Design and In Vitro Evaluation of a Cytotoxic Conjugate Based on the Anti-HER2 Affibody Fused to the Fc Fragment of IgG1

In our previous work we demonstrated that a small protein called affibody can be used for a cytotoxic conjugate development. The anti-HER2 affibody was armed with one moiety of a highly potent auristatin E and specifically killed HER2-positive cancer cells with a nanomolar IC50. The aim of this study was to improve the anti-HER2 affibody conjugate by increasing its size and the number of conjugated auristatin molecules. The affibody was fused to the Fc fragment of IgG1 resulting in a dimeric construct with the molecular weight of 68 kDa, referred to as ZHER2:2891-Fc, ensuring its prolonged half-life in the blood. Due to the presence of four interchain cysteines, the fusion protein could carry four drug molecules. Notably, the in vitro tests of the improved anti-HER2 conjugate revealed that it exhibits the IC50 of 130 pM for the HER2-positive SK-BR-3 cells and 98 nM for the HER2-negative MDA-MB-231 cells. High efficacy and specificity of the auristatin conjugate based on ZHER2:2891-Fc indicate that this construct is suitable for further in vivo evaluation.


Introduction
Safe and efficient delivery of cytotoxic payloads to solid tumors remains one of the main challenges for advanced cancer therapies. An antibody-drug conjugate (ADC) consists of a monoclonal antibody that recognizes tumor-associated antigens covalently linked to a highly cytotoxic agent. This composition confers a limited systematic exposure to the drug minimizing adverse effects of the therapy [1,2]. Since the 1980s, many ADCs have undergone clinical trials and currently, more than 60 are being tested [3][4][5][6]. Nevertheless, only two ADCs, brentuximab vedotin (Adcetris, Seattle Genetics) against Hodgin's lymphoma and ado-trastuzumab emtansine (Kadcyla, Genentech) against metastatic breast cancer, were approved for marketing by the Food and Drug Administration (FDA) and are being used to treat patients.
In our previous study, we demonstrated that the anti-HER2 affibody ZHER2:2891, originally designed by Feldwisch et al. (2010) efficiently delivers a highly cytotoxic antimitotic agent, auristatin E, to HER2-positive cancer cells [19]. However, the small size of this conjugate (8647 Da) would result in its fast renal clearance as the molecules smaller than 25-40 kDa are rapidly removed from circulation via kidney filtration [22,33]. In order to expand the size of our conjugate and increase its half-life in blood, we decided to fuse the ZHER2891 affibody to the Fc fragment of IgG1 [34][35][36]. The resulting ZHER2:2891-Fc fusion homodimerizes via Fc has the molecular mass of 68 kDa, which should contribute to its retention in the circulation. Additionally, the Fc fragment ensures that the conjugate will be recycled from epithelial cells back to blood vessels via interaction with FcRn [37,38]. Moreover, the presence of two interchain disulphide bonds in the Fc part allowed us to conjugate four auristatin molecules to our dimeric construct ( Figure 1) whereas the previous ZHER2:2891-DCS-MMAE conjugate was loaded only with a single auristatin molecule. The cytotoxicity of our improved construct referred to as ZHER2:2891-Fc-MMAE, was evaluated on breast cancer cell lines.

Expression and Purification of the ZHER2:2891-Fc from Mammalian Cells
The conjugate developed here is based on the molecule named ZHER2:2891, which exhibits increased thermal stability and hydrophilicity along with lower liver uptake in animals when compared to the parental ZHER2:342 affibody [30,39]. To expand the size of our construct, we fused the ZHER2:2891 molecule to the Fc fragment of IgG1 ( Figure 1). A similar fusion was previously proposed by Rönnmark et al. (2002) who fused the Tag DNA polymerase specific affibody to the Fc fragment and purified this construct from E. coli cells [40]. However, the ZHER2:2891-Fc fusion protein was expressed in Chinese Hamster Ovary CHO-S cells to ensure its proper folding and glycosylation [41,42]. The highest levels of ZHER2:2891-Fc in the culture medium were observed since day 4 following the transfection with pLEV113-ZHER2:2891-Fc (Figure 2a). Therefore, ZHER2:2891-Fc was purified on day 5 or day 6 using a single-step affinity chromatography on protein A-Sepharose. The purification process was analyzed

