Long-term expression of murine activated factor VII is safe, but elevated levels cause premature mortality

Abstract

Intravenous infusion of recombinant human activated Factor VII (FVIIa) has been used for over a decade in the successful management of bleeding episodes in patients with inhibitory antibodies to Factor VIII or Factor IX. Previously, we showed that expression of murine FVIIa (mFVIIa) from an adeno-associated viral (AAV) vector corrected abnormal hemostatic parameters in hemophilia B mice. To pursue this as a therapeutic approach, we sought to define safe and effective levels of FVIIa for continuous expression. In mice transgenic for mFVIIa or injected with AAV-mFVIIa, we analyzed survival, expression levels, in vitro and in vivo coagulation tests, and histopathology for up to 16 months after birth/mFVIIa expression. We found that continuous expression of mFVIIa at levels at or below 1.5 microg/ml was safe, effective, and compatible with a normal lifespan. However, expression levels of 2 microg/ml or higher were associated with thrombosis and early mortality, with pathologic findings in the heart and lungs that were rescued in a low-factor X (low-FX) mouse background, suggesting a FX-mediated effect. The findings from these mouse models of continuous FVIIa expression have implications for the development of a safe gene transfer approach for hemophilia and are consistent with the possibility of thromboembolic risk of continuously elevated FVIIa levels.

Figure 1. Transgenic construct for mFVIIa (mFVII-2RKR) and mFVII zymogen mice.

The mFVIIa/mFVII cDNA is under control of the mTTR promoter separated by a synthetic intron. A polyadenylation signal (pA) from bovine growth hormone follows the 3′ untranslated region (UTR) for both cDNAs.

Figure 2. Efficacy of the mFVIIa or the mFVII zymogen transgene in HB mice.

(A) Plasma mFVIIa levels (mFVII levels were measured by a separate antigen ELISA resulting in ~1,700 ng/ml), (B) PT, (C) aPTT, and (D) TAT levels in littermate HB (n = 21), WT (n = 21), HB-mFVII (n = 4), HB–low-mFVIIa (n = 10), and HB–high-mFVIIa (n = 7) mice. *P < 0.05 versus HB. Values are mean ± 1 SD.

Figure 3. In vivo tests of coagulation.

(A) Blood loss in mFVIIa transgenic HB mice expressing low (HB-Low mFVIIa, n = 10) or high (HB-High mFVIIa, n = 8) levels of mFVIIa, compared with HB (n = 20) and WT (n = 10) littermates. *P < 0.05 versus HB. Values are mean ± 1 SD. (B) Representative cremaster real-time thrombus formation in WT, HB, or HB-mFVIIa (high and low expresser) mice. (C) FeCl₃ carotid artery injury in WT, HB, or HB-mFVIIa (high and low expresser) mice. The single asterisk denotes the time at which the FeCl₃ patch was applied; the double asterisk denotes vessel occlusion. Results are from 1 indicative mouse per group.

Figure 4. Expression of high mFVIIa levels is associated with increased mortality after crossing the founder C57BL/6–SJL mice into C57BL/6 pure strain background.

Kaplan-Meier survival plot of transgenic mFVIIa mice with low or high levels of transgene expression compared with WT littermates. Curves for high mFVIIa expressers exclude animals dying within 72 h of birth. Squares, WT (n = 10); filled circles, low-mFVIIa N1–N4 (n = 25); triangles, high-mFVIIa F0 (n = 6); diamonds, high-mFVIIa N1 (n = 17); open circles, high-mFVIIa N2 (n = 11). *P < 0.05 versus WT.

Figure 5. Histological findings in lungs and hearts of high-mFVIIa transgenic mice compared with WT littermates.

Intimal/smooth muscle proliferation was observed in the pulmonary veins (B and C) as well as fibrin deposition in the pulmonary bed (E and F) of mFVIIa transgenic mice, but neither was seen in WT mice (A and D). Thrombi were observed in the coronary artery (I and M) and ventricles (H and L) as well as inflammatory infiltrates in the myocardium (J and N) of mFVIIa transgenic mice, but none were observed in WT mice (G and K). Original magnification, ×20 (H and L); ×40 (G and K); ×100 (D, E, I, and M); ×200 (A–C, F, J, and N).

Figure 6. Hemostatic differences in high-mFVIIa transgenic mice.

(A) mFVIIa antigen; (B) PT measurements using hFX-deficient plasma; (C) TAT levels; and (D) Kaplan-Meier survival plot of high-mFVIIa transgenic mice in a background between ~3% and 100% mFX. (A–C) n = 3 (~3% mFX and high-mFVIIa ~3% mFX), 4 (high-mFVIIa), 6 (WT), 8 (50% mFX), 12 (high-mFVIIa 50% mFX). (D) Open squares, WT (n = 10); filled circles, high-mFVIIa F0 (n = 3); stars, high-mFVIIa F2 ~3% mFX (n = 8); diamonds, high-mFVIIa F2 50% mFX (n = 17); crossed-out squares, high-mFVIIa F1 100% mFX (n = 11); open circles, high-mFVIIa F2 100% mFX (n = 7). *P < 0.05 versus WT. Values are mean ± 1 SD.