Nanoparticle conjugates of a highly potent toxin enhance safety and circumvent platinum resistance in ovarian cancer

Advanced-stage epithelial ovarian cancers are amongst the most difficult to treat tumors and have proven to be refractory to most cytotoxic, molecularly targeted, or immunotherapeutic approaches. Here, we report that nanoparticle-drug conjugates (NDCs) of monomethyl auristatin E (MMAE) significantly increase loading on a per-vehicle basis as compared to antibody-drug conjugates (ADCs). Their intraperitoneal administration enabled triggered release of the active MMAE toxin to inhibit tumor growth and to extend animal survival to >90 days in a cell-line xenograft model of disseminated ovarian cancer. In a patient-derived xenograft model of advanced-stage and platinum-resistant ovarian cancer, an MMAE-based NDC doubled the duration of tumor growth inhibition as compared to cisplatin. NDCs of highly potent toxins thus introduce a translatable platform that may be exploited to maximize the safety and efficacy of cytotoxic chemotherapies, combining the best features of ADCs with those of nanoparticle-based therapeutics.

as the elution phase. A refractive index detector (T-REX, Wyatt Technology) was used for copolymer analysis.
For NP characterization studies, DLS and zeta potential measurements were conducted at room temperature, using a ZS90 Malvern Nanosizer (Malevern Instruments, UK) equipped with a He-Ne laser source (633 nm). NP samples were suspended in Milli-Q water at a concentration of ~1 mg/mL. At least 3 measurements were made per sample. TEM images were obtained with a JEOL-1100 Transmission Electron Microscope (JEOL Corporation, Tokyo, Japan) upon phosphate acetic negative staining (Nano Core facility of the David H. Koch Institute for Integrative Cancer Research at MIT (Cambridge, MA)). Cells with a fluorescent signal intensity above the threshold value for untreated cells (i.e. blank) were quantified. All data were analyzed using FlowJo software (Version 7.6.2).

Cellular viability screens to determine the activity of MMAE-conjugated nanoparticles against
established ovarian cancer cell lines and primary "platinum-resistant" HGSOC cells: A2780, COV318, COV362, OVCAR4, OVCAR8, and SKOV3 cells were seeded in 96-well plates (5,000 cells/well) and allowed to adhere overnight; the free drug formulation of MMAE (MMAE), NP(MMAE), or CNP(MMAE) were then incubated with the cells at different concentrations; and, relative cellular viability was assessed after 72 h using the calorimetric MTT assay and in comparison to untreated cells. Primary cells were similarly placed in suspended cultures at a density of 10,000 cells/mL; the same treatments were added to equivalent final concentrations of MMAE; they were allowed to incubate with the cells for 72 h; and, the CCK8 colorimetric assay was used to determine relative viability as a function of treatment administration.  In vitro immunofluorescence staining: OVCAR8 cells were seeded in 6-well plates (5×10 4 cells/well) on top of coverslips and allowed to adhere overnight. The cells were then treated with the free drug formulation of MMAE (MMAE), NP(MMAE), or CNP(MMAE) for 48 h and at a fixed MMAE concentration (5 nM). PBS, empty uncoated NPs (NP) and empty coated NPs (CNP) served as control treatments and were incubated with the cells at equal volumes and/or polymer concentrations. Following the 48 h incubation with each treatment group, the cells were fixed in cold methanol (4 °C) for 30 min and then incubated for an additional 30 min with FITC-labeled anti-α-tubulin antibody (Sigma-Aldrich) at room temperature (1:100 v:v; λ ex = 488; λ em = 570 nm).
DAPI was used to stain the cell nuclei. The coverslips were mounted on slides and the distributions of α-tubulin with respect to cellular nuclei were visualized using an Olympus FV1100 confocal laser scanning fluorescence microscope (Olympus, Tokyo, Japan). Succinic anhydride (60 mg; final 0.6 mmol/mL) was combined with a mixture of Compound 1 (100 mg, 0.3 mmol) and TEA (40 mg, 0.4 mmol) in THF (10 mL). The solution was stirred at room temperature for 24 h and then concentrated by rotary evaporation. The crude material was purified by column chromatography over a silica gel, using a 25:1 v:v mixture of dichloromethane:methanol as the eluent; the purified product consisted of a clear oil and was denoted as Compound 2 (110 mg, 84% yield). 1

Supplementary Figure 3: Synthesis of Compound 3.
MMAE (55 mg, 0.8 equiv.) and hydroxybenzotriazole (HOBt) (2 mg, 0.013 mmol) were added to a stirred solution, consisting of Compound 2 (40 mg, 0. 14 mmol) in a mixture of DMF (0.3 mL) and pyridine (0.1 mL). The reaction was stirred for 24 h at room temperature and monitored by analytical HPLC. The solution was then concentrated under reduced pressure and the crude material was purified by semi-preparative RP-HPLC (using Method 2) to afford Compound 3 (49 mg, 64% yield). 1   for 2H, 5H and 1H per unit, respectively). The polydispersity index (PDI) and the Mw of the copolymers were determined by GPC, using DMF as the eluent.

