sciences, Rudbeck Laboratory, and Division of Nuclear Medicine, Department of Medical Sciences, Uppsala University, Uppsala, Sweden, Department of Molecular Biotechnology, Royal Institute of Technology, Stockholm, Sweden, Department of Radiopharmacy, Faculty of Pharmacy, Mazandaran University of Medical Sciences, Sari, Iran, Hospital Physics, Department of Oncology, Uppsala University Hospital, Uppsala, Sweden, and Affibody AB, Stockholm, Sweden. Received May 20, 2010; Revised Manuscript Received September 23, 2010

Affibody molecules are a class of small (ca. 7 kDa) robust scaffold proteins suitable for radionuclide molecular imaging of therapeutic targets in vivo. A hexahistidine tag at the N-terminus streamlines development of new imaging probes by enabling facile purification using immobilized metal ion affinity chromatography (IMAC), as well as convenient [99mTc(CO)3]+-labeling. However, previous studies in mice have demonstrated that Affibody molecules labeled by this method yield higher liver accumulation of radioactivity, compared to the same tracer lacking the hexahistidine tag and labeled by an alternative method. Two variants of the HER2-binding Affibody molecule ZHER2:342 were made in an attempt to create a tagged tracer that could be purified by immobilized metal affinity chromatography, yet would not result in anomalous hepatic radioactivity accumulation following labeling with [99mTc(CO)3]+. In one construct, the hexahistidine tag was moved to the C-terminus. In the other construct, every second histidine residue in the hexahistidine tag was replaced by the more hydrophilic glutamate, resulting in a HEHEHE-tag. Both variants, denoted ZHER2:342-H6 and (HE)3-ZHER2:342, respectively, could be efficiently purified using IMAC and stably labeled with [99mTc(CO)3]+ and were subsequently compared with the parental H6-ZHER2:342 having an N-terminal hexahistidine tag. All three variants were demonstrated to specifically bind to HER2-expressing cells in vitro. The hepatic accumulation of radioactivity in a murine model was 2-fold lower with [99mTc(CO)3]+ZHER2:342-H6 compared to [99mTc(CO)3]+-H6-ZHER2:342, and more than 10-fold lower with [99mTc(CO)3]+-(HE)3ZHER2:342. These differences translated into appreciably superior tumor-to-liver ratio for [99mTc(CO)3]+(HE)3-ZHER2:342 compared to the alternative conjugates. This information might be useful for development of other scaffold-based molecular imaging probes.

INTRODUCTION Malignant transformation of cells is often associated with overexpression of cell surface proteins, such as different growth factor receptors. The use of antitumor agents with the ability to selectively target such overexpressed proteins increases the specificity of cancer treatment, resulting in improved therapy (1-3). Obviously, only tumors with the relevant alteration on the cell surface respond to the targeted therapy. For this reason, molecular profiling of tumors is necessary to select suitable patients. The commonly used biopsy-based methods are associated with false-negative findings due to inter- and intratumoral heterogeneity of cell surface protein exposure (4, 5). An attractive alternative is in vivo radionuclide molecular imaging (RMI), which is global, minimally invasive, and less sensitive to target expression heterogeneity, and may be used repeatedly to follow the course of the disease (6-8). The precondition of * Corresponding author. Vladimir Tolmachev, Biomedical Radiation Sciences, Rudbeck Laboratory, Uppsala University, S-751 81 Uppsala, Sweden. Phone: +46 18 471 3414, Fax: + 46 18 471 3432, E-mail: [email protected]. † Rudbeck Laboratory, Uppsala University. ‡ Department of Medical Sciences, Uppsala University. § Royal Institute of Technology. | Mazandaran University of Medical Sciences. ⊥ Uppsala University Hospital. # Affibody AB.

