Engineering and validation of peptide-stabilized poly(lactic-co-glycolic) acid nanoparticle for targeted delivery of a vascular disruptive agent in cancer therapy

Abstract

Developing a biocompatible and biodegradable nanoparticle (NP) carrier that integrates drug-loading capability, active targeting, and imaging modality is extremely challenging. Herein, we report a NP with a core of poly(lactic-co-glycolic) acid (PLGA) chemically modified with a drug combretastatin-A4 (CA4), a vascular disrupting agent (VDA) in clinical development for ovarian cancer (OvCA) therapy. The NP is stabilized with a short arginine-glycine-aspartic acid-phenylalanine x3 (RGDFFF) peptide via self-assembly of the peptide on the PLGA surface. Importantly, the use of our RGDFFF coating replaces commonly used polyethylene glycol (PEG) polymer which itself often induces an unwanted immunogenic response. In addition, the RGD motif of the peptide is well-known to preferentially bind to αvβ3 integrin which is implicated in tumor angiogenesis and is exploited as the NP’s targeting component. The NP is enhanced with an optical imaging fluorophore label via chemical modification of the PLGA. The RGDFFF-CA4 NPs are synthesized using a nanoprecipitation method and are ~75 ± 3.7 nm in diameter, where a peptide coating comprises a 2–3 nm outer layer. The NPs are serum stable for 72 h. In vitro studies using human umbilical cord vascular endothelial cells (HUVEC) confirmed high uptake and biological activity of the RGDFFF-CA4 NP. NP uptake and viability reduction were demonstrated in OvCA cells grown in culture and the NPs efficiently accumulated in tumors in a preclinical OvCA mouse model. The RGDFFF NP did not induce an inflammatory response when cultured with immune cells. Finally, the NP was efficiently taken up by patient-derived OvCA cells suggesting a potential for future clinical applications.

Graphical Abstract

INTRODUCTION

Nanopharmaceuticals have been used in the clinic since the early 1990’s, and many are FDA-approved for oral, topical, or systemic administration. Nanoparticles (NPs) can potentially benefit cancer therapy by encapsulating drug molecules, increasing their efficacy and decreasing adverse side effects, as well as facilitating biological imaging studies.¹ Moreover, NP-mediated cancer cell targeting can sharply increase the drug payload inside cells via receptor-mediated endocytosis.¹ NPs usually consist of inorganic materials e.g. silica,² gold,³ iron oxide,⁴ or are formulated from organic components such as polymers,^{5, 6} lipids,^{7, 8} or peptides/proteins.^{9, 10} Some examples of clinically used NPs for cancer therapy include Doxil,¹¹ a liposomal formulation of chemotherapeutic doxorubicin for the treatment HIV-associated Kaposi’s sarcoma, hafnium oxide NPs stimulated with external radiation for electron production to treat squamous cell carcinoma,¹² and Abraxane, an albumin-particle bound paclitaxel for the treatment of advanced non-small cell lung cancer.¹³ In addition, many NP formulations are being evaluated in pre-clinical trials or are in clinical trials, for both therapeutic and diagnostic purposes.

Poly(lactic-co-glycolic acid) (PLGA) polymer has been extensively investigated as a core component of NP formulations,¹⁴ owing to its biocompatibility and biodegradability features. The erosion time of PLGA is well-controlled and depends on the lactide:glycolide ratio, which varies the crystallinity and hydrophobicity of the polymer. The higher ratios give the NPs slow release characteristics, providing a well-suited use in drug delivery applications.⁶ Therefore, PLGA-based NPs have been explored as carriers for many therapeutics, including chemotherapy, antiseptic, anti-inflammatory and antioxidant drugs, as well as antibiotics.^{15, 16}

Surface modifications of PLGA NPs are necessary to achieve targeting and increase serum stability.^{17, 18} The latter is often realized by using highly hydrophilic polyethylene glycol (PEG) as a NP’s coating material that yields stealth particle with improved pharmacokinetics. Nevertheless, a number of studies have demonstrated that cellular uptake and endosomal escape of (PEG)ylated NPs are suppressed due to the steric hindrance conferred by PEG itself, a phenomenon known as “PEG dilemma”,^19–21 which can adversely influence NP-mediated drug delivery.^{22, 23}

The use of short peptides as NP’s building blocks is appealing as they have low toxicity profiles. More than 60 peptides have already been approved for use in the clinic as therapeutics, and many more enter clinical trials yearly.^{24, 25} Peptides are cost-effective, easy to synthesize in large quantities, and offer sequence-dependent tunability to match specific applications.²⁶ As a result, peptides are receiving increasing interest as clinical therapeutics for cancer,^{27, 28} cardiovascular diseases,²⁹ and for certain metabolic disorders³⁰ and infections.³¹

Vascular-targeted therapies represent a clinically relevant option in treating many cancers, including ovarian cancer (OvCA).³² OvCA is the most lethal female reproductive tract malignancy, and in the U.S. there will be an estimated 25,000 new cases and more than 16,000 deaths this year. Most women will present with advanced disease stage and the 5-year survival rate, unchanged for the past four decades owing to the high recurrence rate, remains at ~30%.³³ Targeting existing tumor vasculature with vascular disrupting agents (VDAs) is an attractive strategy to enhance tumor ablation and improve therapeutic outcomes. Combretastatin A4 (CA4), the tubulin-binding agent,^34–37 is one of the VDAs in active development for OvCA therapy.³² The CA4 mode of action involves tubulin binding near the colchicine binding site, which damages endothelial cell cytoskeleton and cell shape change, ultimately leading to compromising blood flow to the tumor and tumor necrosis.³⁸ Although highly potent, CA4 is not soluble in water and a prodrug form, CA4-phosphate must be used in clinical applications. However, the therapy with CA4-phosphate is associated with systemic toxicity, which include cardiovascular adverse events and hypertension.³⁹

Herein, we present a simple methodology to formulate PLGA NPs coated with a short peptide, RGDFFF, and encapsulating CA4 to target angiogenesis in cancer. This work builds upon our previous report on short-peptide stabilization phenomena observed in nanoemulsions.⁴⁰ The RGDFFF peptide self-assembles on the PLGA surface due to the presence of the terminal phenylalanine (F) residues, which are known to facilitate the process.^41–44 Concurrently, the arginine-glycine-aspartic acid (RGD) moiety serves as a vascular and cancer-specific active targeting ligand,^{45, 46} well-established for therapeutic applications^{47, 48} and molecular imaging purposes.^49–53 Importantly, the key feature of RGDFFF-CA4 NP is complete biodegradability. The polymer is FDA approved,⁵⁴ and peptides are inherently biodegradable to amino acids that have generally regarded as safe (GRAS) status, which can collectively reduce barriers to clinical translation of the RGDFFF-CA4 NP.

