Random co-polymerization of NIPAAM with VP and PEG-A was performed by free radical polymerization process of the micellar aggregates of the amphipilic monomers (Figure 1). The polymeric nanoparticles formed in this way also have an amphiphilic character with a hydrophobic core inside the micelles, and a hydrophilic outer shell composed of hydrated amides, pyrrolidone and PEG moieties that project from the monomeric units 2425.

Mid infra-red (IR) spectra of NIPAAM, VP, PEG-A, and "void" (empty) polymeric nanoparticles were obtained to determine whether appropriate polymerization has occurred or whether monomers were present in the physical mixture. As seen in Figure 2, strong peaks in the range of 800–1000 cm^-1 corresponding to the stretching mode of vinyl double bonds disappeared in the spectrum of polymer indicating that polymerization has taken place. The water attached in the process of hydration of the polymer and proton exchange with the solvent gives rise to a broad and intense peak at 3300 cm^-1. The – CH- stretching vibration of the polymer backbone is manifested through peaks at 2936–2969 cm^-1, while peaks at 1642 and 1540 cm^-1 correspond to the amide carbonyl group and the bending frequency of the amide N-H group respectively. The absorptions bands in the region 1443–1457 cm^-1 are due to the bending vibration of CH₃ group and the bending vibration of CH₂ group can be identified in a slightly higher region.

In Figure 3, we illustrate the typical ¹H-NMR spectrum and the chemical shift assignments of the monomers as well as the copolymer formed. Polymerization is indicated by the absence of the proton resonance of the vinyl end groups of the monomers in the spectrum of the formed co-polymeric micelle. Rather, resonance can be observed at the upfield region (δ = 1.4–1.9 ppm), attributable to the saturated protons of the polymeric network. The broad resonance peak at δ = 0.8–1.0 ppm are from the methyl protons of the isopropyl group. The signal peaks for the methyne proton (>CH-) of N-isopropylacrylamide group and methylene protons (-CH₂-) of polyethylene oxide can be observed at 3.81 and 3.71 ppm respectively.

The size and size distribution of the polymeric nanoparticles were measured by means of dynamic light scattering (DLS). In Figure 4A, the typical size distribution of the nanoparticles is illustrated, and the average size corresponds to less than 50 nm diameter at 25°C with a narrow size distribution. Transmission electron microscopy (TEM) of the polymeric nanoparticles is illustrated in Figure 4B, and demonstrates that the particles have spherical morphology and low polydispersity with an approximate size of around 45 nm diameter, which is comparable to the size obtained from DLS measurements.

The entrapment efficiency of curcumin within the nanoparticles was found to be >90%, based on calculations described in the Methods. The in vitro release profile of the loaded curcumin from the nanoparticles at physiological pH is illustrated in Figure 5. Curcumin release occurs in a sustained manner, such that only 40% of the total drug is releaed from the nanoparticles at 24 hours.

An ideal drug delivery platform must be biodegradable, biocompatible and not be associated with incidental adverse effects. The toxicity profile of the void polymeric nanoparticles was studied in vitro and in vivo. In a panel of eight human pancreatic cancer cell lines (Figure 6A), we found no evidence of toxicity in cell viability assays, across a 20-fold dose range of the void nanoparticles. We then studied the effects of these particles in athymic ("nude") mice, a commonly used vehicle for preclinical tumor studies. The mice were randomized to two arms of 4 mice each – control and void nanoparticles (720 mg/kg i.p. twice weekly for three weeks). As seen in Figure 6B, despite the relatively large dosage, the mice receiving void nanoparticles demonstrated no evidence of weight loss, and no gross organ changes were seen at necropsy. No behavioral changes were observed in the mice during the course of administration, or in the ensuing follow up period.

Free curcumin is poorly soluble in aqueous media, with macroscopic undissolved flakes of the compound visible in the solution (Figure 7A); in contrast, nanocurcumin is a clear, dispersed formulation, with its hue derived from the natural color of curcumin (Figure 7B). We performed a series of in vitro functional assays to better characterize the anti-cancer properties of nanocurcumin, using human pancreatic cancer cells as a model system, and directly comparing its efficacy to free curcumin. The choice of the cancer type was based on multiple previous reports confirming the activity of free curcumin against pancreatic cancer cell lines 7826. As seen in Figures 8A and 8B, the polymeric nanoparticles encapsulating curcumin are robustly taken up by pancreatic cancer cells, indicated by the fluorescence emitted from the accumulated intra-cytoplasmic drug. In cell viability (MTT) assays performed against a series of pancreatic cancer lines, nanocurcumin consistently demonstrated comparable efficacy to free curcumin (Figure 9), although some cell lines were resistant to the agent per se. Nanocurcumin was effective in its ability to block clonogenicity of the MiaPaca pancreatic cancer cell line in soft agar assays (Figure 10). In comparison to untreated cells, or cells exposed to void polymeric nanoparticles, both free curcumin and nanocurcumin caused inhibition of clonogenicity at 10 and 15 μM dosages; the effect with nanocurcumin was somewhat more pronounced at the lower dose.