successful application of RMI is high sensitivity and specificity of the imaging agent, which is provided by high accumulation in tumors and low uptake in healthy organs and tissues. High tumor uptake and rapid clearance of nonbound radiolabeled scaffold proteins provide high imaging contrast, increasing the sensitivity of imaging. Previous studies with antibody-based imaging agents have demonstrated that imaging contrast is favored by a small size of the targeting probe (9). This prompted the development of protein-based imaging agents smaller than the smallest binding antibody fragment (10). One example is the Affibody molecules, containing only 58 amino acids (approximately 7 kDa), based on a single polypeptide, three-helix bundle domain derived from staphylococcal protein A. They lack backbone cysteines and are capable of rapid folding. Affibody molecules with high affinity and selectivity for different targets have previously been selected by phage display from libraries made by randomizing 13 surface-exposed residues on helices 1 and 2, i.e., the 2 N-terminal helices (11). The successful development of Affibody-based imaging agents is exemplified by ZHER2:342, which binds to human epidermal receptor type 2 (HER2) with high affinity (KD 22 pM) (12). HER2 is a relevant target for RMI since it is overexpressed in a significant fraction of many carcinomas, and since different HER2-targeted therapeutics are available (13). Technetium-99m (T1/2 ) 6.0 h) is widely used in radionuclide diagnostics because of favorable energy of emitted γ-quanta

10.1021/bc1002357  2010 American Chemical Society Published on Web 10/22/2010

2014 Bioconjugate Chem., Vol. 21, No. 11, 2010

Tolmachev et al.

[99mTc(CO)3]+, and both constructs yielded less radioactivity in the liver in a murine model: 2-fold and more than 10-fold less than the parent tracer, respectively. This was reflected in a significant improvement of tumor-to-liver ratio in tumor xenograft-bearing mice.

MATERIALS AND METHODS Figure 1. Sequences of studied proteins.

and low radiation absorbed dose burden to patients, combined with low price and a wide availability of 99Mo/99mTc generators. An attractive approach for obtaining a protein-based imaging agent suitable for labeling with technetium-99m is to produce the tracer protein recombinantly with a hexahistidine-tag. This tag can be used both for efficient purification by immobilized metal ion affinity chromatography (IMAC) and for chelation of 99mTc-tricarbonyl ([99mTc(CO)3]+) (14). This very elegant labeling technique has been applied on a number of targeting proteins (15-18) to provide stable and site-specific labeling, giving uniform conjugates with reproducible chemical and pharmacological proprieties. Importantly, no additional coupling of chelators is required, and the proteins can be labeled directly after purification. This simple procedure is suitable for development of protein-based imaging agents binding new molecular targets, facilitating comparison of alternative variants of the tracer protein. However, evaluation of the dimeric His6(ZHER2:4)2 anti-HER2 Affibody molecule demonstrated that [99mTc(CO)3]+-His6-(ZHER2:4)2 yielded high accumulation of radioactivity in the liver (19). Since the liver is a major metastatic site for many carcinomas, a high liver uptake is an undesirable property of any imaging agent, and hence, correct assessment of this property is very important when comparing alternative variants of a labeled tracer. Previous studies have shown that different chelators and labeling chemistry may influence the excretion pathway and hepatic uptake of Affibody molecules. For example, increasing the hydrophilicity of the chelator reduced the hepatobiliary excretion of 99mTc-radioactivity, when the peptide-based N3S chelators mercaptoacetyl-X-X-X (maXXX, where X is Gly, Ser, Glu, or Lys) were located at the N-terminus (20-23). In contrast, alternative amino acids in the C-terminal chelating sequence X-X-Cys (where X is Val, Asp, Ser, or Glu) yielded a similar level of hepatobiliary excretion of the radioactivity from 99mTc (24, 25). Taken together, these data suggested that the properties of the N-terminal portion of the molecule are more crucial for hepatic uptake and hepatobiliary excretion than the properties of the C-terminus when chelating 99mTc by an N3S chelator. Furthermore, comparison of the same Affibody molecule with or without an N-terminal hexahistidine tag and sitespecifically labeled at the C-terminus with 111In (by maleimidoDOTA) or 99mTc (by cysteine-containing peptide-based N3S chelator) clearly showed that the presence of the hexahistidine tag led to much higher accumulation of radioactivity in the liver (24, 26). Thus, the presence of an N-terminal hexahistidine tag leads to a higher uptake in the liver. The present study was performed to test the hypothesis that the problem with high accumulation of radioactivity in the liver might be solved by either relocating the hexahistidine tag or making it more hydrophilic. The hypothesis was tested by constructing two novel affibody tracers. In the first construct, the tag was moved from the N-terminus to the C-terminus. In the second construct, the hydrophilicity of the N-terminal tag was increased by replacing three of the six histidine residues by glutamic acid residues, resulting in a HEHEHE-tag (Figure 1). Both molecules denoted ZHER2:342-H6 and (HE)3-ZHER2:342, respectively, could be purified by IMAC and labeled with