RESULTS AND DISCUSSION

We conjugated CA4 to PLGA polymer via an ester bond (Figure 1 top) between the carboxyl group of PLGA and the hydroxyl group of CA4. The PLGA:CA4 ratio in the PLGA-CA4 hybrid (Figures S1–S4) was determined by HPLC and was found to be 90:1. The RGDFFF-CA4 NP was formed using a nanoprecipitation method.⁵⁵ To this end, the PLGA-CA4 conjugate in acetonitrile was added dropwise to an aqueous solution of RGDFFF peptide (Figures S5–S6) at 37°C, at a PLGA-CA4:RGDFFF molar ratio of 1:1.5. The solution was stirred overnight allowing the self-assembly of the RGDFFF-CA4 NPs. Next, the NPs were pre-concentrated and washed in a centrifugal concentrator. The schematic of PLGA-CA4 hybrid, as well as the RGDFFF peptide, are shown in Figure 1 (top), and RGDFFF-CA4 NP is shown in Figure 1 (bottom right) followed by a transmission electron microscopy (TEM) image of the NPs (bottom left).

Figure 1.

Schematic representation of PLGA-CA4 conjugate and RGDFFF peptide (top), and RGDFFF-CA4 NP (bottom right). The TEM image (bottom left) shows a large area with RGDFFF-CA4 NPs.

The size, morphology, and the presence of the peptide shell on the NPs have been evaluated using TEM. The NPs were spherical in shape and the average diameter was 75 ± 3.7 nm (Figure 1). This size would allow NPs to accumulate in the interstitial tumor space through the enhanced permeability and retention (EPR) effect.⁵⁶ The zoomed-in images of the empty RGDFFF (no drug), RGDFFF-CA4, and PLGA NP are presented in Figure 2. The peptide coating is visible as a 2–3 nm rim on the RGDFFF NP’s surface (Figure 2A). Drug encapsulation did not noticeably affect the NP’s morphology, and the peptide coating is apparent on the surface of RGDFFF-CA4 NP (Figure 2B). The thin layer was not present in the plain PLGA NP (Figure 2C), suggesting that the coating formed from the RGDFFF peptide.

Figure 2.

TEM images of the NPs: (A) RGDFFF NP, (B) RGDFFF-CA4 NP, (C) plain PLGA NP (no coating). The XPS spectrum of RGDFFF-PLGA NPs shows the peak of N 1s (D).

To confirm these findings, the empty RGDFFF NP (no drug) was further investigated by X-ray photoelectron spectroscopy (XPS), a surface-sensitive technique capable of providing the composition data of the uppermost 1–10 nm of the surface.⁵⁷ The NPs were deposited on a Si wafer and analyzed for nitrogen presence; nitrogen can originate only from the peptide coating and is not present in the PLGA core. The XPS spectrum of the NPs (Figure 2D) demonstrates a well-pronounced N 1s peak. This finding confirms that the NP’s coating is the RGDFFF peptide. The overall elemental composition of the surface was 1.1% N, 66% C and 32.9% O in the RGDFFF NP. The full XPS spectrum of the RGDFFF NP and control XPS spectrum of the Si wafer only, without evidence of the N 1s peak, is presented in Supporting Information (Figure S7).

The Zeta potential of the NPs was measured by dynamic light scattering at pH 7.4 and was found to be −44.2 ± 4.1 mV, indicating negative surface charge and good colloidal stability. The measured negative surface charge agrees with the isoelectric point (pI) of the RGDFFF peptide, which is 5.84. Thus, at a pH higher than pI, the overall charge of the RGDFFF will be negative while positive at lower pH. Interestingly, this charge conversion could potentially contribute to greater NP uptake by negatively charged endothelial cells in the tumor tissue, where pH is lower than physiological levels.⁵⁸ In addition, negatively charged NPs have prolonged circulating half-lives⁵⁹ and reduced non-specific interactions with plasma proteins that may induce cytotoxicity.⁵⁹

Next, we evaluated the RGDFFF-CA4 NPs for stability in serum. In a test, 1 mg/ml of the NPs was suspended in 10% mouse serum. The NP’s size was measured by DLS over time to determine any cluster formation and aggregation phenomena that would increase the size distribution of the NPs. The size did not vary for the NP preparation, as shown in Figure 3A, indicating that the NPs do not opsonize in serum. Similarly, the NPs’ stability was tested in PBS using the same conditions. The overall diameter of the NPs did not significantly change after 72 h (Figure S8) indicating good colloidal stability of the NP formulation. This result suggests that the NPs are stable and may be suitable for biological applications, which is critical for intravenous drug delivery applications.

Figure 3.

The stability of RGDFFF-CA4 NPs in 10% serum (A). Intracellular production of ROS in the presence of RGDFFF NPs. Error bars represent the mean ± standard deviation of three independent trials (B). NO production of RAW264.7 cell lines in the presence of RGDFFF NPs (C) and plain (uncoated) PLGA NPs (D). In ROS and NO experiments, the concentrations of the NPs are as follows: x 2 – 0.034 mg/ml, x 1 – 0.017 mg/ml, x 0.2 – 0.0034 mg/ml, and x 0.04 – 0.00068 mg/ml. PC – positive control (100 ng/ml LPS), NC – negative control: untreated RAW264.7 cells (PBS).

