Inducing apoptosis of cancer cells using small-molecule plant compounds that bind to GRP78.

BACKGROUND: Glucose regulated protein 78 (GRP78) functions as a sensor of endoplasmic reticulum (ER) stress. The aim of this study was to test the hypothesis that molecules that bind to GRP78 induce the unfolded protein response (UPR) and enhance cell death in combination with ER stress inducers. METHODS: Differential scanning calorimetry (DSC), measurement of cell death by flow cytometry and the induction of ER stress markers using western blotting. RESULTS: Epigallocatechin gallate (EGCG), a flavonoid component of Green Tea Camellia sinensis, and honokiol (HNK), a Magnolia grandiflora derivative, bind to unfolded conformations of the GRP78 ATPase domain. Epigallocatechin gallate and HNK induced death in six neuroectodermal tumour cell lines tested. Levels of death to HNK were twice that for EGCG; half-maximal effective doses were similar but EGCG sensitivity varied more widely between cell types. Honokiol induced ER stress and UPR as predicted from its ability to interact with GRP78, but EGCG was less effective. With respect to cell death, HNK had synergistic effects on melanoma and glioblastoma cells with the ER stress inducers fenretinide or bortezomib, but only additive (fenretinide) or inhibitory (bortezomib) effects on neuroblastoma cells. CONCLUSION: Honokiol induces apoptosis due to ER stress from an interaction with GRP78. The data are consistent with DSC results that suggest that HNK binds to GRP78 more effectively than EGCG. Therefore, HNK may warrant development as an antitumour drug.

Apoptosis of N-type neuroblastoma cells after differentiation with 9-cis-retinoic acid and subsequent washout.

BACKGROUND: The overall survival rate for patients with neuroblastoma has improved over the past two decades, but long-term survival for the subgroup of patients with high-risk disease remains low. In recent years, there has been interest in the potential clinical use of drugs able to induce differentiation of neuroblastoma cells. Since 9-cis-retinoic acid induces better and more sustained differentiation of neuroblastoma in vitro than other retinoic acid isomers, this may be a more appropriate retinoid for use in neuroblastoma therapy. PURPOSE: The purpose of this work was to compare the long-term effects of all-trans- and 9-cis-retinoic acid on neuroblastoma differentiation using an N-type (neuroblastic) cell line, SH SY 5Y, as an in vitro model. In addition, we wanted to find out whether 9-cis-retinoic acid would induce programmed cell death (apoptosis) in these N-type neuroblastoma cells and to determine whether the effects of either 9-cis- or all-trans-retinoic acid are dependent on their continued presence in the culture medium. METHODS: SH SY 5Y cells were incubated in either the continued presence of all-trans- or 9-cis-retinoic acid or for 5 days with retinoic acid followed by culture in the absence of retinoid for up to 13 days. Morphologic changes were observed using phase-contrast and scanning electron microscopy. Apoptosis was determined by flow cytometry of propidium iodide-stained cells and by using terminal deoxynucleotidyl transferase to end-label DNA fragments in situ in apoptotic cells. RESULTS: Culture of SH SY 5Y cells with all-trans- or 9-cis retinoic acid for 5 days induced morphologic differentiation and inhibited cell growth. These effects were maintained in the continuous presence of each retinoic acid isomer but were more profound in cells treated with 9-cis-retinoic acid. The differentiation of cells treated with all-trans-retinoic acid was reversible once retinoic acid was removed from the medium. Conversely, apoptosis was induced in cells treated with 9-cis-retinoic acid for 5 days and cultured for 9 days (4 days after washout) but not in cells cultured in the continuous presence of 9-cis-retinoic acid. This effect was specific to 9-cis-retinoic acid. CONCLUSIONS: Previous studies have demonstrated differential responses to all-trans-retinoic acid in N- and S-type (substrate-adherent or Schwann-like) neuroblastoma cells: Apoptosis is induced in S-type cells, whereas differentiation occurs in N-type cells. The present results show that, unlike all-trans-retinoic acid, 9-cis-retinoic acid induces both differentiation and apoptosis in N-type SH SY 5Y neuroblastoma cells. However, apoptosis was dependent on removal of 9-cis-retinoic acid from the culture medium. IMPLICATIONS: Since both differentiation and apoptosis are involved in tumor regression, 9-cis-retinoic acid may be a more appropriate retinoid for clinical trials in neuroblastoma. The dependence of apoptosis on treatment and subsequent removal of 9-cis-retinoic acid implies that drug scheduling may be an important parameter affecting therapeutic efficacy.

