Growth properties, differentiation capacity and oncogene expression in metastatic and nonmetastatic Friend leukemia cell variants.

The aims of this study were: (1) to characterize the biologic properties of the WGA-resistant (WR) Friend leukemia cells (FLC) as compared to the original nonmetastatic or highly metastatic FLC; (2) to investigate the possible correlations between the expression of some oncogenes (i.e., c-myc, H-ras and K-ras) and the in vitro and in vivo behavior of FLC. The tumorigenic behavior of the different FLC types strongly depended on the site of tumor injection. Both WR FLC and in vitro passaged FLC did not grow as ascites (when injected intraperitoneally) and developed large solid tumors (when injected subcutaneously), without forming any spleen or liver metastasis. In contrast, in vivo passaged FLC rapidly formed hemorrhagic ascites when injected intraperitoneally; the subcutaneous injection of these cells resulted in the development of solid tumors, which were smaller than the other FLC tumors, but capable of metastasizing to the liver and to the spleen. No significant differences were observed in the in vitro growth characteristics and cell cycle parameters among the different FLC types under various experimental conditions (i.e., FCS concentration or cell seeding densities). Similarly to the metastatic in vivo passaged parental cells, WR FLC exhibited a much lower erythroid differentiation after in vitro addition of either dimethyl sulfoxide or hexamethylene bisacetamide than the in vitro passaged FLC. High levels of c-myc oncogene mRNA were expressed in all FLC variants; no major variations in the c-myc expression were observed in FLC cultivated in medium supplemented with different FCS concentrations and/or seeded at various cell densities. In addition, no changes in the expression of H-ras or K-ras were observed between the different FLC types.     

Cell surface sialylation of glycoproteins and glycosphingolipids in cultured metastatic variant RNA-virus transformed non-producer BALB/c 3T3 cell lines.

The sialic acid composition and the display of cell surface sialyl components of several metastatic variant RNA-virus-transformed non-producer BALB/c 3T3 have been studied in culture. The following observations have been made concerning the sialyl components in these lines: (1) the compositions of whole-cell total, protein-bound and lipid-bound sialic acid were not appreciably different; (2) the surface sialic acid studied using the neuraminidase-galactose oxidase method and metabolic labelling followed by neuraminidase hydrolysis showed a positive correlation with the metastatic properties of these lines; (3) the degree of surface sialylation determined by galactose oxidase--sodium borotritide labelling of neuraminidase-treated and untreated cells revealed that 44--89% of exposed galactose and/or N-acetyl galactosamine residues of total cell-surface saccharides were sialylated in highly and intermediately metastatic lines as compared with 11-30% in the poorly and non-metastatic lines; (4) the cell surface glycoproteins and glycosphingolipids contributed equally well in their degree of sialylation and there was a positive correlation with the metastatic properties of the cells in vivo; (5) the cell surface proteins labelled by the lactoperoxidase-catalyzed iodination technique, followed by gel electrophoresis, showed some minor differences between metastatic variant lines. However, glycoproteins detected by the galactose oxidase labelling of neuraminidase-treated and untreated cells showed major differences in composition between the metastatic variant lines. The study of four highly metastatic lines has shown that the cells of these lines were enriched in several sialyl-glycoproteins, whereas three non tumorigenic lines and two poorly metastatic or non-metastatic lines contained unsialylated glycoproteins. The results indicate an enhancement of the degree of sialylation of surface glycoconjugates accompanying the metastatic process in RNA-virus-transformed mouse lines.     

Biologic and biochemical differences between in vitro and in vivo passaged friend erythroleukemia cells. II. Changes in cell surface glycoproteins associated with a highly malignant phenotype

