Sabine Grösch

COX-2 independent induction of cell cycle arrest and apoptosis in colon cancer cells by the selective COX-2 inhibitor celecoxib

The regular use of various nonsteroidal anti‐inflammatory drugs (NSAIDs) was shown to decrease the incidence of colorectal cancer. This effect is thought to be caused predominantly by inhibition of cyclooxygenase‐2 (COX‐2) and, subsequently, prostaglandin synthesis. However, recent studies have suggested that COX‐independent pathways may contribute considerably to these antiproliferative effects. To evaluate the involvement of COX‐dependent and COX‐independent mechanisms further, we assessed the effects of celecoxib (selective COX‐2 inhibitor) and SC560 (selective COX‐1 inhibitor) on cell survival, cell cycle distribution, and apoptosis in three colon cancer cell lines, which differ in their expression of COX‐2. Both drugs induced a G0/G1 phase block and reduced cell survival independent of whether or not the cells expressed COX‐2. Celecoxib was more potent than SC560. The G0/G1 block caused by celecoxib could be attributed to a decreased expression of cyclin A, cyclin B1, and cyclin‐dependent kinase‐1 and an increased expression of the cell cycle inhibitory proteins p21Waf1 and p27Kip1. In addition, celecoxib, but not SC560, induced apoptosis, which was also independent of the COX‐2 expression of the cells. In vivo, celecoxib as well as SC560 reduced the proliferation of HCT‐15 (COX‐2 deficient) colon cancer xenografts in nude mice, but both substances had no significant effect on HT‐29 tumors, which express COX‐2 constitutively. Thus, our in vitro and in vivo data indicate that the antitumor effects of celecoxib probably are mediated through COX‐2 independent mechanisms and are not restricted to COX‐2 over‐expressing tumors.

Chain length-specific properties of ceramides.

Ceramides are a class of sphingolipids that are abundant in cell membranes. They are important structural components of the membrane but can also act as second messengers in various signaling pathways. Until recently, ceramides and dihydroceramides were considered as a single functional class of lipids and no distinction was made between molecules with different chain lengths. However, based on the development of high-throughput, structure-specific and quantitative analytical methods to measure ceramides, it has now become clear that in cellular systems the amounts of ceramides differ with respect to their chain length. Further studies have indicated that some functions of ceramides are chain-length dependent. In this review, we discuss the chain length-specific differences of ceramides including their pathological impact on Alzheimers disease, inflammation, autophagy, apoptosis and cancer.

The polymorphism A118G of the human mu-opioid receptor gene decreases the pupil constrictory effect of morphine-6- glucuronide but not that of morphine

Large individual differences in the clinical response to morphine therapy have been known for a long time by clinicians. The recent advances in genomic research encourage the search for pharmacogenetic causes of that variability. As a measure of central opioid effects, pupil diameters were assessed every 20 min for 18 h after administration of morphine or its active metabolite morphine-6-glucuronide (M6G) in a two-way crossover study. The opioid effects were compared between six subjects with a single-nucleotide polymorphism (SNP) A118G in the mu-opioid receptor gene (five heterozygous, one homozygous) and six control subjects. Non-parametric pharmacokinetic-pharmacodynamic modelling was employed to identify the influence of the A118G SNP on the concentration-response relationship of M6G and morphine, which was described by a sigmoid Emax model. As a measure of potency, the EC50 of the pupil constrictory effects of M6G was 714 +/- 197 nmol/l in wild-type and 1475 +/- 424 nmol/l in heterozygous carriers of the A118G SNP. In the homozygous carrier of the SNP, it had an EC50 of 3140 nmol/l. In addition, the dose-response relationship was flatter in the A118G carriers than in control subjects (shape factor of the sigmoid Emax model: gamma = 3.3 +/- 1.2, 1.7 +/- 0.5 and 1.6 for wild-type, heterozygous and the homozygous A118G carriers, respectively). In contrast, the concentration-response relationship of morphine was not affected by this specific SNP. The A118G SNP in the mu-receptor gene significantly reduces the potency of M6G in humans.

Celecoxib loses its anti-inflammatory efficacy at high doses through activation of NF-κB

Celecoxib, a selective cyclooxygenase‐2 (COX‐2) inhibitor, has recently been approved for the symptomatic treatment of arthritis. In some clinical studies, doses of 400 and 800 mg/day provided somewhat less efficacy compared with 200 mg/day, which suggests an early ceiling effect. Using the zymosan‐induced inflammation model in rats, we show that celecoxib significantly reduces paw swelling at 50 mg/kg but completely loses its anti‐inflammatory efficacy at doses ≥100 mg/kg. To evaluate the underlying mechanisms, we used rat renal mesangial cells as a cell culture model. In these cells, celecoxib (50 μM) increased the interleukin Iβ stimulated nuclear translocation and DNA binding of NF‐κB and facilitated the degradation of I‐κB. Consequently, COX‐2 and tumor necrosis factor α (TNF‐α) expression were increased. The up‐regulation of COX‐2 and TNF‐α also occurred in the spinal cord of rats treated with celecoxib (≥100 mg/kg), indicating that in vitro mechanisms were relevant in vivo. Clinically, the overexpression of COX‐2 might be less important because celecoxib inhibits COX‐2 enzymatically. However, the up‐regulation of TNF‐α and possibly other NF‐κB regulated proinflammatory genes might worsen the pathophysiological processes underlying chronic arthritis.

