Fig 1.
Screening anti-MSLN antibodies for binding to MUC16/CA125.
Antibodies were screened via ELISA for binding to immobilized MSLN and MUC16/CA125 proteins. Human serum albumin (HSA) was used as negative control. Panel A shows that the antibody referred to herein as NAV-001 was the only antibody from the group that did not bind to MUC16/CA125 protein (P = 0.000025). To determine if MUC16/CA125 perturbs anti-MSLN binding to MSLN protein, immobilized MSLN was probed with each antibody with or without MUC16/CA125 (CA). As shown in panel B, MUC16/CA125 had no impact on antibodies binding to MSLN. All data represent a minimum of triplicate experiments.
Fig 2.
Testing MUC16/CA125 binding vs non-binding antibodies for cytotoxicity in ADC format.
Antibodies Ab-1, Ab-2 and NAV-001 were converted into SN-38 ADCs and tested for cytotoxicity using the isogenic MSLN-expressing OVCAR-3 and OV-KD cells, the latter which lacks MUC16/CA125 expression. As shown in Panel A, Ab-1 and Ab-2 had reduced OVCAR-3 target cell (CA125+) killing as compared to NAV-001 (P < 0.00011). All three ADCs killed the isogenic OV-KD cells (CA125-) at a similar magnitude confirming the negative effect of MUC16/CA125 on ADCs containing antibodies that bind it. Shown is the differential killing of each ADC when cultures were treated with 50 ng/mL of each ADC. Panel B is a schematic representation of the potential mechanism by which MUC16/CA125 suppresses ADC killing through antibody binding and reduced internalization. All data represent a minimum of triplicate experiments.
Fig 3.
Retesting MUC16/CA125 binding vs non-binding antibodies for cytotoxicity in an alternative ADC format.
Antibody 2 (Ab-2) and NAV-001 were converted into ADCs by linking the topoisomerase II cytotoxin PNU-159682 and both were tested for cytotoxicity against the isogenic MSLN-expressing OVCAR-3 (panel A and C) and OV-KD (panel B and D) cell lines. As shown in Panels A and B, a similar effect of target cell killing was observed as the SN-38 formatted ADCs shown in Fig 2A, whereby the MUC16/CA125 binding Ab-2-PNU had reduced killing in OVCAR-3 (A) as compared to NAV-001-PNU(P < 0.0033), while similar killing was observed for both against the isogenic non-MUC16/CA125-expressing OV-KD cells (B). Antibodies were tested for cellular uptake using the same cell pairs and as shown in Panels C (OVCAR-3) and D (OV-KD), Ab-2 has reduced internalization as compared to NAV-001 in OVCAR-3 cells (P < 0.00008) but similar internalization kinetics in the OV-KD cells. All data represent a minimum of triplicate experiments.
Fig 4.
NAV-001-PNU has robust bystander cell killing activity.
Panel A, structure of NAV-001-PNU’s linker and cytotoxic payload comprising the following groups: N-maleoyl-B-alanine (MA) as the end group for attachment to antibody (humanized IgG1) via reduced interchain Cys-maleimide chemistry; polyethylene glycol (PEG4) to increase ADC solubility; cathepsin B protease-sensitive valine-citrulline (VC) dipeptide; self-immolative spacer p-aminobenzylcarbamate (PAB) and dimethylaminoethanol (DMAE) for the attachment and release of the intact PNU-159682 cytotoxin. Panel B, MSLN-negative Jurkat cells were treated with NAV-001-PNU, and either co-cocultured with MSLN-positive NCI-N87 cells (red line) or cultured alone (blue line). Jurkat cytotoxicity was observed only when cells were co-cultured with NCI-N87 cells (P < 0.0075), indicating a requirement for NAV-001-PNU processing followed by PNU-159682 liberation and diffusion into the microculture environment after ADC uptake by MSLN-positive target cells. All data represent a minimum of triplicate experiments.
Fig 5.
Serum stability and pharmacokinetics.
Panel A is a bioassay measuring the liberation of the PNU-159682 (PNU) cytotoxin on MSLN-negative A549 cells after 14 days of incubation at 37°C in cynomolgus monkey plasma. NAV-001-PNU toxicity would be observed if >1% of the PNU was liberated from the ADC. As shown, NAV-001-PNU remains intact for >14 days in serum (P = 0.0021). Panel B is an ELISA confirming the bioassay data that NAV-001-PNU remains intact after 14 days of incubation in host plasma at 37°C. Samples were diluted to the theoretical input amounts in serum (50 ng/mL) and tested in ELISA for total IgG and intact ADC that showed no difference between IgG and ADC concentrations after serum incubation as observed in untreated controls (STD). Panel C shows the pharmacokinetic profile of a single-dose of 0.25 or 0.75 mg/kg NAV-001 administered to naïve CD1 mice via ELISA assay using anti-PNU capture and anti-IgG-HRP detector antibodies. All data represent a minimum of triplicate experiments.
Fig 6.
In vivo efficacy and tolerability of NAV-001-PNU in MUC16/CA125 positive and negative PDX models.
Panel 6A, PDFs used for each study were confirmed for MSLN and MUC16/CA125 expression by IHC analyses. As shown, all tumors had robust homogeneous MSLN staining, while the mesothelioma and TNBC tumors also expressed MUC16/CA125. These tumors served as models to confirm NAV-001-PNU in vivo efficacy against MUC16/CA125 expressing tumors. Panel B, tumor growth and tolerability of NAV-001 in PDX models. As shown, single-dose of NAV-001 at 0.25 or 0.75 mg/kg showed significant anti-tumor efficacy (P < 0.005), and long-term regression after 0.75 mg/kg single-dose regimen. Triangles represent day of dosing post randomization. Each cohort of each study employed 6 mice.