EA045298B1 - MONO AND BISPECIFIC ANTIBODIES BINDING TO HERGI AND HERGI/INTEGRIN BETA 1 - Google Patents

Description

Field of technology to which the invention relates

The present invention relates to the field of antibodies and their use for diagnostic and therapeutic purposes in oncology and other areas of medical sciences. In particular, it relates to anti-hERG1 molecules and engineered derivatives thereof, including bispecific antibodies targeting both hERG1 and β1 integrin.

Prior Art

Antibody technology has improved significantly over the past two decades through antibody engineering; The emergence of new technologies in the field of molecular engineering has led to the production of a wide range of genetically engineered antibodies, such as Fab-type fragments, Fv-forms of the simple chain scFv, diabodies, triates, bispecific molecules, minibodies, nanobodies, phage antibodies. In fact, there are a number of applications in which Fc-mediated effects are not required or even desirable due to their associated toxic effects and their ability to induce an immune response capable of neutralizing the effectiveness of an antibody when its Fc is derived from a non-human source.

Among the engineered antibody fragments, single chain variable fragment (scFv) is the most popular and one of the smallest recombinant formats with antigen binding function and has the property of easy manipulation for immunological applications.

ScFv consists of heavy (VH) and light (VL) chain variable regions that are connected to each other by a flexible peptide linker without interfering with the fidelity of the VH-VL pairing sites and antigen binding sites. The choice of linker can affect the solubility, expression, and proper folding of scFv. Peptide linkers can range from 10 to 25 amino acids in length and are typically composed of hydrophilic amino acids such as glycine (G) and serine (S). Hydrophilic sequences prevent peptide intercalation within or between variable domains throughout protein folding. The most common linker used is the (Gly4Ser)3 motif due to its flexibility, neutral charge, and solubility. The use of scFv in diagnostics and therapy provides a number of advantages over whole antibodies, especially in the treatment of solid tumors; in fact, the rate of penetration of the fragment compared to the intact molecule is the most noticeable advantage. In 1988, it was found that an intact IgG molecule took fifty-four hours to penetrate 1 mm into a solid tumor, while a Fab fragment traveled the same distance in sixteen hours. In addition, scFv, like all other antibody fragment formats, can be formulated into multivalent and polyspecific reagents or readily coupled to therapeutics such as radionuclides, toxins or nanoparticles and designed to improve their diagnostic and therapeutic efficacy.

These engineered molecules are easily produced in bacterial or yeast systems and are more efficiently extravasated and have higher tissue penetration capacity than full-length Ig; the only limitation of these molecules is their short half-life due to their small size. Many strategies have been developed to improve pharmacokinetics, such as multimerization of scFvs (shortening their linker sequence) to form tribodies (about 90 kDa) and tetrabodies (about 120 kDa), or conjugation of antibodies to large molecules such as polyethylene glycol (PEG) (Natarajan et al. al. 2005) or human serum albumin (HSA).

Bispecific antibodies (bsAbs) have recently received much attention as potential anticancer therapeutic agents as they offer several advantages:

bsAbs can redirect specific immune cells to tumor cells, thereby enhancing tumor killing;

bsAbs can simultaneously block two different targets in different pathways that have unique or overlapping functions in pathogenesis;

bsAb can potentially increase binding specificity by interacting with two different cell surface antigens instead of one.

The development of bispecific antibodies (bsAbs) has faced many challenges, mainly due to manufacturing problems, poor yield, instability and immunogenicity (Spiess S. et al, 2015).

Regarding the production methodology of bsAbs, they are mainly produced in three ways which include:

Quadrome technology, based on the somatic fusion of two different lines of hybridoma cells; chemical conjugation using chemical cross-linking agents; genetic approaches using recombinant DNA technology.

Bispecific antibodies can be broadly divided into two main subgroups: immunoglobulin G-like (IgG) and non-IgG-like molecules, and there are currently more than 30 bsAbs in clinical development, with two drugs, Catumaxomab and Blinatumomab, already approved for sales.

He-IgG-like include mainly scFv-based bsAbs and nanobodies. It is known that scFv can become dimers, trimers or tetramers depending on the length of the linker, sequentially

- 1 045298 antibodies and other factors (Le Gall F. et al, 1999). This format is preferred and has many potential clinical applications.

Among scFv-based bsAb formats exist.

Tandem scFvs, which consist of two scFvs linked by a flexible peptide linker such as glycine-serine repeat motifs in tandem orientation. The famous bispecific T cell-attracting activator (BiTE) technology is based on this format (Chames P. et al., 2009).

A diabody format in which the variable domains of two different antibodies are linked by two linkers. They have the function of increasing the stability of the diabody.

Single chain diabodies (scDb), the diabody format can be converted to a single chain diabody by adding an additional linker compound between the chains

Tandem diabodies (TandAbs) are formed by two pairs of VL and VH domains linked into one polypeptide chain, forming a tetravalent TandAb.

Dual Affinity Retargeting (DART) DART molecules are created by associating the VH of a first variable region bound to a VL on the second strand and the VH of a second variable region bound to a VL on the first strand in a VLA - VHB + VLB - VHA configuration. Due to their small size, DARTs are prone to destruction (Moore P.A. et al., 2011).

Diabodies and scDbs are also the most efficient way to produce bispecific antibody fragments capable of binding two different antigens and therefore useful for cell cross-linking (e.g., retargeting effector cells of the immune system); recruiting effector molecules (such as toxins, drugs, cytokines, radioisotopes or complement systems) to retarget the host system (for example, viral vectors for gene therapy) (Kontermann, 2005); targeting and inhibition of macromolecular complexes involved in tumor progression.

Antibody engineering has also provided ways to increase avidity (eg, by multimerizing antibody fragments); affinity (for example, mutation in the variable regions of whole Igs or antibody fragments); and enhanced effector functions (for example, by mutation in the constant regions of whole Ig or conjugations of antibody fragments with recombinant Fc, toxin, drugs, cytokines, death ligands, radioisotopes, nanoparticles or complement molecules).

Over the past three decades, the human cardiac specific potassium channel gene 1 (hERG1) has become a target in oncology and other human diseases. However, its therapeutic use has been hampered by the fact that most drugs that cause hERG1 blockade as a primary or side effect can cause cardiotoxicity (prolongation of the electrocardiographic QT interval and the occurrence of ventricular arrhythmias). In a search for biophysical and biomolecular signatures that distinguish hERG1 expressed in the heart from hERG1 expressed in cancer cells and other disease-specific cells, hERG1 was found to form complexes with other plasma membrane proteins, particularly the beta1 integrin receptor subunit , on the plasma membrane of cancer cells. Such a complex is not found in cardiac myocytes (Becchetti A. et al, Sci. Signaling, 10 (473). pii: eaaf3236. doi: 10.1126/scisignal.aaf3236. PMID: 28377405, 2017). Therefore, the hERG1/beta1 integrin complex is configured as an oncogenic unit intrinsic to transformed cells, a fact that differentiates hERG1 in the tumor from the channel expressed in the heart. This discovery means that any molecule (small molecule drug or otherwise protein) targeting the hERG1/beta1 integrin complex can be used for diagnostic and therapeutic purposes without causing cardiotoxicity. Currently, there are no known molecules capable of targeting the hERG1/beta1 integrin complex.

Italian patent IT1367861 describes a hybridoma cell line clone, named A7, capable of secreting an anti-hERG1 monoclonal antibody (mAb) specific for the extracellular portion of the S5 pore of hERG1.

WO 2016020483 (A1) describes the detailed structure of an intact mouse monoclonal anti-hERG1 molecule and the corresponding preparation of an anti-hERG1 scFv obtained after isolation of the VH and VL anti-hERG1 mAbs. This scFv has the same specificity as the corresponding whole antibody and is thus capable of recognizing the same anti-hERG1 protein that is aberrantly expressed in tumors and other diseases. The nucleotide sequences SEQ ID NO: 1 and SEQ ID NO: 3 encoding VH (SEQ ID NO: 2) and VL (SEQ ID NO: 4), respectively, were disclosed together with the nucleotide sequence SEQ ID NO: 5 encoding scFV having SEQ ID NO: 6.

Anti-beta1 integrin mAbs are known in the art for research purposes, for example, TS2/16 (Arroyo et al. J. Cell Biol. 1992, 117 (3), 659-670) and BV7 (Martin Padura et al. J Biol Chem 1994, 269(8), 6124-6132).

The aim of the present invention is to create a bispecific antibody that simultaneously targets hERG1 and beta1 integrin proteins, and which forms a complex on the plasma

- 2 045298 membrane of cancer cells. It is yet another object of the present invention to provide an improved or at least alternative anti-hERG1 antibody.

The essence of the invention

An object of the present invention is a bispecific antibody (bsAb) containing a heavy chain variable domain (VH) and a light chain variable domain (VL) anti-hERG1 Ab that binds the extracellular S5-P domain of hERG1 and the heavy chain variable domain (VH) and variable domain light chain (VL) anti-β1 integrin Ab that binds the extracellular domain of β1 integrin.

Surprisingly, it was found that the bsAb of the invention is able to selectively bind the hERG1 + β1 integrin complex, which is present only in tumor cells. In particular, the bsAb of the invention has demonstrated the ability to inhibit cell growth and migratory, prometastatic activity in vitro in a panel of neoplastic cell lines.

In an aspect, the present invention also provides an anti-hERG1 molecule comprising a heavy chain variable domain (VH) having at least 85% identity with SEQ ID NO: 8, where the residue at position 95 is a Cys, and a light chain variable domain chain (VL) having at least 85% identity with SEQ ID NO: 4, and the specified molecule has specificity for the extracellular part of the S5 pore of hERG1.

Surprisingly, it was found that the anti-hERG1 molecule (anti-hERG1-Cys) of the invention, having a Cys at position 95 of the VH domain (SEQ ID NO: 8), showed better affinity for the immobilized antigen compared to a molecule known in the art having Phe at position 95 of the VH domain (SEQ ID NO: 2). In particular, the scFv molecule of the invention has been shown to inhibit cell growth in vitro in a panel of tumor cell lines.

Detailed Description of the Invention

Preferably, the bsAbs of the invention are those in which the heavy chain variable domain (VH) of the anti-hERG1 Ab has 85% identity with SEQ ID NO: 8 or SEQ ID NO: 2; and the anti-hERG1 Ab light chain variable domain (VL) has 85% identity with SEQ ID NO: 4; and the heavy chain variable domain (VH) and light chain variable domain (VL) of the anti-B1 integrin Ab have at least 85% identity with the VH and VL of TS2/16 (SEQ ID NOs: 26 and 24) or BV7 ( SEQ ID NO: 46 and 48).

Preferably, the bsAb of the invention is a non-IgG-like bsAb, preferably a scFv-based bsAb. According to the invention, the scFV-based bsAb has a format selected from the group consisting of tandem scFvs, diabody format, single chain diabodies, tandem bodies (TandAbs) and dual affinity retargeting molecules (DART), preferably scDbs.

Preferably, the VH domain of the anti-B1 integrin Ab has 90%, 95%, 99%, or 100% identity with SEQ ID NO: 26 or SEQ ID NO: 46, more preferably SEQ ID NO: 26.

Preferably, the VL domain of the anti-B1 integrin Ab has 90%, 95%, 99%, or 100% identity with SEQ ID NO: 24 or SEQ ID NO: 48, more preferably SEQ ID NO: 24.

Preferably, the VH domain of an anti-hERG1 Ab, still retaining the Cys amino acid at position 95, has 90%, 95%, 99%, or 100% identity to SEQ ID NO: 8.

Preferably, the VH domain of the anti-hERG1 Ab has 90%, 95%, 99%, or 100% identity with SEQ ID NO: 2.

Preferably, the VL domain of the anti-hERG1 Ab has 90%, 95%, 99%, or 100% identity to SEQ ID NO: 4.

