The selectin family of adhesion molecules mediates attachment and rolling of neutrophils to stimulated endothelial cells. This step of the inflammatory response is a prerequisite to firm attachment and extravasation. We have reported that microspheres coated with sialyl Lewis(x) (sLe(x)) interact specifically and roll over E-selectin and P-selectin substrates (Brunk et al., 1996; Rodgers et al 2000). This paper extends the use of the cell-free system to the study of the interactions between L-selectin and sLe(x) under flow. We find that sLe(x) microspheres specifically interact with and roll on L-selectin substrates. Rolling velocity increases with wall shear stress and decreases with increasing L-selectin density. Rolling velocities are fast, between 25 and 225 microm/s, typical of L-selectin interactions. The variability of rolling velocity, quantified by the variance in rolling velocity, scales linearly with rolling velocity. Rolling flux varies with both wall shear stress and L-selectin site density. At a density of L-selectin of 800 sites/microm(2), the rolling flux of sLe(x) coated microspheres goes through a clear maximum with respect to shear stress at 0.7 dyne/cm(2). This behavior, in which the maintenance and promotion of rolling interactions on selectins requires shear stress above a threshold value, is known as the shear threshold effect. We found that the magnitude of the effect is greatest at an L-selectin density of 800 sites/microm(2) and gradually diminishes with increasing L-selectin site density. Our study is the first to reveal the shear threshold effect with a cell free system and the first to show the dependence of the shear threshold effect on L-selectin site density in a reconstituted system. Our ability to recreate the shear threshold effect in a cell-free system strongly suggests the origin of the effect is in the physical chemistry of L-selectin interaction with its ligand, and largely eliminates cellular features such as deformability or topography as its cause.

Recently, we showed a correlation between the maturity of hematopoietic stem and progenitor cells during development and rolling efficiency on selectins. These findings motivated us to explore a novel separation that exploits differences in selectin-mediated rolling adhesion between populations of cells. We extend the use of a previously developed cell-free system to study the separation of populations of sialyl Lewis x (sLe(x))-coated microspheres designed to roll with different average velocities on L-selectin chimeric substrates under well-defined flow. Results show that a separation that exploits differences in average rolling velocities between cell or microsphere populations is attainable. Excellent recovery and purity values for the slower rolling, or more desirable, populations are obtained and can be estimated from rolling velocity measurements. We also assess the feasibility of a selectin-mediated separation of adult bone marrow cell populations using previously obtained rolling velocity and rolling flux data for CD34+ and CD34- adult bone marrow cells on L-selectin substrates. We believe that a cell separation mediated by differential rolling adhesion can be used to enrich populations of hematopoietic stem and progenitor cells from an adult bone marrow cell preparation and that this method possesses several major advantages over existing antibody-mediated cell-affinity chromatography technologies.

Current understanding of the adhesion molecules and mechanisms regulating hematopoietic stem and progenitor cell (HSPC) homing to the bone marrow is limited. In contrast, the process by which mature leukocytes are able to home to and extravasate out of blood vessels at sites of inflammation has been well characterized and invites comparison. We studied the interaction of human HSPC from adult bone marrow (ABM) and fetal liver (FL) with E-, P-, and L-selectin immobilized in a flow chamber. CD34+ HSPC from both ABM and FL rolled avidly on E-, P-, and L-selectin across a range of physiologic shear rates, indicating the presence of ligands for all three selectins on HSPC. Results indicate that CD34+ ABM and FL cells roll more efficiently (to a greater extent and more slowly) than more differentiated CD34− cells, especially on P- and L-selectin. In a similar fashion, increased rolling efficiency was seen with CD34+CD38− ABM cells when compared with committed progenitor cells of the CD34+CD38+ phenotype. Rolling of CD34+ ABM cells on P-selectin could be partially inhibited by monoclonal antibody (mAb) against PSGL-1, and was not inhibited by a mAb against CD34, suggesting that HSPC have unique carbohydrate repertoires that facilitate selectin-mediated rolling. Our results provide direct evidence of selectin ligands on HSPC under physiologic flow conditions and are the first to show a correlation between the maturity of HSPC during development and rolling efficiency on selectins, suggesting a mechanism by which HSPC subsets may differentially home to the extravascular spaces of the bone marrow.