Expression and Purification of the Z HER2:2891 -Fc from Mammalian Cells
The conjugate developed here is based on the molecule named Z HER2:2891 , which exhibits increased thermal stability and hydrophilicity along with lower liver uptake in animals when compared to the parental Z HER2:342 affibody [30,39]. To expand the size of our construct, we fused the Z HER2:2891 molecule to the Fc fragment of IgG1 ( Figure 1). A similar fusion was previously proposed by Rönnmark et al. (2002) who fused the Tag DNA polymerase specific affibody to the Fc fragment and purified this construct from E. coli cells [40]. However, the Z HER2:2891 -Fc fusion protein was expressed in Chinese Hamster Ovary CHO-S cells to ensure its proper folding and glycosylation [41,42]. The highest levels of Z HER2:2891 -Fc in the culture medium were observed since day 4 following the transfection with pLEV113-Z HER2:2891 -Fc ( Figure 2a). Therefore, Z HER2:2891 -Fc was purified on day 5 or day 6 using a single-step affinity chromatography on protein A-Sepharose. The purification process was analyzed by Western blotting with the anti-Fc antibody conjugated with HRP ( Figure 2b). The Z HER2:2891 -Fc fusion migrated as a 36-kDa band in SDS-PAGE in the presence of the reducing agent, β-mercaptoethanol (5%), in Laemmli sample buffer. Additionally, the identity and purity of the product were confirmed by mass spectrometry (MS) (Figure 2c). The detected m/z peaks corresponded to a covalent dimer (67,860.2 Da) and monomer (33,929.5 Da) of the Z HER2:2891 -Fc after cleavage of a 19 amino acid-long secretion signal peptide (2273.7 Da). We obtained 1.5 mg of the Z HER2:2891 -Fc protein from 1 liter of CHO-S culture. by Western blotting with the anti-Fc antibody conjugated with HRP ( Figure 2b). The ZHER2:2891-Fc fusion migrated as a 36-kDa band in SDS-PAGE in the presence of the reducing agent, β-mercaptoethanol (5%), in Laemmli sample buffer. Additionally, the identity and purity of the product were confirmed by mass spectrometry (MS) (Figure 2c). The detected m/z peaks corresponded to a covalent dimer (67,860.2 Da) and monomer (33,929.5 Da) of the ZHER2:2891-Fc after cleavage of a 19 amino acid-long secretion signal peptide (2273.7 Da). We obtained 1.5 mg of the ZHER2:2891-Fc protein from 1 liter of CHO-S culture.

Specificity of the ZHER2:2891-Fc Fusion Protein
Our previous study clearly demonstrated that ZHER2:2891 selectively recognized HER2 receptor present on the surface of cancer cell lines that overexpress this receptor [19]. Since the proposed fusion contains the same targeting molecule, we verified the specificity of ZHER2:2891-Fc towards two out of four cell lines that were further used to assess cytotoxicity of monomethylauristatin E (MMAE) coupled to ZHER2:2891-Fc. We chose the SK-BR-3 and MDA-MD-231 breast cancer cells as they express very high and low, physiological levels of HER2 receptor, respectively [19,43,44]. The ZHER2:2891-Fc was labeled with FITC and used for the immunostaining experiment. The fluorescent signal was detected on the surface of SK-BR-3 cells when ZHER2:2891-Fc-FITC or anti-HER2 antibody were used to recognize HER2 receptor (Figure 3a). In the case of the HER2-negative MDA-MB-231 cells, only a weak background signal was observed. We also analyzed ZHER2:2891-Fc binding to SK-BR-3 and MDA-MB-231 cells by Western blotting with anti-Fc antibody-HRP. The band corresponding to ZHER2:2891-Fc was present only in the case of the HER2-positive SK-BR-3 cells (Figure 3b). Taken together, these biochemical and microscopic results are consistent with the analysis of ZHER2:2891 binding to different cell lines carried out in our previous study [19] and confirm that the fusion protein specifically recognizes HER2 receptor.