The interactions between NPs and the innate immune system can induce diverse immunological responses and may lead to immunotoxicity.⁶⁰ We therefore evaluated the compatibility of the RGDFFF NP platform (no drug) with the immune system using macrophage RAW 264.7 cells. In this experiment we measured macrophage production of excessive reactive oxygen species (ROS) and reactive nitrogen species (RNS) upon incubation with the NP. ROS/RNS are produced from cellular activity in the different organelles, such as mitochondria, endoplasmic reticulum (ER), and peroxisomes. These play a pivotal role in various cellular functions such as cellular growth, proliferation, and differentiation.⁶¹ However, an elevated ROS/RNS rate is hazardous and can cause damage to multiple cellular organelles and promote unregulated inflammation.⁶² To evaluate the effect of RGDFFF NP on ROS production, cells were exposed to different NP concentrations, where concentration x1 corresponded to the theoretical plasma concentration of CA4 if the drug was to be present in the NP (see SI for full details). The results are presented in Figure 3B and show no differences between NP treated and blank control suggesting that the RGDFFF NP platform had no effect on ROS formation by the cells. Plain PLGA NPs had a similar effect on ROS production as RGDFFF NPs (Figure S9).

To assess RNS production we measured NO levels in the culture medium of RAW 264.7 cells incubated with the RGDFFF NPs at the same concentrations as in the ROS experiment (Figure S10). The values were compared with lipopolysaccharide (LPS) (positive control) and PBS (negative control). As shown in Figure 3C, high nitric oxide levels in the presence of RGDFFF NPs were not detected. Interestingly, the effect of plain PLGA NP (no coating) on NO production was more pronounced (Figure 3D), suggesting that the RGDFFF coating may diminish NP immunogenicity.

The affinity of the RGD moiety to αvβ3 integrin, as well as the expression of αvβ3 in angiogenic blood vessel cells is known.⁴⁶ In OvCA and breast cancer, αvβ3 integrin expression is associated with malignant progression^63–66 and metastatic potential.⁶⁷ Thus, as an introduction to biological applications of the RGDFFF-CA4 NPs, we first established the expression level of αvβ3 integrin in human umbilical vein endothelial cells (HUVEC), representative of angiogenesis, as well as in selected OvCA cell lines. To this end, we used a Western blot test to verify the presence of αv and β3 integrins in HUVEC, and in an array of immortalized OvCA cell lines, including the isogenic A2780 and CP70 cell lines,⁶⁸ and ES-2, OV-90, SKOV-3, OVCAR-3, TOV-21G cells, all with characterized genomic features and different histotype origins.⁶⁹ GAPDH was used as a reference/housekeeping gene. All cells expressed αv integrin, with HUVEC, TOV-21G, SKOV-3, A2780, and CP70 cells showing slightly higher expression levels than OVCAR-3, ES-2, and OV-90 cells (Figure 4). However, the expression of β3 integrin was more heterogeneous, and the highest signal was detected in HUVEC, as expected, and also in TOV-21G cells. Some signal was detected in other cell lines, such as OVCAR-3, and ES-2, while OV-90, SKOV-3, A2780, and CP70 cells expression was too low to detect.

Figure 4.

Western blots of HUVEC and various OvCA-relevant cell line lysates demonstrating expression of αv and β3 subunits. 80 μg of protein was loaded per lane. The panels shown are representative of N=3 independent trials.

Next, we correlated these expression levels of αv and β3 subunits with the extent of NP uptake. To this end, we used fluorescent RGDFFF-Cy5.5 NPs (no drug), where PLGA was chemically modified with Cy5.5, and the ratio of PLGA:PLGA-Cy5.5 in the core was 4:1. The cells were incubated with RGDFFF-Cy5.5 NPs for 15 min, washed, collected in suspension, and then examined by flow cytometry. The experiment was performed three times in triplicate, and representative results are shown in Figure 5. Approximately 98.0% of HUVEC were demonstrated to have taken up the NP. The highest percent positivity in all other cell lines tested was observed in ES-2 (66.1%) and TOV-21G (46%) cells. This increased level corresponded to the high expression of both αV and β3 integrin subunits. Other OvCA cell lines were also positive for RGDFFF-Cy5.5 NPs, even when the expression of β3 integrin subunit in the cells was below the detection limit. This might stem from the fact that the RGD motif was found to have an affinity to other integrin proteins as well, such as αVβ5 or α5β1.^70–72 Therefore the direct comparison between single integrin expression and the NP’s uptake is challenging.

Figure 5.

Flow cytometry analysis of RGDFFF-Cy5.5 NP uptake. All cell lines were incubated with the NPs and harvested after 15 min. Control cells receiving no NPs, are highlighted in red, and cells treated with NPs are shown in blue. The table characterizes the percent of NP-positive cells.

To corroborate the RGDFFF-Cy5.5 NP uptake by HUVEC and OvCA cells, we imaged the cells using LSM 880 Airyscan Fast Live Cell confocal microscope after 48 h incubation with the NPs. The NPs (red) are visible inside all cells tested (Figure 6 top), further suggesting that the NPs are internalized by the cells. Finally, NP entry into the cells was investigated using time-lapse microscopy. As shown at the bottom of Figure 6, the NPs attach to the cell membranes of CP70 cells over the first 30 min after adding the RGDFFF-Cy5.5 NPs and progressively accumulate inside the cells.

Figure 6.

Top: Cellular internalization of RGDFFF-Cy5.5 NPs (red) in OvCA cell lines after 48 h incubation. The membrane (green) was stained with Alexa fluor 488-WGA, and the nucleus (blue) was labeled with DAPI. The scale bars are 20 μm. Bottom: Time-lapse images of RGDFFF NPs (red) uptake by CP70 cells during the first 30 min of incubation. Arrows point to the NPs. The cellular membrane (green) was stained as above. The scale bar is 10 μm.