The VH domain of the anti-hERG1-Cys Ab of the present invention is preferably encoded by a nucleotide sequence having at least 70%, 80%, 90%, 95%, 99% or 100% homology to SEQ ID NO: 7 wherein is a triplet of residues 283 to 285 may be TGT or TGC.

The VH domain of the anti-hERG1 Ab of the present invention is preferably encoded by a nucleotide sequence having at least 70%, 80%, 90%, 95%, 99% or 100% homology to SEQ ID NO: 1.

The VL domain of the anti-hERG1 Ab of the present invention is preferably encoded by a nucleotide sequence having at least 70%, 80%, 90%, 95%, 99% or 100% homology to SEQ ID NO: 3.

The molecule of the invention may be a fully humanized recombinant antibody.

The molecule of the invention may be a scFv or any other engineered antibody such as Fab, Fv single chain form of scFv, diabodies, tribodies, bispecific molecules, minibodies, phage antibodies; scFv and diabodies (scDb) are preferred.

Preferred linkers are (Gly4Ser)3 motifs.

Particularly preferred is an anti-hERG1-Cys scFv, where VH and VL are linked by a peptide linker; more preferred is the anti-hERG1-Cys scFv having SEQ ID NO: 10.

Particularly preferred is a bsAb that is a single chain diabody (scDb) containing an anti-hERG1-Cys scFv and an anti-β1 integrin scFv, thereby selectively targeting the hERG1 + β1 integrin complex, which is present only in tumor cells.

- 3 045298

Therefore, in a preferred aspect, the present invention provides a single chain bsAb (scDb) diabody comprising a first heavy chain variable domain (VH) having at least 85% identity to SEQ ID NO: 8, wherein the residue at position 95 is Cys and a first VL domain having at least 85% identity with SEQ ID NO: 4, and a second VH domain having at least 85% identity with SEQ ID NO: 26 or SEQ ID NO: 46, and a second VL domain having at least 85% identity with SEQ ID NO: 24 or SEQ ID NO: 48, wherein said bispecific Ab has specificity for the extracellular portion of the hERG1 S5 pore and for the β1 integrin.

The bsAb or anti-hERG1 molecule of the invention is useful as a diagnostic or therapeutic agent.

The diagnostic tool according to the invention is intended for the in vitro detection of hERG1 (for example, in surgical specimens or biopsies), contains an anti-hERG1-Cys scFv and/or a bispecific Ab (preferably as a scDb) according to the invention, either unlabeled or coupled to a fluorophore, preferably Alexa 488 fluorophore.

Thus, an object of the present invention is also an in vitro diagnostic (IVD) kit containing parts for simultaneous, separate or sequential use, said IVD kit comprising:

a container containing anti-hERG1-Cys scFv as described above; and/or container containing bispecific Ab as described above

Preferably, the IVD kit further includes a container containing an intact monoclonal antibody, as described in WO 2016020483, as a reference control.

The IVD kit can be used either on fixed tissue samples for immunohistochemistry methods (with results available in 2-3 weeks), or on fresh biopsy tissue (obtained, for example, by endoscopy or surgery), for use with immunofluorescence methods, and results are available in 1 day.

As described above, an anti-hERG1-Cys scFv labeled with a fluorophore (eg, preferably Alex 750) or radionuclide (eg, preferably Tc99) is a molecule for use in the early diagnosis of hERG1-positive cancers in vivo (in humans ) and is represented by anti-hERG1-Cys scFv. The anti-hERG1-Cys sc-FV antibody of the invention has a specific molecular structure that allows rapid penetration into cancer tissue, rapid binding to the hERG1 biomarker and rapid elimination, making it, when bound to a radionuclide, an ideal molecule for early in vivo detection. people) diagnosed with hERG1-positive cancer. The single dose administration and rapid half-life, 3.5 hour molecule with no systemic toxicity when administered intravenously at a dose of 8 mg/kg and absence of ECG changes (see Fig. 19) allow it to avoid interacting with cardiac cells. Finally, anti-hERG1-Cys scFv was found to have a very good tumor/tissue ratio when administered intravenously at a dose of 1 mg/kg in mice bearing xenograft pancreatic cancer in the pancreas.

Therapeutic agents.

The bispecific antibody molecule of the invention, specifically designed to inhibit the malignant hERG1/beta1 integrin molecular complex for therapeutic purposes, is a single chain bispecific antibody (single chain diabody, scDb) that is capable of selectively binding to hERG1 when expressed with beta1 integrin on cancer cells, without interaction with the heart. ScDb according to the invention is an ideal molecule to be used for repeated (continuous) administration to patients with hERG1 positive cancer, without cardiac safety concerns, with therapeutic potential in both early and late/metastatic stages, as a single agent , and as a combination therapy agent, for addition to chemotherapy, radiation, targeted therapy and immuno-oncology therapy. The rationale for combination therapy is that the pathways constitutively activated in cancer cells by overexpression of the hERG1/beta1 integrin molecular complex are complementary and integrative to the pathways currently targeted by available drugs. Additionally, the tumorigenic pathway involving hERG1/beta1 integrin may represent a tumor escape mechanism compared to currently available treatments. The antihERG1/beta1 integrin scDb was found to have a half-life of approximately 12 hours, with no systemic toxicity when administered intravenously at a dose of 8 mg/kg and no ECG changes (see FIG. 20). Finally, the anti-hERG1/beta1 integrin scDb was found to have very good therapeutic efficacy when administered at a dose of 1 mg/kg twice a week six times in mice with xenografted pancreatic cancer in the pancreas.

Pathologies that can be diagnosed or treated using a bsAb or molecule of the invention are all pathologies characterized by overexpression or misexpression of hERG1 protein. These pathologies include tumors, neurological diseases, endocrine diseases and neuroendocrine diseases.

- 4 045298

The BsAb or molecule according to the invention, in particular anti-hERGl-Cys scFV or scDb, can also be used as a pharmaceutical delivery vector: for example, it can be covalently or non-covalently linked to a radionuclide, enzyme, drugs or toxin.

Thus, it is also a further object of the present invention to provide a pharmaceutical composition comprising a bsAb or molecule according to the invention and at least another pharmaceutically acceptable ingredient.

Another object of the present invention is also the nucleotide sequence encoding the bispecific Ab or anti-hERG1 molecule according to the invention. Suitable degrees of homology (eg, at least 85%) with coding sequences that produce a molecule of the invention are intended to be included.

The molecule of the invention may preferably be prepared using the nucleotide sequences SEQ ID NO: 7 and SEQ ID NO: 2 encoding VH (SEQ ID NO: 8) and VL (SEQ ID NO: 4), respectively.

Particularly preferred according to the invention is a method for producing an anti-hERG1-Cys scFv according to the invention, said method comprising the use of a nucleotide sequence of SEQ ID NO: 9 encoding an anti-hERG1-Cys scFV having SEQ ID NO: 10.

ScDb-hERG1-e1 of the invention is preferably prepared using the nucleotide sequences of SEQ ID NO: 7, SEQ ID NO: 2, SEQ ID NO: 23 or SEQ ID NO: 47 and SEQ ID NO: 25 or SEQ ID NO: 45 encoding respectively, anti-hERG1-Cys VH (SEQ ID NO: 8), anti-hERG1-Cys VL (SEQ ID NO: 4), anti-β1 integrin VL (SEQ ID NO: 24 or SEQ ID NO: 48) and anti -β1-integrin VH (SEQ ID NO: 26 or SEQ ID NO: 46). Preferably, the domains are assembled in the following order: anti-hERG1Cys VH (SEQ ID NO: 8), anti-β1 integrin VL (SEQ ID NO: 24 or SEQ ID NO: 48), anti-β1 integrin VH (SEQ ID NO: 26 or SEQ ID NO: 46) and anti-hERG1-Cys VL (SEQ ID NO: 4).

Particularly preferred according to the invention is a method for producing scDbhERG1-e1 in accordance with the invention, said method comprising using the nucleotide sequence SEQ ID NO: 29 encoding the scFV against hERG1-Cys having SEQ ID NO: 30.

The method according to the invention involves recombinant techniques.

Therefore, the subject of the present invention is also an expression vector or plasmid containing a nucleotide sequence encoding a bispecific Ab or molecule according to the invention, preferably containing SEQ ID NO: 7 and SEQ ID NO: 2, as well as genetically modified microorganisms or a cell containing an expression vector according to invention. The above expression vector or plasmid may also contain SEQ ID NO: 23 or 47 and SEQ ID NO: 25 or 45.

The present invention can be better understood in light of the experimental section below.

Brief description of drawings

Fig. 1 - Electropherograms obtained from DNA sequencing of four colonies after mutagenesis performed on the scFv-hERG1 construct, showing the correct mutation from Phe to Cys.

Fig. 2. (A): yeast expression of scFv-hERG1. Slot blot on supernatants collected by induction of scFv-hERG1-G3 Pichia Pastoris clones at 24 h, 48 h and 72 h.

Clones C7, C12, D9, E8, G3, G7 and the negative control of the untransformed Pichia strain GS115 were analyzed. Staining was performed using DAB chromogen. G3 is shown as the best expressing clone. (B): Western blot analysis performed on purified samples of six clones. G3 showed the highest level of expression, while clone G7 showed almost no expression. (C): Slot blot on supernatants collected by induction of scFvhERG1-D8Cys clones of Pichia Pastoris at 24 h, 48 h and 72 h. Clones B11, C3, D8, D9, G4, G10 and a negative control of an untransformed Pichia strain were analyzed GS115. Staining was performed using DAB chromogen. D8 is shown as the best expressing clone. (D): Western blot analysis performed on purified samples of six clones. D8 was the clone that showed the highest level of expression, while lower expression was found for clone C3.

Fig. 3. (A) SDS-PAGE of purified scFv-hERG1-G3 and scFv-hERG1-D8Cys elution fractions; (B) Gel filtration chromatography of both purified antibodies using Superdex 75 FIR 10/30. (C) scFv-hERG1-D8Cys antibody stability test. SDS-PAGE followed by Coomassie Brilliant Blue staining is shown at different time points (0, 6, 12 and 18 months after protein purification).

Fig. 4. (A) Immunofluorescence performed on fixed and live HEK293Mosk and HEK293-hERG1 cells using both unlabeled (I.IF) and labeled scFv-hERG1-G3 (D.IF) antibodies. Representative images taken at 20x magnification, stained nuclei are represented by blue fluorescence, stained membranes are represented by green fluorescence (Alexa 488).

(B) Immunofluorescence was performed on fixed and live HEK293-Mosk and HEK293 cells

- 5 045298 hERGl, using both unlabeled (I.IF) and labeled scFv-hERG1-D8-Cys (D.IF) antibodies. Representative images taken at 20x magnification, stained nuclei are represented by blue fluorescence, stained membranes are represented by green fluorescence (Alexa 488).

(C) Graphs showing IF intensity (A.U.) calculated using Image J software (ImageJ 1.38, U.S. National Institutes of Health). For each image, the average fluorescence intensity of three different regions was calculated after subtracting the values of the blue channels (which refers to nuclear staining).

In all experiments, the results for HEK 293-hERG1 were significantly higher compared to the results obtained for HEK 293-Mosk. Statistical analysis was performed to evaluate assumptions of normality and homoscedasticity of the data using the Shapiro-Wilk test, and variance was analyzed using ANOVA. Significance of paired waves was assessed using Student's t test or Bonferroni test (*p < 0.05).

Fig. 5. Trypan blue viability assay performed on HCT-116, MDA MB-231, MIA PACA2, HEK 293 HERG1, HEK-MOSK, FLG 29.1, PANC-1, BxPC-3, using monoclonal antibody against hERG1 (100 μg /ml) and scFv-hERG1-D8Cys (10; 20 μg/ml). All experiments were performed in triplicate.