Friend erythroleukemia cells (FLC), serially passaged in vkroor by intraperitoneal injection in DBA/2 mice, exhibit markedly different tumorigenicity and capacity to metastasize. We have attempted to determine whether the differences in tumorigenicity between these two lines of FLC were correlated with any biochemical changes in their cell membranes. Although consistent modifications of FLC membrane gangliosides were detected after FLC multiplied in the peritoneum, the pattern of FLC gangliosides was not a stable characteristic and did not correlate with tumorigenicity. In contrast, analysis of FLC membrane glycoproteins by cell surface labelling techniques (i.e., galactose-oxidase-borohydride techniques and polyacrylamide gel electrophoresis-fluorography) or by metabolic labelling of glycoproteins with ³H-galactose, revealed consistent differences in the high MW region of the gels between parental in vitro passaged FLC (either 745 or 3C1-8 cells) and clones derived from in vivo passaged cells. No significant differences in the membrane proteins were detected between in vitro and in vivo passaged FLC when lactoperoxidase-catalyzed iodination and polyacrylamide gel electrophoresis-autoradiography were used. It is seen that repeated in vivo passages of FLC resulted in the appearance of different patterns of membrane glycoproteins and that these changes appeared to be associated consistently with the capacity of these cells to grow as tumor ascites and to metastasize to the liver and spleen.     

Lectin binding glycoproteins in human melanoma cell lines with high or low tumorigenicity

Lectin binding glycoproteins of 5 human malignant melanoma cell lines (HMMCL), differing in their ability to grow subcutaneously in athymic nude mice, were compared by electrophoresis of total cellular proteins and subsequent incubation of SDS-poly-acrylamide gel with 125I-labelled lectins. Despite the similarity between the protein profiles of the different HMMCL, Concanavalia ensiformis agglutinin (ConA), wheat-germ agglutinin (WGA) and peanut agglutinin (PNA) revealed differences in their glycoprotein expression, in contrast with Ulex europaeus agglutinin I (UEA I). A great diversity was observed in the electrophoretic mobilities and/or staining intensities of ConA and WGA binding glycoproteins of HMMCL. However, neither ConA-reactive glycoproteins nor WGA-reactive glycoproteins could be detected that were characteristic of HMMCL with high tumorigenicity (HT) or low tumorigenicity (LT). In contrast, the expression of two cell-surface PNA binding glycoproteins appeared to be related to the tumorigenic phenotype of HMMCL. One of them, with an apparent molecular weight of 190 kDa, was only detected in the LT cell lines. The other, with an apparent molecular weight of 60 kDa, was detected in all HMMCL but became strongly labelled after neuraminidase treatment only in the HT cell lines. Thus, the expression of glycoproteins rich in terminal galactose residues may characterize human melanoma cells with different tumorigenic behavior.     

Differential expression of glycoproteins containing alpha-D-galactosyl groups on normal human breast epithelial cells and MCF-7 human breast carcinoma cells

Cell surface glycoproteins were isolated from the lysates of 125I-labeled normal human mammary epithelial cells (NHMEC) and from the human breast carcinoma cell line MCF-7, of blood-group O phenotype, by affinity chromatography on Griffonia simplicifolia I lectin-Sepharose. Specific elution of glycoproteins from the column with methyl alpha-D-galactoside suggests the presence of alpha-D-galactosyl groups on these moieties. SDS-PAGE analysis of isolated glycoproteins revealed both quantitative and qualitative differences between glycoproteins from normal and malignant cells. Three major glycoproteins of Mr 180 kDa, 85 kDa and the 44 kDa were obtained from MCF-7 cells. The 180-kDa glycoprotein was absent in NHMEC and the 44-kDa glycoprotein was very weakly expressed in these cells. The only glycoprotein which was found in almost equal amount in the lysate from both normal and malignant cells was the 85-kDa glycoprotein. These results indicate differences between normal human mammary epithelial cells and one kind of malignant human mammary epithelial cells, in the expression of glycoproteins containing alpha-D-galactosyl groups, irrespective of blood-group phenotype; they also demonstrate that alpha-D-galactosyl group are expressed in a very restrictive manner on the surface of this tumor cell line.     

