Inhibition of glioma progression by a newly ... - Wiley Online Library

0 downloads 0 Views 618KB Size Report
tsA58 astrocyte cell line28 was obtained from Isaiah J. Fidler .... S1d i–iii) to Figure 6 (inset) in Graeff et al.29] ..... 3b) (IC50 5 840 6 110 nM; Ki 5 420 6 55.
IJC International Journal of Cancer

Inhibition of glioma progression by a newly discovered CD38 inhibitor Eran Blacher1*, Bar Ben Baruch1, Ayelet Levy1, Nurit Geva1, Keith D. Green2, Sylvie Garneau-Tsodikova2, Micha Fridman3 and Reuven Stein1 1

Department of Neurobiology, George S. Wise Faculty of Life Sciences, Tel Aviv University, Ramat Aviv, Israel Department of Pharmaceutical Sciences, College of Pharmacy, University of Kentucky, Lexington, KY, USA 3 School of Chemistry, Raymond and Beverly Sackler Faculty of Exact Sciences, Tel Aviv University, Tel Aviv, Israel 2

Cancer Therapy

Glioma, the most common cancer of the central nervous system, has very poor prognosis and no effective treatment. It has been shown that activated microglia/macrophages in the glioma tumor microenvironment support progression. Hence, inhibition of the supporting effect of these cells may constitute a useful therapeutic approach. Recently, using a syngeneic mouse glioma progression model, we showed that the ectoenzyme CD38 regulated microglia activation and, in addition, that the loss of CD38 from the tumor microenvironment attenuated glioma progression and prolonged the life span of the tumor-bearing mice. These studies, which employed wild-type (WT) and Cd382/2 C57BL/6J mice, suggest that inhibition of CD38 in glioma microenvironment may be used as a new therapeutic approach to treat glioma. Our study tested this hypothesis. Initially, we found that the natural anthranoid, 4,5-dihydroxyanthraquinone-2-carboxylic acid (rhein), and its highly water-soluble tri-potassium salt form (K-rhein) are inhibitors of CD38 enzymatic (nicotinamide adenine dinucleotide glycohydrolase) activity (IC50 5 1.24 and 0.84 lM, respectively, for recombinant mouse CD38). Treatment of WT, but not Cd382/2 microglia with rhein and K-rhein inhibited microglia activation features known to be regulated by CD38 (lipopolysaccharide/IFN-c-induced activation, induced cell death and NO production). Furthermore, nasal administration of K-rhein into WT, but not Cd382/2 C57BL/6J, mice intracranially injected with GL261 cells substantially and significantly inhibited glioma progression. Hence, these results serve as a proof of concept, demonstrating that targeting CD38 at the tumor microenvironment by small-molecule inhibitors of CD38, for example, K-rhein, may serve as a useful therapeutic approach to treat glioma.

Gliomas account for 50% of primary brain neoplasms. To date, there is no effective treatment against glioma and the median survival of patients suffering from highly malignant glioma is approximately 12 months.1 Microglia are the resident immune cells of the central nervous system (CNS). In response to damage they acquire various features, collectively termed “activation.” Activated microglia have been suggested to play both beneficial and harmful roles in various brain pathologies.2,3 In pathological conditions, macrophages infil-

trate the brain and act together with the resident microglia. The glioma mass contains up to 30% of tumor-associated microglia and macrophages (TMMs),3,4 and there is compelling evidence that demonstrates that TMMs promote glioma progression and invasion.3–7 Therefore, abrogation of the tumor-supporting effect of TMM may offer a useful therapeutic approach for the treatment of glioma. This may be achieved by targeting proteins known to modulate microglia responses to activation signals. We previously demonstrated

Key words: microglia, tumor microenvironment, rhein, brain tumors Abbreviations: ADPR: adenosine diphosphate ribose; AICD: activation-induced cell death; Ba/F3CD38: Ba/F3 cells transfected with wildtype mouse CD38; CNS: central nervous system; CT: computed tomography; FCS: fetal calf serum; K-rhein: rhein tri-potassium salt; LPS: lipopolysaccharide; MM: microglia/macrophage; MTT: 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; NAD1: nicotinamide adenine dinucleotide; E-NAD1: 1,N6-etheno-nicotinamide adenine dinucleotide; PBS: phosphate-buffered saline; PI: propidium iodide; rhein: 4,5-dihydroxyanthraquinone-2-carboxylic acid; TMM: tumor-associated microglia and macrophage; WT: wild-type Additional Supporting Information may be found in the online version of this article. *M.F. and R.S. contributed equally to this work Grant sponsor: Israel Ministry of Trade and Industry—Ramot—KAMIN-Ramot Program, The Chief Scientist Office of the Ministry of Health, Israel; Grant number: 3-7290; Grant sponsors: The Cancer Biology Research Center of Tel Aviv University, The College of Pharmacy at the University of Kentucky DOI: 10.1002/ijc.29095 History: Received 16 June 2014; Accepted 16 Jul 2014; Online 23 Jul 2014 Correspondence to: Reuven Stein, Department of Neurobiology, George S. Wise Faculty of Life Sciences, Tel Aviv, University, Ramat Aviv 69978, Israel, Tel.: 972-3-640-8608, Fax: 972-3-640-7643, E-mail: [email protected]

C 2014 UICC Int. J. Cancer: 136, 1422–1433 (2015) V

1423

Blacher et al.

What’s new? Treatment failure in malignant glioma is attributed in part to the high degree of cellular heterogeneity within individual tumor masses. Contributing to that heterogeneity are tumor-associated microglia and macrophages, which promote glioma progression following activation. Here, in a glioma mouse model, tumor progression was effectively blocked with nasal administration of K-rhein, the water-soluble form of the small molecule-based CD38 inhibitor rhein. CD38 is thought to play a key role in microglia activation in the glioma tumor microenvironment. The results warrant further investigation into K-rhein and CD38 targeting as a new therapeutic approach for glioma.