Fig. 6. Panel A: HEK 293 HERG1 spheroids were tested at 10, 20 and 40 μg/ml scFv-hERG1-D8Cys. The volume of spheroids treated with 20 μg/ml and 40 μg/ml scFv-hERG1-D8Cys is smaller (see dotted and dash-dotted lines, respectively) compared to the control (solid line) at each time point. Instead, panel B shows the growth curve of HEK-MSC spheroids (not expressing hERG1), in which no difference was found for treated spheroids at all three scFv-hERG1-D8Cys concentrations tested compared to the control. Panels A and B also show a representative bright-field image of control HEK293-hERG1 and HEK-MOCK spheroids, respectively, as they appear after 72 h of culture. Panel C. Spheroids of pancreatic ductal adenocarcinoma. Panel C shows the effect on pancreatic ductal adenocarcinoma Mia Race 2 cells. A decrease in spheroid volume was observed for both cells treated with 20 μg/ml and 40 μg/ml scFv-hERG1-D8Cys, with a more pronounced effect obtained at the highest concentration tested (dashed line) compared to controls at each time point. Also, the 72 hour images shown on the right side of the panel show a significant decrease in spheroid volume for cells treated with scFv-hERG1-D8Cys 40 μg/ml antibody (see right image) compared to the control (see left image) . Panel D - Breast Cancer Cells, MDA-MB 231 Spheroids: A noticeable volume reduction effect is observed for all three concentrations of scFvhERG1-D8Cys (10, 20, 40 μg/ml) compared to control. The reduction in volume can also be inferred from the images of the MDA-MB 231 spheroids shown on the right side of the figure.

Fig. 7. Calcein AM cell viability assay performed on spheroids after 72 hours. Green coloring represents living cells while red coloring represents dead cells. The image on the left (panel A) is the control images for each cell line, while the right (panel B) is the images of spheroids treated with 40 μg/ml scFvhERG1-D8Cys. From the image, one can note a decrease in volume for antibody-treated spheroids, especially for Mia Race 2, MDA MB-231 and PANC-1 spheroids and, in addition, an increased number of dead cells, especially for MDA MB-231 and PANC-1 treated spheroids scFv-hERG1D8Cys.

Fig. 8A - Nucleotide sequence of anti-b1 integrin VL and VH domains (TS2/16) (SEQ ID NO: 23 and 25) obtained by automated DNA sequencing service (PRIMM). The underlined italics indicate the VL sequence, and the italicized VH sequence.

Fig. 8B is the nucleotide sequence of the VL and VH domains of anti-b1 integrin (BV7) (SEQ ID NOs: 45 and 47). Primers used to isolate single domains are in bold, while gray is a region of VH sequence that is unknown but ultimately required for correct framing.

Fig. 9. Top panels: schematic structure of two single-chain antibodies, anti-hERG1 and anti-TS2/16. Bottom panels: schematic structure of scDb-hERG1-e1.

Fig. 10. Scheme representing the SOE-PCR method for assembly of anti-b1 integrin scFv in the VL linker-VH order.

Fig. 11. Panel A shows Coomassie staining of purified supernatants obtained from small scale induction of six clones grown after transformation with the anti-hERG1-Phe-β1-scDb construct. One band was detected corresponding to clone G5 with a molecular mass of about 60 kDa, which is consistent with the expected one.

Panel B shows the chromatogram obtained after purification of the supernatant resulting from large-scale expression of clone G5: a single peak is visible and the elutions underlying the blue area were analyzed, and the Coomassie stain shown in

- 6 045298 nel C, shows bands with the correct molecular weight in all elutions tested.

Fig. 12. Panel A. Results of cellular ELISA performed on HEK 293 HERG1 cells using varying amounts of bispecific anti-hERG1-Phe-β1-scDb antibody.

Fig. 13 - indirect IF performed on cells plated on different substrates, BSA and fibronectin (FB). IF shows a stronger signal for FB-coated HEK 293 HERG1 cells compared to the signal obtained for BSA-coated cells.

Fig. 14 - indirect IF performed on HEK 293 HEKG cells (panels A and B). Staining of cells after administration of excess S5PORO peptide is shown in panels D and E. Panel C shows control cells stained with secondary antibodies only.

Fig. 15. Expression and purification of scDb-hERG1-Cys-e1. Panel A. Chromatogram obtained from the purification of Pichia Pastoris supernatants. Panel B. Coomassie stain showing elution analysis of scDb-hERG1-Cys-e1 purification, within the blue peak of the chromatogram.

Fig. 16 - direct IF with scDb-hERG1-Cys-e1-Alexa488. In these particular experiments, cells were incubated with the scDb antibody directly conjugated to the fluorophore Alexa 488. GD25 WT cells (negative for hERG1 and β1 integrin expression) were incubated with the scDbhERG1-Cys-e1 antibody directly labeled with the fluorophore A1exa488. There is no signal. Panel B. HEK 293 HERG1 cells plated on fibronectin (FN) and BSA and stained with scDb-hERG1-Cys-e1Alexa488: as shown in the histogram, the signal is higher (~17 A.U.) in cells plated on FN compared to cells plated on FN. seeded on BSA (~10 A.U.). Panel C. HEK 293 WT cells plated on FN and BSA and incubated with scDb-hERG1-Cys-β1-A1exa488: the signal is lower for cells plated on BSA (A 7 A.U.) compared to cells plated on FN (~ 12 A.U.). As can be seen from panels B and C, fluorescence values are lower for HEK 293 HERG1 cells on FN (~17 A.U.) compared to HEK 293 WT cells plated on FN (~12 A.U.).

Fig. 17 - determination of IC50 on MDA-MB 231 and PANC-1 cells. Effects on cell viability were evident at 24 μg/mL for PANC-1 cells and 42 μg/mL for MDA-MB 231 cells.

Fig. 18. Lateral motility experiments on HCT 116, MDA-MB 231 hERG1, MDA-MB 231 and PANC-1 cells. A clear reduction in MI (motility index) was shown in cells treated with scDb-hERG1-Cys-e1 compared to control cells. A more pronounced effect was reported for MDA-MB 231 hERG1 compared to MDA-MB 231 cells, suggesting an effect of hERG1 on cell motility. Lateral motility experiments performed on PANC-1 cells showed lower MI on treated cells compared to controls. Lateral motility experiments performed on HCT116 cells showed lower MI in treated cells compared to controls.

Fig. 19. (A) Pharmacokinetics of scFv-hERG1-D8Cys in mice by intravenous administration (n=2). Antibody concentrations were determined by ELISA, dosing concentrations in mouse plasma samples taken at 5, 15, 30 min and 1, 2, 6, 24 and 48 h after scFv injection. t1/2 = 3.5 h. Values are the average of two measurements ±SD.

(B) ECG, electrocardiogram recording. ECG measurements are shown in the left panel for a PBS saline control mouse with a corrected QT interval of 86 ms. The right panel shows the ECG plot obtained after administration of scFv-hERG1D8Cys, without significant changes compared to the control with the QT interval set to 90 ms.

(C) In vivo analysis. Each panel reports the fluorescent signal of a representative mouse treated with Alexa750-conjugated scFv-hERG1-D8Cys antibody compared with a control mouse treated with PBS solution. The maximum signal detected was 10 min after injection; and no fluorescent signal was detected 24 h after intravenous administration.

(D) Uptake of scFv-hERG1-D8Cys-Alexa750 and retention of scFv-hERG1-D8Cys-Alexa750 in the MIAPaCa-2-nu/nu PDA mouse model. Mice were treated by tail vein injection with 6.5 μg of scFvhERG1-D8Cys-Alexa750 antibody. Representative photographs of i.v. mice. injected with labeled antibodies (left) were compared with control mice (right). The fluorescence intensity in the abdominal region, the area closest to the tumor, was analyzed. Fluorescence emission spectra were measured using a photon scanner (Biospace Lab), and images were acquired at different time points every 5 min from 5 min postinjection until 60 min postinjection.

Fig. 20. (A) Pharmacokinetics of scDb-hERG1-Cys-e in mice by intravenous administration (n=2). Antibody concentrations were determined by ELISA, dosing plasma concentrations from mice taken at 5, 15, 30 min and 1, 2, 6, 24 and 48 h after scDb injection. t1/2 - 12 h Values are means of two measurements ±SD.

(B) ECG, electrocardiogram recording. The ECG graph obtained after administration of the scDb-hERG1-Cys-e1 antibody does not show significant changes compared to the control graph shown in FIG. 20B, with a corrected QT interval of 83 ms, comparable to control.

- 7 045298 (C) Table showing pancreatic volume (mm3) in MIAPaCa2-nu/nu tumor-bearing mice treated and not treated with the scDb-hERG1-Cys-e1 antibody. Data on metastatic diffusion, % necrotic area on the slide, and number of vessels are also presented.

(D) Images of the pancreas after necroscopy: 1, without processing; 2, treated with three doses of scDb-hERG1-Cys-e1 antibody; 3, treated with six doses of scDb-hERG1-Cys-p1 antibody.

experimental part

1. Mutagenesis of scFv-hERG1.

The amino acid sequence of the scFv hERG1 molecule, as described in WO 2016020483 (A1), contains the amino acid Phe at position 95 of the VH domain. The nucleotide sequence SEQ ID NO: 1 revealed the presence of T at position 283 (c283T) of the VH domain. According to the present invention, a G substitution was introduced instead of a T at position 283 (c283T>G) of the VH domain, resulting in a switch of Phe (TTT) at position 95 to Cys (TGT). This mutation introduced a single amino acid (Cys) at a position between scaffold 3 and CDR3, which unexpectedly resulted in fundamental disulfide bond formation in the immunoglobulin variable domain. Cys was introduced into the original construct to guide the mutagenesis protocol (see Materials and Methods). cDNA obtained from four mutagenized scFv-hERG1 colonies was sequenced, and the sequencing results (Fig. 1) demonstrated the correct TTT to TGT mutation at position c283T>G, indicating the desired Phe to Cys mutation.

2. Protein expression and purification.

The plasmids, scFv-hERG1 and mutagenized scFv-hERG1 (hereafter referred to as scFv-hERG1-Cys), were transformed into P. pastoris host strain GS115 using the spheroplasty technique. Six clones (C7, C12, D9, E8, G3, G7) among scFv-hERG1 transformants and six (B11, C3, D8, D9, G4, G10) from scFv-hERG1-Cys were analyzed. Small scale expression results are shown in FIG. 2 panels A and B for scFv-hERG1 and panels C and D for scFv-hERG1-Cys, respectively. All clones showed protein expression after 72-h induction (top panels), also shown by slot blotting.

After purification, the presence of protein was assessed by Western blotting (panels B and D).

The two highest expressing clones were selected for two antibodies: G3 for scFv-hERG1 (hereafter referred to as scFv-hERG1-G3) and D8 for scFv-hERG1-Cys (referred to as scFv-hERG1-D8-Cys).

Large scale expression assays are shown in Fig. 3. Protein presence was assessed by SDS-PAGE and Coomassie Brilliant Blue staining. Fractions 11, 12, 13 (A, left panel) correspond to scFv-hERG1-G3; fractions 12, 13, 14 (B, right panel) correspond to scFv-hERG1D8Cys. Bands corresponding to the molecular size of both antibodies (about 30 kDa) are visible. When comparing the yields of the two proteins, scFv-hERG1-G3 and scFv-hERG1-D8Cys, significant differences were found: the concentration of scFv-hERG1-G3 was 0.050 μg/μl; The concentration of scFv-hERG1-D8Cys was 0.444 μg/μl.