Cell surface biochemical and metastatic properties of Lens culinaris hemagglutinin-binding variants of a murine large cell lymphoma

Using a low metastatic potential parental (P) line of the murine large cell lymphoma RAW117 and a highly metastatic in vivo-selected liver-colonizing subline (H10), we examined the relationship between cell surface glycoprotein expression and metastasis. The highly metastatic H10 cells showed loss of the major RNA tumor virus envelope glycoprotein gp70 and increased expression of a concanavalin A and Lens culinaris hemagglutinin (LcH)-binding glycoprotein of Mr approximately 15,000 (gp150) by lectin affinity chromatography and 125I-lectin staining of isolated RAW117 glycoproteins. When the amounts of cell surface LcH-binding components were determined on P and H10 cells, the mean amount of cell-bound LcH on H10 cells was significantly greater than on P cells. RAW117-P cells were sorted for low (PLcH-low) or high (PLcH-high) LcH binding using a fluorescence-activated cell sorter, or for binding to immobilized LcH, and the resulting cell sublines were analyzed for their metastatic properties by intravenous injection into BALB/c mice. The parental P cells formed few liver tumor nodules (median 0; range 0-8), as did the PLcH-low cells (median 0; range 0), whereas the high LcH-adherent P cells and the cells sorted for increased LcH binding, PLcH-high, were highly metastatic to the liver (median 200; range 156 to 200+). Analyses of gp150 and gp70 contents indicated higher amounts of gp150 but lower quantities of gp70 on PLcH-high cells than on PLcH-low or P cells. The results suggest that the amounts of cell surface gp150 and gp70 are important in determining the metastatic properties of RAW117 cells.     

Effect of anthracycline analogs on photolabelling of p-glycoprotein by [125I]iodomycin and [3H]azidopine: relation to lipophilicity and inhibition of daunorubicin transport in multidrug resistant cells.

Eight anthracycline analogs that have been shown to modulate multidrug resistance (Friche et al., Biochem. Pharmacol., 39, 1721-1726; 1990) were tested for their inhibitory effect on the photolabelling of P-glycoprotein. We photoaffinity labelled P-glycoprotein in daunorubicin (DNR) resistant Ehrlich ascites tumour cells (EHR2/DNR +) with a [125I]iodinated Bolton-Hunter derivative of daunorubicin ([125I]iodomycin) and with [3H]azidopine. The photolabelling of P-glycoprotein by [125I]iodomycin was inhibited more than 50% by 10 microM (1000-fold molar excess) of DNR (52%), N,N-dibenzyl-DNR (52%), and N-benzyladriamycin-14-valerate (AD-198) (85%). Vincristine at 10 microM inhibited [125I]iodomycin labelling of P-glycoprotein by 95%. Thus vincristine was more potent than any of the eight anthracyclines tested, despite its relatively low lipophilicity. Increasing the concentration of DNR, AD-198 and N,N-dibenzyl-DNR to 40 microM resulted in 90, 99.5 and 99.5% inhibition of P-glycoprotein labelling by [125I]iodomycin, respectively. In comparison with the other anthracycline analogs, N,N-dibenzyl-DNR and Ad-198 were also found to exert the greatest inhibition of [3H]azidopine labelling of P-glycoprotein (about 90% at 100-fold molar excess). The solvents Cremophor EL and Tween 80 (30 micrograms ml-1; 0.003% v/v), which are modulators of multidrug resistance in EHR2/DNR + cells, also inhibited [125I]iodomycin labelling > 90%. We showed earlier that there is a correlation between the lipid solubility within the anthracycline group of MDR-associated drugs and their ability to enhance DNR accumulation in EHR2/DNR + cells but a corresponding correlation to lipophilicity when it comes to the inhibitory effect on the specific photolabelling of Pgp ligand binding sites could not be demonstrated. Neither could a correlation between the modulating effect of the analogs on DNR accumulation and inhibition on the labelling of Pgp be demonstrated. With increasing lipophilicity of the analogs it seems that the chemical structure plays a lesser role, and the degree of lipophilicity becomes a more important feature.     

