In Vivo and In Vitro Studies on Renal Uptake of Radiolabeled Affibody Molecules for Imaging of HER2 Expression in Tumors

Abstract

Affibody molecules (6–7 kDa) are a new class of small robust three-helical scaffold proteins. Radiolabeled subnanomolar anti-HER2 affibody Z_HER2:342 was developed for imaging of HER2 expression in tumors, and a clinical study has demonstrated that the ¹¹¹In- and ⁶⁸Ga-labeled affibody molecules can efficiently detect HER2 expressing metastases in breast cancer patients. However, a significant renal accumulation of radioactivity after systemic injection of a radiolabeled anti-HER2 affibody conjugate is observed. The aim of this study was to investigate the mechanism of renal reabsorption of anti-HER2 affibody at the molecular level. Renal accumulation of radiolabeled anti-HER2 affibody molecules was studied in a murine model and in vitro using opossum-derived proximal tubule (OK) cells. It was found that kidney reabsorption of affibody molecule was not driven by megalin/cubilin. Amino acids in the target-binding side of affibody molecule were involved in binding to OK cells. On OK cells, two types of receptors for anti-HER2 affibody molecule were found: K_D1=0.8 nM, B_max1=71,500 and K_D2=9.2 nM, B_max2=367,000. The results of the present study indicate that affibody molecule and other scaffold-based targeting proteins with a relatively low kidney uptake can be selected using in vitro studies with tubular kidney cells.

Introduction

Radionuclide molecular targeting is a promising method for detection, characterization, and treatment of both localized and metastasized tumors. Several classes of radiolabeled proteins have been used for the delivery of radionuclides to tumors in vivo, utilizing the specific affinity of these agents for phenotypic alterations in malignant cells. In fact, a tumor targeting with radiolabeled antibodies for detection of malignancy or for therapeutic purposes has been reported as early as the 1950's.¹ During the last decades, there was a general trend toward using smaller targeting proteins like antibody fragments, peptide hormones, and scaffold proteins to overcome slow clearance of bulky antibodies. This strategy enables improved imaging with high contrast and reduction of the radiation dose to normal tissues during therapy. Nowadays, peptide-based radionuclide imaging and therapy play a well-established role in the management of patients with neuroendocrine tumors.² The use of somatostatin analogues for imaging and therapy of neuroendocrine tumors³ and RGD (Arg-Gly-Asp) peptides to visualize tumor-associated angiogenesis⁴ are some well-known examples. Selectivity of these radionuclide-loaded peptides to cancer cells, rapid penetration to the tumor tissue in combination with rapid clearance from normal tissues, designates them as very promising targeting agents.

Unfortunately, high kidney reabsorption after glomerular filtration was often a stumbling block for further development of many peptide-based agents as efficient radiolabeled targeting probes. This tubular reabsorption of tracers smaller than the cutoff for glomerular filtration (∼60 kDa) might potentially compromise a reliable diagnostic accuracy in the retroperitoneal, periaortic, and epigastric regions, especially with residualizing labels, such as radiometals.⁵ Moreover, during radionuclide therapy, high accumulation of the radiotracer in the kidneys can be a limiting factor for the successful delivery of appropriate radiation doses to tumors. For example, deterioration of renal functions is the dose-limiting toxicity for treatment of neuroendocrine tumors with ⁹⁰Y-DOTATOC despite a number of kidney-protecting precautions (dose fractionation, coinjection of positively charged amino acids, and pre-estimation of patient's risk factors).⁶

It is known that megalin and cubilin, two endocytic receptors, expressed in the proximal tubule, play a major role in normal reabsorption of proteins in kidneys.⁷ It has been demonstrated that the scavenger receptor megalin, is the main mediator of renal accumulation of somatostatin analogues and other small peptides.^8,9 Melis et al. showed that the reabsorbed somatostatin analogue ¹¹¹In-DTPA-octreotide localized in the renal cortex.¹⁰ Moreover, a profound decrease in radioactivity accumulation in the kidneys of megalin-deficient mice was observed for several radiolabeled small peptides, thus clearly indicating a direct involvement of the megalin/cubilin system in the reabsorption process of these agents.¹¹ Accumulation of the radiotracer in the kidney by the megalin/cubilin system is a complex process. Megalin is a large multiligand receptor (517 kDa) with several domains, each of which might interact with different parts of the ligand. Different binding sites might be involved in the accumulation of different peptides.¹² Moreover, it was proposed that the differences in renal accumulation of radiopeptides may be related to the number of charged amino acids of a molecule.¹³ These data indicate that the kidney reabsorption mechanism for each peptide-based radioactive probe could be individual and the optimal strategy for decreasing renal reabsorption could differ from one protein to another.

