Synthetic curcumin analog EF31 inhibits the growth of head and neck squamous cell carcinoma xenografts

Abstract

Objectives are to examine the efficacy of new synthetic curcumin analogs EF31 in head and neck squamous cell carcinoma in vitro and in vivo, and study their pharmacokinetic and toxicologic effects in vivo. The synthesis of EF31 was described for the first time. Solubility of EF24, EF31 was compared using nephelometric analysis. Human head and neck squamous cell carcinoma Tu212 xenograft tumors were established in athymic nude mice and treated with EF31 i.p. once daily five days a week for about 5 – 6 weeks. The long term effect of EF31 on the NF-κB signaling system in the tumors was examined by Western blot analysis. EF31 at 25 mg/kg, i.p. inhibited tumor growth almost completely. Solubility of EF24 and EF31 are <10, 13 μg/mL or <32, 47 μM, respectively. The serum chemistry profiles of treated mice were within the limits of normal, it revealed a linear increase of Cmax. EF31 decreased the level of phosphorylation of NF-κB p65. In conclusion, the novel synthetic curcumin analogs EF31 is efficacious in inhibiting the growth of Tu212 xenograft tumors and may be useful for treating head and neck squamous cell carcinoma. The long term EF31 treatment inhibited NF-kB p65 phosphorylation in xenografts, implicating downregulation of cancer promoting transcription factors such as angiogenesis and metastasis.

INTRODUCTION

Curcumin (diferuloylmethane), a major component of turmeric, is used as a coloring and flavoring agent in many food items, including curries and mustards. Recent preclinical and clinical studies demonstrate that this phytochemical has a number of anticancer properties [1, 2]. The pharmacological safety of curcumin has been demonstrated by its consumption for centuries at levels of up to 100 mg/day by people in certain countries [3]. One problem with the clinical use of curcumin is its low potency and poor absorption characteristics [4]; however, curcumin remains an ideal lead compound for the design of more effective analogs. Using the shape and size of curcumin as a guide, we have designed and synthesized over 100 analogs. We and others have demonstrated that many of them are more potent than curcumin in a variety of cellular assays [5–9] and in animal models of disease [10].

One compound in particular, EF24 (Fig. 1), is more active than parent curcumin in its anticancer and antiangiogenesis effects [5, 6, 10, 11]. The mechanism of action of EF24, like curcumin, appears to involve multiple targets, the discovery of which is ongoing. We demonstrated that EF24 depolarizes the membrane potential of the mitochondrial membrane, causes G2/M arrest, activates caspase 3 and leads to apoptosis in human breast and prostate cancer cell lines [6]. Others have demonstrated that EF24 inhibits Akt and MDM2, an important negative regulator of the p53 tumor suppressor, by upregulating PTEN and enhancing the level of p53, thereby inducing G2/M arrest and apoptosis in human ovarian cancer cell lines [12]. EF24 significantly suppresses the growth of colon cancer tumor xenografts. In turn, these show decreased tumor angiogenesis, inhibition of Akt and ERK1/2 activation and reduced expression of COX-2, interleukin-8 and VEGF mRNA [13].

Fig. 1. Cytotoxic activity and solubility of EF24 and EF31 against Tu212 SCC cells in vitro.

Chemical structures of curcumin, EF24, EF31 are shown. Molecular weight of curcumin, EF24 and EF31 are 368.38, 311.11 g/mol, respectively (Fig. 1A). Tu212 cells (10⁴ cells/0.2 ml/well) in a 96-well plate were incubated with varying doses of EF24, EF31 (0 – 20 μM) in triplicate for 48 hours. Viable cells were measured by Neutral Red Dye assay. IC₅₀ values of EF24, EF31 were 8, 7μM, respectively (Fig. 1B). The aqueous solubilities of EF24, EF31 were determined by nephelometry to be < 10 μg/mL (32 μM), 13 μg/mL (47 μM), respectively. Solubilities of EF24, EF31 are shown as the concentrations at which the solute began to precipitate out of solution and expressed as counts of refractive nephelometry units (RNU) (Fig. 1C).

