HMEC-1: establishment of an immortalized human microvascular endothelial cell line.

The study of human microvascular endothelial cells has been limited, because these cells are difficult to isolate in pure culture, are fastidious in their in vitro growth requirements, and have a very limited lifespan. In order to overcome these difficulties, we have transfected human dermal microvascular endothelial cells (HMEC) with a PBR-322-based plasmid containing the coding region for the simian virus 40 A gene product, large T antigen, and succeeded in immortalizing them. These cells, termed CDC/EU.HMEC-1 (HMEC-1), have been passaged 95 times to date and show no signs of senescence, whereas normal microvascular endothelial cells undergo senescence at passages 8-10. HMEC-1 exhibit typical cobblestone morphology when grown in monolayer culture, express and secrete von Willebrand's Factor, take up acteylated low-density lipoprotein, and rapidly form tubes when cultured on matrigel. HMEC-1 grow to densities three to seven times higher than microvascular endothelial cells and require much less stringent growth medium. HMEC-1 will grow in the absence of human serum, whereas microvascular endothelial cells require culture medium supplemented with 30% human serum. These cells express other cell-surface molecules typically associated with endothelial cells, including CD31 and CD36 and epitopes identified by monoclonal antibodies EN4 and PAL-E. They also express the cell adhesion molecules ICAM-1 and CD44 and following stimulation with interferon-gamma express major histocompatibility complex class II antigens. HMEC-1 specifically bind lymphocytes in cell adhesion assays. Thus HMEC-1 is the first immortalized human microvascular endothelial cell line that retains the morphologic, phenotypic, and functional characteristics of normal human microvascular endothelial cells.

Possibilities of vaccine manufacture in human diploid cell strains with a serum replacement factor.

Cell lines MDCK (canine kidney), BGM (Buffalo green monkey kidney) and human embryonic lung fibroblast will support viral growth efficiently in medium without serum. Both MRC-5 and WI-38 cell strains have been approved by the Food and Drug Administration for manufacturing viral vaccines against cytomegalovirus and varicella-zoster virus. In this study we examine these two cell lines and viruses for their ability to grow in medium containing a serum replacement. The serum substitute used is LPSR-1 (low protein serum replacement). Using LPSR-1 in defined medium, we demonstrate multipassage cell growth and viral cultivation and replication equivalent to those obtained in medium containing fetal bovine serum (FBS). Viral growth after complete elimination of FBS varies and depends on cell line and virus. Serum substitutes can eliminate FBS in the viral growth phase of vaccine production and reduce costs.

NCI-H292 as an alternative cell line for the isolation and propagation of the human paramyxoviruses.

Primary rhesus monkey kidney (MK) cells have long been the cells of choice for isolation and propagation of the human paramyxoviruses (parainfluenza 1, 2, 3, 4A, 4B, and mumps). However, problems with the supply and cost of MK cells and the presence of endogenous viruses, including herpes B virus and SV-5, necessitated a search for an alternative cell line. Continuous cell cultures of human origin (L132, A-549, HuT-292, HEK, G-293, G-401, A-498, A-704, CAKI-1, RD) and simian origin (LLC-MK2, BSC-1, MA-104, Vero) were evaluated for their capacity to support the growth of the human paramyxoviruses, as followed by cytopathic effect, hemadsorption, hemagglutination, and EIA. NCI-H292 (HuT-292) human lung mucoepidermoid carcinoma cells (ATCC # CRL-1848) proved to be the most sensitive line for cultivating all serotypes and strains of the paramyxoviruses. These cells were also shown to be a suitable substitute for MK in primary isolation of paramyxoviruses from clinical specimens. RPMI-1640 with 1.5 micrograms/ml trypsin was the preferred maintenance medium; alternatively, Eagle's MEM supplemented with 1.5 micrograms/ml trypsin and 0.1% ITS was satisfactory. NCI-H292 cells are a continuous line with excellent growth characteristics, although the genetic polyploidy of the cells may limit the number of passages of usable cells.

Human endothelial cells as an alternative to DH82 cells for isolation of Ehrlichia chaffeensis, E. canis, and Rickettsia rickettsii.