PROTEIN Yield (mg/l) ScFv-hERGl-G3 0200 ScFv-hERGl-D8Cys 1

aYields were normalized to mg protein per liter of Pichia pastoris yeast culture

3. Comparison of antigen affinity between scFv-hERG1-G3 and scFv-hERG1-D8Cys.

The chromatograms presented in Fig. 3 (panels B) show the results obtained from gel filtration. Size exclusion chromatography (SEC) was performed to investigate the possible presence of aggregates that could affect the binding ability of the two antibodies. Several aggregates are detected in the analysis presented in B (left panel), which relates to scFv-hERG1-G3; while scFv-hERG1-D8Cys (B, right panel) is presented in monomeric form.

4. Stability test of scFv-hERG1-D8Cys antibody.

The stability of the scFv-hERG1-D8Cys antibody was assessed directly by analyzing the protein by Coomassie Brilliant Blue SDS-PAGE staining at different time points (6, 12, 18 months) after purification. Data in Fig. 3C show that only one neat single band is visible at all time points, indicating that the protein maintains its stability without showing signs of degradation.

5. Assessment of immunoreactivity of scFv-hERG1-G3 and scFv-hERG1-D8Cys Immunofluorescence analysis was then performed using scFv-hERG1-G3 and scFv-hERG1-D8Cys on fixed cells to determine the immunoreactivity of the two antibodies. HEK 293 transfected with hERG1 cDNA (HEK-hERG1) and, as a control, HEK-MOCK, which does not express hERG1 protein, were used as a cell model. HEK-MOCK cells showed no or weak signal with both antibodies, while hEK 293 hERG1 showed good labeling with scFv-hERG1-G3, and even better with scFv-hERG1-D8Cys (Fig. 4, A and B). Analysis of data obtained using ImageJ software is presented in the graphs shown in panel C. Values obtained from staining

- 8 045298 scFv-hERG1-D8Cys is significantly higher in hERGl-overexpressing cells compared to control values obtained in HEK-MOCK cells.

The immunoreactivity of two antibodies was also tested after direct labeling with the fluorescent molecule Alexa 488. Antibodies scFv-hERG1-G3-Alexa488 and scFv-hERG1-D8Cys-Alexa488 were tested in IF on fixed cells (Fig. 4, A and B), demonstrating retention of recognition ability the antigen in its native conformation even after conjugation with a fluorophore. IF staining was measured using ImageJ software and the results are shown in the graphs presented in panel C. The signal obtained on HEK 293 cells is stronger compared to control HEK-MOCK cells for both scFv-hERG1-G3-Alexa488 and and for scFv-hERG1-D8CysAlexa488.

To evaluate and compare the potential in vivo application of scFv-hERG1-G3-Alexa488 and scFvhERG1-D8Cys-Alexa488 as molecular tools, both antibodies were used for IF on live cells (Fig. 4 A and B).

The experiment confirmed the results obtained by staining on fixed cells; hEKG3 HEK293 cells appear to have a stronger signal compared to the negative control HEK293 MOCK cells. hEK 293 HEK cells appear to have a more specific patchy cell labeling, whereas HEK-MOCK cells have a nonspecific diffuse background. For this reason, a bright-field image of the same area was presented.

6. Inhibition of viability by scFv-hERG1-D8Cys antibodies and spheroid test.

At this stage, the potential of scFv-hERG1-D8Cys to inhibit cell growth in a panel of neoplastic cell lines was further explored. As shown in FIG. 5, significant dose-dependent inhibition of cell proliferation was observed for HCT-116, MDA-MB 231, Mia Race2, HEK 293 HERG1, PANC-1 and BxPc3. Cells were treated with monoclonal antibody against hERG1 (100 μg/ml) and scFv-hERG1-D8Cys (10; 20 μg/ml). As expected, no significant reduction in cell viability was detected in HEK-MOCK cells that do not express hERG1.

To investigate the effect of scFv-hERG1-D8Cys in a 3D cell model, we tested three different concentrations of scFv-hERG1-D8Cys (10; 20; 40 μg/ml) on spheroids.

In fig. Figure 6 is a graph showing the Y-axis volume of spheroids (mm3), while the X-axis reports different time points (24 h, 48 h and 72 h).

In fig. 6, panel A shows the plot obtained for spheroids obtained from HEK293hERG1. The volume of spheroids treated with 20 μg/ml and 40 μg/ml scFv-hERG1-D8Cys is smaller compared to control at each time point.

Instead, panel B shows the growth curve of HEK-MOCK spheroids, in which no differences were found for treated spheroids at all three scFvhERG1-D8Cys concentrations tested compared to the control. Panels A and B also show a representative bright-field image of control HEK293-hERG1 and HEK-MOCK spheroids, respectively, as they appear after 72 hours of culture.

Panel C shows the effect on pancreatic ductal adenocarcinoma Mia Race 2 cells. A decrease in spheroid volume was observed for both cells treated with 20 μg/ml and 40 μg/ml scFv-hERG1-D8Cys, with a more pronounced effect obtained at the highest concentration tested, compared to controls, at each time point. Images taken at 72 hours, shown on the right side of the figure, show an image of the control Mia Race 2 spheroid taken at 4x magnification; while the right image shows a snapshot of a Mia Paca2 spheroid treated with 40 μg/mL scFv-hERG1-D8Cys taken at 10× magnification. In fact, it was not possible to obtain images of Mia Race 2 control spheroids at 72 h at 10× magnification because the volume was too magnified to provide proper focus; while Mia Race 2 spheroids at 72 h treated with scFv-hERG1-D8Cys can be visualized using 10× magnification because their volume is greatly reduced compared to the control.

Panel D shows MDA-MB 231 spheroids: a noticeable volume reduction effect is observed for all three concentrations of scFv-hERG1-D8Cys (10, 20, 40 μg/mL) compared to control. The reduction in volume can also be inferred from the images of the MDA-MB 231 spheroids shown on the right side of the figure.

In fig. 7 shows the results obtained from the Calcein AM cell viability assay performed on spheroids after 72 hours. The green color represents living cells while the red color represents dead cells. The image on the left (panel A) is the control images for each cell line, while the right (panel B) is the images of spheroids treated with 40 μg/ml scFv-hERG1-D8Cys. From the image, one can note a decrease in volume for antibody-treated spheroids, especially for Mia Race 2, MDA MB-231 and PANC-1 spheroids and, in addition, an increased number of dead cells, especially for MDA MB-231 and PANC-1 treated spheroids scFv-hERG1-D8Cys.

- 9 045298

7. scDb-hERGl-βΙ.

A bispecific antibody (bsAb) containing a single chain antibody directed against β1 integrin (scFv-TS2/16) and scFv-hERG1-Cys or scFv-hERG1 (as described above) was developed.

The nucleotide sequence encoding the VL domain of TS2/16 is SEQ ID NO: 23; the nucleotide sequence encoding the VH domain of TS2/16 is SEQ ID NO: 25 (see FIG. 8); accordingly, the amino acid sequence of VL TS2/16 is SEQ ID NO: 24 and the amino acid sequence of VH TS2/16 is SEQ ID NO: 26.

The bispecific antibody format is a single chain diabody (scDb) that includes the variable domains (VH and VL) of two antibodies linked by peptide linkers, as shown in FIG. 9. The top panel of the figure shows two single chain antibodies, anti-hERG1 scFv and anti-β1 integrin TS2/16, scFv.

The bottom panel is a schematic representation of the final structure of the anti-hERG1-β1-integrin bispecific antibody, which was assembled using the variable domains of the two antibodies in the following order: VH scFv-hERG1 antibody (SEQ ID NO: 8 or SEQ ID NO: 2), VL scFv- TS2/16 antibody (SEQ ID NO: 24), VH scFv-TS2/16 antibody (SEQ ID NO: 26), VL scFv-hERG1 antibody (SEQ ID NO: 4).

Antibody VL scFv-TS2/16 (SEQ ID NO: 24) and antibody VH scFv-TS2/16 (SEQ ID NO: 26) are linked by a peptide linker.

Antibody VH scFv-hERG1 (SEQ ID NO: 2 or 8) and antibody VL scFv-TS2/16 (SEQ ID NO: 24) are linked by a peptide linker.

Antibody VH scFv-TS2/16 (SEQ ID NO: 26) and antibody VL scFv-hERG1 (SEQ ID NO: 4) are linked by a peptide linker.

FspI and AvrII restriction sites were inserted at the 5' and 3' ends (underlined below)

VLFspI:

AAAATGCGCAGACTACAAAGATATTGTGATGACACAGAC (SEQ ID No: 27)

VHAvrII:

GGGGCCTAGGATAGACAGATGGGGGTGTCGCGACACCCCCATCTGTCTAT (SEQ ID No: 28).

The following sequence (SEQ ID NO: 29) is the complete nucleotide sequence encoding scDb-hERG1-e1 (SEQ ID NO: 30): the VH sequence of scFv-hERG1-Cys is highlighted in gray, the VL sequence of scFv-TS2/16 is highlighted in underline in italics, the linker sequence A, M, B is highlighted in bold, the VH sequence of the scFv-TS2/16 antibody is underlined in bold italics, the sequence scFv-hERG1-Cys is underlined in gray.

The Myc tag is in bold italic and the His tag is in bold underline. Restriction sites are underlined.

- 10 045298

SEQ ID No. 29:

GAGGCTGAGTGCGCAGACGAGGTCCAACTGCAACAGTCTGGACCTGAACTGG TGAAGCCTGGGGCTTCTGTGAAGATATCCTGCAAGACTTCAGGATACACATTCACTG AATACACCGTTCACTGGGTGAAACAGAGCCATGGAAAGAGCCTTGAATGGATTGGA GGCATTAATCCTAATGGTGGTACTACCTATAATCAGAAGTTCAAGGGCAAGGCCAC ATTGACTATTGACAAGTC CTCCAGCTCAGCCTTCATGGAGCTCCGCAGCCTGACATC TGAGGATTCTGCAGTCTATTACTGTGCAACAGGTTGGGGACCTGACTACTGGGGCCA AGGCACCACTCTCACAGTCTCCTCAGCCAAAACAACACCCCCATCAGTCTATCCACT GGCCCCTGGCTCGAGTGA TA TTGTGA TGACACAGACTCCAACCACCA TGGCTGCA TCTC CCGGGGACAAGA TCACT A TCACCTGCAGTGTCAGTTCAA TTA TAAGTTCCAATTACCTGCA TTGGTA TAGTCAGAAGCCAGGA TTCTCCCCTAAACTCTTGA TTTA TAGGACA TCCAA TCTG GCTTCTGGAGTCCCACCTCGCTTCAGTGGCAGTGGGTCTGGGACCTCTTACTCTCTCACA A TTGGCACCA TGGAGGCTGAAGA TGTTGCCACTTACTACTGCCAGCAGGGTTCTGA TATTC CACTCACGTTCCGGTGA TGGGACCAAGCTGGACCTGAAACGGGCTGA TGCTGCACCAACT GTHTUCGGTGGTGGTGGTTCTGGTGGTGGTGGTTCTGGGCGGCGGCGGCTCCGGT GGAGGAGGAACCGAGGTGAAGGTGGTGGAATCTGGGGGAGGCTTAGTGAAGCCTGG AGGGTCCCTGAAACTCTCCTGTGCAGCCTCTGGATTCACTTTCAGTAGCTATAC CATG TCTTGGGTTCGCCAGACTCCGGAGAAGAGGCTGGAGTGGGTCGCAACCATAAGTAGT GGTGGTTCTTACACCTACTATCCAGACAGTGTGAAGGGCCGATTCACCATTTCCAGAG ACAAAGCCAAGAACACCCTGTATTTTGCAAATGGGCAGTCTGAAGTCTGAGGACACAGC CATGTATTACTGTACAAGAATAGGTTACGACGAAGATTATGCTATGGACCACTGGGGT CAAGGA ACCTCAGTCACCGTCTCCTCAGCCAAAACGACACCCCCATCTGTCTATAGTG CACTGGATATTGTGCTGACACAATCTCCACTCACTTTGTCGGTTAACATTGGTCAACC AGCCTCTATCTCTTGCAAGTCAAGTCAGAGCCTCTTATATACTAATGGAAAACCTA TTTTAATTGGTTATTACAGAGGCCAGCCAGTCTCCAAAGCGCCTAATCTATCTGGT GTCTAAACTGGACTCT GGAGTCCCTGACAGGTTCACTGGCAGTGGATCAGGAACAG ATTTTACACTGAAAATCAGCAGAGTGGAGGCTGAAGATTTGGGAGTTTATTACTGCG CGCAAGGTACACATTTTCCGTGGACGTTCGGTGGAGGGACCAAGCTGGAAATCAAA CGGGCTGATGCTGCACCAACTGTATCCGCGCCGCAGAACAAAAACTCATCTCAGAA GAGGarCrGAATGGGGCCCCTAGGCATCAT CACCATCACCATCATCACTAATAG

The sequence was cloned using the restriction sites indicated above into the commercially available vector pPIC9K (Life Technologies).