NFκB has been linked to many cancers. Although such studies implicate NFκB as a therapeutic target, mechanistically. To the extent it is known, curcumin and EF24 operate similarly on the NFκB signaling pathway. Curcumin inhibits NFκB activation pathway at a step before IκBα phosphorylation [14]. The same group found that curcumin suppresses TNF-induced nuclear translocation of p65, which corresponds to the sequential suppression of IκBα kinase (IKK) activity, phosphorylation and degradation of IκBα, p65-phosphorylation, -nuclear translocation and -acetylation. Curcumin also inhibits TNF-induced Akt activation and its association with IKK. These results indicate that curcumin inhibits NFκB activation and NF-κB regulated gene expression by inhibition of IKK and Akt activation [15]. Curcumin potently inhibits cytokine (TNFα and/or IL-1ß)-mediated NF-κB activation upstream of the NFκB-inducing kinase (NIK) and its activation of the IKKα/IKKß complex [16]. Curcumin reduces nuclear translocation of p65, induces apoptosis in vitro and attenuates hematogenous lung metastasis of human breast cancer MDA-MB-231 cells by intracardiac injection. These effects correlate with reduction of phosphorylated IκB and p65 by Western blot analysis [17]. EF24, as well as curcumin, suppresses the NFκB signaling pathway through direct action on IKKβ and, thereby, inhibits nuclear translocation of p65 [8]. Most of these observations are derived from short term treatment in vitro.

Our laboratory recently synthesized the curcumin analog EF31, we report here that it is slightly more potent than EF24. Human Tu212 squamous cell carcinoma (SCC) xenografts in athymic nude mice were treated with EF31, intraperitoneally (i.p.), daily for 5 days a week for over one month. EF31 exhibited dose-dependent suppression of tumor growth. In addition, we found decreased amounts of phosphorylated IKKβ and p65. Unexpectedly, increased amounts of phosphorylated IκBα were also detected by Western blot analysis. This is the first report of the anticancer activities of this compound in vitro and in vivo, pharmacokinetic characteristics, effects of serum chemistry in mice and the blockade of the NFκB signaling pathways in tumors after being treated with the drugs for over a month.

MATERIALS AND METHODS

Synthesis of curcumin analogs EF31

EF31 (3,5-bis(2-pyridinylmethylidene)-4-piperidone) was prepared by the following procedure. To a 250 mL round bottom flask (RBF) equipped with a stir bar was added 0.25 M aqueous sodium hydroxide (40 mL, 10 mmol). To this stirring base, was added 4-piperidone hydrochloride monohydrate (0.61 g, 4.0 mmol) and then 2-pyridinecarboxaldehyde (0.80 mL, 8.4 mmol). To the yellow solution was added 0.78 M acetyltrimethylammonium chloride (0.51 mL, 0.40 mmol), which gave an orange precipitate. After two hours an additional 0.13 g (0.31 eq more) of 2-pyridinecarboxaldehyde and 2 mL more 1 M NaOH (0.5 eq more) were added. The reaction was allowed to stir at room temperature for 6 hours total. The reaction mixture was diluted with 40 mL of brine and then extracted with dichloromethane (DCM; 3 × 50 mL). The organic layer (bottom) was collected into a 250 mL Erlenmeyer flask. The yellow solution was dried over MgSO₄, filtered into a 500 mL RBF, concentrated in vacuo at ~400 mbar until a minimal amount of liquid remained, and then dried under high vacuum. (mass balance: 1.64 g). The yellow solid was adsorbed onto silica and purified using column chromatography (0 to 20% MeOH in DCM); Rf = 0.43 in 10% MeOH in DCM) to give a yellow solid, which was further recrystallized from boiling absolute ethanol, to give 0.69 g (62% yield) of small yellow crystals. M.p. 183–186 °C; TLC: 19:1 EtOAc-MeOH solvent system (R_f = 0.19); ¹H NMR (400 MHz, CDCl₃, δ): 8.68 (dd, J=4.6, 0.9 Hz, 2 H), 7.69 (td, J=7.6, 1.8 Hz, 2 H), 7.62 (s, 2 H), 7.43 (d, J=7.9 Hz, 2 H), 7.18 (ddd, J=7.6, 4.7, 1.1 Hz, 2 H), 4.48 – 4.57 (m, 4 H), 1.73 (br. s., 1 H); ¹³C NMR (100 MHz, CDCl₃) δ ppm 155.2, 149.9, 139.4, 136.5, 132.6, 127.7, 122.9, 48.8; HRMSAPCI (m/z): [M + H]⁺ calculated for C₁₇H₁₆N₃O, 278.12879; found, 278.12888; Analysis; calculated for C₁₈H₁₇N₃O: C, 74.20; H, 5.88; N, 14.42, and found: C, 74.09; H, 5.75; N, 14.42.