Material and Methods Reagents

Unless otherwise stated, reagents were purchased from Sigma-Aldrich (St. Louis, MO). Rhein was purchased from Suzhou Yacoo Chemical Reagent (Suzhou, China). Mouse IFN-g was purchased from R&D Systems (Minneapolis, C 2014 UICC Int. J. Cancer: 136, 1422–1433 (2015) V

MN). Media were purchased from Invitrogen Life Technologies (Paisley, United Kingdom). Preparation of K-rhein

In brief, 3 mL of ddH2O containing 50 mg of acid-free rhein (0.176 mmol) was added to 5.28 mL of 0.1 M of KOH (total of 0.528 mmol of KOH) and stirred until the yellow-colored slurry turned into a dark crimson transparent solution. The solution was then freeze-dried to yield K-rhein as a brown powder (69 mg, 0.174 mmol, 99% yield). For the indicated experiments, the K-rhein powder was dissolved in ddH2O (2.5 mg/mL). Mice

C57BL/6J CD38-deficient mice (Cd382/2)24 were originally obtained from the Trudeau Institute Breeding Facility (Saranac Lake, NY) and maintained at the Tel Aviv University animal house. WT C57BL/6J mice were purchased from Harlan (Jerusalem, Israel) (catalog # 2BL/ 610) or from The Jackson Laboratory (Bar harbor, ME). Mice were maintained and treated in accordance with all applicable rules and guidelines of the Animal Care and Use Committee of Tel Aviv University. Cell culture

GL261 glioma25 and N9 mouse microglial cells26 were maintained in Dulbecco’s modified Eagle’s medium (DMEM), supplemented with 10% of fetal calf serum (FCS), 1% of penicillin–streptomycin and 4 mM of L-glutamine and grown at 37 C with 5% of CO2. Ba/F3 CD38neg or ectopically expressing WT mouse CD38 (Ba/F3CD38) cells27 were cultured in RPMI 1640 medium supplemented with IL-3 (ProSpec-Tany Technogene Rehovot, Israel), 10% of FCS, nonessential amino acids, sodium pyruvate, L-glutamine, b-mercaptoethanol and antibiotics. Mouse primary microglia were prepared as described in Supporting Information material. The H-2KbtsA58 astrocyte cell line28 was obtained from Isaiah J. Fidler (M. D. Anderson Cancer Center, Houston, TX). The cells were grown at the permissive temperature (32 C) in DMEM supplemented with 10% of FCS, 1% of nonessential amino-acids, 1% of sodium pyruvate, 1% of MEM vitamin solution, 1% of penicillin–streptomycin and 4 mM of L-glutamine, in 5% CO2. The effect of rhein or K-rhein on H-2Kb-tsA58 astrocyte’s viability was examined 24 hr after their transfer to 37 C (to obtain activated astrocytes).

Cancer Therapy

that CD38, which is a nicotinamide adenine dinucleotide (NAD1) glycohydrolase and ADP-ribosyl cyclase, may be such a target. CD38 is expressed by various cell types, including myeloid-derived cells. Its extracellular enzymatic domain uses NAD1 and NADP1 to catalyze the formation of the calcium-mobilizing metabolites adenosine diphosphate ribose (ADPR), cyclic ADPR (cADPR) and nicotinic acid adenine dinucleotide phosphate. Previously, we showed that CD38 regulates microglial activation both in vivo and in vitro8–10 and indicated that loss of CD38 from the tumor microenvironment inhibits glioma.11 Accordingly, in the syngeneic GL261 mouse model of glioma progression loss of CD38 (Cd382/2 mice) substantially attenuates glioma expansion and prolongs the lifespan of glioma-bearing mice. The CD38 deficiency effect was associated with increased cell death in the tumor mass as well as modulation of the TMM properties.11 These results suggest that the inhibition of CD38 catalysis could be used as a therapeutic approach for glioma treatment. To test this hypothesis, we examined the effect of a small-molecule CD38 inhibitor on the GL261 model of glioma progression. Initially, we screened a small collection of water-soluble heteroaromatic compounds. Among them was 4,5-dihydroxyanthraquinone-2-carboxylic acid (rhein); a natural anthranoid that is found in herbs such as Rheum officinale, and is widely used in traditional Chinese medicine.12 Rhein’s proposed medical properties include a laxative effect, pain and fever relief, treatment of pancreatic fibrosis, chronic liver disease, inhibition of inflammation13–17 and anticancer effects.18–23 Here, we show that rhein and its water-soluble derivative rhein tri-potassium salt (K-rhein) are low micromolar noncompetitive CD38 inhibitors that inhibit microglia activation features known to be mediated by CD38. In addition, the treatment of wild-type (WT) mice with K-rhein inhibited GL261 glioma expansion and prolonged the survival of the tumor-bearing mice. These effects were substantially lower in CD38-deficient animals. Our results demonstrate that the inhibition of CD38 by small molecules may lead to the development of a novel glioma treatment.

1424

The measurement of CD38 enzyme activity in Ba/F3 cells

CD38 enzyme activity was measured in Ba/F3 cells as previously described using 1,N6-etheno-nicotinamide adenine dinucleotide1 (E-NAD1) as the CD38 substrate.27 CD38mediated hydrolysis of E-NAD1 results in the generation of fluorescent E-ADPR that is detected at an emission wavelength of 400 nm after excitation at 310 nm. In brief, 5 3 105 of Ba/F3CD38 or Ba/F3 CD38neg cells were plated in triplicate into 96-well black microplates (Corning, Rochester, NY). E-NAD1 (final concentration, 40 lM) was added to each well in the absence or presence of different concentrations of rhein. CD38-mediated hydrolysis of E-NAD1 was determined by measuring the accumulation of the fluorescent reaction product over time using a fluorescence plate reader (Synergy HT MultiDetection Microplate Reader). The determination of substrate kinetics

The kinetic parameters of E-NAD1 were determined by irradiating the molecule at 300 nm and monitoring the fluorescence at 410 nm using a multimode SpectraMax M5 plate reader. Reactions (100 lL) containing Buffer A (MES [50 mM, pH 6.5], MgCl2 [10 mM] and NaCl [50 mM]), recombinant mouse CD38 (R&D Systems, Minneapolis, MN, catalog # 4947-AC-010) (31.3 ng/mL, 1 nM) and varying concentrations of E-NAD1 (0, 1, 5, 10, 20, 25, 50 and 100 lM) in a black 384-well plate were incubated at 25 C and the measurements were taken every 15 sec for 10 min. To convert the fluorescence reading to micromolar values, a factor of 2.32 lM/FU was used.

Cancer Therapy

The determination of rhein and K-rhein’s inhibition kinetics

IC50 values were determined by monitoring the fluorescence as discussed above in black 96-well plates. The inhibitors were dissolved in Buffer A (100 lL) and a fivefold dilution was performed. Concentrations varied from 2 pM to 0.2 mM (for rhein) or 20 pM to 2 mM (for K-rhein). Then, recombinant mouse CD38 (50 lL; final concentration, 31.3 ng/mL) in Buffer A was added and the mixture was incubated for 10 min. To initiate the reactions, a solution of E-NAD1 (final concentration, 31 lM) in Buffer A was added and the reaction progress was monitored by taking measurements every 30 sec for 10 min. Initial rates (first 5 min) were calculated and normalized to the reactions with either DMSO (rhein) or H2O (K-rhein). All assays were performed in triplicate. IC50 values were calculated using a Hill-plot fit with KaleidaGraph 4.1 software. The determination of mode of inhibition of rhein and K-rhein

Mode of inhibition was determined by using the conditions described in the section on inhibition kinetics, with varying concentrations of E-NAD1 (5, 10, 25 and 35 lM) and inhibitors: rhein (1.0, 1.2 and 1.5 lM) and K-rhein (0, 0.3 and 0.5 lM). Lineweaver-Burk analysis was then used to determine the mode of inhibition.