Development of an enzyme-linked immunosorbent assay (ELISA) for detecting IgA antibodies to the Epstein-Barr virus

A 3-step enzyme-linked immunosorbent assay (ELISA) was developed for detecting IgA antibodies to purified Epstein-Barr virus (EBV) polypeptides. The 3-step procedure included the use of a mouse anti-human IgA monoclonal antibody (MAb) to amplify the IgA reaction. The 2 major EBV proteins used in this assay were the 125-kDa component (gp125) associated with the viral capsid antigen (VCA) complex and a major glycoprotein (gp250/200) associated with the membrane antigen (MA) complex. Eighty-two sera were tested on ELISA plates containing either both of the glycoproteins or each one separately. These included 45 IgA antibody-positive sera from patients with nasopharyngeal carcinoma (NPC). With these sera, there was a good correlation, both qualitatively and quantitatively, between results with the immunofluorescence (IF) and ELISA procedures. Although most IgA antibody-positive sera contained antibodies reactive with both gp125 and gp250/200, a number of sera contained antibodies reactive with one of the glycoproteins but not with both. The data indicated that both of these glycoproteins should be used in assays for detecting IgA antibodies to EBV, to avoid false-negative results. This assay should be useful for screening large populations for IgA antibodies to EBV and also possibly for monitoring disease course in patients with NPC.     

Properties of wr2721 (ethiofos) as modulator of Cisplatin-induced neurotoxicity studied at the ultrastructural level in the pond snail lymnaea-stagnalis

Modulating effects of WR2721 were studied on cisplatin-induced histological and ultrastructural changes in the ganglia of the central nervous system of the pond snail Lymnaea stagnalis. The relevance of the study was indicated by showing histochemically that alkaline phosphatase, the enzyme converting WR2721 into the active drug WR1065, is present in the snail brain. Central nervous systems of the snail were either preincubated in 5 mM WR2721 or in snail Ringer for 1 h and then postincubated for 10 or 20 h in: (i) Ringer (control), (ii) WR2721 (5 mM), (iii) cisplatin (0.4 mM), or (iv) cisplatin (0.4 mM) plus WR2721 (5 mM). The following parameters were studied: (i) general morphology, (ii) chromatin (number and mean clump size per unit surface ama, clump size frequency distribution), (iii) nucleoli (ratio of granular/fibrillar areas, (iv), Golgi zones (mean surface area, area containing electron dense material), (v) secretory granules (number), (vi) lysosomes (number per unit surface area of cytoplasm). The focus was on two types of identified neuroendocrine cells. Incubation in Ringer alone (controls) caused slight inactivation of the cells between 10 and 20 h of incubation (e.g. relative decrease of the granular part of the nucleoli). WR2721 alone had comparable effects on the tissue. In addition, in this group, a reduction in the electron dense material of the Golgi zones was observed. No major differences in number of secretory granules were observed in any of the groups. Treatment with cisplatin alone for 20 h caused a disappearance of the orange colour of the ganglia, swelling of axons and distension of intercellular spaces. Such changes were not observed in the group treated with cisplatin plus WR2721 or any of the other experimental groups. Both cisplatin alone and cisplatin plus WR2721 caused an increase in the number of chromatin clumps and a decrease in the mean chromatin clump size after 10 and 20 h of incubation. With regard to these parameters no differences were observed between the two groups. The chromatin clump size distribution curves of both groups were significantly different from the curve of the controls. Compared to that of the cisplatin group, the curve of the cisplatin plus WR2721 group, especially after 10 h, had shifted in the direction of that of the controls. Treatment with cisplatin alone induced drastic changes in nucleolar morphology. After 20 h of incubation the nucleoli had transformed into homogeneous dense structures. After 10 h of incubation in cisplatin plus WR2721 nucleoli still had a normal appearance. After 20 h those of the co-treated group had also become electron dense and appeared to contain numerous dark dots. Cisplatin caused a significant decrease in the extent of the Golgi zones. Co-treatment with WR2721 prevented this decrease to a certain degree. In both groups the electron dense material had disappeared from the Golgi zones. Furthermore, cisplatin alone induced an increase in the number of lysosomes. This increase was slightly (but not significantly) prevented by co-treatment with WR2721. The observations on the snail neurons show that WR2721 at the concentration studied induced only minor morphological changes. The drug modulates to some extent the pathological changes caused by cisplatin. The results support clinical data indicating that cisplatin-induced neurotoxicity is reduced by WR2721.     