Purification and properties of cellular retinoic acid-binding protein from neonatal rat skin.

Cellular retinoic acid-binding protein (CRABP) was detected in cytosolic extracts of dermis and epidermis of neonatal rat skin using high-performance size-exclusion liquid chromatography and was more abundant in dermal tissue. CRABP was purified 1000-fold from an acid-precipitated, 50,000 x g supernatant of neonatal rat skin by ion-exchange chromatography on DEAE-Sephacel, followed by chromatofocussing and hydrophobic-interaction chromatography. The protein had an apparent Mr of 14,800. In chromatofocussing experiments the apoprotein and holoprotein gave different elution profiles, indicating a charge difference between the two forms. The ability of various retinoids to compete with all-trans-retinoic acid for binding to CRABP was assayed: 4-oxoretinoic acid and two synthetic retinoids were effective competitors, but 13-cis-retinoic acid, 3,4-didehydroretinoic acid and the acid derivative of etretinate competed poorly. The binding protein had a Kd for all-trans-retinoic acid of 8 nM using a dextran-charcoal assay, but a higher value was obtained using high-performance size-exclusion liquid chromatography. The holoprotein dissociated rapidly at room temperature and had a half-life of 4.7 min. At 0 degrees C, the holoprotein had a half-life of 200 min.

The role of AP-1 in glucocorticoid resistance in leukaemia.

Glucocorticoids are used in the treatment of acute lymphoblastic leukaemia (ALL) and chronic lymphocytic leukaemia (CLL) but many patients develop glucocorticoid resistance on relapse. The ligand-activated glucocorticoid receptor inhibits activity of the AP-1 transcription factor and the purpose of this study was to test the hypothesis that up-regulation or overexpression of AP1-binding activity may be an important mechanism of glucocorticoid resistance in ALL and CLL. In vitro sensitivity of patient blasts to prednisolone was measured using th

Expression of the alpha and beta retinoic acid receptors in skin.

Retinoic acid receptor (RAR) -alpha and -beta transcripts are expressed in rat and human skin, and in rat and human dermal fibroblasts and keratinocytes in vitro. RAR-alpha transcripts (ca. 2.8 and 3.6 kb) were expressed in all tissues but were more abundant in dermis and dermal fibroblasts than in epidermis or keratinocytes. RAR-beta mRNA was expressed in skin, but patterns of expression differed between human and neonatal rat samples. In human dermal fibroblasts, keratinocytes and whole skin, two RAR-beta transcripts (ca. 3.1 and 3.4 kb) were expressed. Conversely, in neonatal rat skin, dermal fibroblasts, and keratinocytes only the smaller transcript was detectable and was more abundant in cultured cells than in whole tissue. These results suggest that retinoic acid may have complex, as yet undefined, RAR-mediated regulatory functions in both dermis and epidermis.

The N363S polymorphism of the glucocorticoid receptor: potential contribution to central obesity in men and lack of association with other risk factors for coronary heart disease and diabetes mellitus.

Considerable evidence suggests that diabetes mellitus and hypertension are influenced by genetic factors. Studies in humans have associated glucocorticoid receptor (GR) polymorphisms with high blood pressure, insulin sensitivity, body mass index, increased visceral fat, and variations in tissue-specific steroid sensitivity. The N363S polymorphism of the GR results in an asparagine to serine amino acid substitution in a modulatory region of the receptor. Phosphorylation of serine residues in this region has been shown to enhance transactivation of GR responsive genes. The aim of this study was to investigate the association between the 363S allele and risk factors for coronary heart disease and diabetes mellitus in a population of European origin living in the northeast of the United KINGDOM: Blood samples from 135 males and 240 females were characterized for 363 allele status. The overall frequency of the 363S allele was 3.0%, 23 heterozygotes (7 males and 16 females) but no 363S homozygotes were identified. The data show a significant association of the 363S allele with increased waist to hip ratio in males but not females. This allele was not associated with blood pressure, body mass index, serum cholesterol, triglycerides, low-density lipoprotein and high-density lipoprotein cholesterol levels, and glucose tolerance status. The results of this study suggest that this GR polymorphism may contribute to central obesity in men. Further studies are required to elucidate the properties of GR(363S) at a molecular level.