Specificity of the Z HER2:2891 -Fc Fusion Protein
Our previous study clearly demonstrated that Z HER2:2891 selectively recognized HER2 receptor present on the surface of cancer cell lines that overexpress this receptor [19]. Since the proposed fusion contains the same targeting molecule, we verified the specificity of Z HER2:2891 -Fc towards two out of four cell lines that were further used to assess cytotoxicity of monomethylauristatin E (MMAE) coupled to Z HER2:2891 -Fc. We chose the SK-BR-3 and MDA-MD-231 breast cancer cells as they express very high and low, physiological levels of HER2 receptor, respectively [19,43,44]. The Z HER2:2891 -Fc was labeled with FITC and used for the immunostaining experiment. The fluorescent signal was detected on the surface of SK-BR-3 cells when Z HER2:2891 -Fc-FITC or anti-HER2 antibody were used to recognize HER2 receptor (Figure 3a). In the case of the HER2-negative MDA-MB-231 cells, only a weak background signal was observed. We also analyzed Z HER2:2891 -Fc binding to SK-BR-3 and MDA-MB-231 cells by Western blotting with anti-Fc antibody-HRP. The band corresponding to Z HER2:2891 -Fc was present only in the case of the HER2-positive SK-BR-3 cells (Figure 3b). Taken together, these biochemical and microscopic results are consistent with the analysis of Z HER2:2891 binding to different cell lines carried out in our previous study [19] and confirm that the fusion protein specifically recognizes HER2 receptor.

The ZHER2:2891-Fc-MMAE Conjugate
Reduction of the two interchain disulphide bonds between the Fc fragments of the fusion protein resulted in four -SH groups which reacted with the maleimide group present in the drug molecule with a high efficiency (~90%) (Figure 4a). In the MS spectrum recorded for the conjugate, two main m/z peaks were present. The major m/z peak (36,552.9 Da) corresponds to a monomeric ZHER2:2891-Fc loaded with two MMAE molecules whereas the second m/z peak (73,104.9 Da) represents a dimeric ZHER2:2891-Fc carrying up to four auristatin E (MW 1316.7 Da). Our analysis also revealed a trace of the ZHER2:2891-Fc monomer conjugated with one MMAE molecule (m/z = 35,236.2) (Figure 4b). The strong signal detected for the monomeric conjugate loaded with one or two MMAE can be attributed to the breakdown of the non-covalent conjugate dimer since the disulphide bridges were replaced with MMAE molecules. Such noncovalent dimer is prone to decomposition during sample preparation and exposure to a laser beam in the MS experiment. To confirm that the conjugate dimerizes, we subjected ZHER2:2891-Fc-MMAE to a polyacrylamide gel electrophoresis with a low concentration of SDS (0.0375%) in the running buffer under non-reducing conditions [45]. The conjugate migrated as a band of a molecular mass between 50 and 75 kDa, which demonstrates that the dimeric format of ZHER2:2891-Fc is preserved in its auristatin conjugate ( Figure 4c).

The Z HER2:2891 -Fc-MMAE Conjugate
Reduction of the two interchain disulphide bonds between the Fc fragments of the fusion protein resulted in four -SH groups which reacted with the maleimide group present in the drug molecule with a high efficiency (~90%) (Figure 4a). In the MS spectrum recorded for the conjugate, two main m/z peaks were present. The major m/z peak (36,552.9 Da) corresponds to a monomeric Z HER2:2891 -Fc loaded with two MMAE molecules whereas the second m/z peak (73,104.9 Da) represents a dimeric Z HER2:2891 -Fc carrying up to four auristatin E (MW 1316.7 Da). Our analysis also revealed a trace of the Z HER2:2891 -Fc monomer conjugated with one MMAE molecule (m/z = 35,236.2) (Figure 4b). The strong signal detected for the monomeric conjugate loaded with one or two MMAE can be attributed to the breakdown of the non-covalent conjugate dimer since the disulphide bridges were replaced with MMAE molecules. Such noncovalent dimer is prone to decomposition during sample preparation and exposure to a laser beam in the MS experiment. To confirm that the conjugate dimerizes, we subjected Z HER2:2891 -Fc-MMAE to a polyacrylamide gel electrophoresis with a low concentration of SDS (0.0375%) in the running buffer under non-reducing conditions [45]. The conjugate migrated as a band of a molecular mass between 50 and 75 kDa, which demonstrates that the dimeric format of Z HER2:2891 -Fc is preserved in its auristatin conjugate (Figure 4c).