We next sought to evaluate the biological activity of RGDFFF-CA4 NPs in cell culture. As determined by HPLC, the CA4 drug payload in the NP was 0.2 wt.% (Figure S11). First, we established IC₅₀ values for the NPs and free drug in all cell lines (Figure S12–S13), and subsequently performed MTT viability assays in HUVEC and OvCA cell lines using the respective IC₅₀ values for the NP. We compared the RGDFFF-CA4 NPs to cells only without treatment, free CA4, and empty RGDFFF NPs without CA4, as shown in Figure 7. In our experiments, RGDFFF-CA4 NPs showed cytotoxic activity. The most pronounced effect was observed in HUVEC, the primary drug target for the NP, with 20% greater impact on viability when compared to free CA4. In OvCA cell lines, RGDFFF-CA4 NPs lowered the cellular viability at levels comparable to free CA4. We also used optical microscopy to examine the morphological changes in the cells after incubation with RGDFFF-CA4 NP. This was done since CA4 is a microtubule depolymerizing agent that disrupts internal cell architecture. Upon incubation with RGDFFF-CA4 NP, cell morphology was visibly disrupted, changing from elongated and flat to rounded and bloated (Figure S14). The changes were suggestive of CA4 action and were also visible in cells incubated with free CA4. These results suggest that RGDFFF-CA4 NP can be used as efficient transport and drug delivery vehicle, acting upon vascular and OvCA cells.

Figure 7.

In vitro viability with RGDFFF-CA4 NP. Each column represents the mean and standard deviation of N=3 and p<0.005 (HUVEC, OVCAR-3), p<0.04 (TOV-21G, OV-90), p<0.03 (CP70), p<0.02 (A2780), p<0.01 (ES-2), p< 0.12 (SKOV-3). The concentrations correspond to IC₅₀ values of CA4 for each cell line and are as follows: 2 nM (HUVEC), 5 nM (A2780), 4 nM (CP70), 5 nM (ES-2), 100 nM (OV-90), 2.5 nM (OVCAR-3), 10 nM (SKOV-3), 25 nM (TOV-21G).

As a prelude to in vivo biodistribution studies, we examined the pharmacokinetic properties of the RGDFFF-coated NPs (no drugs), where PLGA was modified with Cy7 near-infrared (NIR) fluorophore. In this experiment, the PLGA:PLGA-Cy7 at the ratio of 80:20 constituted the core of the NPs, where the Cy7 was chemically conjugated to the PLGA. We used NIR fluorescence to measure the concentration of the RGDFFF-Cy7 NPs in the blood samples. Nude mice were IV dosed with NPs at a concentration of 2.66 mg/kg. This dose corresponds to NPs carrying a CA4 dose for patients (60 mg/m²),⁷³ converted to its equivalent dose for mice.⁷⁴ Blood samples were drawn at 15, 30, and 90 min. After the experimental fitting of the data, the in vivo half-life of the RGDFFF-Cy7 NPs was determined to be 45 min.

Finally, we evaluated the potential of RGDFFF-Cy7 NPs (no drugs) to target solid tumors in a subcutaneous model of OvCA. The Cy7 labeled NPs, at the same dose as in half-life studies, were IV injected via tail vein into OVCAR-3-tumor bearing nude mice. The NPs biodistribution studies in vivo were performed using an IVIS small animal imaging system. Twenty-four hours post-injection, no fluorescence signal was detected in controls, but a strong NIR signal was observed at the tumor site in the NPs injected mice (Figure 8 top left). Moreover, after four days, the intratumoral NIR signal remained strong, indicating prolonged retention of the NPs in the tumor (Figure 8 top right). At this time point, the tumors were excised and examined ex vivo by NIR fluorescence. Figure 8 (bottom) demonstrates the greater fluorescence in the tumor than in control, as established from the measured regions of interest (ROIs).

Figure 8.

The images of NPs biodistribution in vivo (top) in the OvCA mouse model and excised tumors (bottom) from a NP-injected mouse (left) and a control PBS-injected mouse (right). The images were acquired with the help of the Small Animal Imaging Center in the Translational and Molecular Imaging Institute.

In addition, we imaged tumor cross-sections to investigate the intratumoral distribution of the RGDFFF-Cy7 NPs. Frozen tumor sections were prepared using a cryostat. Standard immunostaining techniques were applied for CD31 (endothelial cells) and DAPI. Briefly, sections were blocked with donkey serum for 20 min at room temperature (RT), washed with PBS, and incubated with rat anti-mouse CD31 antibody for 45 min at RT. Slides were rinsed with PBS and incubated with secondary donkey anti-rat antibody labeled with Cy3 for 30 min, rinsed with PBS, and mounted in mounting medium with DAPI. The stained tumor sections were imaged using LSM 880 Airyscan Fast Live Cell confocal microscope using the appropriate filters and 63X magnification. The images are shown in Figure 9 and in Figure S15. The vascular endothelial cells are shown in red, and the association of the RGDFFF-Cy7 NPs (green) with tumor blood vessels is apparent. Notably, the NPs are also prominently combined with tumor cells (stained with DAPI) suggesting that the NPs effectively bind to tumor-associated vasculature and successfully extravasate into tumor interstitium. These results demonstrate that the RGDFFF-coated NPs exhibit tumor-targeting capabilities and are suitable as drug delivery vehicles for OvCA therapy. The RGDFFF-Cy7 NP biodistribution into organs was also investigated and the NIRF image is presented in Supporting Information (Figure S16). The RGDFFF-Cy7 NPs were found in the liver, spleen, and lungs, but were not detected in the brain, which indicates that the NP did not cross the blood-brain barrier. Complete in vivo efficacy studies using the RGDFFF-CA4 NP in OvCA tumor models are underway.

Figure 9.

Confocal images of frozen sections of OVCAR-3 tumors excised after 4 days from the mouse injected with RGDFFF-Cy7 NPs. The CD31 staining of vascular endothelial cells appears in red, and RGDFFF-Cy7 NPs are in green; the nuclei are stained with DAPI (blue). Scale bars represent 10 μm. The sections clearly show the association of the NPs with the blood vessels and tumor cells.