8. scDb-hERG1-Phe-e1: expression and characterization

The anti-hERG1-Phe-e1-scDb antibody expression construct was cloned into the expression vector pPIC9K, which is a vector suitable for expression in Pichia pastoris yeast cells.

Pichia pastoris strain GS115 was transformed according to the spheroplasty protocol, and 96 clones were screened on YPD agar plates containing G418 for selection. Six clones were then induced on a small scale and purified using Sepharose Ni beads (GE Healthcare) using the histidine tag introduced with the pPIC9K vector. Coomassie staining is shown in Fig. 11, panel A, and shows one band, highlighted by an arrow, with a molecular mass (about 60 kDa) matching that expected for the anti-hERG1-Phe-β1-scDb antibody corresponding to clone G5.

Large-scale expression of the anti-hERG1-Phe-β1-scDb clone G5 was then performed, adapting the induction protocol for large culture volumes.

The supernatant obtained from 1 L of Pichia pastoris cell culture was purified using AKTA Pure (GE Healthcare). The results are presented in Fig. 11, which shows both the chromatogram obtained from antibody purification (panel B) and the Coomassie stain (panel C), where the elutions underlying the blue region were analyzed. Consistent with what was expected, one band corresponding to purified antihERG1-Phe-e1-scDb was detected for each elution.

- 11 045298

AHTu-hERGl-ei-scDb fraction 8; 9; 10; eleven; 12; 13; 14; 15; 16; 18; 20 individuals were pooled together and dialyzed against PBS 1X. Thus, detailed characterization of the antibody was begun.

One of the important steps was the selection of a suitable model for testing the anti-hERG1-Phe-β1scDb antibody. The table below summarizes the expression profile associated with hERG1 and β1 integrin for the cell lines selected for characterization experiments.

Table 4

Expression profile of HEK 293 hERG1, HEK 293 WT and GD25, ________________relative to hERG1 and β1-integrin________________

EXPRESSION HERG1 βΐ-integrin expression HEK 293 HERG1 + + NEK 293 WILD TYPE - + GD25 - -

BsAbs were first analyzed in hEKG1 HEK 293 cells, which express both hERG1 and β1 antigens. A cell ELISA was performed and the results are shown in FIG. 12. Cellular ELISA showed a distinct dose-dependent proportionality for native antigen binding, with a higher OD450 for cells expressing both hERG1 and β 1 antigens, as expected.

Moreover, the bispecific antibody against hERG1-e1-scDb showed the ability to bind antigen under native conditions, as in the case of antigen endogenously expressed by cells. The binding specificity of the anti-hERG1-Phe-β1-scDb bispecific antibody is also confirmed by comparing the same amount (0.5 μg) of anti-hERG1-Phe-β1-scDb and anti-scFv-hERG1-Phe, which is one of two single chain antibodies, which form a bispecific antibody. In fact, the signal obtained after incubation with anti-hERG1-Phe-β1-scDb is higher than the signal obtained using anti-scFv-hERG1-Phe. This result is as expected since the signal obtained by anti-hERG1-Phe-β1-scDb results from the binding of both hERG1 and β1 antigens; whereas the signal obtained using scFv-hERG1 results only from binding to the hERG1 antigen.

The immunoreactivity of the anti-hERG1-Phe-β1-scDb antibody was also assessed by IF on cells grown on BSA (Fig. 13, panel A and B) and fibronectin (FN) (Fig. 13, panel C and D) substrates. In fact, β1 complex formation has been shown to be enhanced by FN-dependent integrin activation. As can be seen from Fig. 13, Panels C and D, display a strong membrane signal in HEK293-hERG1 cells plated on fibronectin due to strict complexation. The signal was analyzed using ImageJ software and the results are plotted.

To further confirm the evidence obtained from previous experiments, the binding of anti-hERG1-Phe-β1-scDb was assessed on HEK293-hERG1 cells given, before antibody incubation, an excess of the S5PORO peptide targeted by the scFv-hERG1 antibody. As can be understood from Fig. 14, panels A and B, in HEK293-hERG1 cells incubated with anti-hERG1-Phe-β1scDb, the signal is confirmed due to binding to both hERG1 and β1-integrin. Panel C shows the negative control, while panels D and E show the results obtained after incubation with the S5PORO peptide; It is clear to see that the signal is decreasing, which is as expected. In fact, HEK293-hERG1 cells that are positive for both antigens show a decrease in staining intensity after incubation with the peptide, likely due to saturation of hERG1 antigen binding sites; thus, the signal that is visible is the signal resulting only from binding to the β1 antigen. Such results are summarized in a graph obtained from fluorescence intensity quantification using ImageJ.

9. scDb-hERG1-Cys-e1: expression and characterization.

A construct expressing the scDb-hERG1-Cys-e1 antibody cloned into the expression vector pPIC9K, which is a vector suitable for expression in Pichia pastoris yeast cells transformed into GS115 yeast cells.

Clones resulting from scDb-hERG1-Cys-e1 transformation were screened according to the protocol previously described for the scDb-hERG1-Phe-e1 antibody.

The supernatant obtained from 1 L of Pichia pastoris cell culture was purified using ACT Pure (GE Healthcare). The results are presented in Fig. 15, which shows both the chromatogram obtained from antibody purification (panel A) and Coomassie staining (panel B). Elutions corresponding to the blue region were analyzed and, consistent with what was expected, a single band with a molecular mass of approximately

- 12,045,298 kDa, corresponding to purified scDb-hERGl-Cys-βΙ.

After successful protein purification, the antibody was tested for direct immunofluorescence (IF) after direct conjugation to Alexa488. The results are presented in Fig. 16 for GD25 WT, HEK 293 WT and HEK 293-hERG1 cells. Images show that GD25 WT cells, panel A (negative for hERG1 and β1-integrin expression), do not show significant staining after incubation with the scDb-hERG1-Cys-e1 antibody, while panel B shows clear membranous staining for HEK cells 293-hERG1 (which express both antigens) plated on fibronectin (FN), which enhances the formation of the hERG1-e1 complex, with a higher fluorescence signal value (~17 A.U.) compared to cells plated on BSA used as a control ( ~10 A.U.). Panel C shows fluorescence staining obtained on HEK 293 WT cells (which express only β1 integrin), showing higher fluorescence signal values for cells plated on FN (~12 A.U.) compared to cells plated on BSA (~7 A.U.). IC50 was determined for both cell lines as shown in FIG. 17, panels A and B. Effects on cell viability were evident at 24 μg/ml for PANC-1 cells and 42 μg/ml for MDA-MB 231 cells. These results are consistent with the expression pattern of hERG1, which is expressed predominantly in PANC- cells. 1, compared to MDA-MB 231 cells.

Accordingly, the effect of scDb-hERG1-Cys-e1 on the migratory behavior of cancer cells was tested through lateral motility assay. Experiments were performed on MDA-MB 231, MDA-MB 231-hERG1, PANC-1 and HCT116 cells. The results are presented in the graphs in Fig. 18. There is a clear decrease in the motility index (MI) in the treated cells compared to the control. This effect is more pronounced in MDA MB 231-hERG1 compared to MDA-MB 231 cells, indicating a hERG1-dependent effect of the antibody on cell migration.

Encouraging results were also obtained in PANC-1 and HCT116 cells, with decreased motility in treated cells compared to controls.

Materials and methods

10. Cloning of the heavy and light chain of the antibody to hERG1. Anti-hERG1 monoclonal antibody heavy and light chain (hERG1-mAb) were isolated from cDNA derived from mRNA purified from hERG1-mAb-secreting hybridomas. To amplify the VH and VL regions, a 5' primer that hybridizes to frame 1 (FR1) of the variable domain of each strand (forward primer) and a primer that hybridizes to the constant region adjacent to the variable domain of each strand (reverse primer) were selected. A degenerate primer was designed for VL that hybridizes to the kappa light chain because it is an immunoglobulin phenotype more pronounced in mice (Honjo and Alt, 1995). The heavy chain (VH) of the antibody was amplified by PHP using the following set of primers:

degVH forward, 5 'GAGGTCCARCTGCAACARTC 3' (SEQ ID NO: 11) and

IgG2 reverse, 5'AGGGGCCAGTGGATAGACTGATGG 3' (SEQ ID NO: 12) (Wang, 2000).

The following set of primers was used to PCR amplify the light chain (VL) of the antibody: degVL (K), 5' GAYATTGTGMTSACMCARWCTMCA 3' (SEQ ID NO: 13) and

K reverse, 5'GGATACAGTTGGTGCAGCATC 3' (SEQ ID NO: 14) (Wang, 2000).

cDNA was amplified using high quality Phusion® DNA polymerase (Finnzymes reagents). Cycling conditions: initial melt at 94°C for 2 min, followed by 25 cycles of a three-stage program (94°C, 30 s; 56°C (VH); 48°C (VL), 1 min and 72°C, 1 minute Reactions were then carried out at 72°C for 10 minutes and cooled to 4°C.

Antibody fragments (VH and VL) isolated from agarose gel electrophoresis were purified using the QIAquick PCR Purification kit (QIAGEN) and then inserted into the pCR™Blunt vector (Invitrogen) following the manufacturer's instructions. The recombinant plasmid was sequenced using the automated DNA sequencing service (PRIMM).

The VH and VL fragments were then cloned into the expression vector pHENIX, which contains a linker sequence (Gly4Ser)3 between two different cloning sites. Primers with appropriate restriction sites were designed to clone antibody fragments into the pHenIX vector.

VL primers:

straight VL-ApaLI, 5' acgcgtgcactgGATATTGTGCTGACACAATCTCCA 3' (SEQ ID NO: 15); reverse VL-NotI, 5'ataagaatgcggccgcGGATACAGTTGGTGCAGCATC 3' (SEQ ID NO: 16).

VH primers:

direct VH-Salk, 5' acgcgtcgacGAGGTCCAACTGCAACAGTC 3' (SEQ ID NO: 17);

reverse VH-XhoI, 5'ccgctcgagccAGGGGCCAGTGGATAGACTGATGG 3' (SEQ ID NO: 18).

PCR products were digested with either ApaLI and NotI (for VH) or SalI and XhoI (for VL) restriction enzymes (New England BioLabs) and ligated into the pHENIX vector at compatible cloning sites. Hydrolysis was carried out for 2 h at 37°C. To avoid re-ligation of compatible ends, the 5' phosphate group was removed from the 5' end of the vector using calf intestinal phosphatase (CIP) as follows.

- 13 045298 using the following protocol: pHENIX vector (50 ng/μl), buffer 3 (New England BioLabs) 1X and CIP (0.5 units/μg vector). The dephosphorylation reaction was incubated for 1 h at 37°C. The phosphorylated vector was purified using a QIAquick PCR purification kit (QIAGEN).