The ZHER2:2891-Fc-MMAE Conjugate
Reduction of the two interchain disulphide bonds between the Fc fragments of the fusion protein resulted in four -SH groups which reacted with the maleimide group present in the drug molecule with a high efficiency (~90%) (Figure 4a). In the MS spectrum recorded for the conjugate, two main m/z peaks were present. The major m/z peak (36,552.9 Da) corresponds to a monomeric ZHER2:2891-Fc loaded with two MMAE molecules whereas the second m/z peak (73,104.9 Da) represents a dimeric ZHER2:2891-Fc carrying up to four auristatin E (MW 1316.7 Da). Our analysis also revealed a trace of the ZHER2:2891-Fc monomer conjugated with one MMAE molecule (m/z = 35,236.2) (Figure 4b). The strong signal detected for the monomeric conjugate loaded with one or two MMAE can be attributed to the breakdown of the non-covalent conjugate dimer since the disulphide bridges were replaced with MMAE molecules. Such noncovalent dimer is prone to decomposition during sample preparation and exposure to a laser beam in the MS experiment. To confirm that the conjugate dimerizes, we subjected ZHER2:2891-Fc-MMAE to a polyacrylamide gel electrophoresis with a low concentration of SDS (0.0375%) in the running buffer under non-reducing conditions [45]. The conjugate migrated as a band of a molecular mass between 50 and 75 kDa, which demonstrates that the dimeric format of ZHER2:2891-Fc is preserved in its auristatin conjugate (Figure 4c).

In Vitro Efficacy of the Z HER2:2891 -Fc-MMAE
Our previous conjugate based on a single Z HER2:2891 with the C-terminal extension, referred to as drug conjugation sequence (DCS), could carry only a single MMAE moiety. The IC 50 of this construct determined for SK-BR-3 cells, which overexpress high levels of HER2 receptor was~5 nM [19]. Since the Z HER2:2891 -Fc-MMAE contained two affibody molecules and was armed with four drug molecules, we had expected to observe a significant decrease in the IC 50 value. The cytotoxicity assays were carried out exactly the same way as for Z HER2:2891 -MMAE and free MMAE in our previous study [19]. In order to more precisely compare HER2 expression levels between used breast cancer cell lines: MDA-MB-231, T-47-D, MDA-MB-453 and SK-BR-3, we carried out a densitometric quantification on the anti-HER2 Western blots published previously [19]. According to our analysis, HER2 levels in MDA-MB-453 were three to five times higher than in T-47-D cells and about six to eight times lower than in SK-BR-3 cells. We did not detect the band corresponding to HER2 in MDA-MD-231, which confirms that this is a strong HER2-negative line. For clarity, we decided to divide the used cell lines

