(530f) Targeted Protein-Drug Conjugate for the Treatment of Triple-Negative Breast Cancer
AIChE Annual Meeting
2024
2024 AIChE Annual Meeting
Food, Pharmaceutical & Bioengineering Division
Biomolecular Engineering II: Disease Diagnosis and Therapeutic Interventions
Wednesday, October 30, 2024 - 2:22pm to 2:40pm
Triple negative breast cancer (TNBC) is the most aggressive breast cancer subset with limited targeted chemotherapies. TNBC is defined by the absence of the estrogen, progesterone, and HER2 hormonal receptors on the cell surface, and targeted hormonal-based therapies are ineffective. The current standard of care for these patients remains untargeted systemic chemotherapy that does not discriminate between healthy and cancerous tissue, resulting in severe dose limiting side effects. Identifying a novel biomarker for this disease remains a top priority in targeted cancer therapy.
Extracellular exposure of phosphatidylserine (PS) is tightly regulated in healthy cells. Typically confined to the inner cytosolic leaflet, once extracellularly exposed, it acts as an âeat meâ signal for macrophage phagocytosis. However, cancer (especially TNBC) cells have high externalized PS but do not undergo apoptosis, making PS a promising biomarker for TNBC therapy. The annexin A5 (ANXA5) protein is the native binding partner to PS and can be used to actively target and deliver chemotherapeutic agents to the tumor microenvironment via PS externalization.
While many advancements have been made on developing targeted chemotherapeutics, protein drug conjugates have changed the way cancer is treated. Protein-drug conjugates consist of a protein linked to a drug and, depending on the conjugation method, a linker that connects the protein and drug together. The protein acts as a delivery vehicle to localize a drug to cancer cells and vasculature through a cancer biomarker. By targeting and localizing the drug to the tumor, less drug is needed to induce cell death, resulting in less side effects.
In this research, the anticancer drug emtansine (DM1) has been conjugated to annexin A5 and evaluated for its cytotoxicity to two mouse TNBC cells (4T1 and EMT6) and to a healthy mammary cell line (MCF10A). DM1 is a potent anticancer drug that inhibits microtubule formation and induces immunogenic death cell (ICD) both in vitro and in vivo when linked to an antibody. The ability of the TNBC cells to induce immune cell death (ICD) was also investigated.
Methods
Production and characterization of the ANXA5-DM1 conjugate
Recombinant ANXA5 was produced in BL21(DE3) E. coli transfected with a pET-30 Ek/LIC/ANXA5 plasmid. The ANXA5 gene sequence was verified by DNA sequencing. The resulting protein was purified using an N-terminal polyhistidine-tag for purification by immobilized metal affinity chromatography (IMAC) with immobilized Ni2+. The histidine tag was removed by HRV-3C protease. DM1 was linked to the primary amines (lysine residues) of ANXA5 with the heterobifunctional crosslinker sulfosuccinimidyl 4-(N-maleimidomethyl)cyclohexane-1-carboxylate. The ANXA5-DM1 bioconjugate was characterized by absorbance spectroscopy, SDS-PAGE gel electrophoresis, and denaturing mass analysis. In the spectroscopic assay, the concentration of DM1 in the ANXA5-DM1 conjugate was determined by measuring the absorbance at 288 nm (peak absorbance for DM1) and then subtracting the absorbance contributed by the protein at this wavelength and taking into account the protein concentration, which was determined by the Bradford protein assay.
In vitro tests
The binding strength of ANXA5 to cells was analyzed via a modified indirect ELISA using biotinylated ANXA5. The level of biotinylation was quantified using streptavidin-horseradish peroxidase and the chromogenic substrate O-phenylenediamine (OPD). For each concentration of biotinylated ANXA5, the specific binding was obtained by subtracting non-specific binding, when no calcium was present (biotinylated ANXA5 with 5 mM of EDTA), from the total binding (biotinylated ANXA5 with 2 mM CaCl2). The dissociation constant was determined using the nonlinear regression one-site total and nonspecific binding model in GraphPad Prism version 9 software.
Cell viability was assayed by a resazurin assay (AlamarBlue assay). When plated for studies, adherent cancer cell cultures were grown to less than 75% confluence, and healthy cell lines cells were grown to 100% confluency to mimic the conditions in healthy tissue. For ANXA5-DM1 cytotoxicity studies, cells were then treated with 0-100 mM of DM1 in the ANXA5-DM1 conjugate or free DM1 for 72 h in fully supplemented growth medium supplemented with 2 mM calcium to promote ANXA5 binding. For ANXA5 cytotoxicity studies, cells were treated with 0-0.7 mM of ANXA5 for 72 h in fully supplemented growth medium supplemented with 2 mM calcium to promote ANXA5 binding. Following incubation with the drug, cell viability was assayed. The concentration to inhibit cell growth by 50% (IC50) was derived from the dose-response curves by using the sum of squared differences to fit a sigmoidal regression.
