Plasma thrombospondin levels in patients with gynecologic malignancies.

BACKGROUND: Thrombospondin is a high molecular weight glycoprotein, originally described as a secretion product of platelets, that functions as an adhesive protein in cell-cell and cell-substratum interactions. It promotes metastases in the murine model. Plasma thrombospondin has been shown to be elevated in patients with disseminated breast, lung, and gastrointestinal malignancies. METHODS: Blood samples were collected by venipuncture into cubes containing ethylenediamine tetraacetic acid as anticoagulant. They were placed on ice immediately and centrifuged under refrigerated conditions. Plasma was removed and frozen until thrombospondin was quantitated by a competitive enzyme-linked immunosorbent assay. Wilcoxon's two-sample rank-sum test was used to evaluate differences between the patient and control groups. RESULTS: The median plasma thrombospondin level was significantly higher in the patient group compared with the control group, and it was directly correlated with stage of disease. There was no correlation between platelet count and thrombospondin level. CONCLUSIONS: Tumor-synthesized thrombospondin could explain the elevated levels in the patent group and also the observation of the correlation between the thrombospondin level and tumor burden. Its function as an adhesive protein may allow it to act as the mediator of metastases. thrombospondin may promote or mediate the metastatic process through its function in cell adhesion.

Thrombospondin levels in patients with malignancy.

Thrombospondin (TSP), a large glycoprotein present in platelets, and various normal and tumor tissues, has recently been shown to promote cell adhesion and platelet aggregation. Most importantly because TSP has been shown to promote metastasis of melanoma tumor cells to the lung in a murine model (1) and since thromboembolic events commonly occur in patients afflicted with metastatic tumors, we explored the role of TSP in human cancer by measuring TSP blood levels in patients with various malignant neoplasms. Blood TSP levels were measured by indirect enzyme-linked immunoadsorbent assay (ELISA) from 20 control subjects, 22 patients with gastrointestinal (GI) cancer, 18 patients with breast cancer, and 17 patients with lung cancer. Control subjects consisted both of healthy subjects and acutely ill patients with no malignancies. TSP levels of both healthy and acutely ill controls were found to range between 245-440 ng/ml with a mean of 365 ng/ml. In contrast, elevated levels of TSP greater than the mean value of 400 ng/ml for controls ranging between 590-3,650 ng/ml were found in 20/22 (91%) patients with GI malignancies, 13/18 (72%) patients with breast cancer, and 15/17 (88%) with lung cancer. Mean TSP levels of GI, breast, and lung cancer patients were 3, 2, and 3 fold greater than controls, respectively. Increased blood TSP levels in patients were not due to increased levels of platelets since both control and patient groups had platelet counts within the normal range. These results suggest that TSP may play a role in tumor cell metastasis in man and could serve as a blood marker for metastasis.

Platelet-specific alpha-granule proteins and thrombospondin in bronchoalveolar lavage in the adult respiratory distress syndrome.

Levels of platelet-specific alpha-granule proteins, PF, BTG, and TSP were measured in BAL fluids of patients with the ARDS, ILD, and normal healthy subjects, comprising two separate cohorts. In both groups BAL showed elevated levels of BTG and thrombospondin in ARDS patients. Low levels of PF4 were found in BAL and did not differ between ARDS and control patients. The BTG:PF4 ratio was 2:1 or greater in BAL of ARDS patients and of control subjects with other lung diseases, suggesting in vivo release. In ARDS patients, the ratio of TSP to BTG exceeded that usually found in plasma. In ARDS patients in group 2, BAL levels of TSP, BTG, and total protein correlated strongly with the composite injury scores that were used to quantitate their degree of lung injury. Elevated levels of platelet-derived proteins, which modulate chemotaxis of inflammatory cells and promote connective tissue reorganization, occur in the alveolar compartment of ARDS and ILD patients but are usually undetectable in BAL of healthy control subjects. Levels in these patients in BAL fluid are nonspecific indices of the severity of lung injury in patients with ARDS.

Temporary inhibition of platelet function with iloprost (ZK36374) preserves canine platelets during extracorporeal membrane oxygenation.