Ligand binding properties of human cellular retinoic acid binding protein II expressed in E. coli as a glutathione-S-transferase fusion protein.

To test the hypothesis that 9-cis-retinoic acid is a ligand for cellular retinoic acid binding protein II (CRABP II), human CRABP II was expressed as a glutathione-S-transferase fusion protein (GST-CRABP II) and a single affinity purification step used to extract it from bacterial lysates. GST-CRABP II bound all trans-retinoic acid with high affinity (Kd 14.2 +/- 6.5 nM), but 9-cis-retinoic acid bound poorly. These studies suggest that 9-cis-retinoic acid is not a ligand for CRABP II. Their ease of purification makes GST-CRABP fusion proteins useful tools for ligand binding studies with different retinoids.

RXR beta isoforms in neuroblastoma cells and evidence for a novel 3'-end transcript.

RXR beta is predominantly involved in retinoid responses in neuroblastoma cells, in particular the N-type SH SY 5Y cells and the S-type SH S EP cells, both derivatives of a mixed phenotype neuroblastoma cell line. The aim of this study was to identify RXR beta isoforms expressed in neuroblastoma cells and to characterise a putative novel RXR beta transcript. RXR beta 1 and RXR beta 2 were expressed in these neuroblastoma cells. An isoform with an insertion into the ligand binding domain, RXR beta(SLSR) (referred to in previous studies as RXR beta 3), was expressed at a similar level to RXR beta. A novel RXR beta transcript was identified by RNase protection assays and was at least as abundant as the expected RXR beta transcript and expressed in other cell types. Evidence suggests that this novel transcript was transcribed from an internal promoter between exons 5 and 6, contained a retained intron (intron 6) and was alternatively spliced with and without the SLSR insertion. These data show that the pattern of RXR beta expression is complex. The relative abundance of the novel RXR beta transcript suggests that it may be an important aspect of RXR beta function or regulation in a range of cell types.

The biological activity of retinoids in melanoma cells. Induction of expression of retinoic acid receptor-beta by retinoic acid in S91 melanoma cells.

The expression of mRNA for retinoic acid receptor beta (RAR-beta) was induced by all trans-retinoic acid in murine S91 melanoma cells. The induction of RAR-beta was dose-dependent, rapid and insensitive to cycloheximide. Both 13-cis-retinoic acid and 3,4-didehydro-all trans-retinoic acid also induced expression of RAR-beta but were only effective at concentrations 100-fold greater than all trans-retinoic acid. The expression of RAR-alpha and RAR-gamma was unaffected by retinoic acid.

Differential effects of retinoic acid isomers on the expression of nuclear receptor co-regulators in neuroblastoma.

Retinoic acid modulates growth and induces differentiation and apoptosis of neuroblastoma cells in vitro, with the all-trans and 9-cis isomers having different biological properties. Transcriptional activation in response to retinoic acid isomers is mediated by retinoic acid receptors and retinoid X receptors. The differential expression of co-activators and co-repressors which preferentially interact with retinoic acid receptors or retinoid X receptors may be a mechanism leading to different cellular responses to 9-cis and all-trans retinoic acid. To test this hypothesis, we have studied the expression of the nuclear receptor co-regulators TIF1alpha, TIF1beta, SUG1 and SMRT in the N-type and S-type neuroblastoma cell lines SH SY 5Y and SH S EP. Transcripts for all four co-regulators were expressed in these neuroblastoma cells. The expression of TIF1alpha, TIF1beta and SUG1 did not change in response to retinoic acid; however, SMRT was induced in both neuroblastoma cell lines, but particularly by all-trans retinoic acid in SH S EP cells. An additional co-activator, Trip3, was isolated by differential mRNA display and shown to be preferentially induced by 9-cis retinoic acid in SH SY 5Y and SH S EP cells. These data suggest that retinoic acid isomer-specific induction of nuclear receptor co-regulators may determine, in part, the differential biological effects of retinoic acid isomers.

