Search

S49 mouse lymphoma cells undergo apoptosis in response to the ALP (alkyl-lysophospholipid) edelfosine (1-O-octadecyl-2-O-methyl-rac-glycero-3-phosphocholine), FasL (Fas ligand) and DNA damage. S49 cells made resistant to ALP (S49(AR)) are defective in sphingomyelin synthesis and ALP uptake, and also have acquired resistance to FasL and DNA damage. However, these cells can be re-sensitized following prolonged culturing in the absence of ALP. The resistant cells show sustained ERK (extracellular-signal-regulated kinase)/Akt activity, consistent with enhanced survival signalling. In search of a common mediator of the observed cross-resistance, we found that S49(AR) cells lacked the PtdIns(3,4,5)P(3) phosphatase SHIP-1 [SH2 (Src homology 2)-domain-containing inositol phosphatase 1], a known regulator of the Akt survival pathway. Re-sensitization of the S49(AR) cells restored SHIP-1 expression as well as phosphoinositide and sphingomyelin levels. Knockdown of SHIP-1 mimicked the S49(AR) phenotype in terms of apoptosis cross-resistance, sphingomyelin deficiency and altered phosphoinositide levels. Collectively, the results of the present study suggest that SHIP-1 collaborates with sphingomyelin synthase to regulate lymphoma cell death irrespective of the nature of the apoptotic stimulus.

The ALP (alkyl-lysophospholipid) edelfosine (1-O-octadecyl-2-O-methyl-rac-glycero-3-phosphocholine; Et-18-OCH3) induces apoptosis in S49 mouse lymphoma cells. To this end, ALP is internalized by lipid raft-dependent endocytosis and inhibits phosphatidylcholine synthesis. A variant cell-line, S49AR, which is resistant to ALP, was shown previously to be unable to internalize ALP via this lipid raft pathway. The reason for this uptake failure is not understood. In the present study, we show that S49AR cells are unable to synthesize SM (sphingomyelin) due to down-regulated SMS1 (SM synthase 1) expression. In parental S49 cells, resistance to ALP could be mimicked by small interfering RNA-induced SMS1 suppression, resulting in SM deficiency and blockage of raft-dependent internalization of ALP and induction of apoptosis. Similar results were obtained by treatment of the cells with myriocin/ISP-1, an inhibitor of general sphingolipid synthesis, or with U18666A, a cholesterol homoeostasis perturbing agent. U18666A is known to inhibit Niemann-Pick C1 protein-dependent vesicular transport of cholesterol from endosomal compartments to the trans-Golgi network and the plasma membrane. U18666A reduced cholesterol partitioning in detergent-resistant lipid rafts and inhibited SM synthesis in S49 cells, causing ALP resistance similar to that observed in S49AR cells. The results are explained by the strong physical interaction between (newly synthesized) SM and available cholesterol at the Golgi, where they facilitate lipid raft formation. We propose that ALP internalization by lipid-raft-dependent endocytosis represents the retrograde route of a constitutive SMS1- and lipid-raft-dependent membrane vesicular recycling process.

The anticancer agent doxorubicin is in certain cases administered as a long-circulating liposomal formulation. Due to angiogenesis-related structural abnormalities in the endothelial lining of many neoplasms, these complexes tend to extravasate and accumulate in the tumor stroma. However, delivery of doxorubicin is still not optimal since liposomes are not taken up directly by tumor cells. Instead, doxorubicin is gradually released into the interstitial space, and the subsequent uptake by surrounding cells is a limiting step in the delivery process. We recently demonstrated that plasma membrane-inserted short-chain sphingomyelin facilitates the cellular uptake of free doxorubicin. Here, we report that N-octanoyl-glucosylceramide acts equally potent but is itself less toxic. When coformulated with liposomal doxorubicin, this short-chain glycosphingolipid administered to cultured A431 epidermoid carcinoma cells led to superior (up to 4-fold) cellular doxorubicin accumulation and cytotoxicity, compared with control doxorubicin liposomes. These results were fully reproducible when N-octanoyl-glucosylceramide was postinserted into Caelyx, a commercial liposomal doxorubicin preparation. The doxorubicin-potentiating effect of N-octanoyl-glucosylceramide-enriched liposomes proved relatively insensitive to high serum concentrations, indicating that in vivo application is a feasible option. N-Octanoyl-glucosylceramide enrichment might thus represent a major improvement of conventional liposomal doxorubicin formulations.

Alkyl-lysophospholipids (ALPs) represent a new class of antitumor drugs that induce apoptotic cell death in a variety of tumor cell lines. Although their precise mechanism of action is unknown, ALPs primarily act on the cell membrane, where they inhibit signaling through the mitogen-activated protein kinase (MAPK) pathway. Because stimulation of the stress-activated protein kinase/c-Jun NH2-terminal kinase (SAPK/JNK) pathway is essential for radiation-induced apoptosis in certain cell types, we tested the effect of ALPs in combination with ionizing radiation on MAPK/SAPK signaling and apoptosis induction. Here, we present data showing that three ALPs, 1-O-octadecyl-2-O-methyl-rac-glycero-3-phosphocholine, hexadecylphosphocholine, and the novel compound octadecyl-(1,1-dimethyl-piperidinio-4-yl)-phosphate (D-21266) induce time- and dose-dependent apoptosis in the human leukemia cell lines U937 and Jurkat T but not in normal vascular endothelial cells. Moreover, in combination with radiation, ALPs strongly enhance the induction of apoptosis in both leukemic cell lines. All tested ALPs not only prevented MAPK activation, but, like radiation, stimulated the SAPK/JNK cascade within minutes. A dominant-negative mutant of c-Jun inhibited radiation- and ALP-induced apoptosis, indicating a requirement for the SAPK/JNK pathway. Our data support the view that ALPs and ionizing radiation cause an enhanced apoptotic effect by modulating the balance between the mitogenic, antiapoptotic MAPK, and the apoptotic SAPK/JNK pathways. This type of modulation of specific signal transduction pathways in tumor cells may lead to the development of new therapeutic strategies.