General. Buffers were prepared using common methods from chemicals supplied by Merck (Darmstadt, Germany). Highquality Milli-Q water (resistance higher than 18 MΩ/cm) was used for preparing solutions. The IsoLink labeling kit was purchased from Covidien. NAP-5 size exclusion columns were from GE Healthcare, Uppsala, Sweden. 99mTc was obtained as pertechnetate from an Ultra-TechneKow generator (Covidien) by elution with sterile 0.9% NaCl. The yield, radiocolloid content, and radiochemical purity of the labeled Affibody constructs were analyzed using 150-771 DARK GREEN, TecControl Chromatography strips from Biodex Medical Systems (New York, USA) as described (20). Cells used during in vitro experiments were detached using trypsin-EDTA solution (0.25% trypsin, 0.02% EDTA in buffer, Biochrom AG, Berlin, Germany). For in vivo experiments, Ketalar (50 mg/mL, Pfizer, NY, USA), Rompun (20 mg/mL, Bayer, Leverkusen, Germany), and Heparin (5000 IE/mL, Leo Pharma, Copenhagen, Danmark) were used. Data on cellular uptake and biodistribution were assessed by an unpaired, two-tailed Student’s t-test using GraphPad Prism (v 4.00 for Windows GraphPad Software, San Diego, California, USA) in order to determine any significant differences (p < 0.05). Instrumentation. The radioactivity was measured using an automated gamma-counter with a 3 in. NaI(Tl) detector (1480 WIZARD, Wallac Oy, Turku, Finland). The distribution of radioactivity along the thin layer chromatography strips and SDS-PAGE gels was measured on a Cyclone Storage Phosphor System and analyzed using the OptiQuant image analysis software. Cells were counted using an electronic cell counter (Beckman Coulter). Production and Purification of Affibody Molecules, ZHER2:342-H6 and (HE)3-ZHER2:342. The gene encoding ZHER2:342 was PCR amplified from vector pAY1069-(ZHER2:342)2 (12) to introduce the amino acid sequence GSHHHHHH at the Cterminus (denoted ZHER2:342-H6) or the amino acid sequence MHEHEHE at the N-terminus (denoted (HE)3-ZHER2:342). PCR fragments were fitted with the restriction sites for HindIII and NdeI, which were subsequently used to subclone the fragments into the expression vector pET21a(+), containing a T7 promoter (Novagen, Darmstadt, Germany). Correct DNA sequences were verified by cycle sequencing using an ABI Prism 3700 Analyzer (Applied Biosystems, CA). The H6-ZHER2:342 Affibody molecule having a His6-tag at N-terminus was produced as described by Orlova and coworkers (12). The molecules ZHER2:342-H6 and (HE)3-ZHER2:342 were produced in Escherichia coli strain BL21(DE3) essentially as previously described for other Affibody molecules (27). After fermentation, cells were harvested by centrifugation, and cell pellets were resuspended in washing buffer (50 mM Na2HPO4, 300 mM NaCl, pH 7). Cells were disrupted by sonication, and cell debris was removed by centrifugation. The ZHER2:342-H6 Affibody molecule was recovered by IMAC purification on a Talon Metal Affinity Resin (BD Bioscience, San Jose, CA) column under native conditions essentially as recommended by the manufacturer, using acid elution (50 mM sodium acetate, 300 mM NaCl, pH 4.8). The cleared cell lysate containing (HE)3-ZHER2:342 was heattreated (60 °C for 10 min) followed by centrifugation (35 000 g, 20 min, 4 °C). The supernatant was subsequently filtered through a 0.45 µm filter (Sartorius Stedim Biotech, Aubagne