Finally, as the first step toward future in vivo therapy studies, we tested the uptake of RGDFFF-Cy5.5 NP in patient-derived OvCA cells. While commercially available cell lines have been used in pre-clinical research and have been extremely useful in advancing the understanding of cancer biology, they do not truly represent the characteristics of original tumors.^75–80 In our studies, we examined the uptake of empty RGDFFF-Cy5.5 NP (no drug) in six PDCLs: four derived from patients with high-grade serous ovarian cancer (PT217, PT927, PT1264, PT1310) and two with serous endometrial cancer (PT1315, PT1332).

Cells were incubated with the NP for 15 min, washed with PBS, and collected in suspension. The fluorescence was measured using flow cytometry, and the results are presented in Figure 10. The analysis of RGDFFF-Cy5.5 NP uptake shows more than 50% positivity for all six PDCLs [range 53 – 91%].

Figure 10.

Flow Cytometry analysis of RGDFFF-Cy5.5 NP uptake by PDCLs. Red shows blank (non-treated cell), and blue shows cells incubated with RGDFFF-Cy5.5 NPs. The X-axis is the Cy5.5 intensity, and Y-axis is the number of cells. The table characterizes the percent of NP-positive cells from three trials.

CONCLUSIONS

In summary, the proposed peptide-coating strategy allows dual action, the stabilization of the NP and targeting, without compromising NP biodegradability features. The RGDFFF-coated NPs have the size and surface charge suitable for drug delivery vehicles, effectively penetrate cells, and are biologically active. Importantly, the NP platform did not demonstrate inflammatory effects on immune cells. Additionally, the NPs effectively target solid tumors associated with tumor’s blood vessels as well as with cancer cells and are taken up by patient-derived cells. Importantly, many different cancer-targeting peptides could be alternatively used,⁸¹ to significantly enhance therapeutic outcomes in other cancers.^{81, 82} Ongoing studies investigate the therapeutic potential of RGDFFF-CA4 NPs.

MATERIALS AND METHODS

Materials.

Ninhydrin 99%; N, N – dicyclohexylurea (DCC) 99%, 4-dimethylaminopyridine (DMAP) 99%, piperidine 99%, trifluoroacetic acid 99% (TFA), and Oakwood Chemical: N,N-diisopropylethylamine (DIPEA) 99.5%, anhydrous tetrahydrofuran (THF), ethyl alcohol 190 proof were purchased from Alfa Aesar. Combretastatin A4 99% was purchased from Biotang. Phenol 99%, sodium hydroxide, phosphotungstic acid, poly(D, L-lactide-co-glycolide) (PLGA) LA: GA (50:50) mol wt. 1,000–5,000, 7,000–17,000 and 24,000–38,000 were purchased from Fisher. Fmoc-L-Phe 4-alkoxybenzyl alcohol resin (0.388 meq/g), Fmoc-L-Arg(Pbf)-OH 99.1%, Fmoc-Gly-OH 99.37%, Fmoc-L-Asp(OtBu)-OH 99.87%, triisopropylsilane (TIPS) were purchased from Chem-Impex INT’L INC. 2-(1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluronium tetrafluoroborate (TBTU) were purchased from ANASPEC INC. Chloroform, ethyl acetate, methylene chloride, anhydrous ethyl ether, 2-propanol (IPA), N,N-dimethylformamide (DMF), methanol, hexanes, were purchased from Fisher. Acetonitrile (HPLC) was purchased from Acros Organics.

Synthesis of the RGDFFF hexapeptide.

The standard solid phase peptide synthesis (SPPS) method was used to synthesize the RGDFFF peptide. Fmoc-L-Phe 4-alkoxybenzyl alcohol resin (0.5 g, 0.388 meq/g) was soaked in 25 mL of DMF for 1 h in a reaction vessel. Next, the amine group of L-phenylalanine amino acid 4-alkoxybenzyl alcohol resin was deprotected with 20% piperidine in DMF solution for 5 min followed by 20 min oscillation cycle. After deprotection, the resin was washed for 1 min with DMF and 1 min with IPA; each a total of three times. The Kaiser test was performed to determine the presence of free NH₂ group. If free NH₂ groups were detected, the coupling with Fmoc-L-phenylalanine amino acid was performed overnight. The following standard coupling conditions were used to attach each amino acid: Wang resin 0.19 mmol, Fmoc protected amino acid (0.38 mmol), TBTU 122.3 mg (0.38 mmol) and DIPEA 98.4 mg (0.76 mmol). The amino acids were dissolved with TBTU in DMF and DIPEA. The resin was washed with DMF, and IPA two times for 1 min each, followed by the Kaiser test. After coupling all amino acids, the resin was washed with 5 mL of DMF, IPA, DMF, methanol, dichloromethane and diethyl ether for 1 min each. The cleavage of RGDFFF from the resin was achieved with a TFA/TIPS/H₂O solution at a ratio of 95/2.5/2.5 for 3 hours. The crude peptide was precipitated out in cold diethyl ether, washed 3 times with cold ether and then dried under the vacuum. The product was washed with 10% HCl 3-times and dried. The peptide was purified in HPLC and characterized by HRMS (see supporting information).

Synthesis of PLGA-CA4 hybrid.

First, 200 mg (0.05 mmol) of 1–5 kDa PLGA polymer was dissolved in 2 mL of anhydrous dichloromethane, followed by addition of 41 mg (0.2 mmol) of DCC and 24 mg (0.2 mmol) of DMAP. Next, 16 mg (0.05 mmol) of CA4 was dissolved in 2 mL of anhydrous dichloromethane and added dropwise to the PLGA solution. The reaction mixture was stirred overnight at room temperature. ¹H NMR was used to monitor the progress of the coupling reaction between CA4 and PLGA. Once the reaction was complete, the crude product was dried using rotary evaporator and dissolved in a small amount of acetonitrile. Next, the solution was added dropwise to cold methanol and kept at - 20°C overnight. After that, the product was centrifuged and dissolved in acetonitrile. Last three steps were repeated 3 times to obtain a pure product. The yield of the reaction was determined to be 32%. NMR characterization of product is presented in supporting information.

