Antiproliferative activity in vitro and in vivo of the spicamycin analogue KRN5500 with altered glycoprotein expression in vitro.

The spicamycin analogue KRN5500 (NSC 650426; SPA) is derived from Streptomyces alanosinicus. The unique structure contains a purine, an aminoheptose sugar, glycine, and a tetradecadiene fatty acid. SPA potently inhibits the growth of certain human tumor cell lines in vitro (IC50 for growth <100 nM) and displays marked activity in vivo in Colo 205 colon carcinoma xenografts. Selective inhibition of labeled precursor incorporation was not evident at 1 or 4 h of exposure to the drug, but at 8 h, [3H] leucine incorporation was inhibited by approximately 40% at or below the IC50 for cell growth. Because of the structural similarity of SPA to inhibitors of glycoprotein processing, we examined the effect of SPA on indicators of glycoprotein synthesis and processing in HL60TB promyelocytic leukemia and Colo 205 colon carcinoma cells. Brief periods of exposure ( approximately 30 min) to SPA at the IC50 for growth increased incorporation of [3H]mannose. When examined by Western blotting after prolonged (40-48 h) incubation with lectins that target mannose-containing carbohydrates, Galanthus nivalis agglutinin and concanavalin A, a qualitative change in the pattern of mannose-containing glycoproteins was observed in HL60TB cells. Significant changes in the pattern of surface glycoprotein expression in intact cells were demonstrated by flow cytometry using fluorescence-labeled lectins. An increase in the number of cells binding G. nivalis agglutinin (indicating terminal mannose) was noted, but a decrease in the amount of lectin bound per cell was noted for wheat germ agglutinin (detecting sialic acid and terminal beta-N-acetyl glucosamine residues). Electron microscopy revealed loss of microvilli, and the Golgi apparatus appeared inflated. Our findings, therefore, raise the possibility that cells exposed to SPA have altered glycoprotein processing after exposure to low concentrations of drug, prior to the occurrence of overt cytotoxicity. These effects are consistent with a prominent early effect of SPA on the enzymatic machinery or organelles important for proper glycoprotein processing and emphasize the novelty of this agent's likely mechanism of action.

Cucurbitacin E-induced disruption of the actin and vimentin cytoskeleton in prostate carcinoma cells.

Cucurbitacin E has been identified by an empiric screening strategy as a sterol with potent growth inhibitory activity in vitro directed against prostate carcinoma explants (IC50 of 7-50 nM in 2- to 6-day exposures). The mechanism of cucurbitacin cytoxicity has not been elucidated previously. In the present study, we observed that cucurbitacin E caused marked disruption of the actin cytoskeleton, and in a series of cucurbitacin analogues, anti-proliferative activity correlated directly with the disruption of the F-actin cytoskeleton. The distribution of vimentin was also altered in cells exposed to cucurbitacin E, as vimentin associated with drug-induced membrane blebs. The appearance of microtubules was unaffected. Western blot analysis of intracellular actin in cells exposed to cucurbitacins and quantitation of rhodamine-phalloidin binding support the hypothesis that cucurbitacin treatment leads to an inappropriate increase in the filamentous or polymerized actin fraction in prostate carcinoma cells. We conclude that cucurbitacins are potent disruptors of cytoskeletal integrity. Prostate carcinoma cells appear notably sensitive to growth inhibition by cucurbitacin E.

Utility of the three-part leukocyte differential count.

The utility of the automated three-part leukocyte differential count was evaluated in an acute care hospital setting. The manual method was specified in 92 of 223 (41%) requests for differentials on one day. For the others, 79 three-part differentials (60%) were completed; 19 (15%) were rejected for histogram abnormalities ("R-flags") or could not be computed; and 33 (25%) were rejected for out-of-range values that were later verified by slide review. The automated method missed 4 of 39 (10%) band elevations and 2 of 2 (100%) cases of eosinophilia reported on manual differentials, but those results had no apparent influence on patient management. In 98 cases of septicemia, automated and manual methods showed similar overall sensitivity (87% and 83%, respectively). Selectively combined with a qualitative slide review, the three-part differential was applicable to 84% of all specimens submitted for a differential count, with acceptable sensitivity and accuracy and substantial savings in personnel time.

Direct delivery of platinum-based antineoplastics to the central nervous system: a toxicity and ultrastructural study.