Inhibition of glioma progression by K-rhein

Measurement of rhein and K-rhein autofluorescence and quenching activity

Autofluorescence. The absorbance spectra of rhein and K-rhein at 100 lM revealed maximum absorbance for both compounds at 430 nm. The fluorescence spectrum of rhein and K-rhein was then determined (excitation, 264 nm; emission, 400 nm) using a fluorescence plate reader (Tecan Infinite M200 pro). Notably, excitation at 264 nm yields an identical E-ADPR fluorescence spectrum as previously reported using excitation at 300 nm [compare the spectrum obtained using excitation 264 nm (Supporting Information Figs. S1b i–iii and S1d i–iii) to Figure 6 (inset) in Graeff et al.29] Quenching effect of rhein or K-rhein on E-ADPR fluorescence. CD38 enzymatic reaction was first performed (31.3 ng/mL, 1 nM recombinant mouse CD38 and E-NAD1 [final concentration, 100 lM]) in the absence of rhein or K-rhein. After 45 min (to allow full conversion of E-NAD1 to E-ADPR) rhein or K-rhein at a final concentration of 5, 25 or 50 lM was added to the reaction tubes. The fluorescence at t 5 0 (before the addition of rhein or K-rhein) and after 1 min was measured (excitation, 264 nm; emission, 400 nm). Cell viability and NO production assays

Cell viability. Cell viability was determined by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay or propidium iodide (PI) flow cytometry analysis. MTT analysis. The different cells types were plated into 96-well plates (5 3 103 cells per well). After treatments, MTT was added to the medium at a final concentration of 0.5 mg/mL, followed by incubation at 37 C in a CO2 incubator. After 2 hr, the medium was aspirated and DMSO (200 lL) was added to the wells. Insoluble crystals were dissolved with mixing and the plates were read on a Synergy HT MultiDetection Microplate Reader, using a test wavelength of 490 nm and a reference wavelength of 690 nm. For the activation-induced cell death (AICD) experiments, primary microglial cultures or N9 cells were treated with lipopolysaccharide (LPS) (100 ng/mL) and IFN-g (100 U/mL) (hereafter LPS/IFN-g) for the indicated time periods. In some experiments, the cells were preincubated with rhein or K-rhein 2 hr prior to the addition of LPS/IFN-g. When NO production was also assessed, upon termination of the experiments, the medium from each well was carefully collected and replaced with an equal volume of fresh medium. Cell viability was then determined by the MTT assay. NO production. Nitrite, measured as a reflection of NO production in culture supernatants, was assayed using modified Griess reagent. To normalize the amount of nitrite measured in each well to the amount of cells present in that well, MTT assay was performed after the new medium was added to the cultured cells and the nitrite value in each well was normalized to the corresponding MTT value. PI flow cytometry analysis. The cells were plated in 60-mm culture dishes (106 cells per dish). After treatments, C 2014 UICC Int. J. Cancer: 136, 1422–1433 (2015) V

1425

Blacher et al.

Intracranial injection of GL261 glioma cells and treatment with K-rhein

Four-month-old male WT C57BL/6J (Harlan) mice were pretreated with vehicle (ddH2O) or K-rhein by nasal administration (5 lL [2.5 mg/mL, equivalent to 10 mg/kg] in each nostril). After 24 hr, GL261 (5 3 103) glioma cells were intracranially injected into the brains of these mice as described previously,11 followed by periodic nasal administration of vehicle or K-rhein three times per week. Computed tomography

Computed tomography (CT) images of the mice brains were acquired at 17 and 22 days post tumor implantation. Briefly, mice were anesthetized by an i.p. injection of ketamine (100 mg/kg) and xylazine (10 mg/kg), injected to the tail vein with 200 lL omnipaque (iohexol 350 mg/mL) and placed in a R Synergy micro-CT scanner at a resolution of TomoScopeV 100 lm, 40 kV X-ray voltage, scan time: 90 sec; three gantry rotations, radiation dose: 322 mGy/cm, each tube current: 1 mAmp. Tumor volume was determined using “3D Doctor” software version 4.0 (Able Software, Lexington, MA) from the Windows 7 Platform software. Scratch-induced migration

GL261 or N9 cells were seeded in six-well plates at a density of 4 3 105 or 6 3 105 cells/well, respectively. After 24 hr, R , and the cells the medium was replaced with Opti-MEMV were treated with vehicles, rhein or K-rhein. After 24 hr, the medium was collected and replaced with PBS, then three areas were scratched creating three gaps of similar widths. The PBS was replaced by the corresponding collected media. At t 5 0, 24, 48 and 72 hr or 3 and 6 hr, for GL261 and N9 cells, respectively, phase-contrast images of the wells were obtained with a CCD camera connected to an Olympus fluorescent microscope (objective, 43). The region imaged at time zero was marked to enable photographing the same area at different times. The widths of gaps at the different time points were measured using the ImageJ software. The data acquired from the three scratches on each plate were averaged to obtain the mean gap width at a given time. Glioma-bearing mice survival analysis

After injection of the GL261 cells, mice were monitored and weighed daily. The endpoint was defined by a lack of physical activity and more than 15% reduction in body weight. The probability of survival was calculated using the Kaplan– Meier method, and statistical analysis was performed using a C 2014 UICC Int. J. Cancer: 136, 1422–1433 (2015) V

log-rank test. Additional information can be found in Supporting Information material.

Results Rhein is an inhibitor of mouse CD38

CD38 enzymatic activity was tested by the assay of hydrolysis of the CD38 substrate E-NAD1 to E-ADPR (Fig. 1a).30 First, the effect of rhein was examined on CD38 activity in live Ba/ F3 cells. As shown previously,27 Ba/F3CD38, but not the parental CD38neg control cells, hydrolyzed E-NAD1 and generated the fluorescent product E-ADPR (Fig. 1b). However, incubation of the cells also with 5, 10 and 50 lM of rhein resulted in dose-dependent inhibition of E-ADPR production; at a concentration of 50 lM rhein almost completely abolished generation of E-ADPR. To better determine the inhibition properties of rhein, we used recombinant mouse CD38. Recombinant CD38 was incubated with E-NAD1 in the presence of increasing concentrations of rhein and E-ADPR generation was monitored. The results show that the recombinant CD38 enzyme obeyed the Michaelis–Menten model (Fig. 1c) and that rhein inhibited E-ADPR production with an IC50 value of 1.24 6 0.10 lM, and a Ki value of 0.62 6 0.05 lM (Fig. 1d). Further kinetic analysis revealed that rhein acts as a noncompetitive inhibitor (Fig. 1e). Notably, under equivalent conditions rhein autofluorescence was negligible (Supporting Information Fig. S1a-ii) and it did not quench E-ADPR fluorescence up to 25 lM. At 50 lM, a concentration that is much higher than the IC50, rhein exhibited a low (5%) quenching effect on E-ADPR fluorescence (Supporting Information Fig S1b-vi). Inhibition of CD38-regulated activation of microglia by rhein