Synthesis of RGDFFF-PLGA-CA4 NPs.

PLGA-CA4 (5 mg, 0.0012 mmol) was dissolved in 2 mL of acetonitrile and added dropwise using a syringe pump (0.2 mL/min flow) into RGDFFF peptide (1.5 mg, 0.0019 mmol) 6 mL solution in nanpure water at 37 °C. The solution was stirred overnight to allow the self-assembly of RGDFFF-CA4 NPs. The NPs were concentrated using Vivaspin centrifugal concentrator (100 k MWCO) and washed 3 times with nanopure water.

UV-Vis Analysis of Combretastatin A4 (CA4).

1 mg of CA4 was dissolved in 1 mL of methanol to prepare a serial dilution of 120, 60, 30 and 15 ppm concentrations. The UV-Vis spectrum of the solutions was taken at 200–400 nm in UV-vis spectrophotometer (CARY 100 Bio VARIAN) in a 3 mL of quartz cuvettes. The wavelength of maximum absorbance for CA4 was selected as 247 nm. This wavelength was used for all further measurements of CA4 concentrations.

Calibration curve for CA4.

Serial dilutions of CA4 of 100, 60, 50, 40, 30, 20, 10, 5 and 2.5 ppm were prepared in 1 mL of methanol each. The HPLC assay was performed using HPLC FLEXAR System (Perkin Elmer). The mobile phase consisted of acetonitrile:water: 0.01%TFA in gradient 10–99% B. 20 μl of the sample was injected on the column at 35°C and a flow rate of 1 mL/min.

HPLC analysis of CA4 concentration in the PLGA-CA4 hybrid.

PLGA-CA4 hybrid (2.7 mg) was dried and added to 6 mL of acetonitrile and 1 mL of 1 M NaOH and hydrolyzed in microwave for 15 min at 85°C. The mixture was then neutralized with 10% HCl and dried using a rotary evaporator. MeOH (2 mL) was added to the sample and stored overnight at −20°C. The sample was centrifuged; the supernatant was collected and dried in rotary evaporator. Next, the sample was dissolved in acetonitrile and 20 μl was injected into HPLC and analyzed for CA4 concentration. The concentration of CA4 was found to be 44.8 μg/mL, corresponding to 3.21 μg of CA4 in the aliquot (71.6 μl). The molar ratio of PLGA:CA4 was calculated to be 0.9 μmol : 0.010 μmol (90:1 PLGA:CA4).

HPLC analysis of CA4 concentration in the RGDFFF-CA4 NP.

The concentration of CA4 in the RGDFFF-CA4 NP was obtained by HPLC using the same protocol as above.

The synthesis of PLGA-Cy5.5 hybrid.

PLGA-NH₂ (101 mg, 0.00363 mmol, MW 28000, PolySciTech) was dissolved in 2.5 mL of dry DMF and mixed with Cy 5.5-NHS ester (2.6 mg, 0.00363 mmol, Lumiprobe) dissolved in 0.5 mL of DMF. Triethylamine (10 μl) was added to the solution and the reaction mixture was stirred overnight. Then, DMF was removed under reduced pressure and crude reaction mixture was dissolved in 1 mL of acetonitrile. Product was precipitated out with cold methanol and stored in the fridge. The reaction provided 49.7 mg of blue solid (47% yield).

The synthesis of RGDFFF-Cy5.5 NP.

PLGA-Cy5.5 (1 mg) was dissolved in 0.1 mL acetonitrile and mixed with PLGA (4 mg, 0.00033 mmol 7–17 kDa) in 1.9 mL acetonitrile. The solution was added dropwise using a syringe pump (0.1 mL/min flow) into 1 mg (0.0013 mmol) of RGDFFF peptide in 6 mL of nanopure water at 37°C. After that the solution was stirred overnight to allow the self-assembly of the RGDFFF-Cy5.5 NPs. The RGDFFF-PLGA-Cy 5.5 NPs were concentrated in Vivaspin centrifugal concentrator (100 k MWCO) and washed 3 times with nanopure water.

The synthesis of PLGA-Cy7 hybrid.

7 mg (0.0097 mmol) of Cy 7-NHS ester (Lumiprobe) were added to 200 mg (0.006 mmol) of 24–38 kDa PLGA, followed by the addition of 2.66 mg (0.013 mmol) of DCC and 3.33 mg (0.026 mmol) of DIPEA. All reagents were dissolved in DMF. Reaction mixture was gently stirred overnight at 4°C. Crude product was dried using rotary evaporator and dissolved in small amount (~ 1 mL) of acetonitrile. The product was added dropwise to cold methanol and kept at −20°C overnight. Next, the product was centrifuged, and dissolved in acetonitrile. Last three steps were repeated 3 times to obtain the pure product. Finally, the product (dark blue) was dried in a lyophilizer. The yield of the reaction was established to be 72%.

The synthesis of plain RGDFFF-Cy7 NPs.

PLGA (4 mg, 0.00013 mmol, 24–38 kDa) and PLGA-Cy7 (1 mg, 0.0000323 mmol) were dissolved in 2 mL acetonitrile and added dropwise using a syringe pump (0.2 mL/min flow) into RGDFFF peptide (1.5 mg, 0.0019 mmol) in 6 mL of nanopure water at 37°C. The solution was stirred overnight to allow the self-assembly of RGDFFF-PLGA-Cy 7 NPs. After that the NPs were concentrated in Vivaspin centrifugal concentrator (100 k MWCO) and washed 3 times with nanopure water.

The synthesis of plain PLGA NPs.

PLGA (5 mg, 0.0013 mmol, 1–5 kDa) in 2 mL of acetonitrile was added dropwise using a syringe pump (0.2 mL/min flow) into 6 mL of nanopure water at 37°C. Next, the PLGA-NPs were stirred overnight. After that, the NPs were concentrated in Vivaspin centrifugal concentrator (100 k MWCO) and washed 3 times with nanopure water.

