Atorvastatin or transgenic expression of TFPI inhibits coagulation initiated by anti-nonGal IgG binding to porcine aortic endothelial cells

Abstract

Background: Intravascular thrombosis remains a barrier to successful xenotransplantation. Tissue factor (TF) expression on porcine aortic endothelial cells (PAECs), which results from their activation by xenoreactive antibodies (Abs) to Galα1,3Gal (Gal) and subsequent complement activation, plays an important role.

Objectives: The present study aimed to clarify the role of Abs directed against nonGal antigens in the activation of PAECs to express functional TF and to investigate selected methods of inhibiting TF activity.

Methods: PAECs from wild-type (WT), α1,3-galactosyltransferase gene-knockout (GT-KO) pigs, or pigs transgenic for CD46 or tissue factor pathway inhibitor (TFPI), were incubated with naïve baboon serum (BS) or sensitized BS (with high anti-nonGal Ab levels). TF activity of PAECs was assessed.

Results: Only fresh, but not heat-inactivated (HI), naïve BS activated WT PAECs to express functional TF. Similarly, PAECs from CD46 pigs were resistant to activation by naïve BS, but not to activation by fresh or HI sensitized BS. HI sensitized BS also activated GT-KO PAECs to induce TF activity. TF expression on PAECs induced by anti-nonGal Abs was inhibited if serum was pretreated with (i) an anti-IgG Fab Ab or (ii) atorvastatin, or (iii) when PAECs were transgenic for TFPI.

Conclusions: Anti-nonGal IgG Abs activated PAECs to induce TF activity through a complement-independent pathway. This implies that GT-KO pigs expressing a complement-regulatory protein may be insufficient to prevent the activation of PAECs. Genetic modification with an 'anticoagulant' gene (e.g. TFPI) or a therapeutic approach (e.g. atorvastatin) will be required to prevent coagulation dysregulation after pig-to-primate organ transplantation.

Figure 1. Surface phenotype of PAECs from CD46 and TFPI transgenic pigs

PAECs were incubated with FTIC-conjugated anti-CD46 and anti-TFPI Abs, respectively, and analyzed by flow cytometry to determine their surface phenotypes (Shaded: isotype control; solid line: WT PAECs; dash line: transgenic PAECs; left: CD46; right: TFPI). Very high expression of CD46 was observed in CD46-transgenic pigs. More modest expression of TFPI was observed in TFPI-transgenic pigs.

Figure 2. Naïve baboon serum activates PAECs to induce TF activity that is complement-dependent

A. WT (whites bars) or CD46 (gray bars) PAECs were incubated with medium (control) or (at a series of concentrations) fresh or HI naïve baboon serum (NBS). The TF activity on the surface of PAECs was determined by the recalcified clotting assay. Fresh, but not heat-inactivated (HI), naïve baboon serum activated WT PAECs to express TF activity. CD46 PAECs were resistant to activation by fresh NBS. (Left panel: fresh naïve baboon serum [NBS]; right panel: HI baboon serum [HI NBS]) ( #p<0.01 compared to control). B. WT (white bars) or CD46 (grey bars) PAECs were incubated with medium (control) or 10% fresh NBS or HI NBS for 4h. TF expression of PAECs was determined by quantitative RT-PCR. (# p<0.01 compared to control). Fresh, but not HI, NBS resulted in an increase in TF expression on WT, but not CD46, PAECs.

Figure 2. Naïve baboon serum activates PAECs to induce TF activity that is complement-dependent

A. WT (whites bars) or CD46 (gray bars) PAECs were incubated with medium (control) or (at a series of concentrations) fresh or HI naïve baboon serum (NBS). The TF activity on the surface of PAECs was determined by the recalcified clotting assay. Fresh, but not heat-inactivated (HI), naïve baboon serum activated WT PAECs to express TF activity. CD46 PAECs were resistant to activation by fresh NBS. (Left panel: fresh naïve baboon serum [NBS]; right panel: HI baboon serum [HI NBS]) ( #p<0.01 compared to control). B. WT (white bars) or CD46 (grey bars) PAECs were incubated with medium (control) or 10% fresh NBS or HI NBS for 4h. TF expression of PAECs was determined by quantitative RT-PCR. (# p<0.01 compared to control). Fresh, but not HI, NBS resulted in an increase in TF expression on WT, but not CD46, PAECs.