Affibody molecules (6–7 kDa) are a new class of small robust three-helical scaffold proteins, based on a modified B-domain of a staphylococcal protein A.¹⁴ Randomization of 13 amino acids on helices 1 and 2 provided a large library that enables a display selection of high-affinity (ranging from low nanomolar to picomolar) binders to different proteins.^14,15 The radiolabeled anti-HER2 affibody molecule Z_HER2:342 and its derivatives have been evaluated preclinically for imaging of HER2 expression in tumors.¹⁶ A clinical study has demonstrated that the ¹¹¹In- and ⁶⁸Ga-labeled affibody molecules can efficiently detect HER2 expressing metastases in breast cancer patients.¹⁷ However, similar to other small proteins and peptides, efficient tumor targeting was associated with significant renal accumulation of the radioactivity. Fusion of Z_HER2:342 with an albumin-binding domain allows noncovalent binding of the conjugate to albumin in the circulation, which resulted in a dramatic decrease of renal uptake,¹⁸ suggesting that the kidney uptake was associated with tubular reabsorption. At the same time, the use of labeling methods yielding nonresidualizing radiocatabolites caused a significant reduction of renal retention.^19

–25

Better understanding of the fate of affibody molecules in the kidneys would open the way for the rational design of new affibody conjugates and/or could facilitate the development of substances that can effectively block kidney reabsorption. Therefore, here we studied the mechanism of kidney reabsorption of affibody molecules in detail both in vitro and in vivo. The opossum-derived proximal tubule (OK) cell line was selected as an in vitro model for this study. In these cells, the expression of scavenger receptors, transporters, and kidney ion channels are preserved.^26
–28 Z_HER2:2395-Cys, a cysteine-containing derivative of the Z_HER2:342 anti-HER2 affibody molecule was selected as a model molecule. This affibody molecule with K_D=27 pM²² was labeled with indium-111 via site-specifically conjugated maleimido derivative of 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA) and with technetium-99m using a peptide-based natural chelating system at the C-terminus.^22,23 Earlier, it was shown that both labeling methods yielded residualizing radiocatabolites and similar biodistribution profiles (kidney radioactivity uptake was 191%±15% injected dose per gram tissue [ID/g] for ^99mTc and 168%±24% ID/g for ¹¹¹In at 4 hours post injection [p.i.]).^22,23

The aim of this study is to investigate the mechanism of renal reabsorption of anti-HER2 affibody molecules at the molecular level, and to determine the role of the megalin/cubilin receptor complex system in this process.

Materials and Methods

The ^99mTc-pertechnetate solution was acquired by elution of an Ultra-TechneKow generator (Covidien) eluted with sterile 0.9% sodium chloride (Covidien). ¹¹¹In-indium chloride was purchased from Covidien. All affibody molecules (Z_HER2:2395-Cys, Z_taq:1154, Z_insulin:810, and Z_EGFR:2377) were kindly provided by Affibody AB. The radioactivity of samples was measured in an automated gamma counter with a 3 inch NaI(Tl) well detector (1480 WIZARD™; Wallac Oy). Cells in in vitro experiments were counted manually using a cell counter (Scepter, Millipore). Labeling of affibody molecules with indium-111 was performed as described previously.²² Labeling with technetium-99m was performed using a two-vial freeze-dried kit method as described previously.²⁹ After purification on the NAP-5 size-exclusion column (Amersham Biosciences) purity of ^99mTc-Z_HER2:2395-C was >95%, whereas the content of hydrolyzed ^99mTc did not exceed 1%.

Animal experiments

All animal experiments were planned and performed in accordance with national legislation on laboratory animals' protection. The studies had been approved by the Local Ethics Committee for Animal Research. In all in vivo experiments indium-labeled Z_HER2:2395-Cys was used.

To evaluate a gender influence of renal reabsorption of radiolabeled affibody molecules, four Naval Medical Research Institute (NMRI) male and four NMRI female mice were intravenous (IV) injected with 1 μg of the ¹¹¹In-labeled anti-HER2 affibody molecule (30 kBq). Biodistribution was determined at 4 hours p.i. Mice were exsanguinated by heart puncture under anesthesia, organs and tissue samples were collected, weighed, and the radioactivity content was determined.

For SPECT imaging, five kidney-specific megalin-deficient mice³⁰ and five wild-type mice were injected IV with 10 μg of the ¹¹¹In-labeled anti-HER2 affibody molecule (20 MBq). SPECT images were acquired at 4 hours p.i., using the USPECT II (MILabs). Mice were scanned in a prone position under general anesthesia (isoflurane/N₂O/O₂) for 45 minutes using the 1.0-mm-diameter pinhole collimator tube. After completion of the acquisition of the images, mice were euthanized and the concentration of the radioactivity in dissected tissues was determined as described above. Scans were reconstructed with MILabs reconstruction software, using an ordered-expectation maximization algorithm, with a voxel size of 0.375 mm.

Cell experiments

The opossum kidney cell line (OK) and ovarian cancer cells (SKOV-3) were purchased from ATCC, via LGC Standards AB, Boras, Sweden. OK cells were cultured in the MEM Earle's and SKOV-3 cells in the RPMI medium complemented with 10% fetal calf serum (FCS), 2 mM L-glutamine, and PEST (penicillin 100 IU/mL, streptomycin 100 μg/mL), all from Biochrom AG, Berlin, Germany. These media are referred to as complete media in the text. In all the experiments with OK cells, cells were incubated in a 1% FCS medium. Cells were detached using the trypsin-ethylenediaminetetraacetic acid (EDTA) solution (0.25% trypsin, 0.02% EDTA in a buffer; Biochrom AG). In vitro studies were performed in triplicate. In all cell experiments, the ^99mTc-labeled affibody molecule Z_HER2:2395-Cys was used. Cells were seeded 2 days before the experiment in Petri dishes (ø 3.5 cm) in amount 2×10⁶ cell/dish.

