John B. Sharefkin

Fluid flow decreases preproendothelin mRNA levels and suppresses endothelin-1 peptide release in cultured human endothelial cells

Endothelin-1, a 21-amino acid peptide secreted by endothelial cells, has constrictor and mitogenic activity for vascular smooth muscle cells, and its mitogenic activity is synergistic with that of platelet-derived growth factor. Endothelial cell-derived endothelin-1 might therefore contribute to intimal hyperplasia in reendothelialized segments of vascular grafts or of endarterectomy and angioplasty sites. Because intimal hyperplasia occurs most often at sites with disordered flow patterns and lower fluid shear stress, we tested the effects of static culture versus high laminar shear stress (25 dyne/cm2) on endothelin-1 precursor (preproendothelin) gene mRNA transcript levels and endothelin-1 peptide release in cultured human endothelial cells. Primary cultures of human umbilical vein endothelial cells were subjected to controlled levels of shear stress in parallel plate flow chambers for 24 hours. To detect preproendothelin mRNA we applied a linked reverse transcriptase-polymerase chain reaction (RT/PCR) to RNA extracted from cultures. Southern blots of RT/PCR reaction products were hybridized with radioactive phosphorous (32P) labeled probes for the amplified preproendothelin complementary deoxyribonucleic acid (cDNA). Detection by RT/PCR of mRNA for glyceraldehyde 3-phosphate dehydrogenase was used to measure a constitutively expressed control signal. Endothelin-1 release into culture medium was measured by radioimmunoassay. Application of 25 dyne/cm2 of shear stress for 24 hours sharply reduced endothelial cell levels of precursor preproendothelin mRNA.(ABSTRACT TRUNCATED AT 250 WORDS)

The cytotoxic effect of surgical glove powder particles on adult human vascular endothelial cell cultures: Implications for clinical uses of tissue culture techniques

Clinical use of autogenous endothelial cell (EC) seeding of vascular prostheses (VP) would require reliable methods for EC harvest for immediate seeding or primary culture in a hospital or operating room setting. Observation of glove powder particles (GPP) in failed primary adult human saphenous vein EC (AHSVEC) cultures led us to study the effect of surgical GPP on cultured AHSVEC. Addition of GPP to the culture medium of growing ASHVEC cultures reduced the cell counts in a dose-dependent fashion; the mean concentration of GPP required to produce a greater than 50% decrease in cell number was 1.5 +/- 0.8 (SD) X 10(4) GPP/ml (N = 10 experiments), equivalent to a mean dose of 36 micrograms glove powder per milliliter. The effect was seen within 24 hr of addition of GPP and was not due to interference with EC attachment and spreading or to changes in medium osmolality, pH, glucose, electrolyte, Ca2+, or Mg2+ content. Instead, the effect appeared to be due to a filterable toxin added during the final rubber-vulcanizing stage of glove manufacture, since pure cornstarch particles and epichlorhydrin-treated pure cornstarch did not prevent culture growth, whereas 0.2 micron filtrates of medium incubated with GPP taken directly from gloves were lethal. We conclude that filterable cytotoxic substances from GPP may be an avoidable cause of failure in EC seeding of VP, and may affect surgical wound healing as well.

Reduced reproductive capacity of freshly harvested endothelial cells in smokers: A possible shortcoming in the success of seeding?

In an attempt to explain the failure of first clinical trials of autologous endothelial seeding in smokers, the initial reproductive capacity of saphenous vein endothelial cells from smokers and nonsmokers was studied by a replicate microwell technique. Endothelial cells were enzymatically harvested from saphenous vein segments of patients with coronary bypasses (21 smokers and 18 nonsmokers). After 15 minutes (group A) and 7 minutes (group B) of collagenase exposure, the endothelial cell harvest from donors who smoked was 41% (p less than 0.02) lower for group A and 30% (p less than 0.2) lower for group B than that from nonsmokers. In analogy, the viable cell yield was 32% (p less than 0.04) and 29% (p less than 0.05) lower for groups A and B, respectively, in cultures from donors who smoked. Daily cell counts over an ensuing 10-day period also revealed a significant difference in the proliferative behavior of endothelial cells from smokers and nonsmokers. Whereas endothelial cells from nonsmokers regularly entered the exponential phase of proliferation on day 4.4 +/- 1.8 (group A) and day 4.6 +/- 1.3 (group B), endothelial cells from smokers reached the logarithmic growth phase either with delay (day 6.8 +/- 2.1, group A) or remained completely quiescent (group B). Lower harvest efficiency and suppressed reproductive capacity of endothelial cells in smokers--on top of an already critically low inoculum in single-staged endothelial cell seeding--might explain the failure of initial clinical trials.