Improved Histidine Tag

Cedex, France) before loading onto a column containing Chelating Sepharose Fast Flow (GE Healthcare, Uppsala, Sweden) with immobilized Ni2+-ions. The protein was recovered by IMAC purification under native conditions using an imidazole gradient (0-500 mM) for elution, according to the manufacturer’s instructions. The process was monitored using ¨ KTA explorer 100 system (GE Healthcare). an A Eluted fractions containing Affibody molecules were pooled and concentrated using Vivaspin2 centrifugal concentrators (Vivascience, Hannover, Germany). The buffer was changed to PBS (137 mM NaCl, 8 mM Na2HPO4, 1.5 mM KH2PO4, 2.7 mM KCl, pH 7.4) by dialysis overnight at 4 °C using SnakeSkin Pleated Dialysis Tubing (Pierce, Rockford, IL). Protein concentrations were determined from the absorbance at 280 nm and by amino acid analysis (Amino Acid Analysis Center, Uppsala University). Sample Purity Analysis. Analytic reverse-phase highperformance liquid chromatography (RP-HPLC) was used to investigate sample purity and the retention times of ZHER2:342H6, (HE)3-ZHER2:342, and H6-ZHER2:342 using a C18 column on an Agilent 1200 HPLC system (Agilent Technologies, Santa Clara, CA). Samples containing 0.2 mg Affibody molecule were injected sequentially onto the column and eluted using a 20 min gradient of 20-65% B (A, 0.1% trifluoroacetic acid (TFA) in H2O; B, 0.1% TFA in CH3CN), with a flow rate of 0.8 mL/ min. Each integrated peak area corresponding to the Affibody molecule was compared to the area of all peaks to calculate the purity. Mass Spectrometry Analysis. The molecular mass of ZHER2: 342-H6 and (HE)3-ZHER2:342 was determined by mass spectrometry analysis using a 6520 Accurate-Mass Q-TOF LC/MS instrument (Agilent Technologies, Santa Clara, CA). The molecular masses were retrieved using Agilent MassHunter B.02.00 software (Agilent Technologies) yielding the isotopic and charge state information. Biosensor Analysis. Affinity constants were measured by real-time biospecific interaction analyses using a Biacore 2000 instrument (Biacore AB, Uppsala, Sweden). The extracellular portion of human HER2 fused to the Fc region of human IgG (R&D Systems, Minneapolis, MN) was immobilized (∼1000 RU) on a flow-cell surface of a CM5 sensor chip (Biacore Life Science, GE Healthcare, Uppsala, Sweden) by amine coupling, according to the manufacture’s instructions. Another sensor chip was activated and deactivated for use in the reference flowcell. Dilution series consisting of five different concentrations were prepared for each Affibody construct and injected in duplicates over the flow-cells with a flow rate of 50 µL/min. The concentration ranges used were 90 pM to 1.5 nM for ZHER2:342-H6, 65 pM to 1 nM for (HE)3-ZHER2:342, and 200 pM to 2.9 nM for H6-ZHER2:342. After each sample injection, the flowcell surface was regenerated by 20 µL of 15 mM HCl. The dissociation equilibrium constant (KD), the association rate constant (ka), and the dissociation rate constant (kd) were calculated using BIAeValuation 3.2 software (Biacore Life Science, GE Healthcare, Uppsala, Sweden), assuming a oneto-one interaction model. Melting Point Analysis. Variable temperature measurements of ZHER2:342-H6 and (HE)3-ZHER2:342 were performed using a JASCO J-810 spectropolarimeter instrument (JASCO, Tokyo, Japan). Samples were diluted to a concentration of 60 µM in PBS, and the absorbance was measured at 221 nm during a temperature gradient increasing 5 °C/min ranging from 20 to 90 °C. Circular dichroism spectra were also recorded from 250 to 195 nm at 20 °C, before and after each variable temperature measurement.