Binding kinetics: cell-associated radioactivity as a function of the incubation time

Cultured OK cells were washed once with the serum-free medium and incubated with 1 mL of 1 nM ^99mTc–Z_HER2:2395-Cys in a 1% FCS medium at 4°C up to 4 hours. At predetermined time points, the cells were washed two times with the serum free-medium and treated with 0.5 mL/dish trypsin-EDTA at room temperature. When the cells were detached, 0.5 mL/dish of the serum-free medium was added and the cells were resuspended and collected. The radioactivity was measured and the percentage of cell-associated radioactivity at different time points was calculated. Data were normalized to maximum cell-associated radioactivity.

Binding of ^99mTc–Z_HER2:2395-Cys to OK cells at different conditions

The goals of this experiment were to investigate if the binding of ^99mTc–Z_HER2:2395-Cys to OK cells is saturable by nonlabeled Z_HER2:342, whether the presence of FCS interferes with binding, and whether internalization contributes to the uptake of ^99mTc–Z_HER2:2395-Cys by OK cells. In total, four sets of cells (12-well plate, Nunclone Surface; Nunc) were prepared. A radiolabeled conjugate was added to cells in a concentration of 1 nM and incubated at 37°C. One set of cells was preincubated for 1 hour at 37°C in a medium without FBS and to another set, a 10,000-fold excess of unlabeled affibody molecule Z_HER2:342 was added 5 minutes before adding the labeled conjugate. One set of cells was incubated for 1 hour at 4°C. After incubation, supernatants were collected and cells were treated as described above. The radioactivity of samples was measured and the percentage of cell-bound radioactivity was calculated.

Competitive binding assay, IC₅₀ determination

The ability of different affibody molecules and human albumin to inhibit binding of ^99mTc–Z_HER2:2395-Cys to OK cells was studied in vitro. The half maximum inhibitory concentration (IC₅₀) of HER2- (Z_HER2:342) and non-HER2- (Z_taq:1154, Z_insulin:810 and Z_EGFR:2377) affibody molecules and of albumin was determined. Cell monolayers were washed once with the serum-free medium and incubated for 2 hours at 4°C with ^99mTc–Z_HER2:2395-Cys (1 nM) in the presence of increasing concentrations (range 0–9.7 10⁻⁴ M) of competitors. After incubation, cells were treated as described above. In addition, the half maximum inhibitory concentration (IC₅₀) of Z_HER2:342 for binding of ^99mTc-Z₂₃₉₅-C to SKOV-3 cells (high HER2 expression) was determined.

Receptor expression and affinity, Scatchard analysis

OK cells were incubated for 2 hours at 4°C with ^99mTc–Z_HER2:2395-Cys (a concentration range of 0.5 to 300 nM). For each data point, four dishes were used, including one presaturated with 30 μM unlabeled Z_HER2:342 to determine unspecific binding. After incubation, the medium was aspirated and the cells were treated as described above. From the cell suspension, 0.1 mL was used for cell counting and the rest for radioactivity measurements. Finally, data were corrected on unspecific binding.

Cellular processing of ^99mTc–Z_HER2:2395-Cys by OK cells

OK cells were incubated with 3 or 30 nM ^99mTc–Z_HER2:2395-Cys in a 1% FCS medium at 37°C (5% CO₂). At various time points, the medium was collected and the cells were washed once with the serum-free medium. The membrane-bound and internalized radioactivity was estimated essentially as described by Wållberg.³¹ Briefly, cells were treated with 0.5 mL of a 4 M urea solution in a 0.2 M glycine buffer, pH 2.5, for 5 minutes on ice. The acid supernatant was collected and the cells were washed twice with an additional 0.5 mL of acid buffer, which were also collected. The radioactivity in the acid fractions was considered as a membrane-bound radioactivity. For lysis, 0.5 mL of a 1 M sodium hydroxide solution was added and the cells were incubated at 37°C for at least 0.5 hours. The alkaline solution containing cell debris was collected. Dishes were washed with an additional 0.5 mL basic solution and the basic fractions were pooled. The radioactivity in alkaline fractions was considered to contain the internalized activity.

Data analysis and statistics

Data for cell analysis were investigated by the unpaired, two-tailed t test using GraphPad Prism (version 4.00 for Windows GraphPad Software) to find significant statistical differences (p<0.05). Binding data were analyzed using the GraphPad Prism software package.

Results

Animal experiments

The biodistribution data of ¹¹¹In-labeled anti-HER2 affibody molecules are presented in Tables 1 and 2. No apparent difference in the radioactivity uptake in kidneys related to the mice gender was found. Remarkably, the uptake of ¹¹¹In-DOTA-Z_HER2:2395-Cys in the kidney of the megalin-deficient mice was identical to the uptake in the kidneys of the wild-type mice. Micro-SPECT images of the kidney in megalin-deficient mice and wild-type mice of ¹¹¹In-DOTA-Z_HER2:2395-Cys confirmed the identical accumulation of the radioactivity in kidneys (Fig. 1), indicating that megalin expression in the kidney is not essential for the renal uptake of ¹¹¹In-labeled anti-HER2 affibody molecules.

FIG. 1.

micro-SPECT images of megalin-deficient mice (A) versus wild-type mice (B) 4 hours p.i. of ¹¹¹In-DOTA-Z_HER2:2395-Cys. p.i., post injection; DOTA, 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid.

Table 1.