Ligation between the scFv-hERG1 fragment and pHENIX was performed in a mixture of buffer 2 (New England BioLabs) and T4 ligase. The vector: scFv ratio was set to 1:3 and 1:10 in the ligation mixture, and incubation was carried out for 15 min at 25°C.

µl of the ligation mixture was electroporated into E. coli TOP10F' and HB2151 cells (pulse 2500 mV). Electroporated cells were recovered using 450 μl of SOC medium (SOB medium supplemented with 1 mM MgSO4, 1 mM MgCl ₂ ) and incubated for 1 h at 37°C with shaking. Bacteria were plated on prewarmed LB agar plates containing antibiotic and incubated, lid down, overnight at 37°C.

11. Cloning of scFv-hERG1-G3 into the pPIC9K expression vector.

The scFv-hERG1 expression cassette was cloned into the transformation vector pPIC9K (kindly provided by Prof. Ermanno Gherardi, University of Pavia), which contains a 6xHis tag. The scFv construct was isolated and amplified from the pHENIX vector by PCR using primers that add FspI and AvrII restriction sites, respectively, at the 3' and 5' ends of the sequence (forward VH-FspI, AAAATGCGCAGAGGTCCAACTGCAACAGTC (SEQ ID NO: 19);

reverse VL-AvrII, GGGGCCTAGGGGATACAGTTGGTGCAGCATC (SEQ ID NO: 20)).

The pPIC9 vector consists of the AOX1 promoter, the 3' AOX1 transcription terminator (TT), and a multicloning site into which the foreign gene is inserted.

The expression cassette was cut with FspI and AvrII and cloned into pPIC9K cut with Eco53K1 and AvrII (New England BioLabs).

12. Mutagenesis of scFv-hERG1

Mutagenesis was performed on the scFv-hERG1 expression cassette cloned into pPIC9K using the QuikChange® XL Site-Site Mutagenesis Kit (Stratagene, Agilent Technologies). Suitable primers for introducing the Cys amino acid were designed according to the manufacturer's instructions and developed by Primm Biotech, left primer: GGATTCTGCAGTCTATTACTGTGCAACAGGTTGGGGACCTG (SEQ ID NO: 21);

right primer: CAGGTCCCCAACCTGTTGCACAGTAATAGACTGCAGAATCC (SEQ ID NO: 22)).

The reaction sample was prepared as follows: 5 μl of 10X reaction buffer; 1 µl scFv-hERG1 dsDNA template (13 ng/µl); 1.84 µl (125 ng) left primer; 1.84 µl (125 ng) right primer; 1 µl dNTP mixture; 3 µl QuickSolution; 36, 32 µl ddH2O. Then 1 μl of PfUTurbo DNA polymerase (2.5 U/μl) was added. The cycling conditions were adjusted: the initial melting was carried out at 95°C for 1 min, followed by 18 cycles of a three-stage program (95°C, 50 s; 60°C, 50 s; and 68°C, 4 min). The reaction mixture was then kept at 68°C for 7 minutes and cooled to 4°C.

After the amplification reaction, 1 μl of restriction enzyme DnpI (10 U/μl) was added directly to the reaction mixture, which was immediately incubated at 37°C for 1 h to hydrolyze the primary.

At this stage, bacterial ultracompetent DH5a cells were transformed by heat shock. Cells were carefully thawed on ice, and 2 μl of DpI-treated DNA was transferred into a separate 200 μl aliquot of ultracompetent cells. The reaction mixture was incubated on ice for 30 min. The tube was then subjected to a heat pulse at 42°C in a dry bath for 45 s. The tube was incubated on ice for 2 minutes. Cells were recovered with 450 μl of SOC medium (SOB medium supplemented with 1 mM MgSO4, 1 mM MgCl2) and incubated for 1 h at 37°C with shaking. Bacteria were plated on prewarmed LB agar plates containing the antibiotic ampicillin (50 μg/ml) and incubated, lid down, overnight at 37°C.

The next day, several colonies had grown and some were selected and the DNA was extracted and sequenced to check for the presence of the desired mutation. The resulting construct was named scFv-hERG1-Cys.

13. Expression of scFv-hERG1-G3 and scFv-hERG1-D8Cys in Pichia pastoris.

Linearized scFv-hERG1 and scFv-hERG1-Cys were digested with SalI and transformed into Pichia Pastoris strain GS115 by spheroplasting to generate Mut + transformants. During transformation, we consulted the directions of the Pichia Expression Kit (Invitrogen).

Five days after transformation, individual colonies were visible to the naked eye, 92 clones and 4 negative controls were selected and transferred to three different 96-well plates with different concentrations of G418: no G418, 5 mg/ml, 15 mg/ml. G418 was selected using the characteristic that pPIC9K contains the bacterial kanamycin gene, which confers resistance to Geneticin® in Pichia. The level of resistance to Geneticin® is approximately dependent on the number of integrated kanamycin genes. One copy of pPIC9K integrated into the Pichia genome confers resistance to Geneticin® to levels of ~0.25 mg/ml. Multiple integrated copies

- 14 045298 pPIC9K can increase the level of resistance to Geneticin® from 0.5 mg/ml (1-2 copies) to 4 mg/ml (7-12 copies). Due to the genetic relationship between the kanamycin gene and the expression cassette (both under the PAOX1 promoter), it can be concluded that Geneticin®-resistant clones contain multiple copies of the gene of interest. For the same reason, secreted protein expression may increase due to a gene dosage effect. Therefore, the presence of pPIC9K was used as a tool to identify pPIC9K transformants that contain multiple copies of the genes of interest, scFv-hERG1 and scFv-hERG1-Cys.

After two days of growth at 30°C, the six best-grown clones from the 15 mg/ml G418 plates were selected and assessed for their ability to express the protein of interest by establishing a small-scale liquid culture according to the Pichia Expression Kit (Invitrogen) protocol.

Samples from each clone culture were collected at different time points: 24 h, 48 h, 72 h. After three days of induction with 100% methanol at 0.5% final concentration, supernatants were collected and tested by blotting.

14. Slot blot analysis.

Yeast supernatants were collected and tested for protein expression by blotting; 200 μl of each supernatant was applied to a PVDF membrane (Amersham) placed in a slot blotting apparatus between two squares of 3 mm Whatman paper. The samples were left in the incubator for 15 min, then a vacuum was used to dry the samples. The membrane was removed and washed with T-PBS. Blocking was done with T-PBS 5% BSA for 45 min and then washed for 10 min with T-PBS. The membrane was incubated for 1 h with anti-6xHis-HRP conjugated antibody (Sigma), diluted 1:2000, in 15 ml T-PBS 5% BSA.

15. Purification on Ni-Sepharose.

Supernatants obtained from yeast transformation clone screening were incubated overnight with Ni Sepharose 6 Fast Flow (Ge Healthcare) according to the manufacturer's instructions. This was followed by two washing steps using 500 μl wash buffer (20 mM sodium phosphate, 500 mM NaCl, pH 7.3) and elution using 250 μl elution buffer (20 mM sodium phosphate, 500 mM imidazole, pH 7. 3).

16. Cleaning ACT.

Purification of 1 liter of yeast supernatant scFv-hERG1-G3 and scFv-hERG1-D8Cys, respectively, was performed by affinity chromatography using the AKTA Protein Purification System (Ge Healthcare Life Sciences) with 1 mL HisTrap HP columns. The washing and equilibration steps were performed according to the manufacturer's instructions using wash buffer (20 mM sodium phosphate, 500 mM NaCl, pH 7.3); elution was performed using a linear gradient of elution buffer (20 mM sodium phosphate, 500 mM NaCl, 500 mM imidazole, pH 7.3). The analysis was performed using UNICORN 7.0 software.

17. Gel filtration.

Samples obtained from purification of both antibodies were subjected to gel filtration using Superdex 75 HR 10/30 (Ge Healthcare Life Sciences). The composition of the wash buffer (20 mM sodium phosphate, 150 mM NaCl, pH 7.3) was adjusted to optimize protocol conditions. Elutions were analyzed by SDS-PAGE.

18. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE).

Each sample was applied with the same volume of 15 μl onto a stacking gel (400 μl acrylamide (40%) - bisacryamide (0.8%), 1 ml 0.5 M Tris-HCl, pH 6.8, 40 μl 10% SDS, 20 µl 10% ammonium persulfate, 4 µl TEMED, 2.54 ml H2O). The stacking gel was added to the separating gel (2.6 ml acrylamide (40%) - bisacryamide (0.8%), 1.75 ml, 1.5 M Tris-HCl, pH 8.8, 70 μl 10% SDS, 35 µl 10% ammonium persulfate, 3.5 µl TEMED, 2.55 ml H2O). Electrophoretic runs were performed at 150 V. Gels were stained with Coomassie Brilliant Blue or transferred to PVDF membranes for Western blotting to assess the presence of protein (∼30 kDa).

19. Western blotting.

After SDS-PAGE, gels were transferred to PVDF membrane (Amersham) in transfer buffer (14.4 g, 3.03 g TrisHCl, 200 ml methanol, 800 ml H2O) at 100 V for one hour. Membranes were washed in TPBS (PBS 0.1% Tween) and then blocked with T-PBS 5% BSA overnight. Membranes were exposed to peroxidation-coupled primary antibody (Sigma) diluted in T-PBS 5% BSA for one hour at room temperature. After washing the membranes three times for ten minutes, signals were visualized using ECL reagent (Amersham).

WB was performed using the following antibodies: anti-myc (1:1000) and anti-6xHis-HRP antibody (Sigma).

20. Quantification of scFv-hERG1-G3 and scFv-hERG1-D8Cys, ELISA assay and Biacore assay.

scFv-hERG1-G3 and scFv-hERG1-D8Cys were pooled together and dialyzed against PBS 1X using Slide-A-Lyzer™ dialysis cassettes (Thermo Fisher). The protein absorbance was measured at 280 nm and the Lambert-Beer equation was applied.

- 15 045298

To evaluate whether the two engineered antibodies, scFv-hERG1-G3 and scFv-hERG1D8Cys, have the ability to bind antigen and then examine the different affinities of the two antibodies, ELISA assays were performed using S5-pore peptide-coated plates (sequence: EQPHMDSRIGWLHN), on which the antibody targets. This peptide is the same as the one we used to screen for the anti-hERG1 A7 antibody.

21. Antibody tagging using Alexa 488.

scFv-hERG1-G3 and scFv-hERG1-D8Cys were conjugated using the Alexa Fluor® 488 Microscale Protein Labeling Kit (Thermo Fisher Scientific) according to protocol guidelines.

22. Immunofluorescence on fixed cells.

HEK 293 hERG1 (HEK293 stably transfected with the pcDNA3.1-hERG1 cDNA construct) and HEK-MOCK (HEK 293 stably transfected with the pcDNA 3.1 cDNA construct) were grown in DMEM with 10% FBS EU serum in an incubator at 37°C with 5% CO2. Cells were seeded on glass coverslips overnight and then washed once with PBS and fixed with 4% paraformaldehyde for 20 min at room temperature. Blocking was performed with 10% BSA for 2 h at room temperature. Antibody incubation was performed using scFv-hERG1-G3, scFv-hERG1D8Cys, diluted 1:20 in blocking solution and incubated for 2 and a half hours, followed by incubation against His (1:250; Abcam) overnight in blocking solution. The next day, cells were washed three times with PBS and incubated with Alexa488 anti-mouse antibody (Invitrogen) for 1 h.

In this case, scFv-hERG1-G3-Alexa488 and scFv-hERG1-D8Cys-Alexa488, diluted 1:20 in a blocking solution, were incubated overnight at 4°C.

For detection, cells were incubated with Hoechst (1:1000) for 30 min and then embedded in propyl gallate. Cells were visualized on a confocal microscope (Nikon, C1).

Immunofluorescence quantification was performed using ImageJ software: for each image, three different areas were measured and the average was calculated.