We then analyzed the mechanisms of action of nanocurcumin on pancreatic cancer cell lines, and compared the functional pathways impacted by nanocurcumin to what has been previously reported for free curcumin 262728293031. A principal cellular target of curcumin in cancer cells is activated nuclear factor kappa B (NFκB), with many of the pleiotropic effects of curcumin being ascribed to inhibition of this seminal transcription factor. In electrophoretic mobility shift ("gel shift") assays to assess for the DNA binding ability of NFκB, we demonstrate that nanocurcumin robustly inhibits NFκB function in pancreatic cancer cell lines BxPC3 and MiaPaca (Figures 11A and 11B). In BxPC3 cells, inhibition of NFκB (assessed by a shift in migration of radio-labeled p65-binding oligonucleotide) can be seen as early as 1–2 hours post exposure to both free and nanocurcumin. In MiaPaca cells, we see persistent activation of NFκB in cells exposed to free curcumin after overnight incubation, while a perceptible gel shift is observed in the nanocurcumin treated cells. Lastly, we examined whether nanocurcumin can inhibit pro-inflammatory cytokines in peripheral blood mononuclear cells (PBMCs); many of these cytokines (IL-6, IL-8, and TNFα) have also been implicated in the carcinogenesis process, including the induction of angiogenesis 32. Incubation of stimulated PBMCs with both free and nanocurcumin decreased steady-state mRNA levels of IL-6, IL-8 and TNFα, compared to DMSO and void nanoparticle-treated cells (Figure 12A–C), with evidence of dose dependent reduction of IL-6 by both agents (Figures 13A–B).

A co-polymer of N-isopropylacrylamide (NIPAAM) with N-vinyl-2-pyrrolidone (VP) poly(ethyleneglycol) monoacrylate (PEG-A) was synthesized through free radical polymerization as shown in the accompanying flowchart (Figure 1). NIPAAM, VP and PEG-A were obtained from Sigma chemicals (St. Louis, MO). NIPAAM was recrystallized using hexane, VP was freshly distilled before use, and PEG-A was washed with n-hexane three times to remove any inhibitors; Millipore water and other chemicals were used as-is. Thereafter, the water-soluble monomers – NIPAAM, VP and PEG-A were dissolved in water in 90: 5: 5 molar ratios. The polymerization was initiated using ammonium persulphate (APS, Sigma) as an initiator in a nitrogen (N₂) atmosphere. Ferrous ammonium sulphate (FAS, Sigma) was added to activate the polymerization reaction, and also to ensure complete polymerization of the monomers. In a typical experimental protocol, 90 mg NIPAAM, 5 μl freshly distilled VP, and 500 μl PEG-A (1% w/v) were added in 10 ml of water. To cross-link the polymer chains, 30 μl of N,N'-Methylene bis acrylamide (MBA, Sigma, 0.049 g/ml) was added to the aqueous solution of monomers. The dissolved oxygen was removed by passing nitrogen gas for 30 minutes. Thereafter, 20 μl of FAS (0.5% w/v), 30 μl of APS and 20 μl of TEMED (Invitrogen, Carlsbad CA, USA) were added to initiate the polymerization reaction. The polymerization was performed at 30°C for 24 hours in a N₂ atmosphere. After the polymerization was complete, the total aqueous solution of polymer was dialyzed overnight using a Spectrapore^® membrane dialysis bag (12 kD cut off) to remove any residual monomers. The dialyzed solution was then lyophilized immediately to obtain a dry powder for subsequent use, which was easily re-dispersible in aqueous media. The yield of the polymeric nanoparticles was typically more than 90% with this protocol.

Curcumin was a kind gift of Indsaff, Inc. (Batala, Punjab, India). Curcumin loading in the polymeric nanoparticles was done by using a post-polymerization method. In this process of loading, the drug is dissolved after the co-polymer formation has taken place. The physical entrapment of curcumin in NIPAAM/VP/PEG-A polymeric nanoparticles was carried out as follows: 100 mg of the lyophilized powder was dispersed in 10 ml distilled water and was stirred to re-constitute the micelles. Free curcumin was dissolved in chloroform (CHCl₃; 10 mg/ml) and the drug solution in CHCl₃ was added to the polymeric solution slowly with constant vortexing and mild sonication. Curcumin is directly loaded into the hydrophobic core of nanoparticles by physical entrapment. The drug-loaded nanoparticles are then lyophilized to dry powder for subsequent use.