Platinum drugs are playing an increasingly major role in cancer treatment, but systemic administration of these agents has resulted in significant toxicity. To examine the effects of cisplatin and two newer agents, iproplatin and carboplatin, we injected the agents directly into the cerebrospinal fluid of rats and found that neurotoxic reactions resulted from doses of cisplatin (10 nmol) much lower than those of iproplatin (40 nmol) or carboplatin (80 nmol). Moreover, central nervous system tissue appeared to be less adversely affected by direct exposure to carboplatin since chronic toxicity was not observed in any of the animals receiving carboplatin until a lethal dose was reached. Furthermore, only the animals receiving cisplatin showed histologic damage in their spinal cords, and ultrastructural studies confirmed that while significant abnormalities were observed in the spinal cords of rats receiving 40 nmol cisplatin, no architectural changes were detected in the spinal cords of animals receiving 240 nmol carboplatin. We conclude that platinum drugs can be delivered intrathecally to achieve a much greater concentration of active drug than can be achieved by intravenous administration and that carboplatin appears to be the most suitable platinum-based drug for use in systems delivering drugs directly to the brain and spinal cord.

Esophagogastric adenocarcinoma in an E1A/E1B transgenic model involves p53 disruption.

We studied tumorigenesis and p53 immunostaining in a murine transgenic model introducing E1A/E1B under the control of the mouse mammary tumor virus-long terminal repeat (MMTV-LTR) promoter in which adenocarcinoma occurs at the squamocolumnar junction in the foregut, predominantly in males, and at no other site. Mutations of p53 are frequent in human esophageal adenocarcinoma and the E1B gene product interferes with p53-mediated apoptosis, inhibiting tumor suppression at the G(1)/S checkpoint. Transgenic animals were generated utilizing a purified linear 6.7 kb fragment of plasmid DNA containing MMTV-LTR/E1A/E1B and were confirmed by dot blot hybridization of tail DNA to (32)P-labeled E1A/E1B probe and polymerase chain reaction (PCR) amplification of E1A. Transgenic and control animals were observed for morbidity and weight changes. Eleven of 45 animals were transgenic (24% efficiency) with an estimated 5 to 57 copies of the gene per genome. Profound weight loss (>20%) led to sacrifice or death of one of five females (at 12 weeks) and four of six males (at 16 to 17 weeks). Grossly visible tumors (2 to 10 mm) were noted in the forestomach at the visible margin between the proximal (squamous-lined) stomach and the distal glandular stomach. Histologic sections confirmed adenocarcinoma arising in each case at the squamocolumnar junction with glandular formation, pleomorphism, and frequent mitotic figures. Immunostaining was positive for p53 indicating accumulation of mutated or altered p53 protein. E1A/E1B transgenic animals developed macroscopic and microscopic adenocarcinoma at the squamocolumnar junction, which corresponds to adenocarcinoma at the human esophagogastric junction. Disruption of p53 was present in the transgenic model as in the human cancer.

Inhibition of fibroblast hyaluronic acid production by suramin.

Hyaluronic acid (HA) is an extracellular matrix glycosaminoglycan localized in the stroma of solid tumors, where it facilitates cell movement and thus tumor invasion and metastasis. This localization of HA is due to its synthesis by stromal fibroblasts in response to paracrine factors produced by the tumor. Such tumor-stromal interactions have been shown to be crucial to the development and progression of prostate cancer. Suramin is an effective antitumor agent in hormone-refractory prostate cancer, but its mechanism(s) of action is not well understood. However, the properties of suramin as an agent which disrupts growth factor action, and the importance of tumor-stroma interactions in prostate tumor development and in HA synthesis led us to study the effect of suramin on HA synthesis. Suramin inhibited HA synthesis by calf serum-stimulated Swiss 3T3 fibroblasts at clinically relevant concentrations (IC50 = 183 micrograms/mL). Increasing the serum concentration from 10 to 20% did not change the IC50 for HA synthesis, but increased the IC50 for [3H]thymidine incorporation from 206 to 342 micrograms/mL, indicating that the antiproliferative effect of suramin can be dissociated from its effect on HA synthesis. Suramin did not alter the cellular concentrations of the two precursors for HA synthesis (UDP-glucuronic acid and UDP-N-acetylglucosamine) at early time points and did not inhibit the HA synthetase activity of isolated membranes at concentrations up to 800 micrograms/mL.(ABSTRACT TRUNCATED AT 250 WORDS)

Evidence of androgen receptor expression in squamous and adenocarcinoma of the esophagus.