Having demonstrated that rhein inhibits the enzymatic activity of CD38, we examined if it can also inhibit CD38regulated biological effects. To this end, we focused on microglia as microglia/macrophages are the main cells that constitute the glioma microenvironment3 and are known to express CD38.8 As the previous studies showed that in some cells rhein has cytotoxic effect,22,31 we first examined the effect of increasing concentrations of rhein on N9 microglia cell viability. The results obtained are shown in Figure 2a. One-way ANOVA analysis with repeated measurements revealed a significant difference within the different treatments (p < 0.001) where rhein significantly (p < 0.001, Bonferroni post hoc test) reduced N9 microglia viability at 50 and 100 lM. Therefore, to avoid interference of rhein’s cytotoxic effect with its other activities, in the following experiments rhein concentrations did not exceed 25 lM. As we previously demonstrated that CD38 enzymatic activity regulates LPS/IFN-g-stimulated AICD and NO production in microglia,8 these two effects were next examined. N9 microglial cells were treated with LPS/IFN-g in the absence or presence of 1, 5 and 10 lM rhein for 24 hr after which cell viability and NO production were determined.

Cancer Therapy

the cells were detached, centrifuged and washed with phosphate-buffered saline (PBS). Then, 50 lg/mL of PI was added and cells were subjected to flow cytometry using Becton Dickinson FaCSort [BD Bioscience (San Jose, CA)]. The data were analyzed using the Flowing software version 2.5.0 supplied by the manufacturer. Dead cells are those that do not exclude the PI dye.

1426

Inhibition of glioma progression by K-rhein

Cancer Therapy

Figure 1. Rhein is a CD38 inhibitor. (a) Structure of rhein. (b) Rhein inhibits the CD38 enzymatic activity in Ba/F3CD38 cells. Ba/F3 CD38neg or Ba/F3CD38 cells were plated in 96-well plates (5 3 105 cells per well) and then untreated or treated with the indicating concentrations of rhein. CD38 enzymatic activity (NAD1 glycohydrolase) was determined. Accumulation of E-ADPR was measured over time and is represented as relative fluorescence units (RFUs). The results are expressed as average RFU (n 5 6). (c–e) Characterization of the effect of rhein on the activity of recombinant mouse CD38. (c) Michaelis–Menten curve of the E-NAD1 reaction. (d) Rhein IC50 curve and (e) its kinetic study.

The results obtained are shown in Figure 2b (viability) and Figure 2c (NO production). One-way ANOVA with repeated measurements revealed a significant difference within the different treatments for each viability and NO production (p < 0.001). As expected, LPS/IFN-g treatment significantly (p < 0.001, Bonferroni post hoc test) reduced cell viability and enhanced NO production compared to untreated cells. Treatment with rhein at all tested concentrations did not substantially affect viability and NO production of LPS/IFN-guntreated cells; however, they significantly (p < 0.001, Bonferroni post hoc test) increased cell viability and reduced NO production of LPS/IFN-g-treated cells compared to LPS/IFNg-alone-treated cells. Next, the effect of rhein was tested on primary WT microglia untreated or treated with LPS/IFN-g in the absence or presence of rhein (1 and 5 lM) for 24, 48 and 72 hr. The results obtained are shown in Figure 2d (viability) and Figure 2e (NO production). Two-way ANOVA

with repeated measurements revealed a significant difference for treatment, time and treatment 3 time for each viability and NO production (p < 0.001). Similar to N9 cells, LPS/ IFN-g treatment significantly (p < 0.001, Bonferroni post hoc test) reduced WT microglia cell viability and enhanced NO production compared to untreated cells, whereas the addition of rhein to the LPS/IFN-g-treated cells significantly (p < 0.001, Bonferroni post hoc test) enhanced cell viability and reduced NO production compared to cells treated only with LPS/IFN-g. To verify that these rhein’s effects are CD38 dependent, we performed the same experiments on CD38deficient (Cd382/2) primary microglia. The results obtained are shown in Figure 2f (viability) and Figure 2g (NO production). Two-way ANOVA with repeated measurements of the viability results did not reveal a significant difference for treatment. In addition, in contrast to the results in WT microglia (Fig. 2d) and as expected from our previous C 2014 UICC Int. J. Cancer: 136, 1422–1433 (2015) V

studies,8 the treatment of Cd382/2 microglia with LPS/IFN-g barely affected cell viability. Furthermore, the addition of rhein to LPS/IFN-g-treated Cd382/2 microglia, in contrast to its effect on WT microglia, did not increase cell viability. Two-way ANOVA with repeated measurements of NO production revealed a significant difference for treatment (p < 0.001). LPS/IFN-g significantly increased NO production (p < 0.031, Bonferroni post hoc test) at all time points examined; however, this effect was much smaller than in WT microglia. Similarly, in contrast to the substantial reduction in NO production in WT microglia, 1 lM of rhein enhanced LPS/IFN-g-induced NO production at 24 hr and modestly reduced it at 48 and 72 hr (p < 0.031, Bonferroni post hoc test). Rhein at a concentration of 5 lM enhanced LPS/IFNg-induced NO production at all time points (p < 0.031, Bonferroni post hoc test). Taken together, these results suggest that treatment with rhein affects microglia in a CD38dependent manner and mimics the effect of CD38 deficiency.8 However, the results showing that rhein affects NO production in LPS/IFN-g-treated Cd382/2 cells suggest that rhein has additional target(s) besides CD38. Previously, it was reported that the CD38 metabolite cADPR controls N9 microglia migration.9 Therefore, the effect of rhein on N9 microglia migration was examined using the scratch assay, which monitors the ability of migrating cells to close a gap formed by a scratch. The normalized results of the gap size in rhein-treated and untreated cells are shown in Figure 2h. Two-way ANOVA with repeated measurements revealed significant effects for time and time 3 treatment (p < 0.001). Assessing the effect of rhein at the 3and 6-hr time points reveled a small but significant inhibition only at 25 lM rhein (p < 0.015, one-way ANOVA followed by Dunnett’s t-test) at 6 hr. K-rhein inhibits glioma progression

As the free acid form of rhein has poor solubility in physiological solutions, K-rhein, a tri-potassium salt (Fig. 3a) was generated. K-rhein is highly soluble in a physiological aqueous solution and has similar mouse CD38 inhibition properties as those of rhein, as indicated by the inhibition kinetics of K-rhein (Fig. 3b) (IC50 5 840 6 110 nM; Ki 5 420 6 55 nM) and its noncompetitive mode of inhibition (Fig. 3c). Under equivalent conditions, K-rhein autofluorescence is negligible (Supporting Information Fig. S1c-ii) and it did not quench E-ADPR fluorescence up to a concentration of 25 lM. At 50 lM, a concentration that is much higher than its IC50 value, K-rhein displayed 10% of quenching effect on EADPR fluorescence (Supporting Information Fig. S1d-vi). To confirm that K-rhein exhibits similar biological effects as rhein, we tested its effects on N9 microglia cell viability, LPS/IFN-g-induced AICD and NO production as well as on cell migration. Figure 3d shows the effect of different concentrations of K-rhein on N9 viability. A significant reduction in cell viability was observed only at 50 and 100 lM of K-rhein (one-way ANOVA, Bonferroni post hoc test, p < 0.001). The C 2014 UICC Int. J. Cancer: 136, 1422–1433 (2015) V