Contact between blood and artificial surfaces results in extensive quantitative and qualitative alterations in platelet function. We evaluated the efficacy of a brief infusion of iloprost (ZK36374), a stable analog of prostacyclin, in preventing these platelet changes during extracorporeal membrane oxygenation. Twelve nonsplenectomized male mongrel dogs (23 to 30 kg) were randomized to treatment (n = 6) and control (n = 6) groups. The treatment animals received an infusion of iloprost at a rate of 150 ng/kg/min with the infusion being terminated 30 minutes after the initiation of extracorporeal membrane oxygenation, despite the fact that all animals were maintained on extracorporeal membrane oxygenation for 3 hours. In the control group, platelet counts dropped to 54% +/- 8.9% (mean +/- standard error of the mean) of initial levels at 30 minutes of extracorporeal membrane oxygenation and gradually rose to 87.2% +/- 6.7% at 3 hours. In contrast, the platelet counts of the iloprost-treated dogs remained stable throughout extracorporeal membrane oxygenation at 98.3% +/- 4.2% of initial counts. Platelet reactivity toward adenosine diphosphate revealed a significant and permanent loss of platelet function in the control group (37.0% +/- 2.1% inhibition). In contrast, the iloprost group demonstrated significant inhibition of platelet reactivity (79.2% +/- 8.3%) during the iloprost infusion but a return to normal function (4.2% +/- 6.7% inhibition) after cessation of drug infusion which persisted throughout extracorporeal membrane oxygenation. Plasma levels of the platelet-specific protein thrombospondin rose progressively from 918 +/- 89 ng/ml to 1465 +/- 239 ng/ml (delta 548 +/- 179 ng/ml) at 30 minutes of extracorporeal membrane oxygenation, which indicates extensive release of platelet granule contents (p less than 0.05). In contrast, plasma thrombospondin levels in the iloprost group demonstrated no additional rise after cessation of the iloprost infusion. In conclusion, iloprost effectively preserves platelet number and function during extracorporeal circulation. The fact that its salutary effects outlast its presence in plasma suggests that prevention of initial platelet-synthetic surface interactions permits the appearance of reduced surface affinity for platelets and, thus, reduced synthetic surface thrombogenicity.

Radioimmunoassay of human platelet thrombospondin: different patterns of thrombospondin and beta-thromboglobulin antigen secretion and clearance from the circulation.

A method for radioimmunoassay of human thrombospondin was developed. Monospecific precipitating anti-human thrombospondin antibody was raised in rabbits after injection of thrombospondin purified by fibrinogen-agarose chromatography and preparative sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The linear portion of the thrombospondin radioimmunoassay standard curve was 0.5 to 20 ng/ml. Normal platelets and platelet-poor plasma contained 28,900 +/- 14,500 ng thrombospondin per 10(9) platelets and 60.6 +/- 10.7 ng/ml (mean +/- SD), respectively. Using radioimmunoassays for beta-thromboglobulin and thrombospondin antigens, we compared platelet location and secretion of these proteins. Both antigens shared similar distributions in platelet subcellular fractions with the largest amount localized to platelet alpha-granules. With thrombin (0.25 U/ml) as a platelet agonist, 62.4% and 19.5% of total beta-thromboglobulin and thrombospondin, respectively, were secreted from suspensions of washed human platelets. Because only 20% of the total platelet thrombospondin was secreted, further studies were initiated to determine whether the remaining thrombospondin became localized on the activated platelets membrane. 125I-Fab antithrombospondin specifically bound to activated platelets but not to unstimulated platelets. In contrast, 125I-Fab anti-beta-thromboglobulin did not bind to activated platelets. Plasma clearance of human beta-thromboglobulin (half-life fast 7.6 minutes, slow 56.6 minutes) and of human thrombospondin (half-life fast 29.9 minutes, slow 190 minutes) followed a biphasic exponential curve. In conclusion, both beta-thromboglobulin and thrombospondin are located in platelet alpha-granules, but they show a different pattern of secretion and expression on the platelet membrane and plasma clearance.

The interaction of human platelet thrombospondin with fibrinogen. Thrombospondin purification and specificity of interaction.

Human platelet thrombospondin (TSP) was purified to homogeneity by chromatography on fibrinogen coupled to cyanogen bromide-activated Sepharose. The yield of TSP was 1.3 mg or approximately 22% of that present in platelet-rich plasma as determined by radioimmunoassay. It analyzed on discontinuous sodium dodecyl sulfate gels as a single band having apparent molecular weights of 180,000 and greater than 400,000 under reducing and nonreducing conditions, respectively. Amino acid analysis gave results similar to previously published values. Antibodies raised in rabbits were monospecific as evaluated by radioimmunoassay. In double immunodiffusion tests, these antibodies gave one line of identity against TSP purified by this procedure and TSP purified by published procedures, confirming the identity of the material isolated. The protein possesses no lectin-like activity. The specificity of the TSP-fibrinogen interaction was investigated. TSP binding to fibrinogen-Sepharose occurred in the presence of EDTA, indicating that calcium and magnesium ions are not required for interaction of TSP with fibrinogen. The binding of TSP to fibrinogen-Sepharose was quantitatively blocked by pretreatment with an antibody to the cyanogen bromide cleavage fragment composed of residues 241-476 of the carboxyl-terminal end of the alpha chain of fibrinogen. Antibodies against the D and E domains of fibrinogen had no effect on the binding. Excess fibrinogen (30 mg/ml) added to platelet extract quantitatively inhibited binding of TSP to fibrinogen-Sepharose. TSP preferentially bound to uncross-linked fibrin, suggesting that the TSP-fibrinogen binding site is unavailable in cross-linked fibrin. These results indicate that TSP binds specifically to immobilized fibrinogen or uncross-linked fibrin through determinants present in the carboxyl-terminal portion of the alpha chain and that these interactions do not require calcium or magnesium ions.