In Vitro Efficacy of the ZHER2:2891-Fc-MMAE
Our previous conjugate based on a single ZHER2:2891 with the C-terminal extension, referred to as drug conjugation sequence (DCS), could carry only a single MMAE moiety. The IC50 of this construct determined for SK-BR-3 cells, which overexpress high levels of HER2 receptor was ~5 nM [19]. Since the ZHER2:2891-Fc-MMAE contained two affibody molecules and was armed with four drug molecules, we had expected to observe a significant decrease in the IC50 value. The cytotoxicity assays were carried out exactly the same way as for ZHER2:2891-MMAE and free MMAE in our previous study [19]. In order to more precisely compare HER2 expression levels between used breast cancer cell lines: MDA-MB-231, T-47-D, MDA-MB-453 and SK-BR-3, we carried out a densitometric quantification on the anti-HER2 Western blots published previously [19]. According to our analysis, HER2 levels in MDA-MB-453 were three to five times higher than in T-47-D cells and about six to eight times lower than in SK-BR-3 cells. We did not detect the band corresponding to HER2 in MDA-MD-231, which confirms that this is a strong HER2-negative line. For clarity, we decided to divide the used cell lines  Table 1). Notably, these results show that we significantly improved in vitro efficacy of our new anti-HER2 affibody-based conjugate in comparison to the previous ZHER2:2891-DCS-MMAE construct (Table 1).

Stability of the Z HER2:2891 -Fc-MMAE Conjugate in Serum
The stability of Z HER2:2891 -Fc-MMAE in human and mouse sera was monitored by Western blotting with an anti-affibody antibody. The strong signal of the Z HER2:2891 -Fc-MMAE conjugate was detected for each day of the 7-day incubation both in human and mouse plasma (Figure 6a,b). This indicates that in the above-mentioned conditions our construct remains relatively stable since the levels of the conjugate detected by the anti-affibody antibody were comparable between different days of the incubation.

Stability of the ZHER2:2891-Fc-MMAE Conjugate in Serum
The stability of ZHER2:2891-Fc-MMAE in human and mouse sera was monitored by Western blotting with an anti-affibody antibody. The strong signal of the ZHER2:2891-Fc-MMAE conjugate was detected for each day of the 7-day incubation both in human and mouse plasma (Figure 6a,b). This indicates that in the above-mentioned conditions our construct remains relatively stable since the levels of the conjugate detected by the anti-affibody antibody were comparable between different days of the incubation.

Discussion
Antibodies are commonly used to generate new cytotoxic conjugates. These proteins are highly specific which ensures that the drug is delivered to target cells overexpressing tumor-associated antigens. However, due to the large size of these molecules (~150 kDa), their ability to penetrate solid tumors is limited [47]. Therefore, other macromolecules are also tested as drug carriers. The anti-HER2 affibodies labeled with radioisotopes have been widely investigated for molecular imaging