The NCI-atlas of dose distributions for regular 125I brain implants

In connection with the development of an optimization method for 125I brain implants in irregularly shaped target volumes, a systematic study was conducted toward optimizing seed configurations for regular target volumes. The intention was to find basic rules for the positioning of strings of seeds in the cylindrical implant pattern of the stereotactic neurosurgical procedure in use, and in accordance with the following criteria: (i) steep dose fall-off outside the target volume; (ii) coverage of the target volume by making the prescribed dose surface coincident with the target volume surface within 1 mm; (iii) uniformity of the dose distribution in the target volume as far as achievable with a seed implant. As a result of this study, an atlas of optimized regular 125I brain implant configurations was compiled. Regular implants were understood as being cylindrical or spherical. Diameters and heights from 2 to 5 cm were covered.     

Cell surface glycoproteins of human tumor cell lines: unusual characteristics of malignant melanoma.

Cell surface glycoproteins of human tumor cell lines (melanomas and astrocytomas, and ovarian, bladder, stomach, cervical, laryngeal, and renal cancers) were studied by labelling with 1) neuraminidase-galactose oxidase-[3H]borohydride, 2) galactose oxidase-[3H]borohydride, and 3) dilute periodate-[3H]borohydride. The labeled components were examined by polyacrylamide gel electrophoresis and fluorography. Each tumor type had a distinctive pattern of labeled glycoproteins when the results from both procedures 1 and 2 were considered. Cell surface glycoproteins of malignant melanoma could not be labeled by procedure 2, whereas the other cell lines had at least two major glycoproteins that could be labeled by this method. Very similar profiles of melanoma glycoproteins were labeled by procedures 1 and 3. From these results the conclusion was reached that cell surface glycoproteins of melanomas are substituted with sialic acid so that their D-galactose and/or N-acetyl-D-galactosamine residues are available for oxidation by galactose oxidase only after neuraminidase treatment. An alternative explanation that these sugars are sterically accessible to galactose oxidase only after neuraminidase treatment also has to be considered. All melanoma lines studied were characterized by having two major cell surface glycoproteins with molecular weights of 110,000 and 90,000, respectively. Lines, however, varied considerably in their expression of other components. In particular, heterogeneity was shown in the expression of gp220, a component identified as fibronectin by immunoprecipitation with a specific antiserum, and in the expression of gp37/32, a pair of glycoproteins having the characteristics of la-like antigens. Of the other cell lines studied, astrocytomas most closely resembled melanoma in their glycoprotein profiles. The brain tumors, however, had two or three glycoproteins, including gp110, which could be labeled by galactose oxidase-[3H]borohydride without neuraminidase treatment.     

Epstein-Barr virus-induced membrane antigens: immunochemical characterization of Triton X-100 solubilized viral membrane antigens from EBV-superinfected Raji cells.

In an attempt to qualitatively identify the membrane antigen (MA) complex induced by Epstein-Barr virus (EBV) infection of lymphoblastoid cells, superinfected Raji cells were surface labelled with 125I by the lactoperoxidase method and solubilized with Triton X-100, then the 125I-labelled membrane proteins were precipitated by sera containing high antibody titers to MA. Analysis of these immune precipitates on sodium dodecyl sulfate polyacrylamide gel eletrophoresis identified four major EBV-specific membrane proteins with molecular weights (mol. wt) of 280,000, 250,000, 170,000 and 90,000. Sera from patients with Burkitt's lymphoma (BL), nasopharyngeal carcinoma (NPC) and infectious mononucleosis (IM) and from EBV-infected disease-free individuals showed differential patterns of reactivity to these antigens with some sera only recognizing three or less of the antigens. The results from invesigations with these sera also indicated that these major proteins were not related to EBV-induced viral capsid antigens (VCA) or the virus-associated early antigen (EA) complexes as defined by immunofluorescence. Metabolic labelling of EBV-infected Raji cells with [14C]glucosamine, followed by Triton X-100 solubilization and radioimmune precipitation, identified the 280,000, 250,000 and 90,000 components as glycoproteins. The lactoperoxidase-labelled 170,000 molecular weight component was not significantly glycosylated and, therefore, could not be categorized as a glycoprotein on the basis of this study. In addition, a glycoprotein with a mol. wt of 130,000 was identified by this approach which also appeared to be specified by EBV. The results from these investigations, therefore, indicated that the EBV-induced MA complex was composed of four major glycoproteins and one nonglycosylated high mol. wt protein.     