Targeting the beta-catenin/APC pathway: a novel mechanism to explain the cyclooxygenase-2-independent anticarcinogenic effects of celecoxib in human colon carcinoma cells

Celecoxib, a cyclooxygenase‐2 (COX‐2) selective nonsteroidal anti‐inflammatory drug, is a new anticarcinogenic agent. Its antitumor effects depend on the one hand on its COX‐2‐inhibiting potency, but on the other hand on COX‐2‐independent mechanisms, which until now have not been fully understood. Here, we investigated whether celecoxib has an impact on the APC/β‐catenin pathway, which has been shown to play a pivotal role in the development of various cancers, especially of the colon. After only 2 h of treatment of human Caco‐2 colon carcinoma cells with 100 µM celecoxib, we observed a rapid translocation of β‐catenin from its predominant membrane localization to the cytoplasm. Inhibition of the glycogen‐synthasekinase‐3β (GSK‐3β) by LiCl prevented this celecoxib‐induced translocation, suggesting that phosphorylation of β‐catenin by the GSK‐3β kinase was essential for this release. Furthermore, the cytosolic accumulation was accompanied by a rapid increase of β‐catenin in the nuclei, starting already 30 min after celecoxib treatment. The DNA binding activity of β‐catenin time dependently decreased 2 h after celecoxib treatment. After this cellular reorganization, we observed a caspase‐ and proteasome‐dependent degradation of β‐catenin after 8 h of drug incubation. Celecoxib‐induced β‐catenin degradation was also observed in various other tumor cell lines (HCT‐116, MCF‐7, and LNCAP) but was not seen after treatment of Caco‐2 cells with either the anticarcinogenic nonsteroidal anti‐inflammatory drug R‐flurbiprofen or the highly COX‐2‐selective inhibitor rofecoxib. These findings indicate that the anticarcinogenic effects of celecoxib can be explained, at least partly, by an extensive degradation of β‐catenin in human colon carcinoma cells.

Inhibition of NF-κB and AP-1 activation by R- and S-flurbiprofen

R-flurbiprofen is considered the ‘inactive’ isomer of the nonsteroidal anti-inflammatory drug (NSAID), flurbiprofen, because it does not inhibit cyclooxygenase (COX) activity. However, previous studies have revealed that it has antinociceptive and antitumor effects not due to epimerization to the cyclooxygenase-inhibiting S-isomer. Here, we show that R-flurbiprofen has additional anti-inflammatory activity comparable with that of dexamethasone in the zymosaninduced paw inflammation model in rats. Different criteria suggest that the observed effects are !

Ceramide synthases and ceramide levels are increased in breast cancer tissue

Several in vitro studies have correlated dysfunction of the sphingolipid-signaling pathway with promotion of tumor cell growth as well as progression and resistance of tumors to chemotherapeutic agents. As ceramides (Cer) constitute the structural backbones of all sphingolipids, we investigated the endogenous ceramide levels in 43 malignant breast tumors and 21 benign breast biopsies and compared them with those of normal tissues using liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS). The total ceramide levels in malignant tumor tissue samples were statistically significantly elevated when compared with normal tissue samples. Upregulation of the total ceramide level averaged 12-fold and 4-fold higher than normal tissue samples, for malignant tumors and benign tissues, respectively. Specifically, the levels of C(16:0)-Cer, C(24:1)-Cer and C(24:0)-Cer were significantly raised in malignant tumors as compared with benign and normal tissue. The augmentation of the various ceramides could be assigned to an increase of the messenger RNA levels of ceramide synthases (CerS) LASS2 (longevity assurance), LASS4 and LASS6. Notably, elevated levels of C(16:0)-Cer were associated with a positive lymph node status, indicating a metastatic potential for this ceramide. Moreover, the levels of C(18:0)-Cer and C(20:0)-Cer were significantly higher in estrogen receptor (ER) positive tumor tissues as compared with ER negative tumor tissues. In conclusion, progression in breast cancer is associated with increased ceramide levels due to an upregulation of specific LASS genes.

Long chain ceramides and very long chain ceramides have opposite effects on human breast and colon cancer cell growth.

