Grabowski Lab

Research topics include: endothelial cell function; imaging platelet thrombosis; blood rheology and thrombosis; hemolytic uremic syndrome


Eric F. Grabowski, MD, ScD
Associate Professor of Pediatrics
Harvard Medical School

Program in pediatric Hemostatsis and Thrombosis, Massachusetts General Hospital for Children

Cardiovascular Thrombosis Laboratory, Massachusetts General Hospital for Children

Research Summary

Substantial evidence suggests that, in vivo, activation of endothelium may induce initial platelet adhesion to endothelium via upregulation of endothelial cell molecules adhesive for platelets (aVß3, ICAM-1, and/or P-selectin), accompanied and followed by tissue-factor (TF)-driven mural platelet thrombosis in arteries or arterioles. We therefore have studied these phenomena under controlled conditions in vitro using a chamber allowing the exposure to flowing blood of monolayers of human glomerular endothelial cells (HGECs) activated with 20 ng/ml TNFa (hrs 0 to 22) and 10 pM Shiga toxin-1 (STX-1; hrs 18 to 22), in simulation of vascular conditions during the childhood epidemic hemolytic uremic syndrome (HUS), an important specific disease entity. Shear rates used are 270 and 650 sec-1, which are characteristic of those found in glomerular arterioles. The chamber allows for real-time digital videomicroscopy of platelets interacting with the monolayers. Our experiments show that HGECs activated with TNFa + STX-1 express 2.7-fold (or more) greater functional TF than HGECs activated with TNFa alone. In further studies, activation with the combination leads to a modest upregulation of aVß3 and a marked upregulation of ICAM-1 on intact monolayers. Flow cytometry of cell suspensions confirms these increases as 1.1- and 60-fold, respectively. In additional experiments, whole blood was incubated for 10 min with both 1.75 µg/ml of the Tab (anti-CD41) monoclonal antibody directed against platelet GPIIb and 1.0 µg/ml of an ALEXA 555-conjugated rabbit anti-mouse second antibody. While only small numbers of single platelets adhere to unactivated monolayers, platelets adhere in strings tethered at the upstream ends and as small aggregates on HGECs activated with TNFa+ STX-1. These strings and aggregates are virtually eliminated following pre-incubation of activated monolayers with either 100 nM of a monoclonal antibody directed against human TF, or 60 nM of a novel antithrombotic agent: active site-inactivated recombinant factor VIIa (irFVIIa, courtesy of Dr. Ulla Hedner, Novo Nordisk Pharmaceuticals). With the latter, in particular, the percent image pixels positive for platelets in four experiments is reduced from 16.1 4.16 to 1.8 0.54 (P<0.01). At the same time, indirect immunofluorescence of monolayers exposed to flowing culture medium shows the presence of ultra large vWF strings. Despite appreciable HGEC detachment induced by TNFa, platelet string and aggregate formation is dependent on the presence of HGECs:

  • the degree of detachment is the same for activation with TNFa + STX-1 as for TNFa alone
  • complete detachment of HGECs with EDTA leads to uniform coverage of the subendothelium with single platelets without strings or aggregates.

These studies demonstrate the potential of therapeutic interventions in HUS directed at blocking HGEC-associated and/or blood-borne TF.

The above model for childhood HUS is also a paradigm for arterial thrombosis on the intact vessel wall more generally. Situations of interest include arterial thrombosis in the antiphospholipid antibody syndrome, and thrombosis in inflammatory states associated with malignancy pregnancy, and deep venous thrombosis.

Group Members

Eric F. Grabowski MD
Principal InvestigatorGroup Members


Back to Top