Combination non-targeted and sGRP78-targeted nanoparticle drug delivery outperforms either component to treat metastatic ovarian cancer

Abstract

Metastatic ovarian cancer (MOC) is highly deadly, due in part to the limited efficacy of standard-of-care chemotherapies to metastatic tumors and non-adherent cancer cells. Here, we demonstrated the effectiveness of a combination therapy of GRP78-targeted (TNP^GRP78pep) and non-targeted (NP) nanoparticles to deliver a novel DM1-prodrug to MOC in a syngeneic mouse model. Cell surface-GRP78 is overexpressed in MOC, making GRP78 an optimal target for selective delivery of nanoparticles to MOC. The NP + TNP^GRP78pep combination treatment reduced tumor burden by 15-fold, compared to untreated control. Increased T cell and macrophage levels in treated groups also suggested antitumor immune system involvement. The NP and TNP^GRP78pep components functioned synergistically through two proposed mechanisms of action. The TNP^GRP78pep targeted non-adherent cancer cells in the peritoneal cavity, preventing the formation of new solid tumors, while the NP passively targeted existing solid tumor sites, providing a sustained release of the drug to the tumor microenvironment.

Keywords: Antitumor immune response; Lipid nanoparticle; Metastasis; Ovarian cancer; Targeted drug delivery; women's health.

Figure 1.. Schematic of proposed in vivo mechanisms of action for combination treatment of MOC by NP[DM1] and TNP^GRP78pep[DM1].

Proposed in vivo mechanisms of action for DM1-prodrug loaded, non-targeted (NP[DM1]) and targeted (TNP^GRP78pep[DM1]) nanoparticles. TNP^GRP78pep[DM1] targets and induces apoptosis of unadhered tumor cells in the peritoneal fluid, inducing immunomodulatory involvement. NP[DM1] passively targets the solid tumor sites, providing sustained release of active drug to kill cancer cells. Created with BioRender.com.

Figure 2.. Cartoon schematics of design and synthesis of targeted nanoparticle (TNP^GRP78pep).

(a) Structures of GRP78-targeting peptide (GRP78pep; left) and GRP78pep-lipid conjugate (right) are shown. The GRP78pep-lipid conjugate contains the GRP78pep (SNTRVAP), an EG₂ linker, oligolysines (K_m), an ethylene glycol linker (EG_n), a tryptophan, and two palmitic acid lipid tails. (b) Structures of DM1 (mertansine; left), and the DM1-prodrug formulation (right) are shown. (c) Cartoon schematics of nanoparticle assembly from specified stoichiometric ratios of purified components are shown. These include the GRP78pep-lipid conjugate, PEG, bulk lipid, cholesterol, and lipophilic dye (DiD or DiR) or DM1-prodrug. (d, e) Dynamic light scattering analysis was used to determine the size of the synthesized nanoparticles. Representative results of non-targeted (NP) and targeted nanoparticles (0.75% TNP^GRP78pep, 1% TNP^GRP78pep, or 1.5% TNP^GRP78pep) with various peptide density.

Figure 3.. Cellular binding studies of GRP78pep and uptake of targeted nanoparticles (TNP^GRP78pep) by ovarian cancer cells.

(a) Cellular binding of fluorescein-labeled GRP78pep to cell surface-GRP78 expressing human ovarian cancer cell lines OVCAR5 and OVCAR8 (left), and murine ovarian cancer cell lines ID8, ID8 Trp53^−/−, and ID8 BRCA Trp53^−/− (right) is shown. All ovarian cancer cell lines demonstrated a K_d of approximately 5 μM for GRP78pep. Raji cell line, which expresses low-to-no cell surface GRP78 receptor (Burkitt lymphoma; GRP78^low) was used as a negative control. (b) Nanoparticles presenting GRP78pep (TNP^GRP78pep) were prepared with varying EG linker length (EG₀-EG₄₅) while holding 1% GRP78pep density and 3 lysines constant, and cellular uptake of TNP^GRP78pep was performed with human (left) and murine (right) ovarian cancer cells. PBS and non-targeted nanoparticles (NP) were used as controls. (c) Cellular uptake of TNP^GRP78pep, with varying oligolysines (0–3 lysines) while holding 1% GRP78pep density and EG₈ linker constant, was performed with human (left) and murine (right) ovarian cancer cells. PBS and NP were used as controls. (d) Cellular uptake of TNP^GRP78pep, with varying GRP78pep density from 0 – 1% (while holding EG8 and 3 lysines constant), by ovarian cancer cells is shown. PBS and NP were used as controls. Raji cells were used as a negative control cell line. (e) Cellular binding of TNP^GRP78pep, with GRP78pep density varying from 0 – 1.5% (while holding EG8 and 3 lysines constant), was performed with human and murine ovarian cancer cell lines (solid bars). PBS and NP were used as controls. To evaluate specificity of nanoparticle binding, competitive binding experiments were simultaneously performed in the presence of excess soluble GRP78pep (free peptide; dashed bars). In all the experiments, cellular binding and uptake was measured by flow cytometry. All binding assays were performed on ice, while uptake assays were performed at 37°C.