Biodistribution of ¹¹¹ In in NMRI Male Versus Female Mice at 4 Hours After Injection of ¹¹¹ In-1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA)-Z_HER2:2395-Cys

	Male	Female
Blood	0.12±0.03	0.12±0.03
Liver	1.28±0.07	1.0±0.2
Spleen	0.05±0.01	0.035±0.008
Kidney	79±2	70±10
Gastrointestinal tract	0.9±0.1^a	1.4±0.1
Carcass	3.7±0.08	2.8±1.0

Significant difference (p<0.05) in a unpaired t-test.

Mice were injected iv (tail vein) with 1 μg of protein per animal. Data are presented as average% ID per organ±SD (n=4).

ID, injected dose; iv, intravenous; NMRI, Naval Medical Research Institute.

Table 2.

Biodistribution of ¹¹¹ In in Megalin-Deficient Mice Versus Wild-Type Mice 4 Hours Post Injection (P.I.) of ¹¹¹ In-DOTA-Z_HER2:2395-Cys

	Megalin-deficient mice	Wild-type mice
Blood	0.080±0.010	0.08±0.02
Muscle	0.09±0.02	0.08±0.02
Lung	0.32±0.05	0.35±0.10
Pancreas	0.15±0.02	0.12±0.03
Spleen	0.51±0.06	0.51±0.08
Kidney	202±16	214±26
Liver	1.6±0.2	1.5±0.3
Stomach	0.25±0.02	0.23±0.05
Duodenum	0.33±0.04	0.29±0.06

Mice were injected iv (tail vein) with 10 μg of protein per animal. Data are presented as average% ID per organ±SD (n=5).

In vitro characterization

The in vitro assay showed that ^99mTc–Z_HER2:2395-Cys at 4°C showed rapid binding to OK cells that reached a plateau almost in 1.5 hours with a cell-associated radioactivity 3.6% from added (Fig. 2). It was also found that incubation of cells at 4°C should not exceed 4 hours, since cells start to detach. Preincubation of OK cells in the serum-free medium to stimulate protein transport did not influence binding of ^99mTc–Z_HER2:2395-Cys to OK cells (Fig. 3). Incubation at 4°C or addition of an excess of the nonlabeled affibody molecule significantly decreased binding.

FIG. 2.

Binding of ^99mTc–Z_HER2:2395-Cys to opossum-derived proximal tubule (OK) cells at 4°C as a function of time. Data presented as an average of three samples with standard deviations, error bars might be not seen because they are smaller than point symbols.

FIG. 3.

Binding of ^99mTc–Z_HER2:2395-Cys (1 nM) to OK cells, 1-hour incubation in different conditions (block—preincubation for 15 minutes with 10 μM Z_HER2:342, prefasting—preincubation for 1 hour with FBS-free media).

The IC₅₀ values of the anti-HER2 affibody molecule Z_HER2:342, the non-HER2 affibody molecules Z_EGFR, Z_insuline, and Z_taq and albumin to OK cells were determined (see Fig. 4A and Table 3). The IC₅₀ value of Z_HER2:342 was in a low nanomolar range (1.46 nM) and was five orders of magnitude lower compared with the other affibody molecules and albumin. Comparison of binding of ^99mTc-Z₂₃₉₅-C to kidney cells (OK) and cancer cells with high HER2 expression (SKOV-3) showed that the inhibition concentration, IC₅₀, for binding of the anti-HER2 affibody molecule Z_HER2:342 to OK cells was more than 30 times lower than for binding to HER2-expressing cells (1.5 nM for OK and 49 nM for SKOV-3 cells). Interestingly, the inhibition curve for OK cells showed a rather flat slope (slope=−0.7), which could be due to the presence of more than one binding site with different affinities for ^99mTc–Z_HER2:2395-Cys on OK cells.

FIG. 4.

(A) Inhibition of ^99mTc-Z₂₃₉₅-C binding to OK cells with different Affibody molecules. (B) Inhibition of ^99mTc–Z_HER2:2395-Cys binding to OK and SKOV-3 (ovarian cancer with high HER2-expression) cells. Data presented as an average of three samples with standard deviations.

Table 3.

Inhibition Concentrations (IC₅₀) for Different Affibody Molecules and Albumin for Binding of ^99m Tc–Z_HER2:2395-Cys to Opossum-Derived Proximal Tubule Cells

Inhibitor	IC₅₀ concentration, M
Z_HER:342	1.46×10⁻⁹
Z_EGFR:2377	1.5–2×10⁻⁴
Z_insuline:810	2.5–3×10⁻⁴
Z_taq:1154	3.5–4×10⁻⁴
Albumin	5×10⁻⁴

The expression of receptors on OK cells capable to bind to ^99mTc–Z_HER2:2395-Cys was quantified in a saturation experiment (Fig. 5). In good agreement with inhibition data, the resulting binding curve fitted to two sites (receptors) binding model with two different affinities for ^99mTc–Z_HER2:2395-Cys, K_D1=0.8 nM (B_max1=71,500 receptors/cell) and K_D2=9.2 nM (B_max2=367,000 receptors/cell). The total amount of receptors capable to bind to the anti-HER2 affibody molecule on OK cells was 440,000 receptors/cell.

FIG. 5.

Nonlinear regression analysis for two sites binding of data from a binding saturation experiment. Increasing concentrations of ^99mTc–Z_HER2:2395-Cys (x-axis) were added to dishes with OK cells.

