1. Introduction
Affibody molecules are a class of small—58 residues and 7 kDa—non-immunoglobulin affinity proteins that have been selected to bind with high affinity to a wide range of different protein targets such as epidermal growth factor receptor (EGFR) [
1], human epidermal growth factor receptor 2 (HER2) [
2], human epidermal growth factor receptor 3 (HER3), [
3] insulin-like growth factor 1 receptor (IGF1R) [
4], and carbonic anhydrase IX (CAIX) [
5].
The Affibody scaffold was originally based on a stability enhanced variant of the B-domain, called the Z-domain, from the IgG-binding protein A of
Staphylococcus aureus. The three-helical bundle and cysteine-free Affibody scaffold has rapid folding kinetics, high solubility, and a relatively high thermal stability, which makes the Affibody technology an attractive research tool for biotechnological and pharmaceutical applications [
6,
7]. One of the therapeutically relevant Affibody molecules binds to the HER2 receptor, which is associated with an aggressive form of breast cancer. This Affibody molecule has been extensively studied as a tracer for molecular imaging of HER2-overexpressing tumors (e.g., by PET), and as therapeutic agent for the targeted delivery of cytotoxic drugs or radionuclides. When used for therapeutic purposes, the HER2-binding Affibody has often been produced as a dimeric construct, to increase the apparent binding affinity to HER2 [
2,
8]. The aim of this current study was to stabilize a dimeric anti-HER2-Affibody against proteolytic degradation and to enhance the thermal stability by taking inspiration from natural head-to-tail circulated proteins.
Ribosomally synthesized and backbone-cyclized proteins have been identified in mammals, plants, fungi, and bacteria, and are often involved in the host defense system. Compared to ordinary linear proteins, head-to-tail cyclized proteins are characterized by their often-high thermal stability, resistance towards many proteases, and stability over broad pH ranges [
9]. The most well-studied group of cyclic peptides/proteins are the small (~30 residues) and ultra-stable plant cyclotides, which are produced in large quantities by certain plant families in defense against pests and pathogens. Cyclic bacteriocins, antimicrobial proteins produced by Gram-negative bacteria, are another group of backbone-cyclized proteins with exceptional thermal and pH stability. One example is the pore-forming cyclic bacteriocin Enterocin AS-48 from
Enterococcus faecalis which has a compact and globular five-helix fold. The highly basic AS-48 has one of the highest reported thermal unfolding temperatures reported in the literature, 102 °C at pH 2.5 and low ionic strength [
10]. In contrast to many eukaryotic cyclic proteins, AS-48 does not contain intramolecular disulfide bonds, but backbone cyclization is essential for the correct fold because the cyclization site is situated in the middle of a helix [
9].
The relative simplicity of tying the peptide backbone together, and the potential large gain in stability that can be achieved, has sparked interest in circular proteins for industrial, biopharmaceutical, and biotechnological applications. Head-to-tail cyclization by chemical methods, such as native chemical ligation [
11], has proven difficult for larger proteins and can suffer by low yields of the circular product even for shorter peptides [
12]. Intein- [
13] and enzymatically based methods, i.e., using sortase A [
14] and butelase-1 [
15], have emerged as efficient ways to generate cyclic peptides and proteins [
16]. Sortase A was originally isolated from
Staphylococcus aureus where it functions to anchor proteins containing an LPXTG-recognition sequence (where X is any amino acid) to penta-glycine cross-bridges in the peptidoglycan cell wall. Recombinantly expressed sortase A has become an important tool for protein engineering, and the enzyme has been used to covalently conjugate proteins to other biomolecules, small synthetic molecules and to surfaces (reviewed in [
17,
18]). Broder and coworkers were the first group to describe sortase A-mediated backbone-cyclization of a recombinant protein when they incubated a bifunctional GFP-variant, containing both N-terminal glycines and a C-terminal LPETG recognition sequence, with the enzyme [
19]. A similar approach was used by Ploegh et al. to cyclize GFP with ≥90% conversion efficiency, and the authors noted that the intra-molecular transpeptidation reaction catalyzed by sortase A is remarkably similar to the cyclization step in the biosynthesis of cyclotides [
15]. Since then, sortase A has been used to catalyze head-to-tail cyclization of short SPPS-synthesized peptides [
20], cytokines [
21], human growth hormone [
22], and even to cyclize proteins in vivo in
Saccharomyces cerevisiae and human HEK293T cells [
23].