Gene expression and neuroblastoma cell differentiation in response to retinoic acid: differential effects of 9-cis and all-trans retinoic acid.

Retinoic acid has considerable potential for the chemoprevention and chemotherapy of cancer. Neuroblastoma cells differentiate in response to retinoic acid in vitro, an observation that has led to clinical trials using either the 13-cis or all-trans isomers of retinoic acid. We review the effects of retinoic acid on neuroblastoma, and the potential involvement of nuclear retinoic acid receptors (RARs) and retinoid X receptors (RXRs). 9-cis retinoic acid is a ligand for RXRs, and we review recent data on the differential effects of 9-cis and all-trans retinoic acid on neuroblastoma differentiation and proliferation in vitro, and possible mechanisms of action via hetero- and homodimers of RARs and RXRs. Although there is uncertainty whether or not 9-cis retinoic acid produces its biological effects primarily via RXR homodimers, in vitro data suggest that this isomer of retinoic acid or stable analogues may have considerable potential for the treatment of resistant, disseminated neuroblastoma.

Retinoid-induced differentiation of neuroblastoma: comparison between LG69, an RXR-selective analogue and 9-cis retinoic acid.

The aim of this study was to investigate in vitro the effects of all-trans retinoic acid (RA), 9-cis RA and the RXR-selective analogue, LG69, on the morphological differentiation, proliferation and gene expression of neuroblastoma cells. Three different cell lines were cultured with the retinoid for either 9 continuous days or for 5 days followed by 4 days without the retinoid and morphological differentiation was assessed both qualitatively and quantitatively. SH SY 5Y cell proliferation was examined by measuring cell numbers after exposure to the retinoids and RAR-beta gene expression was examined by Northern blot analysis. Morphological differentiation was more effectively induced by all-trans and 9-cis RA than by LG69. SH SY 5Y cells, when treated with 9-cis RA for only 5 of the 9 days of culture, underwent apoptosis, but this was not seen with 9 days continuous exposure nor with LG69. Inhibition of SH SY 5Y cell proliferation by all-trans or 9-cis RA was dose-dependent, but LG69 had little effect. Conversely, LG69 induced higher expression of RAR-beta than all-trans RA, but less than that produced by 9-cis RA. These data suggest that 9-cis RA as a single agent is the most effective modulator of neuroblastoma behaviour and may be the most appropriate therapeutic agent.

Retinoids in neuroblastoma therapy: distinct biological properties of 9-cis- and all-trans-retinoic acid.

We investigated the potential for 9-cis-retinoic acid in the differentiation therapy of neuroblastoma using an N-type neuroblastoma cell line, SH SY 5Y, as an experimental model. In these cells, 9-cis-retinoic acid is more effective than other isomers at inducing the expression of RAR-beta. An RAR-alpha-specific antagonist inhibited the induction of RAR-beta in response to all-trans-but not to 9-cis-retinoic acid. This indicates that the mechanism of gene induction by 9-cis-retinoic acid differs markedly from all-trans-retinoic acid. 9-cis-retinoic acid is also better than all-trans at producing sustained morphological differentiation and inhibition of proliferation of SH SY 5Y cells. Although N-type neuroblastoma cells are not thought to undergo apoptosis in response to all-trans-retinoic acid, we observed a significant degree of apoptosis in SH SY 5Y cells treated with 9-cis-retinoic acid for 5 days and then cultured in the absence of retinoid, an effect not observed in cells treated with the all-trans isomer. These results suggest that 9-cis- and all-trans-retinoic acid have distinct biological properties and that 9-cis retinoic acid may be clinically effective in neuroblastoma by inducing both differentiation and apoptosis under an appropriate treatment regimen.

Immunohistochemical markers of human sebaceous gland differentiation.