Nephelometry

Curcumin analog and an internal control were dissolved in 100% dimethyl sulfoxide (DMSO) to obtain a final concentration of 30 mg/mL. The 30 mg/mL stock solution was serially diluted (concentration profile: 30, 20, 15, 10, 7.5, 5, 2.5, 1.25, 0.63, 0.31, and 0.15 mg/mL) in test tubes with 100% DMSO. The concentration profile was transferred to 96 well microplates (Costar black clear bottom) and serially diluted to a final DMSO (EMD) concentration of 1% and a final drug concentration of 300, 200, 150, 100, 75, 50, 25, 12, 6, 3 and 1.5 ~g/mL with phosphate buffered saline, pH 7.4 (Sigma). The microplates were incubated for 90 minutes in the dark at ambient temperature. Laser nephelometry (NEPHELOstar, BMG Lab Technologies) was used to determine the point (concentration) at which the solute began to precipitate out of solution and expressed as counts of refractive nephelometry units (RNU). Laser nephelometry is the measurement of forward scattered light when a laser beam is directed through a solution.

Cell culture of Tu212 cells

Tu212 human SCC cells were maintained in DMEM/F12 (1:1) (Lonza, 12-719F) containing 10% fetal bovine serum (FBS) (Gibco, 26140-079), penicillin (100 units/mL), streptomycin (100 μg/mL) and 2 mM L-glutamine. Cells were incubated at 37–C in 5% CO₂/95% air in a humid atmosphere.

Neutral red assay for cell viability to test the efficacy of EF31 treatment in vitro

This assay was carried out according to the methods published by a modification of the method of Zhang [18]. Tumor cells (Tu212) for the anticancer screen were plated at a concentration of 20,000 cells/200 μL/well in a 96-well plate and incubated overnight to allow the cells to adhere. EF31 or vehicle (0.1% DMSO) (Sigma, St. Louis, MO) was then added to appropriate wells in various concentrations in triplicate. Subsequently, cells were incubated for 72 h. At the termination of culture, medium was removed and 200 μL of fresh, warm medium containing 50 μg of Neutral Red/mL (GIBCO-BRL, 15330-079, 0.33%) was added to each well. Cells were incubated at 37°C for 30 min, followed by two washes with 200 μL of PBS. The Neutral Red taken up by cells was dissolved by adding 200 μL of 0.5NHCl containing 35% ethanol. The plates were placed on a plate shaker until all residues were solubilized. Then, the amount of the dye in each well was read at 570 nm by a Universal Microplate Reader (EL800, Bio-Tek, Winooski, VT). Results from triplicate samples were recorded as optical density units (OD₅₇₀) and averaged after subtraction of the blank

EF31 (i.p.) therapy in vivo

Nude mice were housed in a pathogen-free facility, and all procedures and protocols were approved by the Institutional Animal Care and Use Committee at Emory University. Tu212 cells (2 × 10⁶ cells/mouse) were inoculated subcutaneously into the right posterior back of 7-week-old female athymic nude mice (nu/nu) mice (Harlan) and allowed to form a xenograft. When tumors reached a volume of approximately 150–200 mm³, mice were randomized to receive either vehicle [0.5% carboxymethyl cellulose sodium (CMC) with 10% DMSO in sterile water] or drug (EF31). CMC was dissolved in distilled water and sterilized by autoclaving. Stock solution of the drugs was made fresh each time, weighing out EF31 10 mg or UBS-109 100 mg and dissolved in 1 mL of 100% DMSO by brief heating before making appropriate dilution. EF31 was administered i.p. 5 days per week at a dosage of 12.5 or 25 mg/kg. Tumor size and body weights were measured twice a week. Tumor volume was calculated using the formula (width² × length)/2.

Pharmacokinetic studies

Male ND4 Swiss Webster mice (Harlan) weighing 25 to 30 grams were used for the PK studies. Mice were each given a single dose of EF31 i.p. (12.5 or 25 mg/kg) in 10% DMSO/0.5% CMC. A single blood sample (0.3 ml) was taken from each animal under anesthesia (ketamine/xylazine, 100/10 mg/kg, s.c.) from the submandibular vein at one of eight different time points after dosing: 0.08, 0.25, 0.5, 1, 2, 3, 4 and 6 h (n=3 mice/time point). Samples were collected in heparinized capillary tubes (Fisher), transferred to heparinized microtainers (BD) on ice in the dark and centrifuged within a few minutes at 2000 × g for five minutes in a refrigerated centrifuge to obtain plasma. Plasma was transferred to amber Eppendorf tubes (VWR) on ice, frozen for one hour and stored at −80°C until analysis by LC/MS/MS within two days.