Figure 3. Sensitized baboon serum activates PAECs to induce TF activity that is complement-independent

A. IgM and IgG binding of heat-inactivated naïve baboon serum (HI NBS, n=15) or heat-inactivated sensitized baboon serum (HI SBS; n=3) to PAECs (WT [white bars] and GT-KO [grey bars]) were determined by flow cytometry. Binding to GT-KO PAECs represents binding of anti-nonGal Abs. There was much greater binding of IgG, but not IgM, when PAECs were incubated with HI SBS. B. WT (white bars) or CD46 (grey bars) PAECs were incubated with medium (control) or fresh (left panel) or HI (right panel) sensitized baboon serum (SBS) at a series of concentrations for 8h. The TF activity on the surface of PAECs was determined by the recalcified clotting assay. Both fresh and HI SBS activated WT and CD46 PAECs to induce TF activity. C. GT-KO PAECs were incubated with medium (control) and HI NBS (white bars) or HI SBS (grey bars). HI SBS, but not HI NBS, activated PAECs to express TF activity. D. WT (white bars), CD46 (grey bars), or GT-KO (striped bars) PAECs were incubated with medium (control) or 10% fresh SBS or HI SBS for 4h. TF expression on PAECs was determined by quantitative RT-PCR. (# p<0.01 compared to control.) TF expression was increased on PAEC from all pigs, with greater expression after exposure to fresh SBS than to HI SBS.

Figure 3. Sensitized baboon serum activates PAECs to induce TF activity that is complement-independent

A. IgM and IgG binding of heat-inactivated naïve baboon serum (HI NBS, n=15) or heat-inactivated sensitized baboon serum (HI SBS; n=3) to PAECs (WT [white bars] and GT-KO [grey bars]) were determined by flow cytometry. Binding to GT-KO PAECs represents binding of anti-nonGal Abs. There was much greater binding of IgG, but not IgM, when PAECs were incubated with HI SBS. B. WT (white bars) or CD46 (grey bars) PAECs were incubated with medium (control) or fresh (left panel) or HI (right panel) sensitized baboon serum (SBS) at a series of concentrations for 8h. The TF activity on the surface of PAECs was determined by the recalcified clotting assay. Both fresh and HI SBS activated WT and CD46 PAECs to induce TF activity. C. GT-KO PAECs were incubated with medium (control) and HI NBS (white bars) or HI SBS (grey bars). HI SBS, but not HI NBS, activated PAECs to express TF activity. D. WT (white bars), CD46 (grey bars), or GT-KO (striped bars) PAECs were incubated with medium (control) or 10% fresh SBS or HI SBS for 4h. TF expression on PAECs was determined by quantitative RT-PCR. (# p<0.01 compared to control.) TF expression was increased on PAEC from all pigs, with greater expression after exposure to fresh SBS than to HI SBS.

Figure 3. Sensitized baboon serum activates PAECs to induce TF activity that is complement-independent

A. IgM and IgG binding of heat-inactivated naïve baboon serum (HI NBS, n=15) or heat-inactivated sensitized baboon serum (HI SBS; n=3) to PAECs (WT [white bars] and GT-KO [grey bars]) were determined by flow cytometry. Binding to GT-KO PAECs represents binding of anti-nonGal Abs. There was much greater binding of IgG, but not IgM, when PAECs were incubated with HI SBS. B. WT (white bars) or CD46 (grey bars) PAECs were incubated with medium (control) or fresh (left panel) or HI (right panel) sensitized baboon serum (SBS) at a series of concentrations for 8h. The TF activity on the surface of PAECs was determined by the recalcified clotting assay. Both fresh and HI SBS activated WT and CD46 PAECs to induce TF activity. C. GT-KO PAECs were incubated with medium (control) and HI NBS (white bars) or HI SBS (grey bars). HI SBS, but not HI NBS, activated PAECs to express TF activity. D. WT (white bars), CD46 (grey bars), or GT-KO (striped bars) PAECs were incubated with medium (control) or 10% fresh SBS or HI SBS for 4h. TF expression on PAECs was determined by quantitative RT-PCR. (# p<0.01 compared to control.) TF expression was increased on PAEC from all pigs, with greater expression after exposure to fresh SBS than to HI SBS.