Expression of genes for platelet-derived growth factor in adult human venous endothelium: a possible non-platelet-dependent cause of intimal hyperplasia in vein grafts and perianastomotic areas of vascular prostheses

Neointimal fibromuscular hyperplasia (NFH) in vein grafts and perianastomotic zones of vascular prostheses has been attributed to the effects of platelet-derived growth factor (PDGF) released by platelets interacting with bypass conduits. But inhibition of platelet aggregation often fails to prevent NFH, and recurrent growth of intact, platelet-free endothelium over perianastomotic areas where NFH occurs is inconsistent with the concept of sustained PDGF release from platelets causing NFH progression at late times after surgical procedures. Cultured bovine aortic endothelial cells (ECs) and human umbilical vein ECs have been shown to release a PDGF-like molecule. We report that confluent cultured fourth passage adult human saphenous vein ECs (AHSVECs) grown in the presence of heparin (100 micrograms/ml) and retina-derived growth factor (RDGF) studied by Northern blotting transcribed a messenger ribonucleic acid (mRNA) of 3.9 kb, strongly hybridizing to PDGF B chain probes, and two species of 2.0 and 2.6 kb hybridizing to PDGF A chain probes. Withdrawal of RDGF and heparin from these cultures for 48 hours before mRNA extraction amplified the scanning densitometric mRNA signal per cell by 8.0 +/- 7.6 fold (mean +/- SD) (N = 4 cultures) for B chain mRNA and 5.2 +/- 3.6 fold (N = 3 cultures) for A chain mRNA. In addition, AHSVEC cultures released a PDGF-like substance, because 50% vol/vol AHSVEC-conditioned serum-free medium increased tritiated thymidine uptake elevenfold in PDGF receptor-bearing 3T3 cells whereas an excess (50 micrograms/ml) of nonspecific goat anti-human-PDGF antibody significantly reduced this increase by a mean of 30% to 7.0 +/- 3.4 fold (N = 6 trials, p less than 0.001). Flow cytometry determined AHSVEC cultures to be proliferating with a mean of 6.2% +/- 1.9% (N = 3 culture lines) of ECs in S phase even at confluence when deprived of EC mitogens for 48 hours. Adult human ECs, which proliferate on bypass conduits and host vessels after perioperative injury, may play a role in causing NFH by stimulating proliferation of adjacent smooth muscle cells. Prevention of NFH may require not only antiplatelet agents but also ways to prevent EC release of smooth muscle cell mitogens in response to perioperative EC injury.

Seeding of dacron vascular prostheses with endothelium of aortic origin

Seeding of autologous venous endothelium on Dacron vascular prostheses in dogs results in endothelial coverage of the prosthetic flow surface 4-6 weeks after implantation. Canine aortic endothelium, in contrast, usually fails to completely cover an unseeded prosthesis by pannus ingrowth even over much longer periods. To see if the success of endothelial seeding stems from a difference in the ability of venous and aortic endothelium to grow on prosthetic surfaces, we seeded freshly harvested autologous aortic endothelium on Dacron velour infrarenal aortic prostheses in dogs. Six weeks after surgery these prostheses showed the features reported to be typical of seeded prostheses. Scanning electron micrographs showed a luminal lining of flat polygonal cells without fibrin or adherent formed blood elements, and light microscopy showed an underlying layer containing aligned spindle-shaped cells with elongated nuclei and cell-lined subluminal channels. Control prostheses were covered with fibrin and platelet-rich thrombi everywhere except for limited pannus ingrowth at anastomotic sites. The results suggest that the success of autologous endothelial seeding cannot be ascribed to inherent differences in properties such as mitotic capacity or fibrinolysis between venous and aortic endothelium. The formation of complete endothelial linings by seeding must instead result from a more favorable condition for endothelial cell growth created by the cell harvesting or seeding process itself.