Supplementary Figure 13: Synthesis of mPEG-b-PZLL-b-PASP(DET) (P).
The triblock copolymer of methoxypoly(ethylene glycol)-block-poly(carbobenzyloxy-L-lysine)- . The polymerization ratio of mPEG 114 -b-PZLL 6 was calculated from its 1 H NMR spectrum, using the known molecular weight of mPEG (3.58 ppm, 451H) and by comparing its associated peak to those of the hydrogens of the carbobenzyloxy group of PZLL (5.03 ppm and 7.32 ppm for 2H and 5H per unit, respectively). . The polymerization ratio of mPEG 114 -b-PZLL 6 -b-PBLA 30 was calculated from its 1 H NMR spectrum, using the known molecular weight of mPEG (3.58 ppm, 451H) and by comparing its associated peak to those of the hydrogens of the carbobenzyloxy group of PZLL (5.03 ppm and 7.32 ppm for 2H and 5H per unit, respectively) and to the hydrogens on the benzyloxy ester group and the main chain amide of PBLA (5.03 ppm, 7.32 ppm, and 4.67 ppm for 2H, 5H and 1H per unit, respectively).      Figure 3A in the main manuscript. The cells were plated at a density of 5k cells/well in 96-well plates; they were allowed to adhere for 24 h; and, they were subsequently incubated with each formulation at a fixed MMAE concentration and for an additional 72 h. At the end of the incubation period, relative cellular viability was determined, using the colorimetric MTT assay, and was plotted after normalization to the results obtained with untreated cells.      . PBS, empty uncoated nanoparticles (NP) and empty coated nanoparticles (CNP) served as control treatments and were incubated with the cells at equal volumes and/or polymer concentrations. Following the 48 h of incubation with each treatment group, the cells were fixed and stained with PI. Their DNA content was measured, using flow cytometry; and, cell cycle distribution was determined, using FlowJo software.   1 mg/kg CB17 SCID mice 6 1: Defined as the highest dose that did not induce >20% weight loss, distress, or overt toxicities in any of the treated animals.

Supplementary
2: Resulted in a maximum of 10.5% weight loss at day 6 after injection.
4: Maximum of 14% weight loss by day 6 post-injection; at 60 mg/kg ADC (dose equivalent of 2.1 mg/kg of MMAE), 1 of 3 mice experienced 23% weight loss by day 6 after injection 5: At 40 mg/kg ADC, the mice experienced weight loss of up to 20% within 4 days of injection, at which point 3 of 5 animals were killed. The surviving mice began to regain weight 7 to 10 days after injection and returned to initial body weight by day 16. Doses higher than 40 mg/kg were lethal.
6: Administration of 40 mg/kg ADC resulted in substantial (>20%) weight loss and was not well tolerated.
7: Separate control group used in the corresponding study.  . The mice were monitored and weighed daily; and, they were sacrificed when they exhibited > 15% loss in body weight or at 14 days after administration. The accompanying subfigures depict: A) the daily weights, B) the serology panel for biomarkers of renal function, C) the serology panel for biomarkers of hepatic function, D) the complete blood count (CBC) and E) the white blood cell differential counts; terminal blood draws were performed by cardiac puncture and major organs were harvested for H&E analysis (see Figure S26). White blood cell (WBC), hemoglobin (Hb), hematocrit (Hct), platelets (Plt), neutrophils (Ne), lymphocytes (Ly), monocytes (Mo), eosinophils (Eo), and Basophils (Ba).  introduced via IP injection into female NCr nude mice and were allowed to grow until the LUC signals from their tumors reached 1x10 7 radians (photons/sec/cm 2 /surface area; ~3 weeks). Thereafter, the animals were administered Cy7.5-conjugated CNP(MMAE) (3 mg/kg dose equivalent of MMAE) by IP injection. After 24 h, bioluminescence (LUC) and fluorescence imaging (RFP and Cy7.5) commenced, using an IVIS Caliper LS system (auto exposition mode). The in vivo biodistribution of CNP(MMAE/Cy7.5) was observed by gating on the Cy7.5 channel (λ ex = 740 nm; λ em = 820 nm). The relative locations of the tumors were visualized via in vivo imaging in the RFP channel (λ ex = 540 nm; λ em = 580 nm) and from their LUC signals upon injection of dluciferin (50 mg/kg). Upon completion of in vivo imaging, the mice were sacrificed and their organs were harvested and imaged ex vivo, using the same imaging parameters. The average photon flux in radians for the different reporter signals in each excised organ were quantified by gating on regions of interest, using Living Image Software V.4.5.2, for 3 separate mice that were similarly processed. The relative signal distribution intensities from each organ (after normalization to the signal intensities recorded from the intestines, which were the major organs from which peritoneal tumor implants were explanted) are reported in Figure 4B in the main manuscript.   ) in a PDX model of advanced-stage, platinum-resistant, and high-grade serous ovarian cancer (HGSOC) as assessed by optical imaging. LUC + primary HGSOC cells (4 million cells/mouse) were introduced via IP injection into C.B-17/Icr-SCID/Sed mice and were allowed to grow until the LUC signals from their tumors reached 1x10 7 radians (photons/sec/cm 2 /surface area; ~2 weeks). Thereafter, the animals were administered Cy7.5conjugated CNP(MMAE) (3 mg/kg equivalent dose of MMAE) by IP injection. After 24 h, bioluminescence (LUC) and fluorescence imaging (Cy7.5) commenced, using an IVIS Caliper LS system (auto exposition mode). The in vivo biodistribution of CNP(MMAE/Cy7.5) was observed by gating on the Cy7.5 channel (λ ex = 740 nm; λ em = 820nm). The relative location of the tumors was visualized via in vivo imaging of their LUC signals upon injection of d-luciferin (50 mg/kg). Upon completion of in vivo imaging, the mice were sacrificed and their organs were harvested and imaged ex vivo using the same parameters. The average photon flux in radians for the different reporter signals in each excised organ were quantified by gating on regions of interest, using Living Image Software V.4.5.2, for 3 separate mice that were similarly processed. The relative signal distribution intensities from each organ (after normalization to the signal intensities recorded from the intestines, which were the major organs from which peritoneal tumor implants were explanted) are reported in Figure 5C in the main manuscript. Note that even after resection of peritoneal implants from the serosal surfaces of the intestines of each mouse, a residual signal remained that was attributed to the presence of microscopically infiltrating tumor cells.