Human cutaneous melanoma is heterogeneous with respect to the genetic aberrations involved and the genes altered are potential targets for the immune system. The incidence of cutaneous melanoma is known to be linked to UV peak exposure, and the N-ras oncogene is clearly one of the genes involved in the UV carcinogenesis in melanoma. It is mutated in a significant proportion of melanomas and therefore may serve as a target for T cells. Here, we report that an human leukocyte antigen-A2 binding peptide CLLDILDTAGL, encompassing the frequently found 61-Leu mutation in N-ras, induces cytotoxic T lymphocytes from healthy donor blood that lyse 61-Leu N-ras transfected melanoma cells. Furthermore, we have found an association between the presence of N-ras mutations and clinical response to immunotherapy with interleukin-2 plus interferon in a group of stage IV melanoma patients. Although the overall survival of these patients was not affected by the N-ras status of their melanomas, these studies suggest that mutated N-ras may provide a target for cytotoxic T lymphocytes in melanoma patients.

Intermittent or recreational exposure to sunlight is thought to contribute to development of human cutaneous melanoma. We investigated the incidence of ras oncogene mutation in human cutaneous melanoma in connection to sun-exposed body sites in the patient, using a large series of DNA samples derived from paraffin-embedded material as well as from fresh tumor samples and cell lines. We first show that, of the ras family, predominantly N-ras is activated (15%), whereas rarely H-ras or K-ras are mutated. The occurrence of N-ras mutations correlates with continuous exposure to sunlight of the primary tumor site. Of all tumors initiated on chronically sun-exposed body sites, 26% contained mutated N-ras, in contrast to 0% of sun-protected melanomas. Melanoma lesions obtained from patients from North or Central Europe contained fewer N-ras mutations (12%) as compared with patients from Australia (24%). Mutations were specifically associated with nodular melanoma and to a lesser extent with lentigo malignant melanoma. N-ras mutations did not correlate with metastasis or survival parameters. This study identifies a subset of cutaneous melanomas that contain in the primary lesion ultraviolet-induced N-ras mutations, which are maintained through further progression.

Structural features of Internal Transcribed Spacer 1 (ITS1) that direct its removal from Saccharomyces cerevisiae pre-rRNA during processing were identified by an initial phylogenetic approach followed by in vivo mutational analysis of specific structural elements. We found that S. cerevisiae ITS1 can functionally be replaced by the corresponding regions from the yeasts Torulaspora delbrueckii, Kluyveromyces lactis and Hansenula wingei, indicating that structural elements required in cis for processing are evolutionarily conserved. Despite large differences in size, all ITS1 regions conform to the secondary structure proposed by Yeh et al. [Biochemistry 29 (1990) 5911-5918], showing five domains (I-V; 5'-->3') of which three harbour an evolutionarily highly conserved element. Removal of most of domain II, including its highly conserved element, did not affect processing. In contrast, highly conserved nucleotides directly downstream of processing site A2 in domain III play a major role in production of 17S, but not 26S rRNA. Domain IV and V are dispensable for 17S rRNA formation although an alternative, albeit inefficient, processing route to mature 17S rRNA may be mediated by a conserved region in domain IV. Each of these two domains is individually sufficient for efficient production of 26S rRNA, suggesting two independent processing pathways. We conclude that ITS1 is organized into two functionally and structurally distinct halves.

The rate of ribosomal protein gene (rp-gene) transcription in yeast is accurately adjusted to the cellular requirement for ribosomes under various growth conditions. However, the molecular mechanisms underlying this co-ordinated transcriptional control have not yet been elucidated. Transcriptional activation of rp-genes is mediated through two different multifunctional transacting factors, ABF1 and RAP1. In this report, we demonstrate that changes in cellular rp-mRNA levels during varying growth conditions are not parallelled by changes in the in vitro binding capacity of ABF1 or RAP1 for their cognate sequences. In addition, the nutritional upshift response of rp-genes observed after addition of glucose to a culture growing on a non-fermentative carbon source turns out not to be the result of increased expression of the ABF1 and RAP1 genes or of elevated DNA-binding activity of these factors. Therefore, growth rate-dependent transcription regulation of rp-genes is most probably not mediated by changes in the efficiency of binding of ABF1 and RAP1 to the upstream activation sites of these genes, but rather through other alterations in the efficiency of transcription activation. Furthermore, we tested the possibility that cAMP may play a role in elevating rp-gene expression during a nutritional shift-up. We found that the nutritional upshift response occurs normally in several mutants defective in cAMP metabolism.