The entrapment efficiency (E %) of curcumin loaded in NIPAAM-VP-PEG-A nanoparticles was determined as follows: the nanoparticles were separated from the un-entrapped free drug using NANOSEP (100 kD cut off) membrane filter and the amount of free drug in the filtrate was measured spectrophotometrically using a WALLAC plate reader at 450 nm. The E% was calculated by

E% = ([Drug]_tot - [Drug]_free)/[Drug]_tot × 100

Mid infra red (IR) spectrum of NIPAAM, VP and PEG-A monomers, as well as the void polymeric nanoparticles were taken using Bruker Tensor 27 (FT-IR) spectrophotometer (Bruker Optics Inc., Billerica, MA, USA).

The NMR spectrum of monomers NIPAAM, VP and PA, as well as void polymeric nanoparticles were taken by dissolving the samples in D₂O as solvent using Bruker Avance 400 MHz spectrometer (Bruker BioSpin Corporation, Billerica, MA, USA).

DLS measurements for determining the average size and size distribution of the polymeric micelles were performed using a Nanosizer 90 ZS (Malvern Instruments, Southborough, MA). The intensity of scattered light was detected at 90° to an incident beam. The freeze-dried powder was dispersed in aqueous buffer and measurements were done, after the aqueous micellar solution was filtered with a microfilter having an average pore size of 0.2 mm (Millipore). All the data analysis was performed in automatic mode. Measured size was presented as the average value of 20 runs, with triplicate measurements within each run.

TEM pictures of polymeric nanoparticles were taken in a Hitachi H7600 TEM instrument operating at magnification of 80 kV with 1 K × 1 K digital images captured using an AMT CCD camera. Briefly, a drop of aqueous solution of lyophilized powder (5 mg/ml) was placed on a membrane coated grid surface with a filter paper (Whatman No. 1). A drop of 1% uranyl acetate as immediately added to the surface of the carbon coated grid. After 1 min excess fluid was removed and the grid surface was air dried at room temperature before loaded in the microscope.

A known amount of lyophilized polymeric nanoparticles (100 mg) encapsulating curcumin was dispersed in 10 ml phosphate buffer, pH 7.4, and the solution was divided in 20 microfuge tubes (500 μl each). The tubes were kept in a thermostable water bath set at room temperature. Free curcumin is completely insoluble in water; therefore, at predetermined intervals of time, the solution was centrifuged at 3000 rpm for 10 minutes to separate the released (pelleted) curcumin from the loaded nanoparticles. The released curcumin was redissolved in 1 ml of ethanol and the absorbance was measured spectrophotometrically at 450 nm. The concentration of the released curcumin was then calculated using standard curve of curcumin in ethanol. The percentage of curcumin released was determined from the equation

Release  ( % ) = [ Curcumin ] rel [ Curcumin ] tot × 100 MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacqqGsbGucqqGLbqzcqqGSbaBcqqGLbqzcqqGHbqycqqGZbWCcqqGLbqzcqqGGaaicqGGOaakcqGGLaqjcqGGPaqkcqGH9aqpdaWcaaqaaiabcUfaBjabboeadjabbwha1jabbkhaYjabbogaJjabbwha1jabb2gaTjabbMgaPjabb6gaUjabc2faDnaaBaaaleaacqqGYbGCcqqGLbqzcqqGSbaBaeqaaaGcbaGaei4waSLaee4qamKaeeyDauNaeeOCaiNaee4yamMaeeyDauNaeeyBa0MaeeyAaKMaeeOBa4Maeiyxa01aaSbaaSqaaiabbsha0jabb+gaVjabbsha0bqabaaaaOGaey41aqRaeGymaeJaeGimaaJaeGimaadaaa@6275@

where, [Curcumin]_rel is the concentration of released curcumin collected at time t and [Curcumin]_tot is the total amount of curcumin entrapped in the nanoparticles.

In order to exclude the possibility of de novo toxicity from the polymeric constituents, we utilized void nanoparticles against a panel of eight human pancreatic cancer cell lines (MiaPaca2, Su86.86, BxPC3, Capan1, Panc1, E3LZ10.7, PL5 and PL8). These cells were exposed to void nanoparticles for 96 hours across a 20-fold concentration range (93 – 1852 μg/mL) and cell viability measured by MTS assay, as described below. Further, limited in vivo toxicity studies were performed in athymic (nude) mice by intraperitoneal injection of void polymeric nanoparticles at a considerably high dosage of 720 mg/kg twice weekly, for a period of three weeks. Mice receiving intra-peritoneal nanocurcumin (N = 4) were weighed weekly during the course of therapy and average weight compared to that of control littermate nude mice (N = 4). At the culmination of the three week course, mice were euthanized and necropsy performed to exclude any intraperitoneal deposition of polymers, or gross organ toxicities.