To study the role of the two different receptor types in anti-HER2 affibody molecule binding and cellular processing, OK cells were incubated with two different concentrations of ^99mTc–Z_HER2:2395-Cys (Fig. 6). The binding and internalization patterns at two different concentrations demonstrated different cellular processing of ligand–receptor complexes. At a concentration of 3 nM (same amount of ^99mTc–Z_HER2:2395-Cys bound to each receptor type), the cellular binding of ^99mTc–Z_HER2:2395-Cys reached a maximum at 2 hours. At later time points, cell binding decreased and dropped to 60% of the maximum after 8 hours. In contrast, when OK cells were incubated with a 30 nM concentration of ^99mTc–Z_HER2:2395-Cys (assuming ^99mTc–Z_HER2:2395-Cys more abundantly bound to the receptor with a lower affinity), the cell-associated and internalized radioactivity increased continuously during the experiment (about 80% bound within the first hour).

FIG. 6.

Binding and cellular processing of ^99mTc–Z_HER2:2395-Cys by OK cells at continuous incubation with two different concentrations at 37°C. Data presented as an average of three samples with standard deviations.

Discussion

The mechanism of renal reabsorption and retention of radiolabeled peptides and small proteins has been studied by several groups during the last decade. Possible renal toxicity and tentative mechanisms of its prevention utilizing various blockers were the main driving forces for these studies. The majority of these studies were performed using in vivo models. In particular, the molecular mechanism of kidney reabsorption of radiolabeled peptides and proteins mediated by the megalin/cubilin coreceptors has attracted substantial attention during the last years. Megalin and cubilin are synergistically acting endocytic receptors located mainly on the brush border of the apical parts of proximal tubules.³² Megalin is a large transmembrane receptor; cubilin is a somewhat smaller receptor with no extracellular domain, which directs the trafficking of its ligands toward lysosomes through dimerization with megalin. Both receptors are important for the recovery of proteins, such as albumin, vitamin-binding proteins, hormones, and apolipoproteins, from the glomerular filtrate. Involvement of megalin in renal reabsorption of radiolabeled somatostatin analogues has been demonstrated.^8,26 Later, it has been confirmed that the same receptor pair is partly responsible for the renal reuptake of other peptides and proteins, such as exendin, minigastrin, and neurotensin.¹¹ Interestingly, although the level of expression of megalin/cubilin coreceptors in mice does not differ between the sexes, the same radiolabeled peptide exhibits a higher renal uptake in female than in male mice.¹¹ Coinfusion of substances that are ligands of megalin/cubilin, such as positively charged amino acids and albumin fragments, reduce the kidney uptake of the radiolabeled small protein.^9,33,34 Recently, the involvement of megalin in renal reabsorption of radiolabeled anti-EGFR nanobodies (15 kDa) has been demonstrated,³⁵ despite EGFR expression in a proximal tubule.³⁶ In summary, the studies so far suggest that the most important mechanism responsible for renal reabsorption of small radiolabeled proteins and peptides is megalin-mediated tubular reabsorption.

Affibody molecules appeared as promising imaging probes for in vivo detection of several tumor-associated molecular abnormalities.¹⁶ Their prompt extravasation, high uptake by tumors, low uptake in normal tissues, and a rapid renal clearance provided high-contrast images within a few hours after injection.¹⁷ However, affibody molecules labeled with radiometals consistently show the high radioactivity accumulation in the kidneys. Although the kidneys are a well-defined structure, the high uptake in the kidney might obscure imaging of lesions in the kidneys or in their vicinity and it might cause high-radiation doses during therapeutic application. It is known that HER2 is expressed only in developing proximal tubules.³⁶ Moreover, it was found that high renal reabsorption occurs for anti-HER2 affibody molecules independent of the labeling method used, but in the case of nonresidualizing labels (radiohalogens and some technetium-99m and rhenium-188), the renal retention is low.^25,37 At the same time, the tumor uptake and retention of radioactivity is stable for both residualizing radiometals and nonresidualizing radiohalogens (see Fig. 7). Long retention of the radioactivity in tumors correlates well with slow internalization rates of anti-HER2 affibody molecules in HER2-expressing tumor cells.^31,38 This mismatch indicates that different processes are involved in the cellular processing of radiolabeled affibody molecules in tumor cells and in the kidneys. Knowledge about different chelators and the properties they confer to the targeting agent and optimization of labeling chemistry enabled reduction of renal retention of the radioactivity.^{19,21,24,39,40} However, the molecular mechanism of reabsorption of affibody molecules presumably by tubular cells is not fully understood.

FIG. 7.

Biodistribution over time in SKOV-3 tumor-bearing BALB/c nu/nu mice after administration of ¹¹¹In-DOTA-Z_HER2:342 ³⁷ and ¹²⁵I-PIB-Z_HER2:342.²⁵

In the present study, ¹¹¹In-labeled anti-HER2 affibody molecules were injected in NMRI mice: the same level of the injected activity was retained in the kidneys in both male and female mice. This observation was the first indication that the megalin/cubilin system is not involved in kidney reabsorption of affibody molecules, considering the pronounced difference of the megalin activity in male and female mice. Additional in vivo experiments did not show any difference in the renal uptake of ¹¹¹In-labeled Z_HER2:2395-Cys between megalin-deficient and wild-type mice. This observation indicated that a megalin-mediated mechanism for kidney retention could be excluded.