23. Immunofluorescence on living cells.

Cells were grown overnight on 60 mm dishes (Sarstedt) using an agarose ring (15 g/L) to isolate cells and minimize the volumes of reagents required for incubation. Cells were incubated with scFv-hERG1-G3-Alexa488 and scFv-hERG1-D8Cys-Alexa488 diluted 1:20 in complete culture medium in an incubator at 37°C with 5% CO2 for 4 h. Alignment was performed using Hoechst as described earlier. Cells were visualized using inverted light microscopy (Nikon, Eclipse TE300).

24. Cell viability.

Cell viability was assessed using trypan blue assay; Briefly, HCT116, MDA-MB231, MIA PACA2, HEK293 hERG1, HEK-MOCK, FLG 29.1, PANC-1, BCPC3 cells at a density of 5 χ 103 cells/well were seeded in a 96-well plate. After 24 h, the medium was replaced with 100 μl of fresh medium containing various concentrations of scFv-hERG1-D8Cys antibody (10 μg/ml and 20 μg/ml). Complete immunoglobulin, a monoclonal antibody against hERG1, was tested at a concentration of 100 μg/ml. Cells were cultured for 24 hours in a humidified incubator at 37°C and 5% CO2. After treatment, cells were separated and viable cells were counted. Experiments were carried out in triplicate.

25. 3D spheroid culture and scFv-hERG1 test on spheroids.

HEK-293 hERG1, HEK-MOCK, MDA-MB231, MIA PACA2 and PANC-1 were seeded in a 96-well plate at a density of 103 cells/well for each well on a 1.5% agarose base layer. 100 μl of fresh medium was added to each well and the cells were allowed to grow for 72 h in an incubator at 37°C with 5% CO2. After 72 h, spheroids became visible and the medium was replaced with 100 μl of fresh medium containing various concentrations of scFv-hERG1-D8Cys antibody (10 μg/ml, 20 μg/ml, and 40 μg/ml).

Images were taken using a Nikon, Eclipse TE300 every 24 hours until 72 hours when the Calcein AM cell viability assay was performed.

The volume of the spheroids was assessed by analyzing images taken at 24 h, 48 h, and 72 h using MATLAB software.

The experiment was carried out in triplicate.

26. β1-integrin mAb: reverse transcription of RNA.

TS2/16 and BV7 RNAs were reverse transcribed into cDNA using oligo (dT) primers (Invitrogen) and Superscript® II reverse transcriptase (Invitrogen) according to the manufacturer's instructions in a total volume of 40 μl.

- 16 045298

RNA component Effluent concentration Final concentration 100 Quantity 4 mcg ng/µl Oligo (dT) 12-18 500 µg/ml 25 µg/ml 2 µl dNTPmix 10 mM each 0.5 mm 2 µl H2O category PNR — — 18 µl

The mixture was incubated in a PCR amplifier at 65 °C for 5 minutes, and then quickly cooled on ice and the following components were added: ___________________________________________

Component 5x first thread buffer Stock concentration 5x Final concentration 1x Quantity 8 µl H2O category PNR — — 6 µl

The mixture was then heated at 42°C for 2 minutes after which the enzyme was added: Component Effluent Concentration Final Quantity Concentration

SuperScript™ II RT 200 U/µl 5 U/µl 2 µl

The mixture (40 μl) was then incubated at 42°C for 50 min and then the reaction mixture was inactivated at 70°C for 15 min.

27. β1-integrin mAb: isolation of variable domains using polymerase chain reaction (PCR).

To isolate the antibody variable domains (VL and VH), PCR was performed with the primers described in Wang et al. (2000) as modified (Table 2).

table 2

Primers used to isolate the VL and VH variable domains are from Wang et al., (2000) with modifications

Kappa light chain Textbook title Subsequence Direct primer ' degKappadir ' GAYATTGTGMTSACMCARWCTMCA (SEQ Ш No: 31) Reverse primer Kapparev GGATACAGTTGGTGCAGCATC (SEQ ID No: 32) Heavy chain Name Subsequence Straight degHldir CAGGTTACTCTGAAAGWGTSTG (SEQ ID No: 33) Straight degH2dir GAGGTCCARCTGCAACARTC (SEQ ID No: 34) Straight degH3dir CAGGTCCAAACTUCAGCARCC (SEQ ID No: 35) Straight degH4dir GAGGTGAASSTGGTGGAATC (SEQ ID No: 36) Straight degH5dir GATGTGAACTTGGAAGTGTC (SEQ ID No: 37) Reverse primer IgGlrev GGAAGATCTATAGACAGATGGGGGTGTCGTTTTGGC (SEQ ID No: 38)

There are two classes of light chains: kappa and lambda; but since 95% of mouse antibodies are kappa light chain (Honjo and Alt, 1995), primers specific for the kappa light chain were chosen, ignoring the lambda type. Forward primers were designed using the Framework1 (FRW1) protein sequence alignment of each variable chain region. Reverse primers were designed in the constant region (CH1) near the end of the variable domain of each chain (kappa light chain or IgG1 heavy chain). For the kappa light chain, only one pair of primers was used, whereas for the heavy chain, 5 primer pairs were used, composed of IgG1rev in combination with 5 forward primers.

The primer pair IgG1rev-degH4dir was chosen to isolate VH of both TS2/16 and BV7.

To prevent mutation caused by DNA polymerase, a high fidelity DNA polymerase with proofreading activity was used: KOD Hot Start DNA polymerase (Novagen) using the following protocol:

- 17 045298

Component YUH Buffer Concentration reserve Yuh Final concentration 1x Amount 5 µl MgSO4 25 mm 1.5 mm 3 µl dNTP 2 mM each 0.2 mM each 5 µl Direct primer 10 µM 0.3 µl 1.5 µl Reverse primer 10 µM 0.3 µl 1.5 µl DNA — — 10 ig DNA polymerase for hot start KOD 1 U/µl 0.02 U/µl 1 µl H2O PCR category - — up to 50 µl

The mixture was incubated in a PCR amplifier according to the following protocol:

Stage 1 Temperature 95 °C Time 2 min 2 95°C 20 sec 3 54 °C (VH) or 46 °C (VL) 10 sec 4 70°C 10 sec 5 70°C 5 minutes

Stages 2-4 were repeated 30 times.

28. β1-integrin mAb: cloning of variable domains without the use of restriction enzymes For the VH and VL sequence, the variable domains were cloned without the use of restriction enzymes in a vector suitable for DNA sequencing We used the TA-Cloning Kit or ZeroBlunt Cloning Kit (Invitrogen), following the manufacturer's instructions .

DNA electrophoresis and agarose gel purification.

DNA electrophoresis uses an electric field to move negatively charged DNA to a positive electrode through an agarose gel matrix. PCR products and restriction enzyme-digested DNA were run on an agarose gel (1.5% agarose in TAE buffer (Tris, acetic acid, EDTA)) stained with ethidium bromide, together with the 21og DNA ladder (NEB) marker to separate fragments of different sizes. Electrophoresis was performed at 100 V. Therefore, the band of interest was removed from the gel with a clean scalpel and purified using a QIAquick PCR purification kit (QIAGEN) following the manufacturer's instructions. Purified DNA was eluted with 30 μl of H2O for PLR.

Splicing by overlapping primer PCR (SOE-PCR) The ScFv construct was assembled by splicing by overlap extension PCR (SOE-PCR) in the order VL-linker-VH using the primers described in (Wang et al 2000) with modifications (Table 3). The polypeptide linker connecting the variable domains was designed in the form of four GGGGS repeats.

Table 3 The primer is designed to assemble the VL-linker-VH construct using SOE-PCR (VLREVSOE and VHFORSOE) and clone the sequence into the pHenIX vector (VLFORSFI and VHREVNOT) or into the scFv-hERG-pHenIX vector (VLFORXHO and VHREVAPALI). Italics indicate the regions of the primers that anneal to the template, sequences added to clone the construct in frame with the expression cassette in pHenIX or in the scFv-hERG1-pHenIX vector are highlighted in gray, restriction sites are underlined, gray - sequences added to facilitate enzyme hydrolysis, and bold font represents sequences that overlap in SOE-PCR.

- 18 045298

Table 3

Name Subsequence VLFORSFI VLREVSOE GGAGCCGCCGCCGCCAGAACCACCACCACCAGAA CCACCACCACCGGATACAGTTGGTGCAGCATC VHFORSOE GGCGGCGGCGGCTCCGGTGGTGGTGGATCCGAG GTGAAGGTGGTGGAATC (SEQ ID No: 41) VHREVNOT VLFOXHO VHREVAPALI

The protocol consists of two stages, described in Fig. 10. The first stage allows you to add: at the 3' end of the VL a sequence that encodes the first three GGGGS repeats of the linker, and at the 5' end of the VH a sequence that encodes the last two GGGGS repeats of the linker. At this stage, restriction sites will also be attached at the 5' end of the VL and at the 3' end of the VH, which will be used to clone the VL-linker-VH construct into the expression vector. The second step allows the two PCR products to be combined due to overlapping sequences (15 bp) at the 3' end of VL and at the 5' end of VH.

At the first stage, two parallel PCRs were carried out:

one with primer pair VLFORSFI-VLREVSOE and template pCRII-VL; and another with the primer pair VHFORSOE-VHREVNOT and the template pCRII-VH.

The PCR protocol using KOD DNA polymerase was performed as previously described.

In the second step, we calculated SOE-PCR using 1 µL of each reaction as template

PCR carried out in the first stage, and the primer pair VLFORSFI-VHREVNOT, following the protocol below.

Component YUH Buffer Stock concentration Yuh Final concentration 1x Amount 5 µl MgSO4 25 mm 1.5 mm 3 µl dNTP 2 mM each 0.2 mM each 5 µl primer VLFORSFI 10 µM 0.3 µl 1.5 µl primer VHREVNOT 10 µM 0.3 µl 1.5 µl VL PCR (STEP 1) — — 1 µl VH PCR (STAGE 1) — — 1 µl DNA polymerase for hot start KOD 1 U/µl 0.02 U/µl 1 µl H2O PCR category - — up to 50 µl

The mixture (50 µl) was incubated in a PCR amplifier according to the protocol below.

Stage 1 Temperature 95 °C Time 2 min 2 95°C 20 sec 3 70°C 10 sec 4 70°C 10 sec 5 70°C 5 minutes

Stages 2-4 were repeated 30 times.

29. Preparation and characterization of the single chain anti-hERG1-e1 (scDb) antibody - scDb-hERG1-e1.

The hERG1-31-scDb construct was transformed into Pichia pastoris strain GS115 according to the previously described spheroplasting procedure, and the protein was expressed and purified using the expression and purification protocol previously described for the scFv-hERG1 and scFv-hERG1-Cys antibodies.

- 19 045298

Cellular ELISA.

Cellular ELISA on live cells was performed according to Sette et al., (2013). HEK 293 WT (hERG1-/e 1+) and HEK 293-hERG1 (hERG1+/e 1+) cells were seeded to semi-confluency in a 96-well plate in DMEM plus 10% fetal bovine serum (FBS) and incubated overnight at 37 °C and 5% CO2. After three washes with PBS, anti-hERG1-β1-scDb was diluted at various concentrations in culture medium and added to the cells for two 2 hours at room temperature. The next steps were the same as described above.

Immunofluorescence (IF).

IF was performed according to the protocol that has been described previously. Coverslips were coated with BSA and fibronectin for two hours. IF was performed on HEK 293 WT (hERG1-/e 1+), HEK 293-hERG1 (hERG1+/e 1+) and GD25 WT (hERG1-/e1-) cells.

30. Preparation and preliminary characterization of the anti-hERG1-β1 single-chain diabody (scDb) - scDb-hERG1-Cys-e1 antibody.

Mutagenesis of scDb-hERG1-e1.

