Novel insights into the development of atherosclerosis in hemophilia A mouse models


Valder R. Arruda, The Children’s Hospital of Philadelphia, 3501 Civic Center Boulevard, 5056 Colket Center for Translational Research, Philadelphia, PA 19104, USA.
Tel.: +1 215 590 4907; fax: +1 215 590 3660.


Summary. Background: Cardiovascular diseases in aging people with hemophilia (PWH) represent a growing concern. The underlying hypocoagulability probably provides a protective effect against acute thrombus formation, but the limited data available show no preventive effect against the development of atherogenesis in PWH. Atherosclerosis-prone mice are attractive tools for the study of atherosclerosis development, and may provide insights into disease progression in PWH. Methods: Severe hemophilia A (factor VIII-deficient [FVIIIo]) mice were crossed with mice lacking apolipoprotein E (ApoE−/−) or mice lacking the LDL receptor (LDLR−/−), and then compared to hemostatically normal littermate controls. After mice had received atherogenic diets for 8, 22 or 37 weeks, atherosclerotic lesion size and phenotypic characterization were analyzed in the aortic sinus and whole aortas. Results: ApoE−/−/FVIIIo mice showed a time-dependent protective effect against the development of atherosclerosis, beginning after 22 diet-weeks and persisting to 37 diet-weeks in both the aorta sinus and whole aorta as compared with ApoE−/− mice. Notably, the FVIII deficiency did not influence the progression of atherosclerosis in the FVIIIo/LDLR−/− model as compared with controls at early or late time points. Conclusions: Hypocoagulability ameliorates vascular disease in the ApoE-deficient model in a lipid-independent manner. Interestingly, FVIII deficiency did not affect the development of atherosclerosis in LDLR−/− mice. In contrast to the ApoE model, the LDLR model resembles the lipid profile that is commonly observed in humans with atherosclerosis. These findings, to a certain extent, support the notion of atherosclerosis development in the complete absence of FVIII.


Arterial occlusive diseases (AODs) associated with atherosclerosis remain the main cause of death in North America and Europe [1]. Atherosclerosis is a complex and multifactorial disease, and several coagulation proteins, including factor VIII and FIX, have been implicated as risk factors for atherogenesis and its vascular complications [2]. Retrospective studies suggest that mortality resulting from AOD is lower in hemophilic patients and in female carriers of hemophilia than in age-matched and gender-matched controls [3,4]. However, a series of studies focused on the detection of atherosclerotic lesions by non-invasive imaging techniques or post-mortem analysis in people with hemophilia (PWH) showed that the development of atherosclerotic lesions was comparable to that in non-hemophilia controls [5–7]. Thus, the lower mortality resulting from ischemic heart disease observed in PWH is probably attributable to decreased thrombin generation during plaque rupture, whereas the putative protective effects of FVIII or FIX deficiency on atherosclerosis progression remain conflicting. The relatively low numbers of subjects, variability in the presence of cardiovascular risk factors and underlying infectious diseases and heterogeneous clotting factor concentrates (CFC) replacement regimens are important limitations of these clinical studies [1,8–11]. In this context, the use of genetically engineered mice for severe hemophilia A and atherosclerosis-prone models provides a unique opportunity to investigate the role of FVIII in modifying vascular disease with minimal influences of uncontrolled environmental factors. A previous study with mice double-deficient for apolipoprotein E (ApoE) and FVIII suggested that FVIII deficiency is associated with a protective effect against atherosclerosis development in mice [12]. The basis of these conflicting data from preclinical and clinical studies is not completely understood.

To gain further insights into the impact of FVIII deficiency on the progression of atherosclerosis, we sought to evaluate the long-term progression of atherosclerosis development in two different atherosclerosis-prone models crossed with severe hemophilia A mice. Here, we provide evidence that atherosclerosis lesions developed differently in the absence of FVIII deficiency in LDL receptor (LDLR)-deficient (LDLR–/–) and ApoE-deficient (ApoE–/–) mice.

Materials and methods

Animal models

Mice with FVIII deficiency (FVIIIo) caused by deletion of exon 16 of the murine F8 gene on a C57Bl6/129 background were kindly provided by H. H. Kazazian (University of Pennsylvania). These mice were crossed with ApoE−/− mice or LDLR−/− mice, both on a C57BL background. Littermates from F2 and F3 generations were crossbred to generate FVIIIo/ApoE−/− and FVIIIo/LDLR−/− mice. The control groups consisted of littermate mice matched by age and gender. The genotypes for each model were confirmed by PCR. Four experimental groups were obtained: (i) ApoE−/−; (ii) FVIIIo/ApoE−/−; (iii) LDLR−/−; and (iv) FVIIIo/LDLR−/−. Initially, animals were fed a standard low-fat mouse chow diet. When they were 8 weeks old, LDLR−/− mice were placed on an atherogenic diet with 1.25% cholesterol, whereas ApoE−/− mice received a 0.15% cholesterol Western diet.