Bioconjugate Chem., Vol. 21, No. 11, 2010 2015

Labeling of H6-ZHER2:342, ZHER2:342-H6, (HE)3-ZHER2:342. Labeling of H6-ZHER2:342, ZHER2:342-H6, and (HE)3-ZHER2:342 with [99mTc(CO)3]+ was performed as described by Orlova and coworkers (19). Briefly, 500 µL (350 MBq-2 GBq) of 99mTcO4-containing generator eluate was added to a vial with the IsoLink kit. The mixture was incubated at 100 °C during 20 min. Thereafter, 40 µL of mixture were transferred to a tube containing 50 µg (∼6.8 nmol) Affibody molecule in 40 µL PBS and incubated at 50 °C. The labeling yield after incubation for 10, 20, 40, and 60 min was monitored by ITLC to assess labeling kinetics. When the ITLC strips were eluted with PBS pertechnetate, as well as carbonyl and histidine complexes of 99mTc migrated with the eluent front (Rf ) 1.0), while Affibody molecules do not migrate under these conditions (Rf ) 0.0). To determine the presence of reduced hydrolyzed technetium, ITLC was eluted with pyridine/acetic acid:water (5:3:1.5). When this eluent was used, the technetium colloids stayed at the application point (Rf ) 0.0), while radiolabeled Affibody molecules, as well as pertechnetate and carbonyl complexes of 99mTc, migrated with the solvent front (Rf ) 1.0). The experiments were performed in quadruplicate. In addition, blank experiments were performed, where the Affibody molecules were omitted. For in vitro and in vivo studies, labeling during 60 min was selected. The labeled Affibody molecules were purified using NAP-5 desalting columns (GE Healthcare), pre-equilibrated and eluted with PBS. The purity of each preparation was evaluated using thin layer chromatography. To evaluate storage stability (shelf life), the purified radiolabeled conjugates were kept at ambient temperature (23 °C) and analyzed after 1, 2, 6, and 24 h. To predict in vivo stability of the 99mTc-label, a histidine challenge (14) of conjugates was performed. Briefly, a 500- and 5000-fold excess of histidine was added to the conjugate. After incubation at 37 °C during 1 h, the mixture was analyzed using radio-ITLC. The incubation time was selected on the basis of previous data on the biodistribution of Affibody molecules, showing that more than 95% of Affibody-related radioactivity was cleared from the blood 1 h after injection, irrespectively of label (21, 26, 28). SDS-PAGE analysis was performed to confirm the stability test data. Two samples of each labeled conjugate were incubated at 37 °C with 5000-fold excess of histidine during 1 h. After incubation, samples were analyzed by SDS-PAGE on NuPAGE 4-12% Bis-Tris Gel (Invitogen) in MES buffer (200 V constant). To facilitate interpretation of the result, a sample of 99m Tc-pertechnetate and a sample of the same conjugate not treated with histidine were analyzed in parallel with the test samples on the same gel. Distribution of radioactivity along the SDS-PAGE gels was measured using a Cyclone Storage Phosphor System. In Vitro Studies. The in vitro HER2 specificity of the 99mTclabeled Affibody molecules was evaluated using SKOV-3 ovarian carcinoma cell line and two prostate cancer cell lines, DU-145 and PC-3. All cell lines were purchased from American Type Tissue Culture Collection (ATCC) via LGC Promochem, Borås, Sweden. SKOV-3 was selected because it has a high level of HER2 expression (1.2 × 106 receptors per cell) (29) and was used in our earlier studies on radiolabeled Affibody molecules. Prostate cancer cell lines (both expressing ∼5 × 104 receptors per cell, A. Orlova and J. Malmberg, unpublished results) were selected because of our intention to visualize HER2-expression in prostate cancer in future studies. In vitro specificity testing was performed according to the methods described earlier (24). Briefly, a solution of each conjugate (concentration 0.2 nM) was added to six Petri dishes (ca. 106 cells in each). A 1000-fold excess of nonlabeled recombinant ZHER2:342 was added 5 min before the radiolabeled conjugates to saturate the receptors in blocking experiments. The cells were