Figure 3. Sensitized baboon serum activates PAECs to induce TF activity that is complement-independent

A. IgM and IgG binding of heat-inactivated naïve baboon serum (HI NBS, n=15) or heat-inactivated sensitized baboon serum (HI SBS; n=3) to PAECs (WT [white bars] and GT-KO [grey bars]) were determined by flow cytometry. Binding to GT-KO PAECs represents binding of anti-nonGal Abs. There was much greater binding of IgG, but not IgM, when PAECs were incubated with HI SBS. B. WT (white bars) or CD46 (grey bars) PAECs were incubated with medium (control) or fresh (left panel) or HI (right panel) sensitized baboon serum (SBS) at a series of concentrations for 8h. The TF activity on the surface of PAECs was determined by the recalcified clotting assay. Both fresh and HI SBS activated WT and CD46 PAECs to induce TF activity. C. GT-KO PAECs were incubated with medium (control) and HI NBS (white bars) or HI SBS (grey bars). HI SBS, but not HI NBS, activated PAECs to express TF activity. D. WT (white bars), CD46 (grey bars), or GT-KO (striped bars) PAECs were incubated with medium (control) or 10% fresh SBS or HI SBS for 4h. TF expression on PAECs was determined by quantitative RT-PCR. (# p<0.01 compared to control.) TF expression was increased on PAEC from all pigs, with greater expression after exposure to fresh SBS than to HI SBS.

Figure 4. Sensitized baboon serum anti-nonGal Abs activate PAECs to induce TF activity through IgG binding

A. Heat-inactivated sensitized baboon serum (HI SBS) was pre-incubated with dithiothreitol (DTT, 200μl/ml) to eliminate IgM. IgM or IgG binding to PAECs was determined by flow cytometry. (Shaded: isotype control; solid line: non-treated serum; broken line: serum treated with DTT). B. WT PAECs were incubated with medium (control) or HI SBS, which has been pretreated with DTT for 30min. TF activity on the surface of PAECs was activated by HI SBS was unrelated to the pretreatment of DTT ( the level of IgM) by recalcified clotting assay. C. The experiment performed in B was repeated with HI SBS that had been treated with an anti-IgG Fab Ab at various concentrations (5–20μg/ml) for 30min at 4°C. Anti-IgG Fab Ab blocked the effect of HI SBS (with high levels of anti-nonGal Abs), inactivating WT PAECS (* p<0.05 compared to control).

Figure 4. Sensitized baboon serum anti-nonGal Abs activate PAECs to induce TF activity through IgG binding

A. Heat-inactivated sensitized baboon serum (HI SBS) was pre-incubated with dithiothreitol (DTT, 200μl/ml) to eliminate IgM. IgM or IgG binding to PAECs was determined by flow cytometry. (Shaded: isotype control; solid line: non-treated serum; broken line: serum treated with DTT). B. WT PAECs were incubated with medium (control) or HI SBS, which has been pretreated with DTT for 30min. TF activity on the surface of PAECs was activated by HI SBS was unrelated to the pretreatment of DTT ( the level of IgM) by recalcified clotting assay. C. The experiment performed in B was repeated with HI SBS that had been treated with an anti-IgG Fab Ab at various concentrations (5–20μg/ml) for 30min at 4°C. Anti-IgG Fab Ab blocked the effect of HI SBS (with high levels of anti-nonGal Abs), inactivating WT PAECS (* p<0.05 compared to control).