Mutagenesis was performed on the scDb-hERG1-e1 expression cassette cloned into pPIC9K using the QuikChange® XL Site-Site Mutagenesis Kit (Stratagene, Agilent Technologies). Suitable primers for introducing the Cys amino acid were designed according to the manufacturer's instructions and developed by Primm Biotech, left primer: GGATTCTGCAGTCTATTACTGTGCAACAGGTTGGGGACCTG (SEQ ID NO: 21);

right primer: CAGGTCCCCAACCTGTTGCACAGTAATAGACTGCAGAATCC (SEQ ID NO: 22)).

The reaction sample was prepared as follows: 5 μl of 10X reaction buffer; 1 µl scFv-hERG1 dsDNA template (13 ng/µl); 1.84 µl (125 ng) left primer; 1.84 µl (125 ng) right primer; 1 µl dNTP mixture; 3 µl QuickSolution; 36, 32 µl ddH2O. Then 1 μl of PfuTurbo DNA polymerase (2.5 U/μl) was added. The cycling conditions were adjusted: the initial melting was carried out at 95°C for 1 min, followed by 18 cycles of a three-stage program (95°C, 50 s; 60°C, 50 s; and 68°C, 4 min). The reaction mixture was then kept at 68°C for 7 minutes and cooled to 4°C.

After the amplification reaction, 1 μl of restriction enzyme DnpI (10 U/μl) was added directly to the reaction mixture, which was immediately incubated at 37°C for 1 h to hydrolyze the parent solution. At this point, bacterial ultra-competent DH5a cells were transformed by heat shock. Cells were carefully thawed on ice, and 2 μl of DpI-treated DNA was transferred into a separate 200 μl aliquot of ultracompetent cells. The reaction mixture was incubated on ice for 30 min. The tube was then subjected to a heat pulse at 42°C in a dry bath for 45 s. The tube was incubated on ice for 2 minutes. Cells were recovered with 450 μl of SOC medium (SOB medium supplemented with 1 mM MgSO4, 1 mM MgCl2) and incubated for 1 h at 37°C with shaking. Bacteria were plated on prewarmed LB agar plates containing the antibiotic ampicillin (50 μg/ml) and incubated, lid down, overnight at 37°C.

The next day, several colonies had grown and some were selected and the DNA was extracted and sequenced to check for the presence of the desired mutation. The resulting construct was named scDb-hERG1-Cys-e1.

Expression and purification of scDb-hERG1-Cvs-e1 antibody.

scDb-hERG1-Cys-e1 was transformed into Pichia Pastoris yeast strain GS115 according to a previously described spheroplasting procedure, and the protein was expressed and purified using the expression and purification protocol previously described for the scFv-hERG1 and scFv-hERG1-Cys antibodies. using AKTA Pure (Ge Healthcare). Chromatograms were analyzed using Unicorn 7.0 software.

Immunofluorescence (IF).

IF was performed according to the protocol that was described previously. Coverslips were coated with BSA and fibronectin for two hours. IF was performed on GT25 WT (hERG1-/e 1-), HEK 293 WT (hERG1-/e 1+), HEK 293-hERG1 (hERG1+/e 1+) cells according to a previously described protocol. IF was performed using scDb-hERG1-Cys-e1 conjugated to the fluorophore Alexa488.

Viability analysis.

PANC-1 cells (pancreatic ductal adenocarcinoma) and MDA-MB 231 cells (breast cancer) were seeded at 5x105 in 96-well plates and allowed to grow overnight. The next day, cells were treated with scDb-hERG1-Cys-β1 at different dilutions (0, 10, 20, 40, 100 μg/ml) and incubated with the antibody for 24 hours. Each condition was performed in triplicate.

After incubation, cells were separated and counted, and Origin software was used to determine IC50.

3D spheroid culture.

103 PANC-1 and MDA-MB 231 cells were seeded on an agarose base layer (1.5 g/L) in a 96-well plate and grown for 72 h in a humidified incubator at 37°C and 5% CO2. Then they introduced

-

Claims (21)

scDb-hERGl-Cys-βΙ (40 μg/ml) diluted in culture medium, while fresh medium without antibodies was added to wells containing cells treated as negative controls. Photographs were taken after 24 hours to monitor cell growth using a Nikon, Eclipse TE300 microscope. Lateral mobility analysis. Cells were seeded into Petri dishes with a diameter of 35 mm at an initial density of 5x105 and left for 24 hours. Lateral mobility was assessed using single-layer wound analysis (Silletti et al, 1995; Peck and Isacke, 1996). Wound width was determined immediately thereafter (0 h) by measuring the wound width at 45 fixed points. Cell motility was quantified as the motility index (MI), defined as follows: MI=l-(Wt/Wo) MI=0 indicates no cell movement, whereas MI=1 values indicate complete wound closure Biodistribution of scFv-hERG1-D8Cys and scDb-hERG1-Cys-e1 antibodies. 160 μg of each antibody was injected into two Balb/c mice, i.v. Blood samples were collected from each mouse at different time points after i.v.: 5, 10, 30 min, 1 h, 2 h, 6 h, 24 h after injection. Blood samples were processed and plasma was isolated. ELISA test on plasma samples was performed according to the protocol previously described in this section, t1/2 was calculated using Precise PK Pharmacokinetic software. ECG measurements. ECG measurements were performed before antibody administration and continuously after i.v. antibody injections over 15 min. In vivo experiments. In vivo analysis. Labeling of scFv-hERG1-D8Cys using Alexa 750: 150 μg of scFv-hERG1-D8Cys at a concentration of 2 mg/ml in a solution of PBS and 0.1 M sodium bicarbonate buffer, pH 8.3, incubated for 1 hour at 22°C mixed with 12 μl of Alexa Fluor® 750 NHS ester (succinimidyl ester) (Thermo Fisher Scientific), resuspended in DMSO at a concentration of 10 mg/ml. The reaction was blocked for 5 min on ice, and the labeled protein was purified by size exclusion chromatography on a Sephadex G25 column (Sigma) equilibrated with PBS. In vivo imaging. Three six-week-old immunodeficient athymic females Nude-Foxnl nu/nu were injected intravenously with 50 μl (1 nm dye/mouse) scFv-hERG1-D8Cys labeled with the Alexa 750 fluorophore, and fluorescence was measured 5, 10, 60 min and 24 h after administration of the antibody . One control mouse was treated with sterile PBS. All fluorescence emission spectra were measured using a photon tomograph (Biospace Lab). The thermal imager contained a laser source for fluorescence excitation (=679 nm), an emission filter (=702 nm) for fluorescence detection, and a computer for data analysis. Mouse model: The MIAPaCa-2 cell line was used for tumor cell implantation as described in Lastraioli et al., 2015. Cells were cultured in DMEM supplemented with L-glutamine (4 mM), 10% fetal bovine serum and geneticin (G418) (2 .4 mg/ml) (Gibco) at 37°C in a humidified atmosphere with 5% CO2. MIAPaCa-2-luc cells were injected into the pancreas of nu/nu mice, and the animals were monitored (as described in [17]), and 45 days after cell inoculum, mice were injected with the scFv-hERG1-D8Cys-Alexa750 antibody. CLAIM 1. Bispecific anti-hERG1 and anti-β1 integrin antibody (Ab) containing the heavy chain variable domain (VH) and the light chain variable domain (VL) from the anti-hERG1 antibody that binds the extracellular domain S5-P of hERG1 and the variable domain heavy chain (VH) and the light chain variable domain (VL) of the anti-β1 integrin antibody, which bind the extracellular domain of the β1 integrin, wherein the heavy chain variable domain (VH) of the anti-hERG1 antibody has SEQ ID NO: 8, where the residue in position 95 is Cys; the light chain variable domain (VL) of the anti-hERG1 antibody has SEQ ID NO: 4; the heavy chain variable domain (VH) of the anti-B1 integrin antibody has SEQ ID NO: 26 or SEQ ID NO: 46, and the light chain variable domain (VL) of the anti-B1 integrin antibody has SEQ ID NO: 24 or SEQ ID NO : 48. 2. The bispecific antibody of claim 1 has a format selected from the group consisting of tandem scFv, diabody format, single chain diabodies, tandem diabodies (TandAbs), and dual affinity retargeting molecules (DART). 3. The bispecific antibody of claim 2, comprising a first heavy chain variable domain (VH) having SEQ ID NO: 8, wherein the residue at position 95 is Cys, and a first VL domain having SEQ ID NO: 4, and a second domain having VH SEQ ID NO: 26, and a second VL domain having SEQ - 21 045298 ID NO: 24. 4. Bispecific antibody according to claim 3, where the domains are assembled in the following order: anti-hERG1-Cys VH domain linked by a first linker to the anti-B1-integrin VL domain, linked by a second linker to the anti-B1-integrin VH domain, linked by a third linker to anti-hERG1-Cys VL domain. 5. An anti-hERG1 antibody molecule (Ab) comprising a heavy chain variable domain (VH) having SEQ ID NO: 8, where the residue at position 95 is Cys, and a light chain variable domain (VL) having SEQ ID NO: 4, wherein said molecule has specificity for the extracellular S5-P domain of hERG1. 6. The anti-hERG1 antibody molecule of claim 5, which is a fully humanized recombinant antibody, scFv, Fab, single chain Fv form of scFv, diabody, tribody, minibody or phage display antibody. 7. The anti-hERG1 antibody molecule according to claim 6, which is a ScFv in which VH and VL are linked by a peptide linker. 8. A bispecific antibody according to claim 4 or an anti-hERG1 antibody molecule according to claim 7, where the linkers are (Gly4Ser)3 motifs. 9. Anti-hERG1 antibody molecule according to any one of claims 5 to 8, having SEQ ID NO: 10. 10. Ahtu-OERGI conjugate containing an anti-hERG1 antibody molecule according to any one of claims 5-9, which is labeled with a fluorophore or radionuclide. 11. Use of a bispecific antibody according to any one of claims 1-4 and 8, or an antibody molecule according to any one of claims 5-9, or a conjugate according to claim 10 as a drug. 12. Use of a bispecific antibody according to any one of claims 1-4 and 8, or an antibody molecule according to any one of claims 5-9, or a conjugate according to claim 10 as a diagnostic agent 13. Use according to claim 11 or 12, wherein the bispecific antibody or antibody molecule or conjugate is intended for use in the treatment or diagnosis of all those pathologies that are characterized by overexpression or misexpression of the hERG1 protein, preferably tumors, neurological diseases, endocrine diseases and neuroendocrine diseases. 14. A pharmaceutical composition containing a bispecific antibody according to any one of claims 1-4 and 8 or an anti-hERG1 antibody molecule according to any one of claims 5-9 or a conjugate according to claim 10 and at least another pharmaceutically acceptable ingredient. 15. A sequence of nucleotides encoding a bispecific antibody according to any one of claims 1-4 and 8 or an antibody molecule according to any one of claims 5-9. 16. An expression vector containing the nucleotide sequence according to claim 15. 17. The expression vector according to claim 16, where the vector is a plasmid. 18. A genetically modified microorganism for expressing a bispecific antibody according to any one of claims 1-4 and 8, or an anti-hERG1 antibody molecule according to any one of claims 5-9, containing an expression vector according to any one of claims 16-17. 19. The microorganism according to claim 18, which is a cell. 20. An in vitro diagnostic kit containing parts for simultaneous, separate or sequential use, said kit is used in the diagnosis of pathologies characterized by overexpression or misexpression of the hERG1 protein, characterized in that said kit includes a container containing an anti-hERG1-Cys antibody molecule according to any one of claims 5-9 or a conjugate according to claim 10; and/or a container containing a bispecific antibody according to any one of claims 1-4 and 8. 21. The kit of claim 20, wherein the kit also contains a container containing an intact monoclonal anti-hERG1 antibody as a reference control. -