2016 Bioconjugate Chem., Vol. 21, No. 11, 2010

incubated during 1 h in a humidified incubator at 37 °C. Thereafter, the media were collected, the cells were detached by trypsin-EDTA solution, and the radioactivity in cells and media was measured to calculate the fraction of cell-bound radioactivity. The cellular retention and internalization of radioactivity were evaluated as described by Wållberg and Orlova (22) using the SKOV-3 cell line. Tumor cells were incubated with radiolabeled conjugate (1 nM) for 1 h at 4 °C. Thereafter, the incubation media was discarded and the cells were washed three times using ice-cold serum-free medium. After addition of 1 mL complete medium, the cells were further incubated at 37 °C. Three dishes were analyzed for cell-associated radioactivity at predetermined time points, up to 24 h. Medium was collected, and the cells were washed three times with ice-cold serum-free mediumsthese two steps were omitted for the zero time pointsand treated with 0.5 mL of 4 M urea solution in 0.2 M glycine buffer, pH 2.5, for 5 min on ice. The acid fraction was collected, and cells were washed with another 0.5 mL acid solution. The radioactivity in the acid wash fraction was considered to be membrane-bound radioactivity. After addition of 0.5 mL of 1 M NaOH, the cells were incubated at 37 °C for at least 0.5 h and the basic wash solution collected. Dishes were washed with an additional 0.5 mL of basic solution. The radioactivity in the alkaline fractions was considered to be internalized radioactivity. The antigen-binding capacity (ABC) was analyzed according to the method described and validated by Wållberg and coworkers (30). The SKOV-3 cells were scraped from culture flasks, and three cell pellets containing 107 cells were formed by gentle centrifugation. A solution of radiolabeled Affibody molecules in cell culture medium (1 mL, 0.26 pmol) was added to each pellet to provide approximately a 100-fold molar access of receptor over conjugate. The cells were gently resuspended and incubated for 1 h at 4 °C under gentle shaking. After incubation, cells were pelleted and 0.5 mL of the supernatant was transferred to new tubes. The radioactivity of the samples was measured, and the ABC was calculated as ABC ) (Apellet+media - Amedia)/(Apellet+media + Amedia) × 100% In Vivo Studies. All animal experiments were planned and performed in accordance with national legislation on laboratory animals’ protection and were approved by the Local Ethics Committee for Animal Research. Male NMRI mice (18 weeks old, weight 32.4 ( 1.7 g) were used in the biodistribution studies. The animals were intravenously injected with 520 kBq (1 µg, 0.125 nmol) of [99mTc(CO)3]+-H6-ZHER2:342, [99mTc(CO)3]+-ZHER2:342-H6, or [99mTc(CO)3]+-(HE)3-ZHER2:342 diluted in 100 µL PBS. Mice were sacrificed and dissected at 4 or 24 h after injection. The mice were euthanized at predetermined time points by an intraperitoneal injection of Ketalar-Rompun solution (20 µL of solution per gram body weight: Ketalar, 10 mg/mL; Rompun, 1 mg/mL) followed by heart puncture with a 1 mL syringe rinsed with Heparine (5000 IE/mL). Blood and organ samplesslung, liver, spleen, stomach, kidneys, salivary gland, thyroid, muscle, bone, intestines (with content)sand the remaining carcass were collected, weighed, and the radioactivity measured as described above. The organ uptake values are expressed as percent of injected dose per gram of tissue (% ID/g), except for the intestines and the remaining carcass where values are expressed as % ID per whole sample. Male outbreed NMRI nu/nu mice were used in tumor targeting experiments. Approximately 106 LS174T colon adenocarcinoma cells (ATCC) (expressing ∼6 × 104 receptors per cell; A. Orlova, unpublished results) were subcutaneously implanted in the right hind leg. At the time of the biodistribution experiment, the average tumor size was 0.74 ( 0.39 g.

Tolmachev et al.

Animals were randomized into groups of four and injected intravenously with 1 µg [99mTc(CO)3]+-H6-ZHER2:342, [99mTc(CO)3]+-ZHER2:342-H6, or [99mTc(CO)3]+-(HE)3-ZHER2:342 conjugate (40 kBq) in 100 µL PBS. To verify the specificity of targeting by [99mTc(CO)3]+-(HE)3-ZHER2:342, HER2 receptors were presaturated in one additional group of mice by a subcutaneous injection with 500 µg of unlabeled H6-ZHER2:342 45 min before injection of the labeled conjugate. The animals were euthanized at 4 h p.i., and the biodistribution was studied as described above. For gamma camera imaging, two mice bearing LS174T xenografts were injected with 1.5 MBq (1 µg) [99mTc(CO)3]+H6-ZHER2:342 or an equal amount of [99mTc(CO)3]+-(HE)3ZHER2:342, respectively. The animals were euthanized at 4 h p.i. by overdosing Ketalar/Rompun. After euthanasia, the urinary bladders were dissected. Imaging of both mice was performed simultaneously at the department of Nuclear Medicine at Uppsala University Hospital using a Millenium VG gamma camera (General Electric) equipped with a low-energy, highresolution (LEHR) collimator. Acquisition of static images was performed with a 256 × 256 matrix and a zoom factor of 3. The energy window settings were 140 keV, 15%.