Discussion
Antibodies are commonly used to generate new cytotoxic conjugates. These proteins are highly specific which ensures that the drug is delivered to target cells overexpressing tumor-associated antigens. However, due to the large size of these molecules (~150 kDa), their ability to penetrate solid tumors is limited [47]. Therefore, other macromolecules are also tested as drug carriers. The anti-HER2 affibodies labeled with radioisotopes have been widely investigated for molecular imaging of HER2-positive tumors [31,[48][49][50]. These efforts have recently led to successful clinical trials of the second-generation anti-HER2 affibody loaded with 111 In and 68 Ga [51][52][53].
The anti-HER2 affibody armed with one MMAE molecule specifically targets HER2-positive cells and destroys them with a low nanomolar IC 50 [19]. Here, we improved this conjugate by fusing the Fc fragment of IgG1 to Z HER2:2891 to expand its size and increase the number of the coupled drug molecules (Figure 1). The obtained Z HER2:2891 -Fc construct run on SDS-PAGE gel under reducing conditions as a~36 kD-protein (Figures 2b and 3b) while the mass of the monomeric Z HER2:2891 -Fc determined by MS was 33929.5 Da (Figure 2c). The band shift towards higher molecular weight observed on the SDS-PAGE gel and the~1100 Da discrepancy between the apparent molecular mass of the Z HER2:2891 -Fc monomer (33,929.5 Da) and its calculated molecular mass (32,833 Da) suggest that similar to other Fc fusions [41,54], the Fc fragment was glycosylated in CHO-S cells. Combining the anti-HER2 affibody with Fc allowed us to expand the molecular mass of the affibody from~7 kDa tõ 70 kDa, which is still about half of the antibody size. Such size should contribute to a slower kidney removal of our construct since the kidney filtration cutoff is around 40-60 kDa [33]. Additionally, the Fc domain prolongs the serum half-life of antibodies and Fc-fusion proteins due to pH-dependent binding to the neonatal Fc receptor (FcRn) which protects the proteins from endosomal degradation in the epithelial cells. The Fc-fusion bound to FcRn is then recycled back into circulation [34,36,38]. The ability of the Fc region to improve pharmacokinetics has been verified for numerous proteins including eleven Fc fusions approved by FDA till 2017 [55].
The Z HER2:2891 -Fc construct specifically recognized HER2-positive cells, which was confirmed by microscopic images and a biochemical binding assay (Figure 3). The presence of Fc allowed us to conjugate MMAE molecules to four cysteines obtained by treating Z HER2:2891 -Fc with a reducing agent to break interchain disulphide bonds prior to the conjugation process, as it has been described for many ADCs [56][57][58]. According to the analysis of mass spectra, one or two MMAE molecules were coupled to a single Z HER2:2891 -Fc which resulted in up to four MMAE molecules per dimeric Z HER2:2891 -Fc (Figure 4b). The Z HER2:2891 -Fc-MMAE dimer underwent decomposition to monomer during the MS sample preparation and measurement (Figure 4b). However, the dimeric format of the conjugate was preserved during electrophoretic separation with a low amount of SDS in the running buffer (Figure 4c). This indicates that in physiological conditions Z HER2:2891 -Fc-MMAE should remain dimeric.
We obtained the following IC 50 Table 1). This translates into a significant increase in the sensitivity of HER2-positive cells towards Z HER2:2891 -Fc-MMAE in comparison with their sensitivity to Z HER2:2891 -DCS-MMAE loaded with a single auristatin E molecule (the 38-and 12-fold increase in the case of SK-BR-3 and MDA-MB-453 cells, respectively). Importantly, the sensitivity of the HER2 negative cells to Z HER2:2891 -Fc-MMAE was elevated to a lesser extent ( Figure 5 and Table 1). The higher efficacy of the Z HER2:2891 -Fc-MMAE conjugate may also be caused by an enhanced internalization of this bivalent construct, which is in accordance with the results demonstrating that cross-linking or clustering of transmembrane receptors accelerates endocytosis [59].
Recently, we have generated an auristatin conjugate based on the bivalent Z HER2:4 affibody referred to as (Z HER2:4 ) 2 DCS-MMAE [21]. Z HER2:4 is an earlier version of the anti-HER2 affibody and binds HER2 receptor with the affinity three orders of magnitude lower than Z HER2:2891 [29,30]. The (Z HER2:4 ) 2 DCS-MMAE conjugate killed SK-BR-3 cells with a similar efficacy (IC 50 =~500 pM) as Z HER2:2891 -Fc-MMAE. However, this bivalent construct is highly heterogeneous since it carries from zero to three MMAE molecules. Moreover, the size of (Z HER2:4 ) 2 DCS-MMAE is below 20 kDa which makes it prone to removal through kidney clearance. In this context, the Z HER2:2891 -Fc fusion provides a better platform for a cytotoxic conjugate development than single and bivalent affibodies. Interestingly, the dimeric (Z HER2:4 ) 2 DCS affibody slightly affected the SK-BR-3 cell viability during a 4-day incubation [21]. This finding is in contrast to our preliminary observation that the Z HER2:2891 affibody fused to the Fc fragment enhances viability of the HER2 positive cells (data not shown). This effect of Z HER2:2891 -Fc was not observed for MDA-MB-231 cells that do not overexpress HER2 receptor. We hypothesize that the Z HER2:2891 -Fc dimer induces cell proliferation by activating HER2 signaling pathways. However, this possibility needs to be further investigated.
Ado-trastuzumab emtansine (Kadcyla, Genentech), which is one of the two ADCs currently approved for clinical use by the FDA, similarly to the Z HER2:2891 -Fc-MMAE, targets HER2 receptor. In this ADC cytotoxic payload, DM1 (emtansine), is attached to lysine amines via the SMCC linker [60][61][62]. This conjugate carries 3.2-3.8 DM1 molecules per antibody on average [61,63] which is close to four that is the optimal drug to antibody ratio (DAR) proposed by Hamblett et al. (2004) [57]. However, the distribution of DM1 is not uniform since there are about 40 lysines that can be modified in an antibody affecting pharmacokinetics of the conjugate [64]. Our conjugate relied on MMAE coupling to the interchain cysteines, which results in defined conjugation sites. This strategy ensures batch-to-batch consistency in terms of MMAE location. Importantly, the efficacy of trastuzumab-DM1 tested on the SK-BR-3 cell lines is comparable to the cytotoxicity of (Z HER2:4 ) 2 DCS-MMAE [61].
The Fc part of IgG1 contributes to a cell killing-activity by three mechanisms: antibodydependent cell-mediated cytotoxicity (ADCC), antibody-dependent phagocytosis (ADPh) and complement-dependent cytotoxicity (CDC) [65][66][67]. Notably, a recent study by Mobergslien et al. demonstrated that a fusion of a targeting peptide, referred as WN, with the Fc fragment activated innate immune cells (NK cells, monocytes, dendritic cells). Injection of this construct to 4T1 tumor-bearing BALB/c mice inhibited tumor growth and the tumor tissue was infiltrated by NK cells and T cells [54]. These findings suggest that the cytotoxicity of the proposed Z HER2:2891 -Fc-MMAE conjugate can be further enhanced by its potential to trigger the immunological response directed against HER2-positive cancer cells. This possibility will be addressed in further studies.