Cryostat sections of human skin were stained with monoclonal antibodies to involucrin, a range of cytokeratins, epithelial membrane antigen (EMA), and an ovarian cystadenocarcinoma antibody (OM1) to identify combinations of antibodies that could be used to discriminate between basal and differentiated sebocytes and other cell types present in the pilosebaceous unit. Both the EMA and OM1 monoclonal antibodies specifically recognized differentiated sebocytes. No staining of basal sebocytes or other epidermal cell types was seen. Differentiated (but not basal) sebocytes were also stained by a cytokeratin 10 antibody (LH2). Conversely, the basal sebocytes were recognized by an antibody specific to basal keratinocytes (LH6). Cells of the sebaceous duct stained with both LH2 and LH6 and also with the anti-involucrin monoclonal antibody. Cytokeratin 4 has been detected in sebaceous glands by protein analysis but has not previously been detectable immunohistochemically. We show by immunofluorescence after limited proteolysis that cytokeratin 4 epitopes are distributed in all sebaceous gland cells, including the duct cells.

Characterization of human purified epithelial and stromal cells from endometrium and endometriosis in tissue culture.

OBJECTIVE: To initiate in vitro cultures of separate stromal and epithelial elements from endometriotic tissue and to compare the characteristics of these cells with those of cultured endometrial cells. DESIGN: The study involved testing the viability of a culture system for endometriotic tissue and examination of the phenotype of the cells. SETTING: Fresh tissue samples were collected from the operating theater and transferred to the tissue culture laboratory. PATIENTS, PARTICIPANTS: Twenty patients undergoing laparotomy for endometriosis and patients undergoing surgery for benign conditions were recruited. INTERVENTIONS: Endometrium and endometriotic tissue were separated, cultured in vitro, and labeled by indirect immunofluorescence with monoclonal antibodies against cytoskeletal components and epithelial mucins. MAIN OUTCOME MEASURES: Endometriotic cells have been maintained in vitro and found to resemble endometrial cells closely. RESULTS: With respect to the staining patterns for cytokeratins 18 and 19, vimentin, and three different epithelial mucins, cultured cells from both endometrium and endometriotic tissue had similar properties. Cytokeratins were located in epithelial cells, and vimentin was expressed in both stromal and epithelial cells. The antimucin antibodies all gave distinct patterns of intracellular staining of epithelial cells. CONCLUSIONS: Our results indicate a close similarity between cultured stromal and epithelial cells from endometrium and endometriotic deposits. Culture of these cell populations will permit study of their properties and interactions and may provide some insight into the cause of endometriosis.

Retinoic acid receptor expression during the in vitro differentiation of human neuroblastoma.

Neuroblastoma cells differentiate in response to retinoic acid and cyclic AMP. We have examined the expression of retinoic acid receptors (RARs) in relation to neuroblastoma differentiation and show that short term exposure of SK N SH, SH SY 5Y AND GI LI N cells to physiological concentrations of retinoic acid results in induction of RAR-beta, particularly the lower transcript. Cyclic AMP has no effect on retinoic acid mediated induction and does not alter RAR expression patterns. These data are discussed in the light of evidence that retinoic acid and cyclic AMP act either on different control pathways or at different points within a common pathway.

Receptor mechanisms mediating differentiation and proliferation effects of retinoids on neuroblastoma cells.

The aim of this study was to clarify retinoid receptor mechanisms mediating the effects of 9-cis retinoic acid (RA) and investigate the ability of RAR- and RXR-specific analogues to induce differentiation and inhibit proliferation in neuroblastoma cells. Differentiation and the inhibition of proliferation by 9-cis RA, but not all-trans RA, were inhibited by the RXR-homodimer antagonist LG745. The RXR-specific agonist LGD1069 was ineffective at inducing differentiation or inhibiting proliferation, but showed marked synergism with RAR-specific agonists with respect to inhibiting proliferation. These data suggest that the effects of 9-cis RA are mediated via both RXR-homodimers and heterodimers. However, combinations of RAR- and RXR-selective analogues were not as effective at promoting differentiation. This study indicates that different receptor mechanisms are involved in retinoid-induced differentiation and inhibition of proliferation in neuroblastoma cells.

Concentration-dependent effects of 9-cis retinoic acid on neuroblastoma differentiation and proliferation in vitro.