Mice were killed after 8, 22 or 37 weeks on diet for the analysis of atherosclerosis progression. All procedures were approved by the Ethical Committee for Animal Experiments of the State University of Campinas.

Cholesterol determination

Blood samples were collected after 8, 22 or 37 weeks of the atherogenic diet through the retro-orbital venous plexus, using heparinized capillary tubes. The cholesterol concentration in plasma was determined with an enzymatic assay kit, according to the manufacturer’s instructions (Roche Diagnostics, Mannheim, Germany).

Quantification of atherosclerotic lesions

To quantitatively examine the development of atherosclerosis, we measured atherosclerotic lesions in two distinct anatomical regions: (i) in the aortic sinus, with the Oil-red O method; and (ii) in whole aorta specimens, with the Sudan IV method. Analyses were performed at three time points while mice were fed atherogenic diets, to cover distinct phases of atherogenesis: after 8 weeks (fatty streak lesions), after 22 weeks (early fibrous plaque lesions), and after 37 weeks (complex fibrous plaque lesions). Mice were killed under deep anesthesia, and, after perfusion with sterile phosphate-buffered saline, hearts and aortas were dissected. Serial 8-μm-thick cryosections were obtained from the upper part of the heart, from the aortic sinus (where valve cups become visible) to the aortic arch. These sections were stained with Oil-red O (Sigma Aldrich, St Louis, MO, USA) and counterstained with hematoxylin. Eight sections from each animal, 64 μm apart, were analyzed. For whole aorta specimens, the aorta was removed from the aortic arch to the iliac branches, and fixed in 10% buffered formalin phosphate. Aortas were stained with Sudan IV (Sigma Aldrich) and cut longitudinally. Values for atherosclerosis lesions in aortic root sections (in mm2) and whole aortas (as percentage of total aorta area) corresponded to the mean value obtained by two independent investigators who quantified all lesions in a blinded fashion, using ImageJ (National Institutes of Health, Bethesda, MD, USA) analysis software.

Qualitative analysis of atherosclerotic lesions

To further evaluate the atherosclerosis mouse models, we performed immunohistochemical and histologic analysis of sections from the aortic sinus. For the identification of macrophages in lesions after 8 diet-weeks, slides from the aortic sinus were incubated with a biotin anti-Mac-1 antibody (Pharmingen, San Diego, CA, USA), and antibodies were visualized with an avidin/biotin peroxidase-linked detection system (ABC Vectorstain; Vector Laboratories, Burlingame, CA, USA). The presence of smooth muscle cells was evaluated after 37 diet-weeks by use of a mouse mAb against human α-actin directly coupled to horseradish peroxidase (Dako, Carpinteria, CA, USA). Calcium deposits and collagen fibers within lesions were determined with hematoxylin and Masson’s trichrome staining, respectively. These analyses were performed in a blinded fashion, with two sections from the aortic sinus obtained from three to five animals per group.

Statistical analysis

Data are expressed as means ± standard deviations, unless stated otherwise. Means were compared by use of the Mann–Whitney test, and frequencies were compared by use of Fisher’s exact test. Differences were considered to be statistically significant if the P-value was ≤ 0.05.


Generation of hemophilic A mice with ApoE or LDLR deficiencies

FVIIIo animals were crossed with ApoE−/− or LDLR−/− mice and generated FVIIIo/ApoE−/− and FVIIIo/LDLR−/− mice and littermate hemostatically normal ApoE−/− or LDLR−/− controls, respectively. After the initiation of the high-fat diet, no differences in the cholesterol levels were observed between FVIIIo/LDLR−/− and LDLR−/− controls throughout the experimental period. In contrast, FVIIIo/ApoE−/− mice showed higher cholesterol levels than ApoE−/− littermate controls at all time points tested (P < 0.05), as shown in Table 1. These findings are comparable to previous findings in FVIIIo/ApoE−/− mice receiving a similar diet [12].

Table 1.   Lipid profile in mice on atherogenic diets
GenotypePlasma cholesterol levels (mg dL−1)
8 weeksP*22 weeksP*37 weeksP*
  1. Apolipoprotein E (ApoE)-deficient mice and LDL receptor (LDLR)-deficient mice received a Western diet containing 0.15% or 1.25% cholesterol, respectively. *Mann–Whitney test; n = 9–12 animals per analysis.