1427

effects of K-rhein on AICD and NO production in LPS/IFNg-treated N9 cells are shown in Figure 3e (viability) and Figure 3f (NO production). One-way ANOVA with repeated measurements revealed a significant difference within the different treatments for each viability and NO production (p < 0.001). As expected, LPS/IFN-g treatment significantly (p < 0.001, Bonferroni post hoc test) reduced cell viability and enhanced NO production compared to untreated cells. K-rhein at all tested concentrations significantly (p < 0.001, Bonferroni post hoc test) increased cell viability and reduced NO production (beside at 5 lM) of LPS/IFN-g-treated cells compared to cells treated only with LPS/IFN-g. The effect of K-rhein (1, 5 and 25 lM) on N9 migration is shown in Figure 3g. Two-way ANOVA with repeated measurements revealed significant effects for time and time 3 treatment (p < 0.001). One-way ANOVA analysis of the effect of rhein at the 3- and 6-hr time points revealed a modest but significant inhibition effect of K-rhein at all concentrations at both time points (p < 0.001, Dunnett’s t-test). Having demonstrated that rhein and K-rhein have similar inhibitory effects, we examined the effect of K-rhein on glioma progression in vivo. For this purpose, we used the syngeneic GL261 model of glioma progression that we employed previously to demonstrate that CD38 deficiency in the tumor microenvironment attenuates glioma expansion.11 WT mice were divided into two groups: one was treated with K-rhein and the other with vehicle. K-rhein (0.25 mg/10 lL) and the vehicle were nasally administrated. One day before GL261 cells implantation into mice brains, the mice were treated with Krhein or vehicle followed by 2-day interval of K-rhein or vehicle treatments. To examine to what extent the effect of K-rhein on glioma progression is CD38-dependent, similar experiments were performed on Cd382/2 mice. The tumor volume was assessed 17 and 22 days postinjection by CT scanning. Figures 4a and 4b (CT images) and Figure 4c and 4d (quantification) show the tumor volume in the K-rhein-untreated and -treated WT and Cd382/2 mice, respectively. The results show that at day 17, K-rhein inhibited tumor volume by 84 and 38% in WT and Cd382/2 mice, respectively. At day 22, K-rhein inhibited tumor volume by 74 and 19% in WT and Cd382/2 mice, respectively. Three-way ANOVA analysis of the results of WT and Cd382/2 mice on day 17 or 22 revealed a significant effect for time and for treatment 3 genotype (p < 0.001). These results demonstrate that K-rhein reduced tumor volume in both WT and Cd382/2 mice; however, this effect was substantially and significantly higher in WT mice compared to Cd382/ 2 mice. Kaplan–Meier survival analysis (Fig. 3e) revealed that K-rhein also significantly prolonged the survival of the gliomabearing WT mice (median survival of K-rhein-treated WT mice was 38 vs. 30.5 days of vehicle-treated WT mice, p 5 0.0007). Notably, in glioma-bearing Cd382/2 mice, K-rhein extended median survival by 2 days (30 vs. 28 days); however, this effect was not significant (Fig. 3f). Taken together, these results demonstrate that K-rhein inhibits glioma progression and that this effect is mainly CD38 dependent.

Cancer Therapy

Blacher et al.

1428

Inhibition of glioma progression by K-rhein

effects on GL261 cell viability and migration in vitro. GL261 cells were grown for 24 and 48 hr in the absence or presence of increasing concentrations of rhein or K-rhein. Cell viability was assessed by MTT assay. The results shown in Figures 5a

Cancer Therapy

As the treatment of WT mice with the CD38 inhibitor could affect both the tumor microenvironment and the tumor cells, we tested whether rhein or K-rhein can also directly affect GL261 cells, and therefore examined their

Figure 2.

C 2014 UICC Int. J. Cancer: 136, 1422–1433 (2015) V

1429

Blacher et al.

protein (measured by flow cytometric analysis) (data not shown).

Discussion The results presented demonstrate that rhein and K-rhein are low micromolar inhibitors of mouse CD38 that mimic the effect of CD38 deficiency on microglia activation. In addition, K-rhein substantially inhibited the glioma progression and prolonged the lifespan of the glioma-bearing mice. Hence, our study offers experimental evidence, supporting our hypothesis that targeting CD38 in the tumor microenvironment can be used as a new therapeutic approach for glioma treatment. The effects of rhein and K-rhein on microglia activation and of K-rhein on glioma progression were mainly mediated via CD38 as they were largely abrogated in Cd382/2 cells and mice. However, the observation that K-rhein also modestly inhibited tumor volume in Cd382/2 mice suggests that it has additional targets that also contribute to its effect. This notion is supported by the findings that low concentrations of rhein affect NO production in LPS/IFN-g-treated Cd382/2 primary microglia and that high concentration of rhein and K-rhein was toxic to GL261 cells, which do not express CD38. Previously, we showed that CD38 deficiency inhibited LPS/ IFN-g-induced NO production in primary microglia.8 In agreement with these results, rhein inhibited LPS/IFN-ginduced NO production in WT primary and N9 microglia. However, via its CD38-independent effect, rhein also has an opposite effect on NO production, that is, the enhancement of NO production in LPS/IFN-g-treated microglia. These opposite effects of rhein may explain the partial inhibitory effect of rhein on LPS/IFN-g-induced NO production in N9 cells and WT primary microglia. The observation that rhein alone does