Intramolecular crosslinking and backbone cyclization of Affibody molecules and Z-domain variants have previously been investigated by our group and others [
24,
25,
26]. An intramolecular thioether bond was, for example, shown to thermally stabilize a solid-phase synthesized and monomeric HER2-binding Affibody by 10 °C [
24], and in a chemically synthesized 2-helix variant of the Z-domain, the introduction of a native peptide bond connecting the N-terminus to the C-terminus improved the ability of the protein to refold following thermal denaturation [
25].
In this proof-of-principle study, we have produced and investigated a backbone-cyclized ZHER2:342-dimer produced using sortase A mediated ligation. We designed a dimeric but single-chain construct with two ZHER2-domains connected by a 15-residue linker (–EFGSGSGSCPGSGGG–) containing a unique cysteine residue. The cysteine can be used to label the dimeric constructs with thiol-reactive compounds, such as maleimide-derivatized fluorophores and radiometal chelators, or to immobilize the construct to surfaces or beads. The construct has N-terminal glycines and a C-terminal sortase A recognition sequence, SR, making the dimer a substrate for intramolecular sortase A-catalyzed cyclization. Upon sortase A treatment, this protein construct is designed to form a head-to-tail cyclized 146-residue protein. This dimeric and cyclic ZHER2:342 protein was biophysically characterized using MALDI-MS, surface plasmon resonance (SPR), and circular dichroism (CD) and compared to its linear counterpart. To assess whether backbone-cyclization protects the dimer from proteolytic degradation, the linear and cyclic proteins were treated with both exopeptidases and endopeptidases in separate experiments. Labeling of the unique cysteine residue with a DyLight 594 maleimide dye was performed in order to assess the binding of the cyclic protein to HER2-expressing cells by fluorescence microscopy.
2. Materials and Methods
2.1. Construction of Expression Plasmids, Expression and Purification of Recombinant Proteins
The plasmid coding for the dimeric Z
HER2:342 protein was made in two steps. Firstly, the DNA sequence of Z
HER2:342 was amplified from pAY430-Z
HER2:342-SR-H
6 [
27] using PCR primers that introduced three N-terminal glycines and a C-terminal peptide linker containing a unique cysteine residue followed by BamHI and AccI restriction sites (5′-CCATCCATATGGGCGGTGGCGTAGATAACAAATTCAACAAAGAAATGC-3′ and 5′CGATGGTCTACCGTCCGGATCCCGGGCAAGATCCAGATCCAGAGCCGAATTCTTTCGGCGCCTGAG-3′).
The resulting PCR product was cleaved using NdeI and AccI and subcloned into the pAY430-SR-H6 plasmid cut likewise. To introduce the second Affibody molecule, the ZHER2:342 sequence was amplified a second time from pAY430-ZHER2:342-SR-H6. This time, PCR primers that introduced a 5′-BamHI-site, followed by three N-terminal glycines and a 3′-AccI site (5´- CCATCGGATCCGGCGGTGGCGTAGATAACAAATTCAACAAAGAAATGC-3′ and 5´- GTCTACGTCTACTTTCGGCGCCTGAG-3′), were used. This PCR product was cleaved with BamHI and AccI and subcloned into the plasmid cut likewise, resulting from the first step of the cloning procedure. The final plasmid, pAY430-G3-ZHER2:342-Cys-ZHER2:342-SR-H6, codes for a dimeric ZHER2:342-variant with the following amino acid sequence:
GGGVDNKFNKEMRNAYWEIALLPNLNNQQKRAFIRSLYDDPSQSANLLAEAKKLNDAQAPKEFGSGSGSCPGSGGGVDNKFNKEMRNAYWEIALLPNLNNQQKRAFIRSLYDDPSQSANLLAEAKKLNDAQAPKVDGSGSGSLPETGGHHHHHH The sequence of the final DNA construct was verified using sequencing (Microsynth, Göttingen, Germany).
Expression and subsequent IMAC purification of (Z
HER2)
2:L-Cys was performed following standard protocols for His
6-tagged proteins [
27]. After IMAC purification, the imidazole-containing elution buffer was changed into 10 mM NaOAc, pH 3.6 using PD-10 desalting columns (GE Healthcare, Uppsala, Sweden), and the protein was then lyophilized to a dry protein powder. Matrix-assisted laser desorption ionization mass spectroscopy (MALDI-MS) on a MALDI TOF/TOF analyzer (Sciex, Framingham, MA, USA) and electrospray ionization-mass spectrometry (ESI-MS) (Thermo Ultimate 3000, Thermo Fisher Scientific, Waltham, MA, USA, and Bruker Impact II, Bruker Daltonics, Billerica, MA, USA) were used to verify the molecular weight of the IMAC-purified (Z
HER2)
2:L-Cys. The sortase A-variant used to catalyze the cyclization reaction, P94S/D160N/K196T-sortase A, or sortase A
3* in short, was expressed using a mutated variant of the pGBMCS-SortA vector provided by Addgene (Addgene plasmid no. 21931) [
28], and purified as previously described by our group [
29].