Ehrlichia chaffeensis, etiologic agent of human ehrlichiosis, and Rickettsia rickettsii, etiologic agent of Rocky Mountain spotted fever (RMSF), are both tick-borne agents that cause nonspecific symptoms that may be indistinguishable from each other early in the course of infection. E. canis is a canine pathogen closely related to E. chaffeensis and was initially suspected of being the causative agent of human ehrlichiosis. If a febrile illness is reported, after tick exposure, neither ehrlichiosis nor RMSF can be immediately ruled out. When attempts are made to isolate the agent from blood, a very limited amount of blood is often available; we, therefore, sought a tissue-culture cell line that would support the growth of both R. rickettsii and E. chaffeensis. A newly established human microvascular endothelial immortal cell line (CDC/EU.HMEC-1) was evaluated for supporting the growth of both agents. Our results demonstrate that HMEC-1 supports the growth of R. rickettsii, E. chaffeensis, and E. canis and may be a useful tool for the isolation of these agents.

Intracellular growth of Afipia felis, a putative etiologic agent of cat scratch disease.

The organism Afipia felis, which is though to be an etiologic agent of cat scratch disease, is a gram-negative rod that is clearly seen in infected tissue but is very difficult to isolate from clinical specimens; there has been only one report to date of the successful isolation and maintenance of the bacterium on artificial medium. We have found that A. felis will attach, invade via phagocytosis, and multiply intracellularly within the phagosomes of primary human monocytes and HeLa cells. Once in the cell, the bacterium appears to change morphologically, becoming longer and more pleomorphic, and loses its ability to grow on an artificial medium. Unique proteins have been identified in both the intra- and extracellular variants of A. felis. Convalescent-phase sera from patients with cat scratch disease react poorly with intracellular and extracellular bacteria, suggesting a poor humoral response. The tissue culture protocol presented has been used to isolate 14 new strains of A. felis and has for the first time permitted study of the pathogenesis of this unique organism.

Human microvascular endothelial cell toxicity caused by Brazilian purpuric fever-associated strains of Haemophilus influenzae biogroup aegyptius.

An in vitro cytotoxicity model that uses an immortalized human microvascular endothelial cell line (HMEC-1) differentiates Brazilian purpuric fever (BPF)-associated Haemophilus influenzae biogroup aegyptius (HAE) strains from non-BPF-associated HAE strains. Toxic strains produced a characteristic HMEC-1 phenotype at an MOI of < 1 bacterium/1000 tissue culture cells (TCC). Nontoxic strains required MOIs of > 1000 bacteria/TCC to produce an observable effect. The cytotoxic phenotype was characterized by the presence of large clumps of HMEC-1 cells, which detached from the monolayer within 48 h of inoculation by HAE cells. The cytotoxic phenotype was observed with 100% of BPF-associated HAE (40/40) and 14% of non-BPF-associated HAE (8/57; P < .001). The ability to study a BPF-associated phenotype in vitro using human microvascular cells should enhance our knowledge of BPF pathogenesis.

Tumor necrosis factor-alpha downregulates protein S secretion in human microvascular and umbilical vein endothelial cells but not in the HepG-2 hepatoma cell line.

Protein S deficiency, which is associated with thrombosis, can either be inherited or acquired. Recently, we reported that a decrease in free protein S was observed in 19 of 25 persons with HIV/AIDS. The proinflammatory cytokine, tumor necrosis factor-alpha (TNF-alpha), has been reported to be elevated in human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS) patients and has been shown to induce a procoagulant state on the surface of endothelial cells. We report here that recombinant TNF-alpha (rTNF-alpha) downregulated protein S synthesis in the SV-40T transfected human microvascular endothelial cell line (HMEC-1) model system by approximately 70% and in primary human umbilical vein and dermal microvascular endothelial cell cultures by approximately 50%. Using the HMEC-1 model, Northern blot analysis showed a decrease in protein S RNA at 24 hours that was corroborated by Western blot analysis and enzyme-linked immunosorbent assay (ELISA) quantification. Evidence supporting the specificity of the TNF-alpha effect included the following: (1) TNF-alpha down-regulation of protein S was completely blocked by TNF neutralizing antibody; (2) the effect was transient, and protein S was restored to near normal levels after TNF was removed from cell cultures; (3) an antibody directed to the TNF RI (55-kD receptor) was shown to mimic the action of TNF-alpha on HMEC-1 cells; and (4) other proinflammatory cytokines, interleukin (IL)-1, IL-6, and TGF-beta, had no effect on protein S secretion. However, TNF-alpha showed no regulatory control over protein S synthesis in the human hepatocellular carcinoma cell line HepG-2. We suggest that TNF-alpha downregulation of protein S may be a mechanism for localized procoagulant activity and thrombosis recently reported in some AIDS patients with associated protein S deficiency.