Enzymatic harvesting of adult human saphenous vein endothelial cells: Use of a chemically defined combination of two purified enzymes to attain viable cell yields equal to those attained by crude bacterial collagenase preparations

Seeding vascular prostheses with enzymatically harvested endothelial cells can create endothelial linings that improve small-caliber prosthetic patency. But crude bacterial collagenases used for endothelial harvest contain cytotoxic nonspecific proteases and clostridial cell wall debris which might limit their clinical usefulness. We therefore compared the endothelial cell harvest efficiency of crude bacterial collagenase with that of purified bacterial collagenase alone, purified trypsin alone, and combinations of purified bacterial collagenase and trypsin using concentrations of pure collagenase equal in collagenolytic activity to the crude bacterial collagenase material. The efficiency of harvest from human saphenous vein segments was measured by a microtiter well-growth curve assay of the number of living endothelial cells capable of attachment to fibronectin and subsequent growth obtained per unit area of saphenous vein lumen. Whereas pure collagenase and purified trypsin alone both harvested less than 5% of the baseline endothelial cell density on the veins, a combination of purified collagenase and 0.01% w/v purified trypsin was found to harvest 22% +/- 10% (SD) (n = 8 veins) of the approximately 1.3 x 10(5) endothelial cells/cm2 available on normal saphenous veins. This figure was not statistically different from the harvest efficiency of 19% +/- 10% (N = 4 veins) (p greater than 0.05) obtained by use of 0.1% w/v crude collagenase alone. This result suggests that endothelial harvesting can be done with a defined mixture of pure enzymes which would be clinically preferable to presently used crude extracts of clostridial cultures as a standardized preparation for graft seeding.

Platelet survival and serotonin content after placement of arterial prostheses in dogs: Effects of neointimal coverage and high- and low-dose aspirin

Because prosthetic neointima produces much less prostacyclin (PGI2) than arterial intima and may be more susceptible to cyclooxygenase inhibition, aspirin treatment might enhance surface thrombogenesis. To test this hypothesis, aortic prostheses were placed in eight dogs and measurements of platelet survival and platelet serotonin (5HT) were made under conditions of no treatment and treatment with low-dose (2mg/kg) and high-dose (30 mg/kg) aspirin. These doses equally suppressed platelet function. Measurements were performed preoperatively, 6 to 8 weeks postoperatively (when little neointima was present), and 28 to 32 weeks postoperatively (neointima fully developed). Platelet survival and 5HT levels were markedly reduced 6 to 8 weeks postoperatively and returned to normal at 28 to 32 weeks after implantation. At all times, low-dose aspirin improved platelet survival and this effect was most apparent 6 to 8 weeks postoperatively. Treatment with either aspirin dose decreased platelet 5HT levels at the 28 to 32 week postoperative period but not at other times. At recovery of prostheses, 90% of the luminal surface was covered with endothelialized neointima. Neointimal production of PGI2 was one half to one third that of aortic production. Despite this, low- and high-dose aspirin equally suppressed PGI2 production from both neointima and aorta. Furthermore, aspirin did not increase labeled platelet uptake on neointima. We conclude that (1) aspirin treatment does not render prosthetic neointima thrombogenic and (2) aspirin alters platelet survival and 5HT levels by mechanisms other than inhibition of platelet and neointima cyclooxygenase.

Heparin and dibutyryl cAMP modulate gene expression in stimulated human saphenous vein smooth muscle cells

SummaryIncreased expression of basic fibroblast growth factor (bFGF), platelet-derived growth factor (PDGF) A chain, and tissue plasminogen activator (tPA) by smooth muscle cells (SMC) has been postulated to mediate the progression of intimal hyperplasia. We tested whether heparin would suppress the expression of these genes in stimulated human saphenous vein SMC. Quiescent cultured human saphenous vein SMC were stimulated for 4 h with heat-inactivated fetal bovine serum (10% by vol) in the presence or absence of heparin (1 to 250µg/ml). Heparin (50µg/ml) attenuated the induction by serum of bFGF mRNA, tPA mRNA, and tPA secretion. Nonanticoagulant heparin also attenuated serum induction of bFGF and tPA mRNA levels. To further study the role of second messenger signaling, a more specific mode of SMC stimulation was used with thrombin (3 U/ml) in the presence or absence of dibutyryl cyclic AMP (Bu2-cAMP; 0.5 mM). In contrast to heparin, which had no effect on PDGF expression, Bu2-cAMP decreased the induction by thrombin of PDGF-A chain mRNA levels. In thrombin-stimulated SMC, Bu2-cAMP significantly decreased secretion of PDGF-AA protein. Thrombin, however, caused an increase in bFGF mRNA levels which was potentiated by Bu2-cAMP with associated potentiation by Bu2-cAMP of intracellular bFGF protein levels. The induction of tPA mRNA and tPA secretion by thrombin was sharply blocked by Bu2-cAMP. These results suggest that heparin reduces intimal hyperplasia at least partly via partial inhibition of SMC gene expression.