Further experiments were carried out in cultured immortal tubular opossum kidney OK cells, which retained the molecular properties of kidney tubular cells. The time course of ^99mTc–Z_HER2:2395-Cys binding to OK cells showed that the equilibrium was reached fast and binding of ^99mTc–Z_HER2:2395-Cys to OK cells was significantly reduced by preincubation with an excess of an unlabeled conjugate, which was an indication that binding of the radiolabeled anti-HER2 affibody molecule to tubular cells is mediated by a saturable mechanism. Preincubation of OK cells in a serum-free medium before the incubation with radiolabeled proteins had been used to remove serum albumin from megalin/cubilin to examine the megalin/cubilin driven uptake of albumin.²⁷ Because prefasting had no effect on the binding of radiolabeled affibody ^99mTc–Z_HER2:2395-Cys to OK cells, megalin/cubilin is not involved.

Incubation of cells with ^99mTc–Z_HER2:2395-Cys at 37°C provided a significantly higher cell uptake than incubation at 4°C, suggesting internalization of the radiolabeled tracer. To exclude the influence of internalization, the IC₅₀ was determined at 4°C. In the competition experiments, five orders of magnitude higher concentration of anti-EGFR, anti-insulin, and anti-Taq-polymerase affibody molecules was required to inhibit the binding of ^99mTc–Z_HER2:2395-Cys to OK-cells by 50%, compared to the unlabeled parent affibody molecule (Z_HER2:342). Affibody molecules consist of a scaffold part, which is common for all tested affibody molecules, and a target-binding area determined by 13 amino acids on the surface of helices 1 and 2. The target-binding region is unique for Z_HER2:2395-Cys and Z_HER2:342. The results of the inhibition experiments clearly showed that amino acids in the target-binding parts of Z_HER2:2395-Cys and Z_HER2:342 were involved in the binding to OK cells. Furthermore, albumin was as inefficient in blocking of ^99mTc–Z_HER2:2395-Cys to OK cells as non-HER2-binding affibody molecules. As albumin is a ligand of megalin, this is another strong indication that the megalin/cubilin system is not involved in binding of the anti-HER2 affibody molecule that correlates with in vivo findings.

To evaluate a possibility that opossum ErbB2 (homologue of human HER2) is directly involved in the binding of ^99mTc–Z_HER2:2395-Cys by OK cells, an inhibition profile of ^99mTc–Z_HER2:2395-Cys was measured on the HER2-expressing human ovarian carcinoma SKOV-3 cell line. The half inhibition concentration on OK cells was 30-fold lower than that on SKOV-3 cells, where the uptake is HER-2 mediated. These data suggest that another receptor, different from the targeted antigen HER2, is responsible for this extensive uptake of ^99mTc–Z_HER2:2395-Cys by tubular OK cells.

Further determination of the equilibrium dissociation constant K_D indicated the presence of two types of receptors on the OK cells that bind the anti-HER2 affibody molecule ^99mTc–Z_HER2:2395-Cys: one with a high affinity of 0.8 nM and one with a lower affinity of about 9.2 nM. Cellular processing of ^99mTc–Z_HER2:2395-Cys by OK cells was studied by a continuous incubation at physiological conditions at two different concentrations of the radiolabeled affibody molecule. The lower concentration (3 nM) was chosen at such a level that a contribution of a receptor with K_D1 would be substantial (about 50% of bound radioactivity). At this concentration, 4·K_D1 or 0.3·K_D2, ∼80% of high-affinity receptors and 25% of low-affinity receptors will be occupied. At the higher concentration (30 nM), the more abundant low-affinity receptor should predominantly contribute to the binding and cellular processing pattern of ^99mTc–Z_HER2:2395-Cys. We found that binding kinetics was very rapid for both receptors and internalization was slow. After 2 hours incubation at the lower concentration, the cell-associated activity started to decline, and decreased to 60% of the maximum value at 8 hours of incubation. It should be noted that no depletion of radiolabeled affibody molecules occurred. Such a pattern points to internalization of the receptor–ligand complex without subsequent repopulation of the receptors. This pattern has not been observed for anti-HER2 affibody molecules bound to HER2 expressing cells.²¹ In HER2 expressing tumor cells, monomeric affibody molecules do not trigger intracellular signaling,⁴¹ which is reflected in slow internalization of a ligand–receptor complex.²¹ The binding and processing pattern after binding to the low-affinity receptor was different: after rapid binding during the first 2 hours, the cell-associated radioactivity increased slowly during the 8-hour period. Taking into account that the high-affinity receptor-bound activity decreased, it can be assumed that the low-affinity receptor is constantly repopulated. After 8 hours, the internalized radioactivity was at a level that was twofold higher than that observed in HER2 expressing tumor cells.²¹

We started this project with the aim to identify the molecular mechanism of renal reabsorption of the anti-HER2 affibody molecule Z_HER2:2395-Cys, in particular, and affibody molecules as the scaffold protein in general. The development of efficient predictive in vitro models can help to reduce a number laboratory animals used for the development of new imaging agents. We hypothesized that the mechanism of renal reabsorption for affibody molecules was similar to that of other small proteins, for example, megalin/cubilin mediated. Unexpectedly, we have found that this coreceptor system is not involved in renal reabsorption of anti-HER2 affibody molecules. Moreover, in our in vitro model, the anti-HER2 affibody molecule binds to two types (or subtypes) of receptors with high affinity in a low nanomolar range. The nature of these receptors on OK cells still needs to be identified.