Construction of the Z HER2:2891 -Fc Expression Plasmid
The nucleotide sequence of Z HER2:2891 was amplified by PCR (Polimerase Chain Reaction) from the bacterial expression plasmid pDEST-Z HER2:2891 -DCS [19] using the primers A: 5 -CTCTTCT TCCTGTCAGTAACGACTGGTGTCCACTCCGCCGAAGCCAAATATGCAAAA-3 and B: 5 -TTGCGT CCGGATTTCGGTGCCTGGCTATCAT-3 . This PCR resulted in the DNA fragment encoding the last 12 amino acids of the signaling peptide, followed by the sequence encoding of the affibody and Kpn2I (BspMII) restriction site at the 3 of the PCR product. The second round of PCR was carried out with primer C: 5 -CTCCAAGCTTTGAACCACCATGGAATGGAGCTGGGTCTT TCTCTTCTTCCTGTCAGTAACG-3 and B and resulted in the DNA fragment comprising HindIII restriction site at the 5 , the sequence encoding all amino acids of the signal peptide, the affibody and Kpn2I (BspMII) restriction site at the 3' of the PCR product. The PCR product and the pLEV113-ECD_FGFR2-Fc construct [41] were digested with HindIII and Kpn2I (BspMII) and subjected to ligation using T4 ligase. The resulting construct pLEV113-Z HER2:2891 -Fc was verified by sequencing, propagated in DH10α E. coli cells and purified using Plasmid Maxi Kit (Qiagen, Hilden, Germany).

Protein Expression
The Z HER2:2891 -Fc expression was carried out according to Sokolowska-Wedzina et al. (2014) with minor modifications [41]. Briefly, CHO-S cells (Invitrogen, Carlsbad, CA, USA) were cultured in PowerCHO-2CD medium (Lonza, Basel, Switzerland) supplemented with 8 mM L-glutamine and 1× penicillin/streptomycin solution (Biowest, Nuaillé, France). Cells were grown at 37 • C in a shaking incubator (140 rpm) with 8% CO 2 . The cells were diluted with fresh medium to density of 0.3 × 10 6 cells/mL when the culture density reached 1-1.2 × 10 6 cells/mL. One day before transfection the CHO-S cells were diluted to 0.6 × 10 6 cells/mL. On the day of transfection culture was centrifuged and cell pellet were resuspended in ProCHO4 medium (Lonza, Basel, Switzerland) at 2 × 10 6 cells/mL. Appropriate amount of plasmid (1.25 µg of DNA per 1 × 10 6 cells/mL) was resuspended in 150 mM NaCl. The final volume of the DNA solution was 5% of the cell suspension volume. Appropriate amount of 1 mg/mL linear PEI (Polyscience, Warrington, PA, USA) was added to 150 mM NaCl aiming for 5 µg PEI per 1 × 10 6 cells/mL. The final volume of the PEI solution was 5% of the cell suspension. The DNA and PEI solutions were mixed together, incubated at room temperature for 10 min and then added dropwise to the cell solution. The cells were incubated for 4 h in standard conditions (37 • C, 140 rpm, 8% CO 2 ). After this time, the culture was diluted with an equal volume of PowerCHO-2CD supplemented with 8 mM L-glutamine and 2× penicillin-streptomycin solution (Biowest, Nuaillé, France) and incubated at 31 • C. On the second day following the transfection the culture was supplemented with an additional 4 mM L-glutamine and harvested on sixth day.