2.2. Template-Based Structure Prediction
A homology model of the linear dimer was generated by submitting the amino acid sequence to the fully automated SWISS-MODEL workspace on the ExPASy server. Available online:
https://swissmodel.expasy.org (accessed on 26 May 2020) [
30].
2.3. Reduction, Capping and Purification of (ZHER2)2:L
Lyophilized protein powder (10 mg) was dissolved in 1.5 mL of 200 mM ammonium bicarbonate, 3 M guanidine hydrochloride (GdnHCl); pH 8.5. The reducing agent, tris(2-carboxyethyl)phosphine (TCEP), was added to a final concentration of 50 mM from a frozen stock solution (500 mM TCEP in water, pH 7.0) to reduce any intermolecular disulfide bridges, and the sample was left to reduce on a thermo block set at 55 °C for 1 h. Following reduction, the protein denaturant GdnHCl was removed using a PD-10 column equilibrated with 200 mM ammonium bicarbonate, pH 8.5, and TCEP to a final concentration of 10 mM was immediately added to the eluted protein to maintain the cysteines in reduced condition. To cap the reduced cysteines, iodoacetamide (IAA) was added to a final concentration of 17 mM from a newly prepared stock solution (375 mM IAA in 200 mM ammonium bicarbonate, pH 8.5). After a 30 min reaction time, at RT and shielded from light, excess iodoacetamide was removed using PD-10 columns equilibrated in sortase A reaction buffer without Ca2+ (50 mM HEPES, 150 mM NaCl, pH 7.5). Half of the carbamidomethylated protein batch (~5 mg) was purified using reversed phase-high performance liquid chromatography (RP-HPLC) on a Zorbax C18 column (300SB-C18, 9.4 × 250 mm, 5 μm particle size; Agilent, Santa Clara, CA, USA) with an elution gradient rising from 25 to 55% B in 30 min (A: 0.1% trifluoroacetic acid (TFA) in milli-Q water; B: 0.1% TFA in acetonitrile) and a flow rate of 3 mL/min. The molecular weight of the iodoacetamide-reacted and linear (ZHER2)2:L was verified using MALDI-MS and ESI-MS. HPLC-fractions containing the correct product were pooled and lyophilized.
2.4. Optimization of Sortase A-Mediated Head-to-Tail Cyclization
Reducing SDS-PAGE analysis was used to evaluate the optimal protein and enzyme concentrations for intramolecular head-to-tail cyclization. In an effort to minimize the formation of higher molecular weight species, (ZHER2)2:L was diluted to a concentration range between 68 and 1.7 μM, while the sortase A3* concentration was kept at 5 μM. After a 30 min reaction at 37 °C, samples were added to a 15-well NuPAGE 4–12% Bis-Tris gel (Life Technologies, Carlsbad, CA, USA). The gel was run under reducing conditions and stained using GelCodeTM Blue Safe Protein Stain (Pierce, Rockford, IL, USA) according to the manufacturer’s instructions. The optimal (ZHER2)2:L concentration was chosen as the highest protein concentration that did not produce higher molecular weight species visible on a Coomassie-stained SDS-PAGE gel.
In the next step, the concentration of (ZHER2)2:L was maintained constant at 1.7 μM, while the concentration of sortase A3* was altered between 0.08 and 5 μM. The samples were left at 37 °C for 20–24 h. Approximately 2.5 μg (ZHER2)2:L from each sample was applied to an SDS-PAGE gel, and the gel was run, stained, and inspected as described for the previous gel.