The synthesis of RGDFFF NPs.

PLGA (5 mg, 0.0013 mmol, 1–5 kDa) in 2 mL of acetonitrile was added dropwise using a syringe pump (0.2 mL/min flow) into RGDFFF peptide (1.5 mg, 0.0019 mmol) in 6 mL of nanopure water at 37°C. The solution was stirred overnight to allow the self-assembly of the RGDFFF NPs. After that, the NPs were concentrated in Vivaspin centrifugal concentrator (100 k MWCO) and washed 3 times with nanopure water.

Dynamic light scattering (DLS) measurements.

The NP solution (100 μl) was mixed with 2900 μl of nanopure water and a hydrodynamic diameter was measured by DLS. The average NP diameter and polydispersity was established from three independent measurements.

Transmission electron microscopy (TEM) imaging.

The NP sample was mixed with acetate buffer (0.125 M CH₃COONH₄, 0.6 mM (NH₄)₂CO₃ and 0.26 mM tetrasodium EDTA at pH 7.4). The sample (10 μl) was negatively stained with 10 μl of 2% (w/v) of phosphotungstic acid, cast on a 200-mesh carbon coated copper grid (Electron Microscopy Sciences) treated in NanoClean 1070, and dried in air. The TEM imaging was performed using a FEI Titan Themis 200 transmission electron microscope.

X-ray Photoelectron Spectroscopy analysis.

The RGDFFF NP was analyzed by the Advanced Science Research Center (ASRC) – CUNY, Surface Science Facility. Shortly, the RGDFFF NP sample was cast on a Si wafer, dried under vacuum, and analyzed for elemental composition using Physical Electronics Versaprobe II XPS.

Stability measurements in serum.

The RGDFFF-CA4 NP solution (100 μl) was mixed with 2900 μl of 10% FBS in nanopure water and incubated at 37 °C. The hydrodynamic diameter of the NP sample was measured after 0h, 2h, 4h, 6h, 24h, 48h, and 72h by DLS. The average NP diameter and polydispersity was established from three independent measurements.

Stability measurements in PBS.

The RGDFFF-CA4 NP solution (100 μl) was mixed with 2900 μl of phosphate buffer at pH 7.4 and incubated at 37 °C. The hydrodynamic diameter of the NP sample was measured after 0h, 2h, 4h, 6h, 24h, 48h, and 72h by DLS. The average NP diameter and polydispersity was established from three independent measurements.

Confocal imaging.

A2780, CP70, SKOV-3, OV-90, TOV-21G, ES-2, OVCAR-3, and HUVECs were seeded in 3 × 10⁴ cells/chamber in 4 chamber confocal-ready plates (154536, Lab-Tek II Chamber Slide). After 24 hours, cells were incubated with RGDFFF-Cy5.5 NPs (0.25 mg/mL) for 48 hours. Next, cells were washed two times with PBS and incubated with 10 μg/mL WGA-Alexa flour 488 for 15 minutes at 37 °C. Then, cells were washed two times with PBS and fixed with 4% PFA for 20 minutes at 37 °C. Next, the cells were mounted with mounting media containing DAPI and imaged with LSM 880 Airyscan Fast Live Cell confocal microscope.

In time-lapse experiment, CP70 cells were seeded at 1.9 × 10⁴ cells/chamber in 8 chamber plates (155409, Lab-Tech II Chambered Coverglass). After 24 h, cells were washed two times with PBS and incubated at 37 °C with 10 μg/mL of Alexa fluor 488-WGA for 15 minutes. RGDFFF-Cy5.5 NPs (0.25 mg/mL) in fresh media was added to each chamber. Using appropriate filters, the cells were imaged live using LSM 880 Airyscan Fast Live Cell confocal microscope.

Flow Cytometry (FACS).

A2780, CP70, SKOV-3, OV-90, TOV-21G, ES-2, OVCAR-3, HUVEC, cells were seeded at 3.6 × 10⁵ cells/well in 6 well plates. After 48 hours, 100 μl of RGDFFF-Cy5.5 NPs in fresh media were added to each well and incubated for 15 min. Next, the cells were washed 3 times with PBS, trypsinized, and washed 3 times with PBS. Then, the cells were fixed with 1% paraformaldehyde. The cells were kept at 4°C overnight, and fluorescence was measured using fluorescence-activated cell sorting (FACS) Calibur instrument.

Western blot.

Cells were lysed in RIPA lysis buffer (50 mM Tris, pH 8, 150 mM NaCl, 1.0% NP - 40, 0.5 % sodium deoxycholate, 0.1 % SDS) including 1 % protease inhibitor cocktail (ThermoFisher Scientific). The protein concentrations were determined using BCA (bicinchoninic acid) protein assay kit (ThermoFisher Scientific). 80 μg of protein in NuPAGE^™ LDS Sample Buffer (ThermoFisher Scientific) was boiled at 100°C for 5 min. Next, the samples were resolved by SDS-PAGE (Polyacrylamide gel electrophoresis) and transferred to nitrocellulose membrane (ThermoFisher Scientific). The blots were proved by primary alphaV (Santa Cruz, #sc-376156) and β3 (Santa Cruz, #sc-46655) antibodies and horseradish peroxidase conjugated secondary antibodies (ThermosFisher Scientific, #31430), and then visualized by enhanced chemiluminescence (ECL) Western blotting detection reagents.

IN VITRO EXPERIMENTS

Cell culture methods.