In conclusion, we have shown that megalin/cubilin coreceptors do not play a significant role in the renal uptake of radiolabeled anti-HER2 affibody molecules in mice. Experiments with opossum tubular OK cells indicated that binding of the anti-HER2 affibody molecule ^99mTc–Z_HER2:2395-Cys is mediated by scavenger receptors cross-reacting with the binding site of the anti-HER2 affibody molecules, but it is unlikely that HER2 is one of them. We are considering approaches to further identify these receptor types. The results of this study indicate that affibody molecules and other scaffold-based targeting proteins with a relatively low kidney uptake can be selected using in vitro studies with tubular kidney cells.

Footnotes

Acknowledgment

This work was supported by a grant from the Swedish Research Council (Vetenskapsrådet). The authors express their gratitude to Affibody AB for providing affibody molecules for this study. Affibody AB holds intellectual property rights and trademarks for affibody molecules. A. Orlova is a member of the Scientific Advisory Board of Affibody AB.

Disclosure Statement

There are no existing financial conflicts.

References

Presman

. The development and use of radiolabeled antitumor antibodies. Cancer Res, 1980; 40:2960.

Pool

, Krenning

, Koning

et al. Preclinical and clinical studies of peptide receptor radionuclide therapy. Semin Nucl Med, 2010; 40:209.

van Essen

, Krenning

, Kam

et al. Peptide-receptor radionuclide therapy for endocrine tumors. Nat Rev Endocrinol, 2009; 5:382.

Dijkgraaf

, Boerman

. Radionuclide imaging of tumor angiogenesis. Cancer Biother Radiopharm, 2009; 24:637.

Behr

, Goldenberg

, Becker

. Reducing the renal uptake of radiolabeled antibody fragments and peptides for diagnosis and therapy: Present status, future prospects and limitations. Eur J Nucl Med, 1998; 25:201.

Valkema

, Pauwels

, Kvols

et al. Long-term follow-up of renal function after peptide receptor radiation therapy with 90Y-DOTA0,Tyr3-Octreotide and 177Lu-DOTA0,Tyr3-Octreotate. J Nucl Med, 2005; 46:83S.

Verroust

, Christensen

. Megalin and cubilin–the story of two multipurpose receptors unfold. Nephrol Dial Transplant, 2002; 17:1867.

de Jong

, Barone

, Krenning

et al. Megalin Is Essential for Renal Proximal Tubule Reabsorption of 111In-DTPA-Octreotide. J Nucl Med, 2005; 46:1696.

Vegt

, Eek

, Oyen

et al. Albumin-derived peptides efficiently reduce renal uptake of radiolabelled peptides. Eur J Nucl Med Mol Imaging, 2010; 37:226.

10.

Melis

, Krenning

, Bernard

et al. Localisation and mechanism of renal retention of radiolabelled somatostatin analogues. Eur J Nucl Med Mol Imaging, 2005; 32:1136.

11.

Vegt

, Melis

, Eek

et al. Renal uptake of different radiolabelled peptides is mediated by megalin: SPECT and biodistribution studies in megalin-deficient mice. Eur J Nucl Med Mol Imaging, 2011; 38:623.

12.

Vegt

, de Jong

, Wetzels

JFM

et al. Renal toxicity of radiolabeled peptides and antibody fragments: Mechanisms, impact on radionuclide therapy, and strategies for prevention. J Nucl Med, 2010; 51:1049.

13.

Gotthardt

, van Eerd-Vismale

, Oyen

et al. Indication for different mechanisms of kidney uptake of radiolabeled peptides. J Nucl Med, 2007; 48:596.

14.

Nygren

. Alternative binding proteins: Affibody binding proteins developed from a small three-helix bundle scaffold. FEBS J, 2008; 275:2668.

15.

Löfblom

, Feldwisch

, Tolmachev

et al. Affibody molecules: Engineered proteins for therapeutic, diagnostic and biotechnological applications. FEBS Lett, 2010; 584:2670.

16.

Ahlgren

, Tolmachev

. Radionuclide molecular imaging using Affibody molecules. Curr Pharm Biotechnol, 2010; 11:581.

17.

Baum

, Prasad

, Müller

et al. Molecular imaging of HER2-expressing malignant tumors in breast cancer patients using synthetic ¹¹¹In- or ⁶⁸Ga-labeled affibody molecules. J Nucl Med, 2010; 51:892.

18.

Tolmachev

, Orlova

, Pehrson

et al. Radionuclide therapy of HER2-positive microxenografts using a 177Lu-labeled HER2-specific Affibody molecule. Cancer Res, 2007; 67:2773.

19.

Ekblad

, Tran

, Orlova

et al. Development and preclinical characterisation of 99mTc-labelled Affibody molecules with reduced renal uptake. Eur J Nucl Med Mol Imaging, 2008; 35:2245.

20.

Orlova

, Tran

, Ekblad

et al. (186)Re-maSGS-Z (HER2:342), a potential Affibody conjugate for systemic therapy of HER2-expressing tumours. Eur J Nucl Med Mol Imaging, 2010; 37:260.

21.

Wållberg

, Orlova

, Altai

et al. Molecular design and optimization of ^99mTc-labeled recombinant affibody molecules improves their biodistribution and imaging properties. J Nucl Med, 2011; 52:461.

22.

Ahlgren

, Orlova

, Rosik

et al. Evaluation of maleimide derivative of DOTA for site-specific labeling of recombinant affibody molecules. Bioconjug Chem, 2008; 19:235.