The plasma samples (50 μL) were deproteinated with acetonitrile (500 μL) and analyzed by LC/MS/MS using an internal standard spiking technique with an AB SCIEX 4000 QTRAP system equipped with an Agilent 1200 series HPLC using a developed and verified bioanalytical assay method to determine the plasma concentrations of EF31. The mean, standard deviation, and %CV of plasma levels were determined using Microsoft® Office Excel 2007. Mean concentration data was plotted against time to create plasma concentration versus time profiles.

Pharmacokinetic analysis of the plasma concentration data was conducted using non-compartmental analysis with WinNonlin Version 5.2.1. Areas under the curve (AUC) were estimated using the linear trapezoidal rule. The first order rate constant (Lambda Z) associated with the terminal (log-linear) elimination phase was estimated using linear regression of at least three non-zero concentrations versus time points in the late phase of the log concentration versus time profile. The half-life (t½) was calculated as ln(2)/Lambda Z. Nominal times were used in all the data analysis.

Western blotting

Tumor tissue extracts were obtained by incubating tissue in RIPA buffer (50 mM Tris-HCl, pH 7.4, 1% NP-40, 0.25% sodium-deoxycholate, 150 mM NaCl, and 1 mM EDTA) supplemented with protease inhibitor cocktail (Complete; Roche) for 30 min on ice, then cleared by centrifugation at 13,000 rpm for 15 min at 4°C. Total protein (100 μg) was separated by 10% SDS-PAGE gels and transferred onto polyvinylidene fluoride (PVDF) membranes (Hybond; GE Healthcare) for Western blotting. For Western blot detection, tissue extracts were boiled with 4:1 volume loading buffer. Primary antibodies for Western blotting were: IKKα (Cell Signaling Technology, #9936) at 1:1000, IKKα [phospho T23] (Abgent, AP3374a) at 1:1000, IKKβ (Novus, NB100-92040) at 1:1000, IKKβ [phospho Y199] (Abgent, AP3127a) at 1:500, NF-κB p65 (Novus, NB100-2176) at 1:1000, NF-κB p65 [phospho Ser276] (Novus, NB100-82086) at 1:1000, IκBα (Cell Signaling Technology, #4814) at 1:2000, IκBα [phospho Ser32/36] (Novus, NB100-92563) at 1:1000, or Actin (Santa Cruz Biotechnology, sc-1616) at 1:3000, anti-mouse IgG, anti-rabbit IgG, or anti-goat IgG horse radish peroxidase (HRP)-conjugated secondary antibodies (Cell Signaling Technology) were used at 1:5000, and HRP activity was visualized using the enhanced chemiluminescence (ECL) plus system (GE Healthcare).

Immunohistochemistry (IHC)

Tissues were fixed in 10% formaldehyde. After paraffin embedding, 6 μm sections were used for immunohistochemical studies. The paraffin-embedded sections of the tumor specimens were treated in xylene and dipped in a gradient of ethanol (once in 95% ethanol, once in 70% ethanol, once in 50% ethanol and once in water). The sections were heated in Tris-EDTA buffer (100°C, 20 minutes). Endogenous peroxidase activity was blocked with 3% hydrogen peroxide. The sections were then incubated with a polyclonal antibody specific for each antigen. Primary antibodies were: IKKβ (Novus, NB100-92040) at 1:50, IKKβ [phospho Ser177/181] (USBio,13000-12) at 1:100, NF-κB p65 (Novus, NB100-2176) at 1:100, NF-κB p65 [phospho Ser276] (Novus, NB100-82086) at 1:100, IκBα (Cell Signaling Technology, 4814) at 1:50, IκBα [phospho Ser32/36] (Novus, NB100-92563) at 1:100, Incubation times were overnight at 4°C, The sections were then incubated with anti-mouse & rabbit (abcam, ab64264;Mouse and Rabbit Specific HRP/DAB detection IHC kit), then follow the kit's instraction. Hematoxylin staining was used for histological diagnosis. Coverslips were mounted on glass slides with mounting medium. Controls using normal rabbit, or rat serum were run to exclude nonspecific staining.