Figure 2. The effect of rhein on N9 and primary microglia. (a) The effect of rhein on N9 cell viability. N9 cells plated in 96-well plates were treated with the indicated concentrations of rhein or with vehicle (DMSO, 0.05%) alone (0 rhein). After 24 hr, cell viability was determined by the MTT assay. Viability is expressed as a percentage of the MTT values of the vehicle alone treatment. The values shown are means 6 SEM (bars) (n 5 3). ***p < 0.001, significantly different from vehicle alone-treated cells (Bonferroni post hoc test). (b, c) Rhein inhibits AICD and NO production in N9 cells. N9 cells plated in 96-well plates were treated without (2) or with LPS/IFN-g (1) in the presence of the indicated concentrations of rhein or vehicle (DMSO, 0.05%). After 24 hr, cell viability and NO production were measured. Cell viability (b) was determined by the MTT assay and expressed as a percentage of the MTT values of the vehicle-alone treatment. The values shown are means 6 SEM (bars) (n 5 3). ***p < 0.001, significantly different from vehicle or LPS/IFN-g alone-treated cells (Bonferroni post hoc test). NO production (c) was measured by Griess reaction. Nitrite levels were normalized to the amount of live cells in the corresponding treatment, as determined by MTT assay, and the results shown are expressed as percentage of the normalized nitrite values of vehicle-treated cells. The values shown are means 6 SEM (bars) (n 5 3). ***p < 0.001, significantly different from vehicle or LPS/IFN-g alone-treated cells (Bonferroni post hoc test). The nitrite values in LPS/IFN-g-treated samples were 0.44 6 0.009 mM (non-normalized) and 4.58 0.15 mM/MTT (normalized to MTT). (d–g) Rhein inhibits AICD and NO production in primary WT microglia but not primary Cd382/2 microglia. Primary microglia from WT (d, e) and Cd382/2 mice (f, g) were treated with LPS/IFN-g together with the indicated concentrations of rhein or vehicle (DMSO, 0.05%) (LPS/IFN-g 1 DMSO) or were treated with vehicle alone (control). At the indicated time periods, cell viability (d, f) and NO production (e, g) were determined as described above. The values shown are means 6 SEM (bars) (n 5 4) ***p < 0.001 significantly different from control or LPS/IFN-g 1 DMSO-treated cells (Bonferroni post hoc test). The nitrite values in LPS/IFN-g-treated samples were 4.73 6 1.5, 6.68 6 2.9 and 8.28 6 0.01 mM (non-normalized) and 7.83 6 2.3, 12.04 6 3.6 and 16.32 6 3.5 (normalized to MTT), and 1.09 6 0.1, 1.19 6 0.1 and 1.3 6 0.05 mM (non-normalized) and 0.99 6 0.05, 1.26 6 0.14 and 0.91 6 0.06 (normalized to MTT) at 24, 48 and 72 hr in WT primary microglia and Cd382/2 primary microglia, respectively. (h) The effect of rhein on cell migration. N9 cells were treated with DMSO (0.25%) or the indicated concentrations of rhein, scratch wounds were inflicted and the resulting gaps were imaged at the indicted time points. The results are expressed as the percentage of gap width after treatment from the original gap width. The values shown are means 6 SEM (bars) (n 5 3) *p 5 0.015 (Dunnett’s t-test) significant difference compared to vehicle alone-treated cells.

C 2014 UICC Int. J. Cancer: 136, 1422–1433 (2015) V

Cancer Therapy

and 5b, for rhein and K-rhein, respectively, show that after 24 hr both rhein and K-rhein significantly inhibited GL261 cell viability at 50 and 100 lM (p < 0.05, Student’s t-test). After 48 hr, these compounds significantly inhibited GL261 cell viability also at lower concentrations although this effect was small (up to 10%). Cell viability was also assessed by PI flow cytometry, which directly measures cell death. In agreement with the MTT assay results, treatment with 100 lM rhein (Supporting Information Fig. S2a) or K-rhein (Supporting Information Fig. S2b) substantially and significantly increased cell death (p < 0.05, Student’s t-test). Notably, rhein and K-rhein were also toxic for astrocytes at high concentrations (100 lM; p < 0.05, Student’s t-test) (Supporting Information Figs. S3a and S3b, respectively) although this effect was small (15%). Next, the effect of rhein and K-rhein on migration was examined by the scratch assay. GL261 cells were treated with rhein or K-rhein at concentrations of 0.5–25 lM which only slightly affected cell viability (Figs. 5a and 5b). The results are shown in Figures 5c and 5d for rhein and K-rhein, respectively. Two-way ANOVA with repeated measurements revealed significant effects for treatment, time and time 3 treatment (p < 0.001). After 24 hr, only 25 lM of rhein significantly inhibited the closure of the gap, whereas at 48 and 72 hr, all tested concentrations significantly inhibited gap closure (p < 0.006, Dunnett’s ttest). K-rhein significantly inhibited closure of the gap at all concentrations after 48 hr, whereas after 72 hr, only 5 and 25 lM K-rhein had a significant effect (p < 0.006, Dunnett’s t-test). Hence, rhein and K-rhein inhibit migration of GL261 cells. Taken together, these results suggest that rhein and K-rhein can also act on the tumor cells directly. This effect, however, is not mediated via CD38 as GL261 cells did not exhibit a detectable CD38 enzymatic activity (examined using the E-ADPR production assay) or plasma membrane

1430

Inhibition of glioma progression by K-rhein

Cancer Therapy

Figure 3. Biochemical and biological features of K-rhein. (a–c) K-rhein inhibits CD38 enzymatic activity. (a) Structure of K-rhein. (b) K-rhein IC50 curve. (c) K-rhein kinetic study. (d–g) The effect of K-rhein on N9 microglia. N9 cells were treated with the indicated concentrations of K-rhein without or with LPS/IFN-g. The results of cell viability of K-rhein-treated cells (d); viability (e) and NO production (f) of LPS/IFN-gtreated cells or cell migration were expressed as shown in Figures 2a–c and 2h. The values shown are means 6 SEM (bars) (n 5 3). (d) ***p < 0.001, significantly different from vehicle alone-treated cells (Bonferroni post hoc test) (e and f) ***p < 0.001, significantly different from vehicle or LPS/IFN-g alone-treated cells (Bonferroni post hoc test). (g) ***p < 0.001, significantly different from vehicle alone-treated cells (Dunnett’s t-test).

not enhance NO production (Fig. 2c) suggests that this effect is LPS/IFN-g dependent. Our results are in agreement with the previous studies that showed that rhein inhibited LPSinduced NO production in macrophages32,33 and that rhein can promote opposite (pro- and antiinflammatory) effects in LPS-treated macrophages.33 Taken together, our results show that the effect of rhein and K-rhein is both CD38 dependent and independent. However, with respect to the effect of these molecules on glioma progression and on some aspects of microglia activation, these effects are mainly CD38 dependent. Analogous to CD38 deficiency, the mechanism whereby K-rhein attenuates glioma progression probably involves the modulation of microglia activation features in the tumor microenvironment, which, in addition to NO production and AICD, may also include cytokine production.11 Indeed, a trend to an inhibitory effect of rhein on IL-6 and TNF-a mRNA expression was observed (data not shown). Furthermore, our results suggest that K-rhein may also act by inhibiting microglia migration toward the tumor and even by killing microglia

at high concentrations. The effects of rhein and K-rhein on microglia activation probably involve cADPR-mediated increase in intracellular-free Ca21 as we showed previously using CD38 deficiency.8 This in turn may culminate into NFjB signaling as other studies showed that Ca21 can regulate NFjB activity34,35 and NFjB activation was shown to play a major role in transcriptional regulation of inflammation.36 Alternatively, inhibition of CD38 enzymatic activity can lead to an increase in NAD1 levels, which in turn may increase the activity of the NAD1 consuming enzyme SIRT2 that was shown to inhibit microglia activation.36 K-rhein was delivered into the mice by nasal administration. This procedure was used as it was demonstrated as an effective drug delivery route to the CNS37,38 and because it is believed to bypass the blood-brain barrier.39 In addition, it was suggested that nose-to-brain transport of drugs has a therapeutic potential to combat glioblastoma.40 In WT mice, treatment with K-rhein reduced tumor size by 84 and 74% at day 17 and 22, respectively, and increased C 2014 UICC Int. J. Cancer: 136, 1422–1433 (2015) V