Biochemical identification of primate lymphoid cell-surface glycoproteins.

We have employed the galactose oxidase-tritiated sodium borohydride labelling method to examine the surface glycoproteins of cotton-topped marmoset and other primate cell lines either established from tumors or transformed in vitro by different lymphotropic herpesviruses. The labelled surface glycoproteins were separated on acrylamide gels in the presence of sodium dodecyl sulfate (SDS) and analyzed by fluorography. Our results indicate that (1) lymphocytes of the same class from different primate species are similar but can be distinguished; (2) T and B lymphocytes of the same species can be differentiated; (3) cotton-topped marmoset lymphocytes of the same class show marked similarities regardless of tumor or in vitro origin or virus used for transformation; (4) three cell lines established from different EBV-induced tumors of the same marmoset show essentially the same labelling pattern, supporting the hypothesis that they originated from a single clone.     

A targeted glycoproteomic approach identifies cadherin-5 as a novel biomarker of metastatic breast cancer

Aberrant glycosylation has long been recognised as a hallmark of cancer, and is increasingly being exploited in biomarker discovery studies. Helix pomatia agglutinin (HPA) is known to bind aberrant glycans associated with metastatic breast cancer, and was used here to isolate glycoproteins from pooled breast cancer serum samples of (i) patients with recurrent breast cancer and (ii) patients with no sign of recurrence 5 years after diagnosis of their primary tumour. Pregnancy zone protein, the polymeric immunoglobulin receptor and cadherin-5 emerged as potential markers of metastasis following proteomic identification of HPA binding glycoproteins. ELISAs were developed to verify these findings, and to assess protein glycosylation, in individual patient sera. The cadherin-5 ELISA discriminated serum samples of patients with recurrent breast cancer from those with no sign of recurrence, and analysis of cadherin-5 glycosylation by HPA also showed a significant difference between the two sample groups. The targeted glycoproteomic and validatory approach developed here has shown that when taking into account both the protein levels and HPA binding, serum cadherin-5 discriminated patients with recurrent breast cancer from those with no sign of recurrence with 90% specificity.     

Time dependence of the selective modulation of cisplatin-induced nephrotoxicity by WR2721 in the mouse.

2-(3-Aminopropylamino)ethylphosphorothioic acid (WR2721; ethiofos) was shown to selectively protect nontumor tissues from cis-diamminedichloroplatinum(II) (cisplatin)-induced toxicity, when administered 30 min prior to the platinum drug. Selectivity of protection by WR2721 is probably due to the preferential formation and uptake of the thiol metabolite 2-(3-aminopropylamino)ethanethiol (WR1065), which can inactivate toxic platinum-species inside the cell. We investigated the protective potential of WR2721, when administered at different time points relative to cisplatin. BALB/c mice treated with WR2721 (200 mg/kg i.p.) either 30 min or 5 min prior to cisplatin (i.p.) allowed a 2.2-fold increase in cisplatin dose to 19 mg/kg before the occurrence of nephrotoxicity as expressed by an increase in plasma urea. A small part of the protection could be ascribed to the mannitol (200 mg/kg), present in the formulated WR2721. WR2721 (200 mg/kg) 30 min after 14.5-16-mg/kg cisplatin did not offer any protection against the rise in plasma urea. WR2721 (200 mg/kg) 5 min before 19-mg/kg cisplatin did not cause liver toxicity (increase in serum glutamic pyruvic transaminase or serum glutamic oxaloacetic transaminase). Furthermore, WR2721 (200 mg/kg) 5 min prior to cisplatin did not reduce antitumor activity in nude mice bearing well-established human ovarian cancer xenografts. Under protection of WR2721, the dose of cisplatin could be increased by a factor of 1.6 to 8 mg/kg (administered twice weekly), resulting in an increased antitumor activity.     