Statistical analysis

All the experiments in the study were repeated at least three times and the data are presented as the mean ± standard deviation. The significance of the difference between groups was evaluated with a two-tailed Student's t-test, p <0.05 was considered significant.

RESULTS

Cytotoxic activity and solubility of EF24, EF31

Cytotoxic activities in Tu212 SCC cells in vitro were assayed in triplicate using Neutral Red dye as described previously [5, 6]. The IC₅₀ values of EF24 and EF31 were similar with apparent IC₅₀ values of 8μM and 7μM, respectively (Figs. 1A & 1B). The aqueous solubilities of EF24 and EF31 were determined by nephelometry to be < 10 μg/mL (32 μM), 13 μg/mL (47 μM) respectively.

Efficacy of EF31 in 0.5% CMC and 10% DMSO by i.p. injection into female athymic nude mice bearing Tu212 SCC xenografts

Animals bearing Tu212 SCC xenografts were treated with EF31 i.p. once/day for 5 days/week (M-F) with the following regimens, including control group: Vehicle (0.5% CMC and 10% DMSO) (n = 10), Group 1 : EF 31, 12.5mg/kg (in vehicle) (n = 10), and Group 2 : EF 31, 25mg/kg (in vehicle) (n = 10), Asterisks indicate significant difference in tumor size from the control (p < 0.05–0.01). Intraperitoneal administration provided the adequate plasma concentration of the drug to inhibit xenograft tumors. EF31 at 12.5 mg/kg and 25 mg/kg reduced tumor size significantly, but the tumors continued to grow very slowly. Both doses of EF31 reduced body weight temporarily by ~10% during treatment days 7–11; however, the animals regained body weight upon continued treatment. In the EF31 (25mg/kg) group, three animals died, apparently due to treatment. Thus, this i.p. dose may be above the maximally tolerated dose (MTD). Repeated experiments demonstrated similar efficacy for EF31 (Fig. 2A & 2B).

Fig. 2. Efficacy of EF31 in 0.5% carboxymethyl cellulose sodium (CMC) and 10% DMSO by i.p. in Tu212 SCC xenografts in female athymic nude mice.

Tu212 cells (2 × 10⁶/0.1 ml medium) were inoculated subcutaneously (s.c.). EF31 (12.5 mg/kg and 25 mg/kg) or vehicle were administered intraperitoneally (i.p.) to mice (n = 10/group) daily for 5 days a week (Monday through Friday) for 5 weeks. Treatment was started when tumors grew to approximately 0.2–0.3 cm in diameter. (A) Tumor size. Asterisks indicate significant difference from the control (p < 0.05–0.01). (B) EF31 at 25 mg/kg reduced body weight temporarily by ~10% on average but animals regained body weight as treatment continued.

Effects on serum chemistry profiles and normal organs

The serum chemistry profile was all within the normal values based on data from the University of Minnesota. Pathologic examination of hematoxylin-eosin staining of the liver, kidneys, lungs and spleen from six groups (n = 5/group) including control, EF31 12.5 and 25 mg/kg, i.p. revealed normal. Analysis of serum chemistry revealed that both drugs significantly reduce alanine transaminase (ALT), alkaline phosphatase and phosphorus as compared with the controls, although even the lowest values are within the normal limits.

PK studies of EF31 in mice

After i.p. administration of EF31 at 12.5 and 25 mg/kg in a DMSO/0.5% CMC (10%/90%) formulation, peak mouse plasma concentrations were reached at 0.25 and 0.5 hours post-dose (Tmax) with average concentrations (Cmax) of 1067 and 2127 ng/mL, respectively. The terminal elimination half-life (t½) averaged 2.2 and 2.4 hours, with an average AUC(0-∞) of 3386 and 7769 hr × ng/mL, respectively. Following i.p. administration of EF31, the mean concentration versus time profiles suggested that the compound absorption rate was rapid and clearance from blood occurred at a moderate rate following Cmax (Fig. 3A & 3B).

Fig. 3. Pharmacokinetic study of EF31 administered by i.p.

(A, B) After i.p. administration of EF31 at 25 mg/kg in a DMSO/0.5% CMC (10%/90%) formulation, peak mouse plasma concentrations were reached at 0.5 hours post-dose (Tmax) with an average concentration (Cmax) of 2127 ng/mL. The terminal elimination half-life (t½) averaged 2.4 hours, with an average AUC(0-∞) of 7769 hr × ng/mL. Both Cmax and Tmax were dose proportional following i.p. administration of EF31 to mice at 12.5 and 25 mg/kg.