23.

Ahlgren

, Wållberg

, Tran

et al. Targeting of HER2-expressing tumors using a site-specifically ^99mTc-labeled recombinant Affibody molecule Z_HER2:2395 with C-terminal engineered cysteine. J Nucl Med, 2009; 50:781.

24.

Tran

, Ekblad

, Orlova

et al. Effects of lysine-containing mercaptoacetyl-based chelators on the biodistribution of ^99mTc-labeled anti-HER2 Affibody molecules. Bioconjugate Chem, 2008; 19:2568.

25.

Orlova

, Magnusson

, Eriksson

et al. Tumor imaging a using picomolar affinity HER2 binding Affibody molecule. Cancer Res, 2006; 66:4339.

26.

Barone

, van der Smissen

, Devuyst

et al. Endocytosis of the somatostatin analogue, octreotide, by the proximal tubule-derived opossum kidney (OK) cell line. Kidney Int, 2005; 67:969.

27.

Gekle

, Knaus

, Nielsen

et al. Transforming growth factor-β1 reduces megalin- and cubilin-mediated endocytosis of albumin in proximal-tubule-derived opossum kidney cells. J Physiol, 2003; 552:471.

28.

Cihlo

, Melicharová

, Petrik

et al. Comparison of ¹¹¹In-DOTA-NOC and ¹¹¹In-DOTA-TATE distribution in the target and dose-limiting tissues: Conflicting results in vitro and in vivo. Anticancer Res, 2008; 28:2189.

29.

Ahlgren

, Andersson

, Tolmachev

. Kit formulation for ^99mTc-labeling of recombinant anti-HER2 Affibody molecules with a C-terminally engineered cysteine. Nucl Med Biol, 2010; 37:539.

30.

Leheste

, Rolinski

, Vorum

et al. Megalin knockout mice as an animal model of low molecular weight proteinuria. Am J Pathol, 1999; 155:136.

31.

Wållberg

, Orlova

. Slow internalization of anti-HER2 synthetic affibody monomer ¹¹¹In-DOTA-Z_HER2:342-pep2: Implications for development of labeled tracers. Cancer Biother Radiopharm, 2008; 23:435.

32.

Birn

, Fyfe

, Jacobsen

et al. Cubilin is an albumin binding protein important for renal tubular albumin reabsorption. J Clin Invest, 2000; 105:1353.

33.

Behr

, Becker

, Sharkey

et al. Reduction of renal uptake of monoclonal antibody fragments by amino acid infusion. J Nucl Med, 1996; 37:829.

34.

de Jong

, Rolleman

, Bernard

et al. Inhibition of renal uptake of indium-111-DTPA-octreotide in vivo. J Nucl Med, 1996; 37:1388.

35.

Gainkam

, Caveliers

, Devoogdt

et al. Localization, mechanism and reduction of renal retention of technetium-99m labeled epidermal growth factor receptor-specific nanobody in mice. Contrast Media Mol Imaging, 2011; 6:85.

36.

Melenhorst

WBWH

, Mulder

, Xi

et al. Epidermal growth factor receptor signaling in the kidney. Key roles in physiology and disease. Hypertension, 2008; 52:987.

37.

Orlova

, Tolmachev

, Pehrson

et al. Synthetic Affibody molecules: A novel class of affinity ligands for molecular imaging of HER2-expressing malignant tumors. Cancer Res, 2007; 67:2178.

38.

Malmberg

, Tolmachev

, Orlova

. Imaging agents for in vivo molecular profiling of disseminated prostate cancer. Cellular processing of ¹¹¹In-labeled CHX-A′′-DTPA-trastuzumab and anti-HER2 ABY-025 Affibody molecule by prostate cancer cell lines. Exp Ther Med, 2011; 2:523.

39.

Tran

, Rosik

, Abrahmsén

et al. Design, synthesis and biological evaluation of a multifunctional HER2 specific Affibody molecule for molecular imaging. Eur J Nucl Med Mol Imaging, 2009; 36:1864.

40.

Altai

, Wållberg

, Orlova

et al. Order of amino acids in C-terminal cysteine-containing peptide-based chelators influences biodistribution of ^99mTc-labeled recombinant Affibody molecules. Amino Acids, 2012; 42:1975.

41.

Ekerljung

, Steffen

, Carlsson

, Lennartsson

. Effects of HER2-binding affibody molecules on intracellular signaling pathways. Tumour Biol, 2006; 27:201.

In Vivo and In Vitro Studies on Renal Uptake of Radiolabeled Affibody Molecules for Imaging of HER2 Expression in Tumors

Abstract

Introduction

Materials and Methods

Animal experiments

Cell experiments

Binding kinetics: cell-associated radioactivity as a function of the incubation time

Binding of 99mTc–ZHER2:2395-Cys to OK cells at different conditions

Competitive binding assay, IC50 determination

Receptor expression and affinity, Scatchard analysis

Cellular processing of 99mTc–ZHER2:2395-Cys by OK cells

Data analysis and statistics

Results

Animal experiments

In vitro characterization

Discussion

Footnotes

Acknowledgment

Disclosure Statement

References

Binding of ^99mTc–Z_HER2:2395-Cys to OK cells at different conditions

Competitive binding assay, IC₅₀ determination

Cellular processing of ^99mTc–Z_HER2:2395-Cys by OK cells