Cell lines A2780, CP70, SKOV-3, OV-90, TOV-21G, ES-2, OVCAR-3, HUVEC were purchased from ATCC. The cell lines were cultured in Dulbecco’s Modified Eagle Medium (Sigma Aldrich) supplemented with 10% (A2780, CP70, SKOV-3, ES-2, OVCAR-3) or with 15% (TOV-21G, OV-90) fetal bovine serum (FBS, HyClone) with L-Glutamine (HyClone) and penicillin/streptomycin (HyClone). HUVECs were cultured in Endothelial Basal Growth Medium (Lonza EGM^™−2 BulletKit^™) supplemented with 2% FBS with VEGF. All cells were grown in a 5% CO₂, water saturated atmosphere at 37°C. For in vitro incubation experiments, 3×10⁵ cells were seeded in each well of a 96-well plate and pre-cultured overnight. The RGDFFF-CA4 NP was diluted in growth medium. The stock solution for CA4 was prepared in DMSO and diluted in the growth medium to obtain the final concentration of DMSO lower than 0.02%. The incubation time with the cells was 48 h. The cell viability/cytotoxicity was evaluated using TACS MTT Cell Proliferation assay (Trevigen) according to the manufacturer’s instructions, and analyzed using the plate reader (SpectraMax M3 by Molecular devices).

ROS assay.

The murine macrophage cell line RAW 264.7 cells (purchased from ATCC) were seeded in a 96-well black, clear bottom plate (5 × 10⁴ cells per well) with standard DMEM (10%) culture media. After 24 h, the cells were washed 3 times with PBS, and 100 μL of 20 μM concentration of DCFH-DA (2’,7’-dichlorofluorescein diacetate; Milipore Sigma 287810) dye was added to each well. The cells were incubated with the dye for 45 min at 37°C, 5% CO₂ incubator. Next, the cells were washed 3 times with PBS and exposed to different concentrations of RGDFFF NPs made in serum-free media (Gibco RPMI Medium 1640 1X; 11835–030). Fluorescence was measured every hour by 6 hours total at excitation/emission 480/530 nm, respectively.

RNS (NO) Assay.

The experiment was carried out following the guidance of the Nanotechnology Characterization Laboratory at NIH, NCL Method ITA-7. Shortly, RAW 264.7 cells were cultured in complete RPMI-1640 Medium (Gibco) supplemented with 10% heat-inactivated fetal bovine serum with L-Glutamine (HyClone), β-mercaptoethanol and penicillin/streptomycin (HyClone). Cells were grown in a 5% CO₂ at 37°C.

Culture Plate.

For the experiment, 2×10⁶ cells per well were seeded into a 24-well plate and pre-cultured overnight. The stock solution of all samples, controls, and blank were added to appropriate wells and incubated for 48 h at humidified 37°C, 5% CO₂ incubator.

Test Plate.

50 μl of each sample from the Culture Plate was transferred to a 96-well plate in duplicates. Other wells were refilled with stock solutions of quality controls and calibration standards in culture medium (4 wells per concentration). Next, 100 μl of the Greiss reagent was added to each well and shaken for 2 minutes, and the absorbance was measured at 550 nm.

The concentration of RGDFFF NPs was adjusted to obtain the theoretical plasma concentration of combretastatin A4, calculated from the following equation:

$$\mathit{\text{Theoretical}}\mathrm{ }\mathit{\text{plasma}}\mathrm{ }\mathit{\text{concentration}}=\frac{\mathit{\text{human}}\mathrm{ }\mathit{\text{dose}}\mathrm{ }of\mathrm{ }CA4}{\mathit{\text{human}}\mathrm{ }\mathit{\text{blood}}\mathrm{ }\mathit{\text{volume}}}=\frac{60\frac{mg}{m^2}\mathrm{*}1.6\mathrm{ }{m}^2}{5600\mathrm{ }mL}=0.017\mathrm{ }\frac{mg}{mL}=1x\mathrm{ }\mathit{\text{concentration}}$$

The reported human dose for CA4 is in a range of 50–65 mg/m².

IN VIVO EXPERIMENTS.

All in vivo experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of the Icahn School of Medicine at Mount Sinai. Five week-old female Nude mice (Charles River, MA), weighing 18–20 g and maintained in specific pathogen-free conditions, were used to generate the ovarian cancer tumor model and in vivo half-life experiments.

In vivo half-life of RGDFFF-Cy7 NP.

To establish the in vivo half-life, the NPs were administered to Nude mice via tail vein injection (n=2). Retro-orbital blood samplings were drawn at 15, 30, and 90 min after injection. Blood samples were analyzed using fluorescence.

Mouse Tumor Model.

To establish the mouse tumor model, 1×10⁷ OVCAR-3 cells were inoculated subcutaneously into the right hind limbs of Nude mice (n=3). When resultant tumors reached approximately 500 mm³, as measured using external calipers, mice were injected with RGDFFF-Cy7 NPs via tail vein. At 24 h after injection, the in vivo fluorescence imaging was performed for all mice, as further described below. Mice were then euthanized and subcutaneous tumors and organs (brain, lung, liver, kidney, spleen and heart) were collected for ex vivo fluorescence imaging.

In vivo fluorescence imaging.

In vivo and ex vivo NIRF imaging experiments were performed using the IVIS-200 System (Xenogen, CA). To enable detection of the RGDFFF-Cy7 NPs, a 745 nm excitation filter and 800 nm emission filter were used. A field of view (FOV) of 17.6 and an excitation time of 4 s were chosen.

Tumor tissue slide.

Frozen tumor sections were prepared using cryostat. Standard immunostaining techniques were applied for CD31 and DAPI. In short, the sections were blocked with donkey serum for 20 min at room temperature, washed with PBS (Dulbecco’s phosphate buffered saline, Corning), and incubated with rat anti-mouse CD31 antibody (BD Pharmigen #557355) for 45 min at room temperature. The slides were rinsed with PBS and incubated with secondary donkey anti-rat antibody (Jackson Immuno Research Laboratories, Inc. #712–165-150) labeled with Cy3 for 30 min, rinsed with PBS, and mounted in mounting medium with DAPI. The stained tumor sections were imaging using LSM 880 Airyscan Fast Live Cell confocal microscope using the appropriate filters and magnification 40 and 63 times.

ABBREVIATIONS

NP

nanoparticle

PLGA

poly(lactic)co-(glycolic) acid

HPLC

high-performance liquid chromatography

DLS

dynamic light scattering

TEM

transmission electron microscopy

OvCA

ovarian cancer

ROS

reactive oxygen species

RNS

reactive nitrogen species

FDA

Food and Drug Administration