Both Cmax and Tmax in mice were dose proportional following i.p. administration of EF31 at 12.5 and 25 mg/kg. However, the terminal elimination half-lives (t½) and volumes of distribution (Vz/F) were similar for both doses. The apparent terminal volumes of distribution (Vz/F) (corrected for weight) for EF31 were 11.5 and 11.1 L/kg for the 12.5 and 25 mg/kg doses, respectively. This may indicate a significant distribution of EF31 into total body water or tissues.

EF31 decreases the level of phosphorylated NF-κB p65

To explore the in vivo effects of EF31 on the NF-κB signaling pathway, we used Western blotting to analyze tumors from mice treated with EF31 at 12.5 mg/kg and 25 mg/kg, and vehicle control, i.p., for 6 weeks. This experiment was performed since the curcumin analog EF24 was demonstrated to inhibit NF-κB by inhibiting IκBα kinaseβ (IKKβ) in vitro [8]. EF31 at 25 mg/kg reduced the levels of phosphorylated IKKα but not the total IKKα as compared to the control. However, while the drug reduced the level of phosphorylated IKKβ, it also increased the level of total IKKβ in a dose-dependent manner. Levels of the down-stream target of IKKβ, IκBα phosphorylated at serine 32 and 36, were increased, but the total IκBα decreased in a dose-dependent manner. EF31 decreased both levels of the phosphorylated NF-κB p65 at Ser276 as measured and the total p65 in a dose-dependent manner (Fig. 4).

Fig. 4. Changes in the NF-κB pathways in tumors treated with EF31 (Western blots).

Tumor bearing animals were treated with EF31 0, 12.5 mg/kg and 25 mg/kg i.p. daily for 5 days a week for 5 weeks as shown in Fig. 2. Tumor extracts were analyzed by Western blotting using various antibodies for NF-κB pathways.

In the Hemotoxylin and Eosin (H & E) staining experiments, tumor tissues treated at 25 mg/kg are almost entirely replaced with pink, ring-shaped structures that surround a small nest of tumors in the center. It appears that the tumor cells are replaced with fibrous scar tissue (Fig. 5). The rank order of degree of the IHC staining of the phosphorylated IκBα is stronger in tumors treated at 25 mg/kg than 12.5 mg/kg, which is stronger than that in the control (Fig. 5). All of IKKβ, phosphorylated IKKβ, IκBα, phosphorylated IκBα, NF-κB p65 and phosphorylated NF-κB p65 results correspond with that of the Western blot (Fig. 4 and 5). In an in vitro short term culture of cells treated with cytokines, it is likely that phosphorylated IκBα will be decreased in the presence of EF31 [30]

Fig. 5. Hemotoxylin and Eosin staining (×100) and corresponding immunohistochemistry patterns of xenografts mouse tumor tissues.

(A) H & E staining, tumor bearing animals were treated with EF31 0, 12.5 mg/kg and 25 mg/kg i.p. daily for 5 days a week for 5 weeks. (B) IKKβ and phospho-IKKβ staining of mice tumor tissues serial sections.(C) IkBα and phospho- IkBα staining of mice tumor tissues serial sections.(D) NF-κB p65 and NF-κB phospho-p65 staining of mice tumor tissues serial sections. The scale bars indicate 100 μm.

DISCUSSION

In this report, we demonstrate that the synthetic curcumin analog EF31 significantly inhibit the growth of human head and neck squamous cell cancer xenografts. The drugs were efficacious when diluted in 0.5% CMC and 10% DMSO.

A pharmacokinetic (PK) study was done in mice to characterize the drug levels and PK parameters associated with i.p. dosing of EF-31 under the conditions that were used in the tumor efficacy experiment (Fig. 2). Examination of the results shows that peak plasma levels were reached within 30 minutes and that measureable drug levels were seen for at least six hours. Both Cmax and AUC were dose-related, with the Cmaxes values falling in the 1 to 2 μg/mL range, consistent with the micromolar drug concentrations (4–8 μM) that were required for tumor cell growth inhibition in vitro (Fig 1). Estimated volumes of distribution were ~ 11 L/kg, suggesting significant penetration of tissues. Several factors might influence the efficacy of EF31, including IC₅₀ values, the aqueous solubility, and the terminal elimination half-life (t½) of EF31 12.5 and 25 mg/kg, which averaged 2.2 and 2.4 hours.