1431

Figure 4. K-rhein inhibits tumor volume and prolongs survival of glioma-bearing mice. WT (n 5 78, three independent experiments) and Cd382/2 (n 5 49, two independent experiments) male mice were injected with GL261 cells and treated with vehicle (ddH2O) (n 5 39 and 25 for WT and Cd382/2, respectively) or K-rhein (n 5 39 and 24 for WT and Cd382/2, respectively). Mice brains were scanned by CT at 17 and 22 days post-injection. Images of representative WT (a) and Cd382/2 (b) mice treated with vehicle (upper panel) or K-rhein (lower panel) taken at 22 days post-injection are shown. (c, d) Quantification of the tumor volumes of vehicle- or K-rhein-treated WT (c) and Cd382/2 (d) mice. Results are presented as mean 6 S.E.M (bars). Kaplan–Meier survival curves of the same WT (e) or Cd382/2 (f) mice treated with vehicle or K-rhein.

median survival by 7.5 days, whereas CD38 deficiency reduced tumor size by 38 and 19% at day 17 and 22, respectively, and increased median survival by 3.5 days. It is possible that what appeared as a stronger effect of K-rhein relative to CD38 deficiency is owing to its effect on the tumor cells. Indeed, our findings show that rhein and K-rhein have inhibitory effects on GL261 cell viability and migration and other showed that rhein inhibits proliferation,18 migration and invasion,41 and induces apoptosis in various cancer cells in vitro.20,21,41,42 The effects of rhein and K-rhein on the GL261 cells are probably CD38 independent as neither CD38 expression nor enzymatic activity was detected in these cells. The corresponding mechanism is currently unknown yet may C 2014 UICC Int. J. Cancer: 136, 1422–1433 (2015) V

involve the NFjB pathway,33 expression of the p53, CD95,22 MMPs43,44 and JNK and p38 phosphorylation. Regardless of the possibility that rhein or K-rhein may act in a CD38-independent manner, our results show that the effect of K-rhein on glioma progression is mainly CD38 dependent. Therefore, our study strongly supports our hypothesis that targeting CD38 by small-molecule inhibitors may become a useful therapeutic approach to treat glioma. This view is supported by the observation that Cd38 is expressed in tumor samples of human glioblastoma multiform patients and that increased the expression of Cd38 significantly correlates with a reduction in patients’ survival (Supporting Information material).

Cancer Therapy

Blacher et al.

1432

Inhibition of glioma progression by K-rhein

Cancer Therapy

Figure 5. Rhein and K-rhein inhibit viability and migration of GL261 cells. Rhein (a) or K-rhein (b) effect on cell viability. GL261 cells were treated with the indicated concentrations of rhein, K-rhein or with vehicle [DMSO, 1.8% (0 rhein) or ddH2O, respectively] for 24 and 48 hr after which cell viability was examined. Viability was expressed as a percentage of the MTT values without rhein or K-rhein. The values shown are means 6 SEM (bars) (n 5 3). [*p < 0.05, **p < 0.005, ***p < 0.005, a significant difference between vehicle alone-treated cells and rhein or K-rhein-treated cells at 24 or 48 hr (Student’s t-test)]. The effect of rhein (c) or K-rhein (d) on cell migration. GL261 cells were treated with DMSO (0.25%), ddH2O (vehicles) or the indicated concentrations of rhein or K-rhein. A scratch wound assay was performed, analyzed and expressed as shown in Figure 2h. The values shown are means 6 SEM (bars) (n 5 3). **p < 0.005 (Dunnett’s t-test), significant difference between vehicle alone-treated cells and rhein or K-rhein-treated cells.

Acknowledgements The authors thank Prof. Isaiah J. Fidler for providing H-2KbtsA58 astrocytes, Amit Frishberg for helping with the human glioma bioinformatics analysis, Dr. Maria Kramer and Mr. Ori Green (School of Chemistry) and Dr. Yael Zilberstein (Sackler Cellular & Molecular Imaging Center) for their excellent techni-

cal support. This study was supported by Israel Ministry of Trade and Industry—Ramot—KAMIN-Ramot Program (for R.S. and M.F.), The Chief Scientist Office of the Ministry of Health, Israel (grant no. 3-7290) (for R.S.), The Cancer Biology Research Center of Tel Aviv University (for R.S.) and The College of Pharmacy at the University of Kentucky (for S.G.-T.).

References 1.

2.

3.

4.

Krex D, Klink B, Hartmann C, et al. Long-term survival with glioblastoma multiforme. Brain 2007;130:2596–606. Rivest S. Regulation of innate immune responses in the brain. Nat Rev Immunol 2009;9: 429–39. Charles NA, Holland EC, Gilbertson R, et al. The brain tumor microenvironment. Glia 2012;60: 502–14. Badie B, Schartner J. Role of microglia in glioma biology. Microsc Res Tech 2001;54:106–13.

5.

6.

7.

Watters JJ, Schartner JM, Badie B. Microglia function in brain tumors. J Neurosci Res 2005;81: 447–55. Yang I, Han SJ, Kaur G, et al. The role of microglia in central nervous system immunity and glioma immunology. J Clin Neurosci 2010;17: 6–10. Markovic DS, Vinnakota K, Chirasani S, et al. Gliomas induce and exploit microglial MT1MMP expression for tumor expansion. Proc Natl Acad Sci USA 2009;106:12530–5.

8.