Figure 4.. In vitro cytotoxicity of DM1, NP[DM1], and TNPGRP78pep[DM1] against ovarian cancer cells.

(a) Cytotoxicity of DM1 against OVCAR5, OVCAR8, and ID8 Trp53^−/− ovarian cancer cell lines was determined at 72 h (left). Standard-of-care chemotherapeutics Paclitaxel (center) and Doxorubicin (right) showed similar order-of-magnitude IC50, compared to DM1, at 72 h. (IC50_DM1 ≈ 50 nM, IC50_Paclitaxel ≈ 10 nM (on OVCAR5 & OVCAR8), IC50_paclitaxel ≈ 500 nM (on ID8 Trp53^−/−), and IC50_Doxorubicin ≈ 50 nM) (b) Cytotoxicity of DM1-prodrug loaded NP (NP[DM1] with 0% GRP78pep density) and DM1-prodrug loaded TNP^GRP78pep[DM1] (with 0.75% or 1.5% GRP78pep density) was determined at 48 h (left) and 72 h (right) using ID8 Trp53^−/− cell lines. At 48 h, the IC50s were as follows: IC50_DM1 ≈ 120 nM, IC50_NP ≈ 120 nM, IC50_0.75%TNP ≈ 60 nM, and IC50_1.5%TNP ≈ 60 nM. At 72 h, the IC50s were: IC50_DM1 ≈ 110 nM, IC50_NP ≈ 40 nM, IC50_0.75%TNP ≈ 40 nM, and IC50_1.5%TNP ≈ 40 nM. (c) Pulsed cytotoxicity of NP[DM1] and TNP^GRP78pep[DM1] (with 0.75% or 1.5% GRP78pep density) against ID8 Trp53^−/− cell lines was determined at 48 h (left) and 72 h (right). At 48 h, the IC50s were as follows: IC50_DM1 ≈ 550 nM, IC50_NP ≈ 2100 nM, IC50_0.75%TNP ≈ 650 nM, and IC50_1.5%TNP ≈ 350 nM. At 72 h, the IC50s were: IC50_DM1 ≈ 350 nM, IC50_NP ≈ 900 nM, IC50_0.75%TNP ≈ 700 nM, and IC50_1.5%TNP ≈ 350 nM. For pulse assays, cells were washed after 3 hours of drug treatment, and fresh media was added for the remainder of the 48- or 72-h period.

Figure 5.. In vivo biodistribution of NP and TNP^GRP78pep.

(a) Mice were injected i.p. with red fluorescent protein (RFP)-tagged ID8 Trp53^−/− ovarian cancer cells. DiR-tagged nanoparticles (NP, 0.75% TNP^GRP78pep, 1.5% TNP^GRP78pep, or NP+1.5%TNP^GRP78pep combination treatment) were injected i.p. 6 weeks after injection of cancer cells (when sufficient peritoneal tumor burden was observed via live mouse RFP imaging). To determine colocalization and uptake of DiR-tagged nanoparticles by RFP+ metastatic solid tumors and metastatic ascites cells, mice were dissected 24 h post-nanoparticle treatment. Timeline created with BioRender.com. (b) To study colocalization of DiR-labeled nanoparticles and metastasized RFP+ solid tumors on the indicated organs, the peritoneal organs were collected, imaged for RFP and DiR fluorescence, and analyzed with ImageJ. Representative images of organs from a mouse in the combination treatment group are shown. (c) To further analyze if TNP^GRP78pep preferentially colocalized to solid tumors, relative to NP, the fluorescence of DiR-labeled nanoparticles (NP and TNP^GRP78pep) and RFP+ solid tumors on peritoneal organs was quantified via ImageJ. RFP fluorescence (x-axis) was plotted against DiR fluorescence (y-axis), and linear fit analysis was applied to the data. A greater positive slope indicates enhanced colocalization. Analysis of this quantified fluorescence demonstrated that colocalization (indicated by positive slope) occurred with all of the nanoparticle treatment groups to differing degrees. Slope for NP = 1.7; slope for 0.75% TNP^GRP78pep = 2.6; slope for 1.5% TNP^GRP78pep = 0.59; or slope for combination = 0.88. Slope determined via linear regression. As metastatic tumors did not grow on all studied organs, only organs which demonstrated RFP+ tumor are shown (omentum/pancreas, ovaries/uterus, mesentery, and fat). (d) To demonstrate the varying colocalization trends between nanoparticle formulations, the slopes identified in (c) are shown. (e) To study DiR-labeled nanoparticle uptake by RFP+ metastatic solid tumors, tumors obtained from peritoneal organs were disaggregated, and cellular uptake was detected via flow cytometry (gated for RFP+ cells). (f) To study DiR-labeled nanoparticle uptake by RFP+ metastatic ascites cells, peritoneal lavage (ascites with metastatic cells) was collected, and uptake was detected via flow cytometry (gated for RFP+ cells). Parametric Welch’s t-test, * p-value < 0.05; n = 5.