Western blot analyses were performed on Tu212 tumor xenografts from animals treated for 6 weeks with EF31. The results reflect the effects of long-term drug treatment on the NFκB signaling pathway, in contrast to short term treatment (hours) of cells in vitro. EF24, an analog of EF31, was previously shown to inhibit IKKβ activity by means of TNF-α induced IκBα phosphorylation during short term incubation (30 min- 3 h) in vitro [8].

The level of p-IKKα and p-IKKβ decreased in a dose-dependent manner, whereas the level of unphosphorylated IKKα and IKKβ increased in a dose-dependent manner. This reciprocal response of tumors to a long-term treatment of EF31 dose at 0, 12.5, and 25 mg/kg may suggest a compensatory reaction of tumors to the inhibition of IKKα and IKKβ phsophorylation. The reciprocal response in the level of increased p-IκBα and the level of decreased unphosphorylated IκBα in a dose-dependent manner also seen but this is counter intuitive, since the both levels of p-IKKα and p-IKKβ that phosphorylate IκBα were decreased. This suggests the presence of other IκBα kinases that phosphorylate IκBα, leading to the process of ubiquination, degradation of IκBα. This degradation of IκBα from the IκBα/NFκB p65-p50 complex promte the increased nuclear translocation of NFκB p65-p50. the IκB kinases IKKα/β complex, and casein kinase II (CK II), and the mitogen-activated 90 kDa ribosomal S6 kinase (p90rsk1) can phosphorylate one or both of the two serines (Serine 32 and Serine 36) in IκBα [19]. Inhibition of CKII significantly suppressed tumor growth in HNSCC xenograft models [20]. EF31 inhibited HNSCC xenografts, suggesting that EF31 also inhibited CKII. It is unknown whether EF31 inhibits the p90rsk1. Thus, there are two potential candidates that increased the level of the phosphorylated IκBα. First, it has been demonstrated that the p38 MAPK stimulated by transforming growth factor (TGF)-β1 phosphorylated IκBα [21]. Since EF24 does not inhibit p38 MAPK activity [9], it is unlikely that EF31 inhibits the p38 MAPK. Second, tyrosine phosphorylation of IκBα activated NFκB without proteolytic degradation of IκBα [22]. Western blot analysis revealed that the both levels of phosphorylated NFκB p65 and unphosphorylated NFκB p65 were decreased in an EF31 dose-dependent manner. The mechanism of homeostatic maintenance of the NFκB level is well documented. NFκB (e.g., p65-p50 heterodimer) bind to the κB site on the promoter region of IκBα and induce its re-synthesis [23]. The newly synthesized IκBα is translocated into the nucleus, where it binds and transports NFκB to the cytoplasm [24]. The autoregulatory loop between NFκB and IκBα is responsible for maintenance of the homeostasis of NFκB and IκBα. We demonstrated that EF31 decreased the level of total IκBα, and the level of phosphorylated and total p65 NFκB in a relatively dose-dependent manner. It is unclear why the phosphorylated IκBα was present abundantly in the fibrous structure by the IHC staining. It could be that in the process of continuous chemotherapy with EF31, the tumor cells are near death and cannot produce the adequate levels of ubiquitin-proteaosome to degrade the phosphorylated IκBα, leaving it to be absorbed and deposited onto the continuously growing, fibrous scar tissue in the tumors.

NFκB p65 is phosphorylated at serine 276 (p-NFκB p65), the major site of phosphorylation. Two protein kinases phosphorylate NFκB p65 at serine 276. These are the catalytic subunit (PKAc) of cyclic AMP-dependent protein kinase (A-PK) [25, 26] and mitogen-and stress-activated protein kinase (MSK1), which is a kinase activated in cells downstream of both the ERK1/2 (extracellular-signal-regulated kinase) and p38 MAPK (mitogen-activated protein kinase) cascades [27, 28]. Inhibition of MSK1 by EF31 is unlikely because EF24 was demonstrated to increase phosphorylation of ERK1/2, JNK and p38. [9]. Thus, it is possible that EF31 inhibits PKAc in tumors undergoing chronic treatment. Indeed, curcumin was shown to inhibit PKAc [29].

Scheme 1.