2.5. Cyclization and Purification of (ZHER2)2:C
In a typical cyclization reaction, 300 nmol (~5 mg) of alkylated (Z
HER2)
2:L was diluted to 1.7 μM in 176 mL sortase A reaction buffer containing 3 μmol NiCl
2. The cyclization reaction was initiated by the addition of sortase A
3* to a final concentration of 0.63 μM, and the reaction was left to proceed at 37 °C for 17–20 h. Following cyclization, the protein sample was concentrated, and buffer was exchanged using Amicon Ultra-15 filtration units with a 3 kDa cut-off (Merck Millipore, Darmstadt, Germany) down to ~10 mL in 50 mM HEPES, 150 mM NaCl; pH 7.5. The cyclic protein was purified using a “reversed IMAC step” on a HisPur Cobalt resin (Pierce, Rockford, IL, USA) equilibrated with 50 mM HEPES, 150 mM NaCl; pH 7.5. Unreacted linear protein and His-tagged sortase A
3* binds to the column matrix, while cyclized protein, lacking the C-terminal His
6-tag, is eluted in the flow through. Fractions absorbing at 280 nm were pooled and buffer was exchanged to 10 mM NaOAc, pH 3.6 using PD-10 columns before protein lyophilization. The cyclic protein was purified using RP-HPLC, using the same conditions as previously described for the linear (Z
HER2)
2:L protein but with a slightly different gradient (35–45% B over 20 min). For representative chromatograms of the RP-HPLC purification, see
Figure S1. The molecular weight of the purified cyclic iodoacetamide-capped Z
HER2-dimer, hereafter called (Z
HER2)
2:C, was verified using MALDI-MS, and HPLC fractions containing the correct product were pooled and lyophilized.
SDS-PAGE analysis was used to determine the efficiency of the sortase A-mediated cyclization reaction, and the apparent molecular weight of the linear and cyclic IAA-capped Z
HER2-dimers. The samples were mixed with 5.2 µL glycerol and 4 µL 5× loading buffer containing β-mercaptoethanol, and 1× sortase reaction buffer was added to reach a volume of 20 µL. PageRuler
TM Plus (Thermo Fischer Scientific, Vilnius, Lithuania) pre-stained protein ladder was used as molecular weight standard. The samples contained approximately 2.5 µg of protein and were heated for 5 min at 95 °C and loaded on to a 10-well NuPAGE 4–12% BT 1.0 gel (Life Technologies, Carlsbad, CA, USA). The gel was run in an Novex Mini-Cell (Life Technologies, Carlsbad, CA, USA). for 35 min at 180 V in 1× MES buffer. The gel was stained with GelCode
TM Blue Safe Protein Stain (Pierce, Rockford, IL, USA) and scanned. Gels used for protein band intensity measurements were color-adjusted using the Black & White adjustment tool in Adobe Photoshop 2021 (Adobe), and protein band pixel intensities were analyzed using the open-source software ImageJ, version 1.8.0_172 Available online:
http://imagej.nih.gov (accessed at 10 May 2021). The relative mobility (Rf; migration distance of the protein/migration distance of the dye front) of the proteins in the molecular weight standard was plotted against the logarithm of their molecular weights, and a linear equation was fitted to the data points using Excel for Mac, version 16.48 (Microsoft, Redmond, WA, USA). The linear calibration curve was used to determine the apparent molecular weights for the IAA-capped linear and cyclic Z
HER2-dimers.
2.6. Production and Purification of (ZHER2)2:C-DL594
A mass of 5 mg of (Z
HER2)
2:L-Cys was cyclized and purified using the optimized protocol outlined for (Z
HER2)
2:L in
Section 2.5. above, with minor changes described here. To keep the free cysteine in (Z
HER2)
2:L-Cys in a reduced state, 10 mM of TCEP was added to the sortase reaction buffer during the 22 h cyclization reaction. Following cyclization, the protein was concentrated on Amicon Ultra-15 filtration units with 10 kDa MWCO (Merck), and after a “reversed IMAC” treatment, as described in
Section 2.5., the cyclic Z
HER2 dimer, (Z
HER2)
2:C-Cys, was buffer-exchanged to 10 mM NaOAc, pH 3.6 using PD-10 columns and lyophilized to a dry protein powder.
