GMI-1070, a novel pan-selectin antagonist, reverses acute vascular occlusions in sickle cell mice

Abstract

Leukocyte adhesion in the microvasculature influences blood rheology and plays a key role in vaso-occlusive manifestations of sickle cell disease. Notably, polymorphonuclear neutrophils (PMNs) can capture circulating sickle red blood cells (sRBCs) in inflamed venules, leading to critical reduction in blood flow and vaso-occlusion. Recent studies have suggested that E-selectin expression by endothelial cells plays a key role by sending activating signals that lead to the activation of Mac-1 at the leading edge of PMNs, thereby allowing RBC capture. Thus, the inhibition of E-selectin may represent a valuable target in this disease. Here, we have tested the biologic properties of a novel synthetic pan-selectin inhibitor, GMI-1070, with in vitro assays and in a humanized model of sickle cell vaso-occlusion analyzed by intravital microscopy. We have found that GMI-1070 predominantly inhibited E-selectin-mediated adhesion and dramatically inhibited sRBC-leukocyte interactions, leading to improved microcirculatory blood flow and improved survival. These results suggest that GMI-1070 may represent a valuable novel therapeutic intervention for acute sickle cell crises that should be further evaluated in a clinical trial.

Figures

Figure 1

Schematic representation of intravital microscopy protocols. (A) In protocol 1, after anesthesia and surgical preparations, SCD mice were injected intraperitoneally (i.p.) with murine TNF-α (time 0) and then immediately with GMI-1070, or an antibody mixture containing anti–P- and –E-selectin monoclonal antibodies, or PBS through an intracarotid artery catheter (i.c.). To ensure activities of the selectin antagonist during the entire course of studies, SCD mice received a second dose of antagonists or vehicle controls 70 minutes after the first injection (T70). Images of the cremasteric venules under intravital microscopy were recorded between the time points of 90 and 150 minutes (T90 → T150). During filming, the hemodynamic parameters, including centerline velocity, venular diameter, and shear rate, were measured. (B) Protocol 2 was designed to assess the therapeutic effects of GMI-1070 on SCD mice with ongoing acute veno-occlusive crisis primed by TNF-α. SCD mice were infused with GMI-1070 at 110 minutes after animals were challenged with TNF-α (time 0). Images were recorded in the 2-hour interval between 120 and 240 minutes (T120 → T240) after the administration of GMI-1070.

Figure 2

Chemical structure of GMI-1070. GMI-1070 is a glycomimetic pan-selectin inhibitor that was rationally designed on the basis of the bioactive conformation of sLex in the binding site of E-selectin as determined empirically by nuclear magnetic resonance methods. Incorporation of a sulfate-binding domain is important for interactions with P- and L-selectins.

Figure 3

GMI-1070 inhibits selectin-mediated rolling. Studies in a parallel plate flow chamber showed a dose response of GMI-1070 to inhibit selectin-mediated rolling of isolated human neutrophils on monolayers of HUVECs activated to express E-selectin or P-selectin. Inhibition was greater on E-selectin–expressing HUVECs than on P-selectin–expressing HUVECs. RI indicates Rolling Index.

Figure 4

GMI-1070 alters leukocyte recruitment. (A) The rolling flux fraction was increased by nearly 2-fold in SCD mice treated with GMI-1070, whereas it was completely inhibited when both endothelial selectins were blocked by antibody administration. (B) The average number of leukocytes adherent to endothelium was significantly reduced in SCD mice treated with GMI-1070 but completely inhibited by antiselectin antibodies. (C) Representative images of venules from sickle mice treated with PBS, GMI-1070, or anti–P- and –E-selectin antibodies. Each still frame was taken at the 30-minute time point after TNF-α injection. Both small molecule selectin antagonists and antiselectin antibodies significantly reduced the number of adherent leukocytes (circles) and RBCs interacting with adherent leukocytes (black arrows; see data in Figure 5). The small white arrow indicates an RBC interaction with the endothelium, and the large white arrow shows the direction of blood flow.

Figure 5

Leukocyte rolling velocity histograms. Leukocyte rolling velocities in (A) PBS-treated control SCD mice and (B) GMI-1070–treated SCD mice. The arrowheads indicate the means. (C) Overlay of histograms of panels A and B shows the increased average leukocyte velocities in GMI-1070–treated mice. (D) Cumulative frequency histogram of leukocyte rolling velocities for these 2 groups. Median values are indicated by vertical lines; *** P < .001.

Figure 6

GMI-1070 inhibits RBC captures and improves blood flow in SCD mice. (A) Number of RBC interactions with adherent WBCs per minute. Both GMI-1070 and antiselectin antibodies reduced the capture rates of RBCs per adherent leukocytes. (B) The blood flow rates in SCD mice treated with GMI-1070 or anti–P- and –E-selectin antibodies were significantly higher than in mice treated with PBS control. This difference was not due to venular size because the average venular diameter was nearly identical (∼ 21 μm) among the 3 groups (Table 1); ***P < .001.

Figure 7

Administration of GMI-1070 after the inflammatory challenge produces the same biologic effects. Mice were treated with GMI-1070 or PBS as per protocol 2. (A) Leukocyte rolling flux (cells rolling per minute) was significantly elevated greater than 4-fold increase in SCD mice treated with GMI-1070; ***P < .001. (B) Recruitment of adherent leukocytes was severely inhibited in GMI-1070–treated SCD mice compared with PBS-treated SCD mice; **P < .01. (C) Scatter plots of the number of adherent leukocytes during the experimental period. Each dot represents data obtained from a single venule. GMI-1070 rapidly reduced the number of adherent leukocytes, compared with the PBS control group, but the number of adherent leukocytes remained almost unchanged over time after TNF-α exposure in both PBS-treated control (gray regression line) and GMI-1070–treated mice (black regression line). (D) Interactions between circulating RBCs and adherent leukocytes were abrogated (∼ 95% reduction) in SCD mice infused with GMI-1070. ***P < .001. (E) Scatter plots of the number of interactions between RBCs and adherent leukocytes during the experimental period. Each symbol represents data from a single venule. The increase in RBC-WBC interactions over time after TNF-α exposure was abrogated by the GMI-1070–treated compared with PBS-treated control mice. The gray and black regression lines represent PBS-treated and GMI-1070–treated mice, respectively.

Figure 8

GMI-1070 sustains blood flow rates and prolongs survival in TNF-α–exposed SCD mice challenged by surgical trauma. (A) Blood flow rates were 2-fold higher in GMI-1070–treated SCD mice compared with control SCD mice; ***P < .001. (B) The Kaplan-Meier survival curves for GMI-1070–treated or control SCD mice. Survival was significantly improved in the GMI-1070–treated group. Log-rank (Mantel-Cox) test, P = .007.