Figure 6.. In vivo efficacy of NP[DM1] and TNP^GRP78pep[DM1].

(a) Mice were injected i.p. with RFP-tagged ID8 Trp53^−/− ovarian cancer cells. Nanoparticle treatments with varying GRP78pep density (NP[DM1], 0.75% TNP^GRP78pep[DM1], 1.5% TNP^GRP78pep[DM1], or a combination treatment with NP[DM1] + 1.5%TNP^GRP78pep[DM1]) were administered i.p. on days 16, 20, 23, 27, 30, 34, 37, and 41. PBS was used as a vehicle control. All treatments contained the equivalent of 3 mg DM1-prodrug per kg mouse weight. Mice were observed for 58 days. Timeline created with BioRender.com. (b) The average weight of mice in each treatment group is shown as a marker of systemic toxicity. No statistically significant toxicity (i.e. weight loss) was observed in the treatment groups, relative to the PBS control. (c) At end of study, organs were collected and imaged for RFP to detect metastatic tumors. Final tumor burden was determined via ImageJ quantification of RFP+ tumor fluorescence. The NP[DM1]+1.5%TNP^GRP78pep[DM1] combination treatment showed statistically lower tumor burden, relative to the PBS control. (d) Organs were weighed to check for systemic toxicity. Organs from noncancerous healthy mice were also weighed, as an additional control. None of the treatments demonstrated any significant organ toxicity relative to the controls. Nonparametric, unpaired t-tests, * p-value < 0.05; n = 6.

Figure 7.. In vivo effect of NP[DM1] and TNP^GRP78pep[DM1] on peritoneal anti-tumorigenic immune cell involvement.

The levels of immune cell subsets in the peritoneal cavity were determined for the indicated treatment groups. All nanoparticle treatment groups were dosed at 3 mg DM1-prodrug per kg. PBS was vehicle control. To determine levels of immune cells in the peritoneal fluid, peritoneal lavage was performed on mice immediately after sacrifice. After processing lavage samples to remove red blood cells, immune cellspecific antibody panels were used to identify the various immune cells via flow cytometry. (a) The effect of nanoparticle treatment groups on T cell subsets (T cytotoxic, T helper, NKT, TCRɣδ+ T, and T regulatory cells) are shown. T cells were gated based on CD3+ marker, and subsets were gated as follows: cytotoxic T cells based on CD8+, helper T cells on CD4+, NK T cells on CD56+, TCRɣδ T cells on TCRɣδ+, and regulatory T cells on CD25+. Frequencies of the indicated T cell subsets were determined as a percentage of total cells collected from mice peritoneal fluid. (b) The effect of nanoparticle treatment groups on other anti-tumorigenic lymphocytes, B cells (left) and NK cells (right), are shown as a percentage of cells in peritoneal fluid. B cells were gated based on marker CD19+, and NK cells based on CD56+. (c) To evaluate the impact of nanoparticle treatment on monocyte & granulocyte cell populations, neutrophils/MDSCs (myeloid-derived suppressor cells; left), dendritic cells (DCs; center), and macrophages (right) were studied. Neutrophils/MDSCs were gated on Ly6G+Ly6C+, DCs on Ly6G− CD11b−/lo CD11c+, and macrophages on CD11b+ CD11c−. (d) To further study macrophage levels, a macrophage-specific panel of antibodies was used. Overall, macrophages were identified via F4/80+ marker (left), while macrophage subset M1 (anti-tumorigenic; center-left) was gated on CD86+CD206-, subset M2 (pro-tumorigenic; center-right) gated on CD86-CD206+, and subset M1M2 (transitioning between M1 and M2; right) gated on CD86+CD206+, shown as a percentage of cells from peritoneal fluid. Parametric unpaired Welch’s t-test, * p-value < 0.05; n = 6. Arrow indicates all groups in the direction of the arrow have noted p-value, relative to the group at the arrow’s start.