ApoE−/−884 ± 243< 0.01919 ± 3260.02848 ± 349< 0.01
FVIIIo/ApoE−/−1391 ± 3411229 ± 3891406 ± 347
LDLR−/−1239 ± 3550.401574 ± 3320.071001 ± 3250.06
FVIIIo/LDLR−/−1351 ± 1031343 ± 1551235 ± 219

FVIII deficiency ameliorates atherosclerosis progression in ApoE−/− mice

At an early time point (8 weeks on the Western diet), there was no difference in the development of atherosclerotic lesions in the aortic sinus between FVIIIo/ApoE−/− mice (0.051 ± 0.04 mm2; n = 11) and ApoE−/− controls (0.058 ± 0.02 mm2; n = 12, P = 0.16). However, after 22 diet-weeks on an atherogenic diet, FVIIIo/ApoE−/− mice showed a significant reduction in the development of atherosclerotic lesions (0.328 ± 0.09 mm2; n = 12) as compared with ApoE−/− controls (0.423 ± 0.08 mm2; n = 11; P = 0.01). These differences persisted and became more evident after 37 diet-weeks, when lesion sizes in FVIIIo/ApoE−/− mice (0.446 ± 0.11 mm2; n = 11) were significantly lower than in ApoE−/− controls (0.650 ± 0.20 mm2; n = 10; P = 0.006), as illustrated in Fig. 1A,B.

Figure 1.

 Influence of factor VIII on atherosclerosis development in apolipoprotein E-deficient (ApoE−/−) mice. (A, C) Representative microphotographs taken after 37 weeks on atherogenic diets of FVIII-deficient (FVIIIo)/ApoE−/− mice and ApoE−/− littermate controls illustrate the significant reduction of atherosclerosis development in the aortic sinus (A) and aorta (C). (B, D) Statistical comparison of atherosclerotic lesion size between ApoE−/− and FVIIIo/ApoE−/− mice after 8, 22 and 37 weeks on atherogenic diets in the aortic sinus, where lesions are measured in mm2 (B), and the aorta, where the percentage area of the whole aorta covered by lesions was evaluated (D). *Mann–Whitney test; bars indicate standard errors of the mean (n = 10–12 animals per group).

We observed a similar time-dependent protective effect of FVIII deficiency on analysis of whole aorta samples. After 22 diet-weeks, the proportion of lesion area was lower in FVIIIo/ApoE−/− mice (3.87% ± 3.03%; n = 12) than in ApoE−/− controls (6.92% ± 2.83%; n = 13; P = 0.01), and this difference was further enhanced at 37 diet-weeks, when lesions in FVIIIo/ApoE−/− mice (11.5% ± 5.53%; n = 11) were about 50% smaller than in ApoE−/− controls (22.6%± 7.67%; n = 13; P < 0.001), as shown in Fig. 1C,D. Thus, despite an unfavorable lipid profile, FVIII deficiency is associated with an antiatherosclerosis phenotype in ApoE–/–mice.

FVIII deficiency does not influence the development of atherosclerosis in LDLR−/− mice

At similar time points after initiation of a high-fat diet as described above, there were no differences in the atherosclerosis lesions in the aortic sinus between FVIIIo/LDLR−/− mice and LDLR−/− littermate controls. Specifically, after 8 diet-weeks, lesion sizes were 0.05 ± 0.04 mm2 in FVIIIo/LDLR−/− (n = 11) mice and 0.08 ± 0.04 mm2 in LDLR−/− mice (n = 10). After 22 diet-weeks, lesion sizes were 0.49 ± 0.14 mm2 in FVIIIo/LDLR−/− mice (n = 11) and 0.38 ± 0.17 mm2 in LDLR−/− mice (n = 12). After 37 diet-weeks, lesion sizes were 0.61 ± 0.07 mm2 in FVIIIo/LDLR−/− mice (n = 11) and 0.69 ± 0.08 mm2 in LDLR−/− mice (n = 10; P = 0.08), as shown in Fig. 2A,B. In the whole aorta, the progression of atherosclerotic lesions was also similar between FVIIIo/LDLR−/− mice and LDLR−/− littermate controls after 22 diet-weeks (10.0% ± 5.4% vs. 10.8% ± 3.9%, respectively; n = 12 per group; P = 0.6) and after 37 diet-weeks (20.2% ± 6.8% vs. 23.4% ± 6.6%, respectively; n = 12 per group; P = 0.3), as shown in Fig. 2C,D.

Figure 2.

 Influence of factor VIII on atherosclerosis development in LDL receptor-deficient (LDLR−/−) mice. (A, C) Representative microphotographs taken after 37 weeks on atherogenic diets of FVIII-deficient (FVIIIo)/LDLR−/− mice and LDLR−/− littermate controls illustrate that no significant difference in atherosclerosis progression could be observed either in the aortic sinus (A) or in the aorta (C). (B, D) Statistical comparison of atherosclerotic lesion size between LDLR−/− and FVIIIo/LDLR−/− mice after 8, 22 and 37 weeks on atherogenic diets in the aortic sinus, where lesions are measured in mm2 (B), and the aorta, where the percentage area of the whole aorta covered by lesions was evaluated (D). *Mann–Whitney test; bars indicate standard errors of the mean (n = 10–12 animals per group).