Differential effects of the glycolysis inhibitor 2-deoxy-D-glucose on the activity of pro-apoptotic agents in metastatic melanoma cells, and induction of a cytoprotective autophagic response

2-Deoxy-D-glucose (2DG) is a synthetic glucose analogue that inhibits glycolysis and blocks cancer cell growth. In this report, we evaluated the role of 2DG in the induction of cell death in human metastatic melanoma cells. We have also examined the effects of 2DG in combined treatments with four different pro-apoptotic agents: (i) Temozolomide (TMZ), a chemotherapic drug commonly used to treat metastatic melanoma, (ii) Pyrimethamine (Pyr), a pro-apoptotic antifolate drug recently reappraised in cancer therapy, (iii) Cisplatin (CisPt), a drug capable of directly binding to DNA ultimately triggering apoptosis of cancer cells and (iv) the kinase inhibitor Staurosporine (STS), a prototypical inducer of mitochondria-mediated apoptosis. We found that 2DG per se: (i) induced a cell cycle arrest in G(0) /G(1) , (ii) promoted autophagy, (iii) was ineffective in inducing apoptosis in association with the chemotherapic drug TMZ, whereas (iv) it was synergistic with CisPt and STS pro-apoptotic drugs through a mechanism involving changes of mitochondrial homeostasis. Conversely, (v) 2DG hindered the pro-apoptotic effects of Pyr via a mechanism involving either the block of cell cycle in G(0) /G(1) or the modification of the free radical production of the cell, i.e., decreasing the production of reactive oxygen species (ROS) and increasing the production of reactive nitrogen species (RNS). Moreover, a clear-cut autophagic response involving endoplasmic reticulum remodelling was detectable. Since autophagic cytoprotection has been suggested to contribute to the induction of chemoresistance, these results could provide useful clues as concerns the use of 2DG as anticancer agent in combinatory protocols.     

Type-II insulin-like growth-factor receptor in conditioned medium from HT-29 human colon carcinoma cell line.

The HT-29 human colon cancer cell line has previously been shown to secrete high amounts of insulin-like growth factor II (IGF-II). The recent demonstration that soluble IGF-II/mannose 6-phosphate receptor was present in fetal serum prompted us to search for a release of type-II IGF receptor by these human colonic carcinoma cells. Serum-free conditioned medium from the HT-29 cell line was gel filtered on Sephadex G-200. There was significant binding of [125I]IGF-II to the void volume fractions in addition to binding to the 40-kDa IGF-binding protein (IGF-BP) fractions. Competitive binding studies using [125I]IGF-II and the void volume pool showed a pattern typical of the type-II receptor. It exhibited a high affinity for IGF-II (KD = 0.4 nM), but had a low affinity for IGF-I (KD = 6.8 nM), and no detectable affinity for insulin. Additional evidence was provided by affinity cross-linking of [125I]IGF-II to the same high-molecular-weight material which demonstrated a major specific band at 250 kDa after reduction of disulfide bonds. In contrast, the type-I IGF receptor was undetectable. The extracellular type-II IGF receptor was not a significant carrier for IGF-II since virtually all IGF-II secreted by HT-29 cells was associated with IGF-BP. The presence of a soluble IGF-II/mannose 6-phosphate receptor in the culture medium from colonic cancer cells suggests that it may play an important role in tumor pathogenesis.     

Differential response of adherent and unanchored melanoma cells to bromodeoxyuridine evidenced by specific lectin-binding protein changes

The possible differential response of adherent and nonadherent cells of the same tumor type to pyrimidine analogues has been investigated. We show that bromodeoxyuridine (BUdR) increases interactions of attached cells with their substrate without markedly affecting the cell adhesion properties of the same cells when these are not anchored. However, evidence for an adhesion-independent response of both cell types to BUdR has been obtained with lectin binding assays using 125I-labelled Lens culinaris agglutinin (LCA). This revealed a greatly increased binding of LCA to a large glycoconjugate in all cultures exposed to the halogenated pyrimidine. Attachment-dependent effects of BUdR were manifested in flattened cells by a greater LCA-binding to a 240-kDa protein and by increased interaction of 125I-labelled wheat-germ agglutinin (WGA) with a 200-kDa protein and a large glycoconjugate sharply defined in electrophoresis. Although both tumor cell aggregates and anchored cells exhibit detectable responses to pyrimidine analogues such as BUdR, the corresponding effects are thus manifested unequally in cells with different adhesion properties.     