Figure 4. Sensitized baboon serum anti-nonGal Abs activate PAECs to induce TF activity through IgG binding

A. Heat-inactivated sensitized baboon serum (HI SBS) was pre-incubated with dithiothreitol (DTT, 200μl/ml) to eliminate IgM. IgM or IgG binding to PAECs was determined by flow cytometry. (Shaded: isotype control; solid line: non-treated serum; broken line: serum treated with DTT). B. WT PAECs were incubated with medium (control) or HI SBS, which has been pretreated with DTT for 30min. TF activity on the surface of PAECs was activated by HI SBS was unrelated to the pretreatment of DTT ( the level of IgM) by recalcified clotting assay. C. The experiment performed in B was repeated with HI SBS that had been treated with an anti-IgG Fab Ab at various concentrations (5–20μg/ml) for 30min at 4°C. Anti-IgG Fab Ab blocked the effect of HI SBS (with high levels of anti-nonGal Abs), inactivating WT PAECS (* p<0.05 compared to control).

Figure 5. Human sera with high levels of natural anti-nonGal Abs activate PAECs to induce TF activity that is complement-independent

A. The relationship between the titer of anti-nonGal IgG Abs in 20% heat-inactivated human sera (n=15) and induced TF activity on PAECs. IgG binding to GT-KO PAECs is indicated. The same HI HS were used to activate WT PAECs. TF activity on PAECs was determined by the recalcified clotting assay. B. WT and CD46 PAECs were incubated with HI HS with high-titer anti-nonGal Abs (n=2) at various concentrations (2.5–10%). TF activity on WT and CD46 PAECs was induced after incubation with these two HS by the recalcified clotting assay. (#p<0.01 compared to control). C. The experiment performed in Figure 4C was repeated with 10% HI HS (with high-titer anti-nonGal Abs) that had been treated with an anti-Fab Ab at 20μg/ml. Anti-IgG Fab Ab blocked the effect of this human serum to activate WT PAECs to express TF.

Figure 5. Human sera with high levels of natural anti-nonGal Abs activate PAECs to induce TF activity that is complement-independent

A. The relationship between the titer of anti-nonGal IgG Abs in 20% heat-inactivated human sera (n=15) and induced TF activity on PAECs. IgG binding to GT-KO PAECs is indicated. The same HI HS were used to activate WT PAECs. TF activity on PAECs was determined by the recalcified clotting assay. B. WT and CD46 PAECs were incubated with HI HS with high-titer anti-nonGal Abs (n=2) at various concentrations (2.5–10%). TF activity on WT and CD46 PAECs was induced after incubation with these two HS by the recalcified clotting assay. (#p<0.01 compared to control). C. The experiment performed in Figure 4C was repeated with 10% HI HS (with high-titer anti-nonGal Abs) that had been treated with an anti-Fab Ab at 20μg/ml. Anti-IgG Fab Ab blocked the effect of this human serum to activate WT PAECs to express TF.

Figure 5. Human sera with high levels of natural anti-nonGal Abs activate PAECs to induce TF activity that is complement-independent

A. The relationship between the titer of anti-nonGal IgG Abs in 20% heat-inactivated human sera (n=15) and induced TF activity on PAECs. IgG binding to GT-KO PAECs is indicated. The same HI HS were used to activate WT PAECs. TF activity on PAECs was determined by the recalcified clotting assay. B. WT and CD46 PAECs were incubated with HI HS with high-titer anti-nonGal Abs (n=2) at various concentrations (2.5–10%). TF activity on WT and CD46 PAECs was induced after incubation with these two HS by the recalcified clotting assay. (#p<0.01 compared to control). C. The experiment performed in Figure 4C was repeated with 10% HI HS (with high-titer anti-nonGal Abs) that had been treated with an anti-Fab Ab at 20μg/ml. Anti-IgG Fab Ab blocked the effect of this human serum to activate WT PAECs to express TF.

Figure 6. Activation of PAECs by naïve and sensitized baboon sera to induce TF activity is inhibited by expression of TFPI

WT PAECs (white bars), TFPI PAECs (grey bars), and TFPI PAECs pretreated with anti-TFPI Ab at 50μg/ml for 30min (lined bars) were incubated with medium (control), 10% naïve (NBS), sensitized (SBS), or heat-inactivated sensitized (HI SBS) baboon serum for 8h. TF activity was determined by the recalcified clotting assay (A) and quantitative RT-PCR (B). A. TFPI PAECs inhibited TF activity induced by all 3 baboon sera, and the effect was countered by the addition of anti-TFPI Ab. (# p<0.01 WT vs TFPI PAECs) B. NBS, SBS or HI SBS increased the expression of TF mRNA in TFPI PAECs (*p<0.05, # p<0.01 vs control).