Curcumin is naturally fluorescent in the visible green spectrum. In order to study uptake of curcumin encapsulated in nanoparticles, BxPC3 cells were plated in 100 mm dishes, and allowed to grow to sub-confluent levels. Thereafter, the cells were incubated with nanocurcumin for 2–4 hours, and visualized in the FITC channel.

Growth inhibition was measured using the CellTiter 96^® A_queous Cell Proliferation Assay (Promega), which relies on the conversion of a tetrazolium compound (MTS) to a colored formazan product by the activity of living cells. Briefly, 2000 cells/well were plated in 96 well plates, and were treated with 0, 5, 10, 15 and 20 μM concentrations of free curcumin and equivalent nanocurcumin, for 72 hours, at which point the assay was terminated, and relative growth inhibition compared to vehicle-treated cells measured using the CellTiter 96^® reagent, as described in the manufacturer's protocol. A panel of ten human pancreatic cancer cell lines were examined (BxPC3, AsPC1, MiaPaca, XPA-1, XPA-2, PL-11, PL-12, PL-18, PK-9 and Panc 2.03) in the MTT assays; the sources and culture conditions of these ten lines have been previously described [58]. All experiments were set up in triplicates to determine means and standard deviations.

Colony formation in soft agar was assessed for therapy with free curcumin and equivalent dosage of nanocurcumin. Briefly, 2 ml of mixture of serum supplemented media and 1 % agar containing 5, 10 or 15 μM of free curcumin and equivalent nanocurcumin was added in a 35 mm culture dish and allowed to solidify (base agar) respectively. Next, on top of the base layer was added a mixture of serum supplemented media and 0.7 % agar (total 2 mL) containing 10,000 MiaPaca2 cells in the presence of void polymer, free or nano-curcumin, and was allowed to solidify (top agar); a fourth set of plates contained MiaPac2 cells without any additives. Subsequently, the dishes were kept in tissue culture incubator maintained at 37°C and 5 % CO₂ for 14 days to allow for colony growth. All assays were performed in triplicates. The colony assay was terminated at day 14, when plates were stained and colonies counted on ChemiDoc XRS instrument (Bio-Rad, Hercules, CA).

Nuclear extracts were prepared as described [59]. Briefly, double-stranded oligonucleotides containing a consensus binding site for c-Rel (5'-GGG GAC TTT CCC-3') (Santa Cruz Biotechnology) were 5' end-labeled using polynucleotide kinase and [³²P]dATP. Nuclear extracts (2.5–5 μg) were incubated with ≈1 μl of labeled oligonucleotide (20,000 c.p.m.) in 20 μl of incubation buffer (10 mM Tris-HCl, 40 mM NaCl, 1 mM EDTA, 1 mM β-mercaptoethanol, 2% glycerol, 1–2 μg of poly dI-dC) for 20 min at 25°C. DNA-protein complexes were resolved by electrophoresis in 5% non-denaturing polyacrylamide gels and analyzed by autoradiography.

IL-6, IL-8 and TNF-alpha mRNA levels were assessed as described previously 55. Briefly, peripheral blood mononuclear cells (PBMC) from a healthy donor were isolated by centrifugation on a Ficoll Hypaque density gradient (GE Healthcare Biosciences) and washed twice with phosphate buffered saline (PBS; Invitrogen, Carlsbad, CA). Next, 500,000 cells per well of a 24 well plate in 1 ml of RPMI (Invitrogen) supplemented with 10% FBS (Invitrogen) and 1× Pen/Strep (Biofluids, Camarillo, CA) were co-stimulated with 2% phytohaemagglutinin (PHA M-Form, liquid; Invitrogen) and 1 μg/ml lipopolysaccharide (LPS; Sigma-Aldrich, St. Louis, MO) in the presence of free or nanocurcumin, using solvent and void nanoparticles as controls, respectively. Cells were lysed after 24 hours incubation at 37°C and 5% CO₂ and RNA extracted using the RNeasy Mini Kit (Qiagen, Valencia, CA). Relative fold steady-state mRNA levels were determined on a 7300 Real time PCR System (Applied Biosystems, Foster City, CA,) by RT-PCR as described 56.