The observation that neuroblastoma cells differentiate in response to retinoic acid (RA) in vitro has led to clinical trials using either 13-cis or all-trans RA. Since 9-cis RA may also have important biological functions, we have compared the potential of RA isomers to induce differentiation and inhibit cell proliferation of SH SY 5Y neuroblastoma cells. 9-cis RA at high concentrations is better at inducing morphological differentiation than either all-trans or 13-cis RA and as effective at inhibiting proliferation. Hence, 9-cis RA or stable analogues may have important therapeutic potential in the treatment of neuroblastoma.

Changes in oncogene mRNA expression during human keratinocyte differentiation.

The nuclear proto-oncogenes are involved in transcriptional regulation and control many cell processes. The role of changes in proto-oncogene expression in controlling the balance between proliferation and differentiation was studied in cultured keratinocytes. Normal human keratinocytes were grown in the serum-free medium MCDB153 with an extracellular calcium concentration of 70 microM. After treatment with different differentiation conditions, cellular RNA was size-fractionated on agarose gels and transferred to nylon membranes which were subsequently hybridized with c-myc, c-jun, and H-ras 32P-labelled probes. Relative RNA loading was assessed using probes for beta-actin and ribosomal 18s RNA. Inducing differentiation by increasing the calcium concentration of the medium from 70 microM to 1.5 mM resulted in a marked decrease in c-myc RNA levels to 26% of control levels within 8 h. After 48 h in 1.5 mM calcium, c-myc levels had recovered to approximately 50% of control levels. There was a gradual reduction in c-jun levels to 56% of control levels by 4 days. Treatment with 10 nM TPA, which also induces keratinocyte differentiation, reduced c-myc RNA levels to 70% of control levels during the first 4 h, but thereafter c-myc levels remained approximately constant for a further 20 h. TGF beta (2 ng/ml), which inhibits keratinocyte growth without inducing differentiation, did not alter c-myc RNA levels over a 4-day period. There were no changes in c-myc levels following the addition of retinoic acid and none of the conditions altered H-ras levels.(ABSTRACT TRUNCATED AT 250 WORDS)

Glucocorticoid resistance and the AP-1 transcription factor in leukaemia.

Glucocorticoids have been used in the treatment of acute lymphoblastic leukaemia (ALL) for many years, initially as the only agent and then as part of multiagent chemotherapy. In ALL 20% of patients are resistant to glucocorticoids at presentation but this rises to greater than 70% on relapse. It has recently been reported that the glucocorticoid receptor inhibits activity of the AP-1 transcription factor by the ligand-dependant binding of glucocorticoid receptor (GR) to the fos and jun components of AP-1. Since AP-1 is necessary for cell proliferation, the upregulation or over-expression of AP-1 may be a mechanism of resistance to glucocorticoids. The aim of the study was to investigate whether AP-1 levels correlate with in vitro glucocorticoid resistance. In vitro sensitivity to glucocorticoids was measured using the MTT assay. AP-1 levels were quantified using gel shift analysis: a consensus sequence for the AP-1 binding site was synthesised, labelled with 32P and incubated with nuclear extracts of leukaemic blasts from 14 ALL and 26 CLL patients. Leukaemic blasts were treated with prednisolone or with vehicle alone before preparation of nuclear extracts. The gels were dried and bands quantified using a phosphorimager, using an appropriate internal standard and correcting for protein loading and cytoplasmic contamination of nuclear extracts. The patient samples fell into two distinct groups with respect to their sensitivity to glucocorticoids: AP-1 levels were significantly higher (p < 0.02) in sensitive blasts than resistant ones. There was no significant change in AP-1 levels after treating blasts for 4 hours with 0.2 mM prednisolone. No change was seen in CLL samples. These data show that glucocorticoid resistance is not associated with increased AP-1. Conversely, glucocorticoid resistance in these samples was apparently associated with decreased AP-1 levels in ALL samples. Whether this has any causal relationship to glucocorticoid resistance is unknown. Clearly, further studies on the role of AP-1 and related transcription factors is essential for understanding the control of proliferation and apoptosis in ALL.