Figure 6. Activation of PAECs by naïve and sensitized baboon sera to induce TF activity is inhibited by expression of TFPI

WT PAECs (white bars), TFPI PAECs (grey bars), and TFPI PAECs pretreated with anti-TFPI Ab at 50μg/ml for 30min (lined bars) were incubated with medium (control), 10% naïve (NBS), sensitized (SBS), or heat-inactivated sensitized (HI SBS) baboon serum for 8h. TF activity was determined by the recalcified clotting assay (A) and quantitative RT-PCR (B). A. TFPI PAECs inhibited TF activity induced by all 3 baboon sera, and the effect was countered by the addition of anti-TFPI Ab. (# p<0.01 WT vs TFPI PAECs) B. NBS, SBS or HI SBS increased the expression of TF mRNA in TFPI PAECs (*p<0.05, # p<0.01 vs control).

Figure 7. Atorvastatin inhibits TF activity on PAECs induced by heat-inactivated sensitized baboon serum

WT PAECs were pretreated with atorvastatin (ATS) for 16h before incubation with heat-inactivated sensitized baboon serum (HI SBS). In one experiment, atorvastatin pretreatment was combined with mevalonic acid (MVA). TF activity was determined by the recalcified clotting assay (A) and quantitative RT-PCR (B). Rho A activation and Akt dephosphorylation were determined 8h after incubation with HI SBS (C). A. Atorvastatin (at 10μM) inhibited 50% TF activity of PAECs after incubation with HI SBS. The effect was abolished by the addition of MVA. B. After incubation with HI SBS, atorvastatin (at 10μM) inhibited the expression of TF mRNA. The effect was abolished by the addition of MVA (at 10 μM). (# p<0.01 compared to untreated control). C. Atorvastatin led to dephosphorylation of Rho A and phosphorylation of Akt induced by HI SBS. The effect was reversed by the addition of MVA.

Figure 7. Atorvastatin inhibits TF activity on PAECs induced by heat-inactivated sensitized baboon serum

WT PAECs were pretreated with atorvastatin (ATS) for 16h before incubation with heat-inactivated sensitized baboon serum (HI SBS). In one experiment, atorvastatin pretreatment was combined with mevalonic acid (MVA). TF activity was determined by the recalcified clotting assay (A) and quantitative RT-PCR (B). Rho A activation and Akt dephosphorylation were determined 8h after incubation with HI SBS (C). A. Atorvastatin (at 10μM) inhibited 50% TF activity of PAECs after incubation with HI SBS. The effect was abolished by the addition of MVA. B. After incubation with HI SBS, atorvastatin (at 10μM) inhibited the expression of TF mRNA. The effect was abolished by the addition of MVA (at 10 μM). (# p<0.01 compared to untreated control). C. Atorvastatin led to dephosphorylation of Rho A and phosphorylation of Akt induced by HI SBS. The effect was reversed by the addition of MVA.

Figure 7. Atorvastatin inhibits TF activity on PAECs induced by heat-inactivated sensitized baboon serum

WT PAECs were pretreated with atorvastatin (ATS) for 16h before incubation with heat-inactivated sensitized baboon serum (HI SBS). In one experiment, atorvastatin pretreatment was combined with mevalonic acid (MVA). TF activity was determined by the recalcified clotting assay (A) and quantitative RT-PCR (B). Rho A activation and Akt dephosphorylation were determined 8h after incubation with HI SBS (C). A. Atorvastatin (at 10μM) inhibited 50% TF activity of PAECs after incubation with HI SBS. The effect was abolished by the addition of MVA. B. After incubation with HI SBS, atorvastatin (at 10μM) inhibited the expression of TF mRNA. The effect was abolished by the addition of MVA (at 10 μM). (# p<0.01 compared to untreated control). C. Atorvastatin led to dephosphorylation of Rho A and phosphorylation of Akt induced by HI SBS. The effect was reversed by the addition of MVA.