For maleimide labeling, 1 mg (~60 nmol) of lyophilized protein was dissolved in 0.5 mL of 200 mM ammonium bicarbonate, 3 M GdnHCl; pH 8.5 with 50 mM TCEP, and the sample was left to reduce on a thermo block set at 55 °C for 1 h. The reduction buffer was removed using a NAP-5 column (GE healthcare, Uppsala, Sweden), and the protein was eluted in 1 mL of 20 mM Tris buffer, pH 7.0, containing 10 mM TCEP. One milligram of DyLight 594 maleimide fluorescent dye (Pierce, Rockford, IL, USA) was dissolved in 100 μL of
N,
N-dimethylformamide (DMF) and added to the reduced protein at 18 molar excess relative to the sulfhydryl group in the protein. Following a 17 h reaction at room temperature, the sample was dialyzed in a dialysis tube with a 6–8 kDa MWCO (Spectra/Por Membrane; Spectrum Laboratories Inc., Rancho Dominguez, CA, USA), against 4 L of 10 mM potassium phosphate buffer with 15 mM KCl, pH 7.2 at 4 °C for 24 h. A final RP-HPLC step using the same HPLC setup described in
Section 2.3. but using a slightly different gradient (30–45% B over 15 min) was added to remove remaining unreacted fluorescent dye and sortase A from the cyclic DyLight 594-labeled Z
HER2-dimer. For a representative chromatogram of the RP-HPLC purification, see
Figure S2. The molecular weight of (Z
HER2)
2:C-DL594 was verified using MALDI-MS, and HPLC-fractions containing the correct product were combined and lyophilized. The protein concentration and the degree of labeling were estimated by measuring the absorbance at 280 nm and 593 nm, following the manufacturer’s protocol.
2.7. Circular Dichroism Spectroscopy
Circular dichroism (CD) studies of the linear and cyclic proteins were performed using a Chirascan CD spectrometer (Applied Photophysics, Leatherhead, UK) equipped with a Peltier temperature-controlled cuvette holder and a thermal sensor that records the temperature in the sample cuvette. Both proteins were dissolved in 20 mM potassium phosphate buffer with 100 mM KCl (pH 7.4) at a concentration of about 0.2 mg/mL, and all CD measurements were performed in a capped quartz cuvette with a 0.1 cm path length. Far-UV spectra of the proteins were recorded by scanning the ellipticity in the 195–260 range, and were converted to molar ellipticity before data processing. The mean residue ellipticity (MRE) is given by:
where θ
obs is the measured ellipticity in degrees, l is the cuvette path length in centimeters, c is the protein concentration in molar quantities, and n is the number of amino acids in each protein. The fraction helix F
H was calculated from the mean residue ellipticity at 222 nm, MRE
222, by the method of Scholtz et al. [
31]:
Complete helix [θ]
H and complete random coil [θ]
C are expressed in deg cm
2 dmol
−1, and are given by:
T is the temperature expressed in °C, and n is the number of residues in the protein [
31]. Thermal unfolding was monitored by recording the ellipticity at 222 nm, θ
222, in millidegrees as a function of temperature. The temperature was increased between 20 and 90 °C in 0.1 °C increments at a speed of 5 °C/min.
Fraction folded (F
N) as a function of temperature was calculated using the following equation, assuming a two-state unfolding behavior:
where θ
N and θ
D are the ellipticity at 222 nm, in millidegrees, of the native and the denatured state, respectively.
After denaturation, the temperature in the cuvette was returned to 20 °C and the far-UV spectra of the proteins were rescanned to ensure folding reversibility.
2.8. Surface Plasmon Resonance (SPR)-Based Binding Analysis
The binding kinetics of the interactions between the cyclic and linear proteins and HER2 were analyzed on a Biacore T200 system (GE Healthcare, Uppsala, Sweden). A carboxymethylated, dextran-coated CM5 chip was activated with EDC/NHS and HER2-Fc (Her2/ERBB2 Protein, Human, Recombinant (hFc Tag); Sino Biological Inc., Beijing, China) was immobilized on the surface with an immobilization level of 760 RU. The analytes were diluted in phosphate-buffered saline with 0.5% Tween-20 (PBS-T), pH 7.4, to 5 different concentrations, 500, 167, 56.6, 18.5 and 6.2 nM, and were allowed to flow over the surfaces at 50 µL/min in a single-cycle setup with a 5 min association phase and a final 120 min dissociation time for 2 replicates. The equilibrium dissociation constant (Kd) was determined from a 1-to-1 binding model.