Androgen receptors (AR) are known to stimulate cellular proliferation in certain tumors. We have assessed the androgen receptor status of esophageal carcinoma in surgically resected specimens as well as in established human esophageal carcinoma cells lines. In these initial studies we sought to characterize the frequency of expression of androgen receptors in squamous versus adenocarcinoma and in male versus female patients, and to assess the possible influence of AR expression on survivaL We analyzed androgen receptor expression utilizing immunohistochemistry in adenocarcinoma and squamous cell carcinoma of the esophagus in surgical specimens from 25 patients treated at Johns Hopkins Bayview Medical Center. Tumors in 7 of 21 males (33%) and 1 of 4 females (25%) showed positive androgen receptor staining with the monoclonal body antibody AR 441. There was no suggestion of a difference in expression of AR between males and females. Five of 11 adenocarcinomas (45%) and 3 of 14 squamous carcinomas were positive. Survival was similar in AR+ and AR- patients. Studies with established tissue culture cell lines showed AR expression by RT-PCR, with stronger expression of AR in adenocarcinoma lines than in squamous carcinoma lines. The presence of AR in human esophageal cancer is an impetus for further studies to assess anti-androgen therapy for treatment and or prevention of these tumors.

Malignant hyperthermia-like reaction secondary to ingestion of hops in five dogs.

Five dogs, 4 of which were Greyhounds, suffered adverse effects secondary to the ingestion of spent hops. Mean time to onset of clinical signs was 3 hours, and clinical signs included marked hyperthermia, restlessness, panting, vomiting, signs of abdominal pain, and seizures. Four of the 5 dogs died despite aggressive therapeutic measures, and there was rapid onset of rigor mortis in 3. The overrepresentation of Greyhounds, coupled with the clinical signs, was suggestive of a malignant hyperthermia-like response to the ingestion of hops. It also is possible that hops contain an uncoupler of oxidative phosphorylation.

Cucurbitacin E targets proliferating endothelia.

Tumor vasculature offers a target for drugs directed against proliferating endothelia. We hypothesized that cucurbitacin E (CuE), an actin-disrupting agent, would preferentially inhibit proliferating vs. quiescent endothelia. Log-phase and confluent ECV304 and HUVEC human endothelial cells were exposed to 10-200 nM CuE for 1-96 hr, and toxicity was determined at 2-6 days by sulforhodamine B assay. Confluent ECV cells were scraped by Q-tip to allow proliferation at the margin and exposed to CuE at LD50, and the actin cytoskeleton was stained by rhodamine-phalloidin. Cell cycle analysis was performed by flow cytometry. Log-phase cells were significantly inhibited compared to confluent. LD50's of log-phase cells were much less than for confluent cells (12 vs. 170 nM, ECV; 13 vs 76 nM, HUVEC). Persistent growth inhibition required 24-96 hr exposure to LD50. Followed less than 6 hr exposure. CuE induced F-actin to accumulate centrally in clumps in newly proliferating cells; actin appeared normal in quiescent cells. CuE treated endothelial cells were not blocked at cytokinesis. CuE preferentially potently inhibits proliferating human endothelia compared to quiescent cells in vitro. CuE induces depolymerization of F-actin in proliferating cells. At low concentrations, cucurbitacin E may potently inhibit vascular proliferation by disrupting actin.

Jasplakinolide, a cytotoxic natural product, induces actin polymerization and competitively inhibits the binding of phalloidin to F-actin.

Jasplakinolide, a naturally occurring cyclic peptide from the marine sponge, Jaspis johnstoni, has both fungicidal and antiproliferative activity. We now report that this peptide is a potent inducer of actin polymerization in vitro. The peptide has a much greater effect on Mg(2+)-actin than on Ca(2+)-actin. Competitive binding studies using rhodamine-phalloidin suggest that jasplakinolide binds to F-actin competitively with phalloidin with a dissociation constant of approximately 15 nM. This compares favorably to the previously reported IC50 of 35 nM for the antiproliferative effect of jasplakinolide on PC3 prostate carcinoma cells. The binding curve suggests that nearest neighbor positive cooperativity influences the binding of jasplakinolide (and perhaps also phalloidin) to F-actin. These results imply that jasplakinolide may exert its cytotoxic effect in vivo by inducing actin polymerization and/or stabilizing pre-existing actin filaments.

Role of insulin-like growth factor-I in esophageal mucosal healing processes.

The current study examines the stimulation of healing processes and signal transduction that is mediated by insulin-like growth factor-I (IGF-I) in an ex vivo esophageal explant model when using tyrphostin inhibition of receptor tyrosine kinase. The explant model provides a 3-dimensional cellular environment of multiple interacting cells isolated from the neural and vascular supply. Tyrphostins previously characterized for their interactions with epithelial growth factor (EGF) receptor-associated protein tyrosine kinases were tested for their potential effects on IGF-I growth-promoting activity. Explants of rabbit esophagus were incubated in media with or without IGF-I. Tyrphostins 1, 23, 25, 46, 47, 51, and 63 were added. We assessed DNA synthesis by tritiated thymidine incorporation. Outgrowth from the edge of the primary mucosa of the explant was evaluated on histologic sections, and cell proliferation was confirmed with immunohistology. IGF-I increased the incorporation of tritiated thymidine by 50% to 100%. Tyrphostins 23 and 47 eliminated IGF-I-induced proliferation in a dose-dependent manner. Tyrphostins 25, 46, and 51--along with negative controls tyrphostin 1 and tyrphostin 63--were ineffective, inasmuch as IGF-I-stimulated growth remained unchanged in their presence. Proliferative activity demonstrated by PCNA staining was confined to new mucosa. Two of 5 tyrphostins originally developed as EGF receptor protein tyrosine kinase inhibitors were effective in inhibiting the actions of exogenous IGF-I. We conclude that IGF-I stimulation may play an important role in repair processes in the esophagus and that this stimulation can be inhibited by using specific tyrphostins.