Immunogenic cell death tests were evaluated by ATP release and calreticulin externalization as follows: To determine if ANXA5-DM1 induces ATP release from cells after treatment, an ATP luminescence kit was used with cells grown in 96 well plates. Cells were treated with 0 or 10 nM of DM1 in the ANXA5-DM1 conjugate or free DM1 for 24 h. To determine if ANXA5-DM1 induces calreticulin surface expression, flow cytometry was utilized. EMT6 and 4T1 cells were seeded in 24 well plates and were treated with 0 or 10 nM of DM1 in the ANXA5-DM1 conjugate or free DM1 for 24 h. Suspended cells were incubated with FITC-labeled anti-calreticulin 30 minutes at 4°C in the dark. After incubation, cells were analyzed on a BD Biosciences Accuri C6 flow cytometer with excitation at 488 nm and a 533/30 bandpass filter.
Results
SDS-PAGE analysis showed a molecular weight increase of 6 ± 3 molecules following DM1 (MW 740 Da; MW 960 Da with crosslinker) addition (Fig. 1b, left). By deconvoluting the ANXA5-DM1 conjugate in the mass range of 35-45 kDa of denaturing intact mass analysis, a left-shifted bell curve distribution of ANXA5-DM1 is observed (Fig. 1b, right). The weighted average of the drug-to-protein ratio obtained was 3.9 molecules of DM1 to 1 molecule of ANXA5, which is within the range previously found with SDS-PAGE analysis. Purified ANXA5 was found to have high specificity and affinity for externalized PS on TNBC cells as indicated by the low dissociation constant in the low nanomolar range, while no ANXA5-PS binding was observed on healthy MCF10A breast cancer cells.
Next, we examined the cytotoxicity using a 72 h in vitro assay of the ANXA5-DM1 conjugate and free DM1 in mouse 4T1 and EMT6 triple-negative breast cancer cells and in MCF10A healthy mammary cells. For TNBC cells, the antineoplastic activity of DM1 was significantly enhanced as part of an ANXA5 bioconjugate. Measuring the inhibitory concentration of 50 percent (IC50) in a 72 h in vitro viability assay, the ANXA5-DM1 bioconjugate was more than 2 orders of magnitude more potent than free DM1 in the EMT6 and 4T1 cell lines (Fig. 1c, top). The ANXA5-DM1 conjugate was significantly less cytotoxic to healthy MCF10A cells grown to 100% confluence than the free DM1 at doses up to 48 mM. The IC50 of the ANXA5-DM1 could not be determined and exceeded 48 mM; however, the IC50 for free DM1 was 160 nM (Fig. 1c, bottom left). Compared to the two TNBC cell lines studied, the 48 mM dose of DM1 in the ANXA5-DM1 conjugate is 56,000 to 229,000 times larger than the IC50 for the noncancerous mammary cells. This indicates the ANXA5-DM1 conjugate is considerably less toxic to healthy breast cells than it is to cancer cells.
To ensure DM1 is causing the cytotoxic action of the conjugate, cytotoxicity studies of ANXA5 were carried out on TNBC cells as well as healthy mammary cells. The presence of ANXA5 on two TNBC cell lines and one healthy mammary cell line did not significantly impact viability with doses up to 0.7 mM (Fig. 1c, bottom right). At the higher IC50 of ANXA5-DM1 for the TNBC cells (0.85 nM DM1), the equivalent IC50 based on the ANXA5 concentration is 0.21 nM, assuming 4 moles DM1 per mole ANXA5; this concentration is 3300 times less than the highest ANXA5 concentration tested for free ANXA5 (690 nM). This indicates that the cytotoxic action of the conjugate is a result of the addition of the drug and not the protein alone.
Finally, we examined ANXA5-DM1's ability to induce ICD by evaluating the release of ATP into the extracellular space and calreticulin surface expression (Fig. 1d). At 10 nM, ANXA5-DM1 significantly increased ATP release. ANXA5-DM1 caused nearly 10 times more ATP released than free DM1. The untreated control had essentially 0 ATP in the extracellular space. Incubation with ANXA5-DM1 significantly increased calreticulin surface expression in comparison to the untreated control as determined by median fluorescent intensity. Although both ANXA5-DM1 and free DM1 significantly increased cell death, only ANXA5-DM1 significantly increased ICD. Additionally, ANXA5-DM1 significantly increased cell death in comparison to free DM1.
In summary, we have successfully conjugated DM1 to ANXA5 for the targeted treatment of TNBC. We find that the ANXA5-DM1 conjugate quickly initiates cell death and induces two hallmarks of ICD.
Implications
ICD has been characterized as a unique pathway to activate the adaptive immune system against cancer. Novel use of the protein-drug conjugate ANXA5-DM1 for inducing ICD in breast tumors to activate the adaptive immune system could be combined with stimulation of the immune system with a checkpoint inhibitor and an immune adjuvant to treat metastatic TNBC.