RESULTS Production and Purification of Affibody Molecules with Different Histidine Tags. The affibody tracers were produced recombinantly in Escherichia coli followed by cell disruption by sonication. ZHER2:342-H6 was directly purified from the lysate by IMAC using a column with immobilized Co2+. Purification of (HE)3-ZHER2:342 was performed by heat-treating the lysate at 60 °C for 10 min to precipitate endogenous E. coli proteins followed by centrifugation and purification by IMAC using a column with immobilized Ni2+. Purification of ZHER2:342-H6 resulted in an essentially pure protein of the expected molecular mass (Figure 2). Samples taken after the different steps in the purification of (HE)3-ZHER2:342 were analyzed by SDS-PAGE and showed that the heat-treatment resulted in reduction of endogenous E. coli proteins and that the pooled and concentrated material after IMAC was essentially pure with only one band having the expected molecular mass (Figure 2). Analysis of Physical Properties of Purified Affibody Constructs. Analytic RP-HPLC analysis showed a purity for both ZHER2:342-H6 and (HE)3-ZHER2:342 of over 90% (Table 1). The molecular masses of ZHER2:342-H6 and (HE)3-ZHER2:342 were determined by mass spectrometry analysis and showed a good correlation with the theoretical values (Table 1). The KD value for H6-ZHER2:342 interacting with HER2 was determined to be 23 pM, essentially identical to the previously determined 22 pM (12). The KD values obtained for ZHER2:342-H6 and (HE)3ZHER2:342 were similar, 20 and 18 pM, respectively (Table 1), indicating that neither the location nor the composition of the histidine tag influenced the HER2 binding. ZHER2:342-H6 and (HE)3-ZHER2:342 were also analyzed by circular dichroism, showing that both molecules had a melting temperature of 65 °C (Table 2), correlating well with the previously reported value for H6-ZHER2:342 (63 °C) (12). Also, both Affibody molecules gave CD spectra showing high R-helical content, similar to those previously determined for H6-ZHER2:342. The spectra recorded for both constructs before and after heating to 90 °C were highly similar, indicating that both molecules refolded efficiently (Figure 2). Labeling of H6-ZHER2:342, ZHER2:342-H6, (HE)3-ZHER2:342. The use of labeling conditions optimized by Orlova and co-workers (19) enabled efficient labeling, resulting in a radiochemical yield of more than 80% after 60 min of incubation with [99mTc(CO)3(H2O)3]+ for all conjugates (Figure 3). Labeling of H6-ZHER2:342 was somewhat more efficient than labeling of

Improved Histidine Tag

Bioconjugate Chem., Vol. 21, No. 11, 2010 2017 Table 2. In Vitro Stability of [99mTc(CO)3]+-H6-ZHER2:342, [99mTc(CO)3]+-ZHER2:342-H6, and [99mTc(CO)3]+-(HE)3-ZHER2:342a stability in PBS at ambient temperature Tc(CO)3]+- [99mTc(CO)3]+- [99mTc(CO)3]+H6-ZHER2:342 ZHER2:342-H6 (HE)3-ZHER2:342

99m

[ 0 1h 2h 6h 24 h

99.9 ( 0.9 99.8 ( 0.2 99.0 ( 0.2 99.6 ( 0 97.2 ( 0.2

control 500-fold excess histidine 5000-fold excess histidine

99.1 98.5 ( 0.2 98.2 ( 0.2

98.2 ( 0.2 97.8 ( 0.2 98.1 ( 0.2 97.7 ( 0.6 96.7 ( 0

99.3 ( 0.1 98.6 ( 0.1 99.7 ( 0.2 99.5 ( 0.3 98.5 ( 0.8

Histidine Challenge (1 h at 37 °C) 96.5 95.2 ( 0.7 95.1 ( 0.9

99.0 98.4 ( 0.5 98.4 ( 0.2

a Each data point represents an average from three samples ( standard deviation.

Figure 3. Labeling kinetics of Affibody molecules using IsoLink at 50 °C. [99mTc(CO)3(H2O)3]+ was added to 50 µg of Affibody molecules in 40 + l PBS. After 10, 20, 40, and 60 min of incubation, samples of the reaction mixture were analyzed using thin-layer chromatography. Labeling yields are presented as an average from four independent experiments with the error bars corresponding to (1 standard deviation.