Ceramides are known to be key players in intracellular signaling and are involved in apoptosis, cell senescence, proliferation, cell growth and differentiation. They are synthesized by ceramide synthases (CerS). So far, six different mammalian CerS (CerS1-6) have been described. Recently, we demonstrated that human breast cancer tissue displays increased activity of CerS2, 4, and 6, together with enhanced generation of their products, ceramides C(16:0), C(24:0), and C(24:1). Moreover, these increases were significantly associated with tumor dignity. To clarify the impact of this observation, we manipulated cellular ceramide levels by overexpressing ceramide synthases 2, 4 or 6 in MCF-7 (breast cancer) and HCT-116 (colon cancer) cells, respectively. Overexpression of ceramide synthases 4 and 6 elevated generation of short chain ceramides C(16:0), C(18:0) and C(20:0), while overexpression of ceramide synthase 2 had no effect on ceramide production in vivo, presumably due to limited substrate availability, because external addition of very long chain acyl-CoAs resulted in a significant upregulation of very long chain ceramides. We also demonstrated that upregulation of CerS4 and 6 led to the inhibition of cell proliferation and induction of apoptosis, whereas upregulation of CerS2 increased cell proliferation. On the basis of our data, we propose that a disequilibrium between ceramides of various chain length is crucial for cancer progression, while normal cells require an equilibrium between very long and long chain ceramides for normal physiology.

The transfer half-life of morphine-6-glucuronide from plasma to effect site assessed by pupil size measurement in healthy volunteers.

Background Clinical and experimental data suggested a long delay between the plasma concentration versus time course of morphine-6-glucuronide and the time course of its central opioid effects. This study was aimed at the quantification of the transfer half-life (t1/2,ke0) of this delay. Methods Pupil size was used as a measure of central opioid effect. Eight healthy volunteers (four men, four women) participated in that single-blind randomized crossover study. Median dosages administered intravenously were 0.5 mg morphine as loading dose followed by 10.7 mg given as infusion over a period of 4.7 h, and 10.2 mg M6G as loading dose followed by 39.1 mg M6G given over a period of 3.7 h. The duration of the infusion was tailored to achieve submaximum pupil constriction. The pupil diameter was assessed every 20 min for approximately 18 h. Values of t1/2,ke0 were obtained by semiparametric pharmacokinetic-pharmacodynamic modeling. Results The estimated median t1/2,ke0 of M6G was 6.4 h (range, 2.9–16.2 h), and that of morphine was 2.8 h (range, 1.8–4.4 h). The individual t1/2,ke0 of M6G was always longer than that of morphine. Judged by the concentration at half-maximun effect (EC50) values of the sigmoid pupil size at maximum constriction (Emax) model describing concentration–response relation, M6G was apparently 22 times less potent than morphine (EC50 = 740.5 nm [range, 500–1,520 nm] for M6G and 36.2 nm [range, 19.7–43.3 nm] for morphine). The steepness of the sigmoid Emax model did not significantly differ between morphine and M6G (&ggr; = 1.9 and 2.6, respectively). To produce similar pupil effects, the M6G dose had to be 2.8 times greater than the morphine dose. Conclusions The reported numerical value of the t1/2,ke0 of M6G in humans obtained after direct administration of M6G is a step toward a complete modeling approach to the prediction of the clinical effects of morphine. The study raises questions about the high interindividual variability of the transfer half-life between plasma and effect site (ke0) values and the apparent low potency of M6G.

Sphingosine kinase 2 deficient tumor xenografts show impaired growth and fail to polarize macrophages towards an anti-inflammatory phenotype

A challenging task of the immune system is to fight cancer cells. However, a variety of human cancers educate immune cells to become tumor supportive. This is exemplified for tumor‐associated macrophages (TAMs), which are polarized towards an anti‐inflammatory and cancer promoting phenotype. Mechanistic explanations, how cancer cells influence the macrophage phenotype are urgently needed to address potential anti‐cancer strategies along this line. One potential immune modulating compound, sphingosine‐1‐phosphate (S1P), was recently highlighted in both tumor growth and immune modulation. Using a xenograft model in nude mice, we demonstrate a supportive role of sphingosine kinase 2 (SphK2), one of the S1P‐producing enzymes for tumor progression. The growth of SphK2‐deficient MCF‐7 breast tumor xenografts was markedly delayed when compared with controls. Infiltration of macrophages in SphK2‐deficient and control tumors was comparable. However, TAMs from SphK2‐deficient tumors displayed a pronounced anti‐tumor phenotype, showing an increased expression of pro‐inflammatory markers/mediators such as NO, TNF‐α, IL‐12 and MHCII and a low expression of anti‐inflammatory IL‐10 and CD206. These data suggest a role for S1P, generated by SphK2, in early tumor development by affecting macrophage polarization.