Protein Purification
The purification of Z HER2:2891 -Fc was carried out according to Sokolowska-Wedzina et al.

Immunostaining
SK-BR-3 and MDA-MB-231 cells were allowed to grow in appropriate media on cover slips that were placed in a 6-well plates for two days. Cells were washed twice with PBS and then fixed with 4% paraformaldehyde for 10 minutes. After paraformaldehyde removal cells were washed 3 times with PBS and then incubated with PBS, 2% BSA, 0.01% Triton X-100 for 15 min at RT. Cells were washed twice with PBS, 0.01% Triton X-100 and incubated for 1 h with a mouse anti-HER2 (e2-4001) monoclonal antibody (Thermo Fisher Scientific, Waltham, MA, USA) or Z HER2:2891 -Fc-FITC diluted in PBS, 0.01% Triton X-100. Following the incubation, cells were washed three times with PBS, 0.01% Triton X-100. Then a 1 h incubation with a secondary fluorescein (FITC) AffiniPure F (ab') fragment donkey anti-mouse IgG (Jackson ImmunoResearch, West Grove, PA, USA) was carried out for the cells stained with the mouse anti-HER2 antibody. Following three washes with PBS, 0.01% Triton X-100 cover slips were mounted onto slides.

Cell Lines
The following breast cancer cell lines were used in this study: SK-BR The cells were cultured at 37 • C in 5% CO 2 and humidified atmosphere. Upon reaching 90% confluence, the cells were passaged using Trypsin-EDTA solution (Thermo Fisher Scientific, Waltham, MA, USA)

Cytotoxicity Assays
The early-passage cells were seeded at the density of 5000 cells per well in a 96-well plate. The cells were allowed to attach and after 24 h different concentrations of the conjugate were added to the wells. After 96-incubation of continuous exposure to the Z HER2:2891 -Fc-MMAE conjugate, the medium was replaced with a fresh medium containing 10% of AlamarBlue reagent (Thermo Fisher Scientific, Waltham, MA, USA). After 4 h, formation of a fluorescent reduced form of the dye was determined using EnVision Multimode Plate Reader (PerkinElmer, Waltham, MA, USA). The IC 50 values were determined using the OriginPro8 software v8.0724 (OriginLab, Northampton, MA, USA).

Monitoring Stability of the Z HER2:2891 -Fc-MMAE Conjugate in Mouse and Human Sera
For the stability assay, 10 µL of conjugate (0.5-1 mg/mL solution) was added to 90 µL of mouse (Invitrogen, Carlsbad, CA, USA) or human serum (Sigma-Aldrich, Saint Louis, MO, USA) and incubated for 6 days at 37 • C. The eppendorf tubes were sealed with parafilm to minimize evaporation. The 10 µL samples were taken every 24 h, mixed with reducing Laemmli sample buffer, boiled (96 • C, 10 min) and separated on 12% SDS-PAGE gel. Upon electrotransfer the membrane was stained with Ponceus S solution and subjected to Western blotting with the goat anti-affibody antibody (Affibody AB, Solna, Sweden) and secondary anti-goat antibody-HRP (Jackson ImmuoResearch, West Grove, PA, USA).