Mayo L, Jacob-Hirsch J, Amariglio N, et al. Dual role of CD38 in microglial activation and activation-induced cell death. J Immunol 2008; 181:92–103. 9. Partida-Sanchez S, Iribarren P, Moreno-Garcia ME, et al. Chemotaxis and calcium responses of phagocytes to formyl peptide receptor ligands is differentially regulated by cyclic ADP ribose. J Immunol 2004;172:1896–906. 10. Levy A, Bercovich-Kinori A, Alexandrovich AG, et al. CD38 facilitates recovery from

C 2014 UICC Int. J. Cancer: 136, 1422–1433 (2015) V

1433

Blacher et al.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

23. Ip SW, Weng YS, Lin SY, et al. The role of Ca21 on rhein-induced apoptosis in human cervical cancer Ca Ski cells. Anticancer Res 2007;27:379– 89. 24. Cockayne DA, Muchamuel T, Grimaldi JC, et al. Mice deficient for the ecto-nicotinamide adenine dinucleotide glycohydrolase CD38 exhibit altered humoral immune responses. Blood 1998;92:1324– 33. 25. Szatmari T, Lumniczky K, Desaknai S, et al. Detailed characterization of the mouse glioma 261 tumor model for experimental glioblastoma therapy. Cancer Sci 2006;97:546–53. 26. Righi M, Mori L, De Libero G, et al. Monokine production by microglial cell clones. Eur J Immunol 1989;19:1443–8. 27. Lund FE, Muller-Steffner H, Romero-Ramirez H, et al. CD38 induces apoptosis of a murine pro-B leukemic cell line by a tyrosine kinase-dependent but ADP-ribosyl cyclase- and NAD glycohydrolase-independent mechanism. Int Immunol 2006;18:1029–42. 28. Langley RR, Fan D, Guo L, et al. Generation of an immortalized astrocyte cell line from H-2KbtsA58 mice to study the role of astrocytes in brain metastasis. Int J Oncol 2009;35:665–72. 29. Graeff RM, Walseth TF, Hill HK, et al. Fluorescent analogs of cyclic ADP-ribose: synthesis, spectral characterization, and use. Biochemistry 1996;35:379–86. 30. Partida-Sanchez S, Cockayne DA, Monard S, et al. Cyclic ADP-ribose production by CD38 regulates intracellular calcium release, extracellular calcium influx and chemotaxis in neutrophils and is required for bacterial clearance in vivo. Nat Med 2001;7:1209–16. 31. Chang CY, Chan HL, Lin HY, et al. Rhein induces apoptosis in human breast cancer cells. Evid Based Complement Alternat Med 2012;2012: 952504. 32. Wang CC, Huang YJ, Chen LG, et al. Inducible nitric oxide synthase inhibitors of Chinese herbs III. Rheum palmatum. Planta Med 2002;68:869–74. 33. Gao Y, Chen X, Fang L, et al. Rhein exerts proand anti-inflammatory actions by targeting IKKbeta inhibition in LPS-activated macrophages. Free Radic Biol Med 2014;72C:104–12.

34. Lilienbaum A, Israel A. From calcium to NFkappa B signaling pathways in neurons. Mol Cell Biol 2003;23:2680–98. 35. Zhang H, Li Y, Wang S, et al. LPS-induced NFkappa B activation requires Ca21 as a mediator in isolated pancreatic acinar cells of rat. Chin Med J 2003;116:1662–7. 36. Lawrence T. The nuclear factor NF-kappaB pathway in inflammation. Cold Spring Harb Perspect Biol 2009;1:a001651. 37. Born J, Lange T, Kern W, et al. Sniffing neuropeptides: a transnasal approach to the human brain. Nat Neurosci 2002;5:514–6. 38. Illum L. Transport of drugs from the nasal cavity to the central nervous system. Eur J Pharm Sci 2000;11:1–18. 39. Badhan RS, Kaur M, Lungare S, et al. Improving brain drug targeting through exploitation of the nose-to-brain route: a physiological and pharmacokinetic perspective. Curr Drug Deliv, 2014;11:458–71 40. van Woensel M, Wauthoz N, Rosiere R, et al. Formulations for intranasal delivery of pharmacological agents to combat brain disease: a new opportunity to tackle GBM? Cancers 2013;5: 1020–48. 41. Chen YY, Chiang SY, Lin JG, et al. Emodin, aloe-emodin and rhein inhibit migration and invasion in human tongue cancer SCC-4 cells through the inhibition of gene expression of matrix metalloproteinase-9. Int J Oncol 2010;36: 1113–20. 42. Fernand VE, Losso JN, Truax RE, et al. Rhein inhibits angiogenesis and the viability of hormone-dependent and -independent cancer cells under normoxic or hypoxic conditions in vitro. Chem Biol Interact 2011;192:220–32. 43. Tamura T, Kosaka N, Ishiwa J, et al. Rhein, an active metabolite of diacerein, down-regulates the production of pro-matrix metalloproteinases-1, 3, -9 and -13 and up-regulates the production of tissue inhibitor of metalloproteinase-1 in cultured rabbit articular chondrocytes. Osteoarthritis Cartilage 2001;9:257–63. 44. Sanchez C, Mathy-Hartert M, Deberg MA, et al. Effects of rhein on human articular chondrocytes in alginate beads. Biochem Pharmacol 2003;65: 377–88.

Cancer Therapy

22.

traumatic brain injury. J Neurotrauma 2009;26: 1521–33. Levy A, Blacher E, Vaknine H, et al. CD38 deficiency in the tumor microenvironment attenuates glioma progression and modulates features of tumor-associated microglia/macrophages. Neuro Oncol 2012;14:1037–49. Peigen X, Liyi H, Liwei W. Ethnopharmacologic study of Chinese rhubarb. J Ethnopharmacol 1984;10:275–93. Franchi-Micheli S, Lavacchi L, Friedmann CA, et al. The influence of rhein on the biosynthesis of prostaglandin-like substances in-vitro. J Pharm Pharmacol 1983;35:262–4. Guo MZ, Li XS, Xu HR, et al. Rhein inhibits liver fibrosis induced by carbon tetrachloride in rats. Acta Pharmacol Sin 2002;23:739–44. Gorji A. Pharmacological treatment of headache using traditional Persian medicine. Trends Pharmacol Sci 2003;24:331–4. Ma BL, Ma YM, Yan DM, et al. Effective constituents in Xiexin Decoction for anti-inflammation. J Ethnopharmacol 2009;125:151–6. Tsang SW, Zhang H, Lin C, et al. Rhein, a natural anthraquinone derivative, attenuates the activation of pancreatic stellate cells and ameliorates pancreatic fibrosis in mice with experimental chronic pancreatitis. PloS One 2013;8:e82201. Cichewicz RH, Zhang Y, Seeram NP, et al. Inhibition of human tumor cell proliferation by novel anthraquinones from daylilies. Life Sci 2004;74: 1791–9. Huang Q, Lu G, Shen HM, et al. Anti-cancer properties of anthraquinones from rhubarb. Med Res Rev 2007;27:609–30. Hsia TC, Yang JS, Chen GW, et al. The roles of endoplasmic reticulum stress and Ca21 on rheininduced apoptosis in A-549 human lung cancer cells. Anticancer Res 2009;29:309–18. Lai WW, Yang JS, Lai KC, et al. Rhein induced apoptosis through the endoplasmic reticulum stress, caspase- and mitochondria-dependent pathways in SCC-4 human tongue squamous cancer cells. In Vivo 2009;23:309–16. Kuo PL, Hsu YL, Ng LT, et al. Rhein inhibits the growth and induces the apoptosis of Hep G2 cells. Planta Med 2004;70:12–6.

C 2014 UICC Int. J. Cancer: 136, 1422–1433 (2015) V