Impaired microvascular perfusion in sepsis requires activated coagulation and P-selectin-mediated platelet adhesion in capillaries

Abstract

Purpose: Impaired microvascular perfusion in sepsis is not treated effectively because its mechanism is unknown. Since inflammatory and coagulation pathways cross-activate, we tested if stoppage of blood flow in septic capillaries is due to oxidant-dependent adhesion of platelets in these microvessels.

Methods: Sepsis was induced in wild type, eNOS(-/-), iNOS(-/-), and gp91phox(-/-) mice (n = 14-199) by injection of feces into the peritoneum. Platelet adhesion, fibrin deposition, and blood flow stoppage in capillaries of hindlimb skeletal muscle were assessed by intravital microscopy. Prophylactic treatments at the onset of sepsis were intravenous injection of platelet-depleting antibody, P-selectin blocking antibody, ascorbate, or antithrombin. Therapeutic treatments (delayed until 6 h) were injection of ascorbate or the glycoprotein IIb/IIIa inhibitor eptifibatide, or local superfusion of the muscle with NOS cofactor tetrahydrobiopterin or NO donor S-nitroso-N-acetylpenicillamine (SNAP).

Results: Sepsis at 6-7 h markedly increased the number of stopped-flow capillaries and the occurrence of platelet adhesion and fibrin deposition in these capillaries. Platelet depletion, iNOS and gp91phox deficiencies, P-selectin blockade, antithrombin, or prophylactic ascorbate prevented, whereas delayed ascorbate, eptifibatide, tetrahydrobiopterin, or SNAP reversed, septic platelet adhesion and/or flow stoppage. The reversals by ascorbate and tetrahydrobiopterin were absent in eNOS(-/-) mice. Platelet adhesion predicted 90% of capillary flow stoppage.

Conclusion: Impaired perfusion and/or platelet adhesion in septic capillaries requires NADPH oxidase, iNOS, P-selectin, and activated coagulation, and is inhibited by intravenous administration of ascorbate and by local superfusion of tetrahydrobiopterin and NO. Reversal of flow stoppage by ascorbate and tetrahydrobiopterin may depend on local eNOS-derived NO which dislodges platelets from the capillary wall.

Figures

Fig. 1

Sepsis increases blood flow stoppage, platelet adhesion, and fibrin deposition in capillaries in mouse skeletal muscle. Sepsis was induced by injection of feces into peritoneum (FIP). Flow stoppage was determined in 18 control and 16 septic mice at 6–7 h, adhesion in 9 control and 7 and septic mice at 7 h, and fibrin deposition in 2 control and 4 and septic mice at 7 h post-FIP, *Difference from control, p < 0.05

Fig. 2

Injections of P-selectin blocking antibody at 0.5 h, eptifibatide at 6 h (glycoprotein IIb/IIIa inhibitor), and antithrombin at 0.5 h inhibited platelet adhesion and blood flow stoppage in capillaries at 6–7 h post-FIP. *Difference from sepsis, p < 0.05, n = 5–9 for adhesion groups, and 5–16 for stopped-flow groups

Fig. 3

Sepsis increases thrombosis in capillaries. The thrombosis potential was tested by flooding the muscle with FeCl3 at 6 h post-FIP. Left *Difference between post-FeCl3 and pre-FeCl3 percentage of stopped-flow capillaries, p < 0.05, n = 5–18. Right Change in stopped flow percentage = (post-FeCl3 percentage) − (pre-FeCl3 percentage). *Difference from control, p < 0.05, n = 5 per group

Fig. 4

Ascorbate injection at 0 h and gp91phox knockout prevent platelet adhesion and blood flow stoppage in capillaries at 6–7 h post-FIP, *Difference from sepsis wild type (wt) group, p < 0.05, n = 6–7 for adhesion groups, and 5–16 for stopped-flow groups

Fig. 5

Ascorbate injection at 6 h reverses platelet adhesion and blood flow stoppage in capillaries at 7 h post-FIP in wild type but not eNOS−/− mice, *Difference from sepsis group, p < 0.05, n = 6–17 for adhesion groups, and 8–16 for stopped-flow groups

Fig. 6

Effects of tetrahydrobiopterin (BH4) and NO donor SNAP on platelet adhesion and blood flow stoppage in capillaries of wild type and eNOS−/− mice (top), and effect of iNOS knockout on adhesion/stoppage (bottom). At 6 h, 0.1-mL boluses of BH4 (repeated over 1 h period) flooded the muscle surface to determine platelet adhesion and flow stoppage at 7 h post-FIP. Alternatively, a single 0.1-mL bolus of SNAP flooded the surface, and platelet adhesion and stoppage were determined 15 min later (when the temporary SNAP-induced vasodilation had ended). BH4 reversed platelet adhesion and flow stoppage in septic capillaries of wild type but not eNOS−/− mice. SNAP reversed platelet adhesion and flow stoppage in both types of mice. iNOS knockout inhibited septic platelet adhesion at 6 h post-FIP. *Difference from appropriate sepsis group, p < 0.05, n = 5–10 for adhesion groups, and 5–16 for stopped-flow groups in top row, and 6 and 10 for adhesion, and 6 and 10 for stopped-flow groups at bottom, respectively