2.9. Protease Digestion Assay
The difference in stability towards protease degradation of the cyclic and linear proteins was demonstrated with a mixture of the endopeptidases α-chymotrypsin (≥40 U mg−1) and trypsin (12,443 U mg−1), called pancreatin, and the exopeptidase carboxypeptidase A (70 U mg−1). All proteins were extracted from bovine pancreas and bought from Sigma-Aldrich (Saint Louis, MO, USA). The sample proteins were diluted in 0.1 M Tris-Cl (pH 8.5) to a concentration of 50 µM. The cleavage was measured after 0, 5, 15, 30 and 60 min. In the pancreatin assay, 54 µL protein was mixed with 102 µL of pancreatin in PBS (pH 7.4) to a final concentration of 33 µM protein, 0.4 µg/mL α-chymotrypsin and 0.9 µg/mL trypsin. The digestions were performed at 37 °C and the reactions were stopped by the addition of 9 µL 10% TFA. The samples were analyzed by RP-HPLC (1200 series, Agilent Technologies, San Diego, CA, USA) at 35 °C on a Zorbax C18 analytical column (300SB-C18, 4.5 × 15 mm, 3.5 µM particle size; Agilent, Santa Clara, CA, USA) with a flow rate of 1 mL/min over 20 min with a gradient of 0–80% B (A: 0.1% TFA in milli-Q water; B: 0.1% TFA in acetonitrile). The elution peaks were integrated and compared to the undigested sample at t = 0. The digestions were performed as technical triplicates, except for the 0 time point and the time points where 100% of the protein had been digested, where only one data point was gathered.
In the carboxypeptidase A assay, 10 µL of protein sample in 0.1 M Tris-Cl (pH 8.5) was mixed with 10 µL of carboxypeptidase A in 0.1 M Tris-Cl (pH 8.5) to a final concentration of 25 µM protein and 0.5 µM carboxypeptidase A. The reactions were stopped with the addition of 1 µL of 10% TFA, and the samples were analyzed by MALDI-MS (MALDI TOF/TOF analyzer, Sciex) to determine the amount of undigested protein.
2.10. Cell Culture and Treatment
A human ovarian cell line (SKOV-3) that presents a high expression of HER2 surface receptors and a human breast cancer cell line (MCF-7) with a low expression of HER2 receptor were cultured at 37 °C in a 5% CO2 humidified atmosphere in Dulbecco’s modified Eagle medium (DMEM) (Gibco, Thermo Fisher Scientific, Uppsala, Sweden) supplemented with 10% fetal bovine serum, 0.1 mM non-essential amino acids (Gibco, Thermo Fisher Scientific, Sweden), 2 mM L-glutamine (Gibco, Thermo Fisher Scientific, Sweden) and 1% penicillin/streptomycin (Gibco, Thermo Fisher Scientific, Sweden). Coverslips were added in a 12-well plate and treated with poly-L-lysine (Sigma-Aldrich, Stockholm, Sweden) (0.1 mg/mL) for 15 min at room temperature, and 100,000 SKOV-3 cells were seeded on the coverslips. After an overnight incubation, the media was removed and replaced with fresh DMEM containing (ZHER2)2:C (1.2 µM), and the cells were incubated for 5 min at 37 °C. (ZHER2)2:C-DL594 was then added to the cells (2.4 nM) which were then incubated at 37 °C for 1 h. A positive control was also prepared with cells incubated with only (ZHER2)2:C-DL594 (2.4 nM) for 1 h at 37 °C. The medium was then removed, and after 3 washes with PBS, a solution of 4% paraformaldehyde (PFA) in PBS (Alfa Aesar, Thermo Fisher Scientific, Sweden) was added to each well (15 min, room temperature) to fix the cells. The PFA solution was removed, and the cells were washed once with PBS. All the cells were then treated with 4′,6-diamidino-2-phenylindole (DAPI) (Roche, Merck, Sweden) (1 µg/mL, 5 min, room temperature) for nuclear staining. After a wash with PBS, the coverslips were then collected and mounted on a microscopy slide before sealing them with nail polish.
On an 8-well microscopy slide (Millicell EZ, Merck, Sweden), 20,000 SKOV-3 cells were seeded and allowed to grow overnight. On the day of treatment, the media were removed and replaced with fresh DMEM containing trastuzumab (Herceptin®, Roche, Sweden) (1 mg/mL) and incubated for 5 min at 37 °C before the addition of (ZHER2)2:C-DL594 (2.4 nM) and another hour of incubation at 37 °C. The cells were then washed, fixed, and mounted in the same fashion as previously. On another 8-well microscopy slide, 20,000 MCF-7 cells were seeded and allowed to grow for 2 days. The medium was then replaced with fresh DMEM containing (ZHER2)2:C-DL594 (2.4 nM) and the cells were incubated for 1 h at 37 °C. The cells were then again washed, fixed, and mounted in the same fashion as the SKOV-3 cells.
All cells were visualized using a Nikon Plan Fluor 10× Ph1 DL optic (BergmanLabora AB, Danderyd, Sweden) and the images were captured with an Andor Zyla VSC-05780 camera (BergmanLabora AB, Danderyd, Sweden) coupled to the software NIS-Element. The images were then processed with ImageJ, version 1.8.0_172.