Figure 2. Analysis of ZHER2:342-H6 and HEHEHE-ZHER2:342 Affibody molecules. SDS-PAGE analysis of samples taken during the purification of HEHEHE-ZHER2:342 is shown in panel A. Lane 1, sample after cell lysis and initial clarification; lane 2, sample after heat-treatment; lane 3, sample after IMAC purification. Panel B: an overlay of CD-spectrum recorded before and after thermal unfolding (as in panel E) of HEHEHE-ZHER2:342. Panel C: SDS-PAGE analysis of a sample of the eluted material after IMAC purification of ZHER2:342-H6. Panel D: an overlay of CD-spectrum recorded before and after thermal unfolding (as in panel E) of ZHER2:342-H6. Panel E: overlay of the relative ellipticity measured by CD-spectrometry at 221 nm as a function of temperature for HEHEHE-ZHER2:342 and ZHER2:342-H6. Table 1. Characteristics of Affibody Molecules Affibody molecule

calculated Mw (Da) found Mw (Da) sample purity (%) RP-HPLC retention time (min) Tm (°C) KD (pM) Ka (M-1 s-1) Ks (s-1)

(HE)3-ZHER2:342

ZHER2:342-H6

H6-ZHER2:342

7835 7836 92.6 15.6 65 18 7.0 × 106 1.2 × 10-4

7872 7873 95.6 15.3 65 20 2.2 × 107 4.5 × 10-4

8215 8215 98.0 15.6 63 23 4.0 × 106 8.79 × 10-5

ZHER2:342-H6 and (HE)3-ZHER2:342. A simple purification using a disposable desalting column provided isolated yield of 75 (

7%, 71 ( 12%, and 77 ( 4% for [99mTc(CO)3]+-H6-ZHER2:342, [99mTc(CO)3]+-ZHER2:342-H6, and [99mTc(CO)3]+-(HE)3-ZHER2:342, respectively (mean ( standard deviation, n ) 3). The radiochemical purity was over 95%, and the specific radioactivity was in the range 1.7-2 MBq/µg (14-15.8 GBq/µmol). Radiocolloid formation was very low, less than 1.5% with all three constructs. The results from the stability tests are presented in Table 2 and Figure 4, showing that all conjugates demonstrated a remarkably high stability. The loss of radioactivity was about 1% after 24 h incubation in PBS at room temperature. All constructs showed a high stability of label attachment in the histidine challenge; the release of technetium-99m was only minor and independent of the excess of free histidine. The results were comparable with previously published data for 99mTc Histag-labeled compounds (14, 19). Figure 4 shows a representative radioactivity pattern after SDS-PAGE analysis of a sample from 60 min incubation of ([99mTc(CO)3]+-(HE)3-ZHER2:342) with 5000fold excess of histidine at 37 °C. 99mTc-pertechnetate and a control sample incubated in PBS at room temperature were included for comparison. The main peak corresponds to migration of a monomeric Affibody molecule. The only other radioactivity peak corresponds to low-molecular-weight compounds, such as 99mTc-pertechnetate. This peak did not contain more than 1.1% of the total activity after the challenge, showing high stability of the complex. In Vitro Studies. Binding specificity tests (Table 3) indicated that the binding of all 99mTc-labeled Affibody constructs to living HER2-expressing cells was receptor mediated; saturation of the

2018 Bioconjugate Chem., Vol. 21, No. 11, 2010

Tolmachev et al.

Figure 4. SDS-PAGE analysis of [99mTc(CO)3]+-(HE)3-ZHER2:342 stability under histidine challenge. Distribution of radioactivity along lanes was visualized and quantified using Cyclone Storage Phosphor System. 1, [99mTc(CO)3]+-(HE)3-ZHER2:342 control sample, which was incubated in PBS at room temperature for 1 h; 2, [99mTc(CO)3]+-(HE)3-ZHER2:342 sample, which was incubated in 5000-fold excess of histidine at 37 °C for 1 h; 3, 99mTc-pertechnetate was used as a marker for low-molecularweight compounds. Table 3. Specificity of Binding of Technetium-99m-Labeled Affibody Molecules to HER2-Expressing Cells in Vitroa nonblocked

blocked

p-value

2.3 ( 0.5 3.0 ( 0.1 0.9 ( 0.1