Surface glycoprotein patterns of normal and malignant human lymphoid cells. II. B cells, B blasts and Epstein-Barr virus (EBV)-positive and -negative B lymphoid cell lines.

Surface glycoproteins of normal human B lymphocytes, B blasts and various types of lymphoid cell lines were labelled by the galactose oxidase-tritiated sodium borohydride method. The labelled glycoproteins were separated by polyacryUmide slab gel electrophoresis in the presence of sodium dodecy! sulfate and visualized by modified autoradiography (fluorography). The battery of examined hematopoietic cell lines included Epstein-Barr virus (EBV) carrying lines of proven malignant origin (Burkitt's lymphoma) and presumed non-neoplastic origin (lymphoblastoid cell lines), and EBV-genome-negative lymphocytic lymphoma, histiocytic lymphoma, myeloma and myeloid leukemia lines. The presence of possible EBV-associated surface glycoproteins, detectable by the labelling method, was studied by use of two EBV-negative cell lines (BJAB and Ramos) and their EBV-converted sublines. The six Burkitt lymphoma and three lymphocytic lymphoma lines had all the basic surface glycoprotein pattern of resting B cells and, in addition to individually distinct bands, two characteristic pairs of glycoproteins sf apparent molecular weights of 87,000/85,000 and 71,000/69,OOO. These glycoproteins were not detected on the normal B celts, B blasts or non-neoplastic lymphoblastoid lines. Neither were they found on the other types of neoptastic line. No consistent difference in the surface glycoprotein patterns was detected between the EBV-genome-negative and EBV-con-verted BJAB and Ramos sublines. The glycoprotein pattern of the six lymphoblastoid lines resembled that of the B blasts. The histiocytic lymphoma, myeloma and leukemia lines all had distinct patterns. These results confirm that the Burkitt's lymphoma and the lymphoblastoid cell lines represent two biologically distinct EBV-carrying B lymphoid celts and that the galactose oxidase NaB[³H]₄ surface labelling technique can be used as a reliable molecular mapping method to distinguish between these two and other types of lymphoid cell lines.     

In vitro selectivity, in vivo biodistribution and tumour uptake of annexin V radiolabelled with a positron emitting radioisotope

The availability of a noninvasive method to detect and quantify apoptosis in tumours will enable tumour response to several cancer therapies to be assessed. We have synthesised two radiotracers, annexin V and the N-succinimidyl-3-iodobenzoic acid (SIB) derivative of annexin V, labelled with radio-iodine ((124)I and (125)I) and provided proof of the concept by assessing specific binding and biodistribution of these probes to apoptotic cells and tumours. We have also assessed the tumour uptake of [(124)I]annexin V in a mouse model of apoptosis. RIF-1 cells induced to undergo apoptosis in vitro showed a drug concentration-dependent increased binding of [(125)I]annexin V and [(125)I]SIB-annexin V. In the same model system, there was an increase in terminal deoxynucleotidyl transferase-mediated nick end labelling (TUNEL)-positive cells and a decrease in clonogenic survival. Radiotracer binding was completely inhibited by preincubation with unlabelled annexin V. In RIF-1 tumour-bearing mice, rapid distribution of [(125)I]SIB-annexin V-derived radioactivity to kidneys was observed and the radiotracer accumulated in urine. The binding of [(125)I]SIB-annexin V to RIF-1 tumours increased by 2.3-fold at 48 h after a single intraperitoneal injection of 5-fluorouracil (165 mg kg(-1) body weight), compared to a 4.4-fold increase in TUNEL-positive cells measured by immunostaining. Positron emission tomography images with both radiotracers demonstrated intense localisation in the kidneys and bladder. Unlike [(124)I]SIB-annexin V, [(124)I]annexin V also showed localisation in the thyroid region presumably due to deiodination of the radiolabel. [(124)I]SIB-annexin V is an attractive candidate for in vivo imaging of apoptosis by PET.