Vanadate inhibition of hyaluronic acid synthesis in Swiss 3T3 fibroblasts.

Vanadate (50 microM) increases the incorporation of 3H from D-[1,6-3H]glucosamine into hyaluronic acid (HA) in serum stimulated Swiss 3T3 fibroblasts. This increase is not a result of increased HA synthesis, but is due to a significant increase in the specific activity of the cellular UDP-N-acetyl-glucosamine. When the 3H incorporation is corrected for the change in precursor specific activity, vanadate is shown to inhibit the synthesis of HA, a result that is confirmed by measuring the HA with a competitive binding assay. The inhibition of HA synthesis by vanadate is also shown to be dependent on the media used in the experiment, with a substantial increase in the inhibition seen when the media contains the pH indicator, phenol red. The HA synthetase activity of isolated membranes is not inhibited by vanadate at concentrations up to 500 microM.

Antiproliferative effect in vitro and antitumor activity in vivo of brefeldin A.

PURPOSE: An empiric in vitro screen of human tumor cell lines found brefeldin A inhibited the growth of immortalized human cell lines, with particular sensitivity to brefeldin in a series of immortalized melanoma cell lines and nonimmortalized prostate carcinoma explants. Brefeldin A alters the morphology and function of the Golgi apparatus, endosomal, and trans-Golgi compartments in different cell types. The studies presented here sought to obtain evidence of in vivo antitumor activity by brefeldin A. METHODS: Antiproliferative activity was studied in prostate carcinoma cells in vitro using cell counts, protein, and viable stains. Activity was also studied in vivo against subcutaneous and subrenal capsule melanoma models. RESULTS: Protracted exposures in vitro (between 24 and 72 hours) are necessary to cause persistent growth inhibition of immortalized PC3 prostate carcinoma cells. In human melanoma athymic mouse xenografts, brefeldin A showed antitumor activity in vivo when given 16 to 64 mg/kg/injection intraperitoneally q 7 h x 2, daily for 5 days. Activity was also observed in the intraperitoneal LOX IMVI (65%-100% increase in life span, with 17%-50% day 60 survivors); early-stage subcutaneous LOX IMVI and SK-MEL-5 (86%-100% growth inhibition), and subrenal capsule SK-MEL-5 and M19-MEL models. CONCLUSIONS: Brefeldin A possesses noteworthy antitumor activity in vivo and antiproliferative effects in vitro in certain cell types. Strategies to allow protracted exposure of tumor cells to brefeldin A while preserving a therapeutic index are needed to assess the clinical potential of brefeldin A.

Toxicity of cordycepin in combination with the adenosine deaminase inhibitor 2'-deoxycoformycin in beagle dogs.

For 3 consecutive days, the nucleoside cordycepin (3'-deoxyadenosine) was administered as 1-hr iv infusions (0, 1, 4, 8, 10, or 20 mg/kg/day) to dogs. These doses were given 1 hr after a bolus iv injection (0.25 mg/kg/day) of 2'-deoxycoformycin (dCF), a potent inhibitor of adenosine deaminase. The hypothesis was that dCF would affect the toxicity of cordycepin. Plasma adenosine deaminase activity was strongly inhibited during the dose period and for 5 days following the final dose of dCF. Dogs given cordycepin alone showed no drug-related toxicities. In dogs given only dCF, drug-related toxicity to lymphoid tissue (lymphopenia and thymus lymphoid depletion), thrombocytopenia, and decreases in food consumption were observed. Cordycepin in combination with dCF produced symptoms associated with severe gastrointestinal toxicity (decreased body weights, emesis, diarrhea, decreased food consumption, and necrosis of the gastrointestinal tract) and bone marrow toxicity (lymphopenia, thrombocytopenia, and depletion of hematopoietic cells). The gastrointestinal tract and bone marrow were sites associated with dose-limiting toxicities. In surviving dogs, most of the effects were reversible by Day 30. The maximum tolerated dose of cordycepin administered in combination with dCF was 8 mg/kg/day (160 mg/m2/day) given daily for 3 days.