Cellular and phenotypic characterization of atherosclerotic lesions

FVIIIo/ApoE−/− mice presented a lower frequency of calcium deposits in atherosclerotic lesions than ApoE−/− controls after 22 diet-weeks (P = 0.05; Fisher’s exact test). However, no additional differences could be observed between FVIIIo/ApoE−/− and FVIIIo/LDLR−/− mice as compared with their controls regarding the number of macrophages, proportion of smooth muscle fibers, and collagen in atherosclerotic lesions (data not shown).


A series of retrospective studies have suggested that acute ischemic cardiovascular complications are less common in PWH [3,4], and that these protective effects may result exclusively from a lower extent of thrombin generation in the acute phase of AOD. In contrast, the development of atherosclerotic disease in PWH seems to be minimally affected by inherited deficiency of clotting factors [5–7]. Although prospective studies enrolling a large cohort of PWH are needed to confirm these early findings, animal models can be valuable tools with which to identify modulators of atherosclerosis and to test new drugs or therapeutic interventions for the treatment of AOD in hemophilia.

However, it has been recently shown that different atherosclerosis-prone mouse models may differ from each other in their response to specific experimental manipulations, so the choice of a model is critical for the validity of the study results [13]. A previous study using hemophilia A mice crossbred with ApoE−/− mice suggested that FVIII deficiency results in a protective effect against atherosclerosis progression [12]. In mice with severe von Willebrand disease (residual levels of FVIII of ∼ 27% of normal [14]) crossed with ApoE−/− or LDLR−/− mice showed a reduction of atherosclerosis at the aorta sinus and at restricted areas of the aorta associated with high shear stress [15]. Thus, partial deficiency of FVIII does not systemically prevent atherosclerosis progression in these models.

In murine models of defective procoagulant or anticoagulant activity, typically just one atherosclerosis-prone model is used (reviewed in [13,16]). Therefore, we evaluated the long-term follow-up of atherosclerosis progression in two atherosclerosis-prone mouse models interbred with mice with severe FVIII deficiency.

We first evaluated atherosclerosis progression in ApoE−/− mice, a model characterized by spontaneous development of atherosclerotic lesions and very high cholesterol levels that resembles the rare human type III hyperlipidemia (cholesterol profile of very low-density and intermediate-density lipoproteins) [17]. ApoE−/− mice have been widely used in the study of atherosclerosis progression, mainly because of the rapid development of atherosclerosis [13]. Our findings showed that FVIII deficiency is associated with a time-dependent sustained protection against atherosclerosis development. We found that FVIIIo/ApoE−/− mice exhibited low rates of atherosclerotic lesions at both aortic sinus and abdominal aorta sites, and, notably, these effects were sustained at weeks 22–37 (the end of the experiment). These effects seemed to be lipid-independent, as plasma cholesterol levels in these FVIIIo/ApoE−/− mice were approximately 50% higher than those in the littermate ApoE−/− mice. Our findings are in agreement with a previous report on a similar model, in which FVIIIo/ApoE−/− mice exhibited fewer atherosclerotic lesions, with concomitantly increased levels of plasma cholesterol [12]. In the ApoE−/− model, the use of an agonist of liver X receptors (LXRs), regulators of cholesterol homeostasis, was successful in limiting the progression of atherosclerosis, and these effects were independent of their effect on cholesterol metabolism [18]. Recently, insights into the underlying mechanism of the vascular protective effects of LXR agonists have emerged from evidence of their antiplatelet and antithrombotic properties in mice [19]. LXR agonists inhibit the ability of platelets to modulate the size, stability and growth of thrombus formation upon laser-induced injury in the microcirculation. Using this injury model, we previously found that platelet accumulation and thrombus formation were severely hampered in murine models of hemophilia [20]. Thus, we hypothesize that the hypocoagulability of hemophilia models mimics, to a certain extent, the protective effect of LXR agonists in the ApoE−/− model, such as a low extent of thrombin generation and its effects on endothelial cells and platelets [15]. Collectively, these data suggest that the development of atherosclerosis in the ApoE−/− model is modulated, at least in part, by cellular and protein components of hemostasis.

Next, we sought to further investigate the impact of FVIII deficiency in FVIIIo/LDLR−/− mice. Atherosclerotic lesions in LDLR−/− mice are morphologically comparable to those observed in ApoE−/− mice, but lesions in the former model tend to develop more slowly, and mostly when animals are put on atherogenic diets [21]. In addition, this model is thought to resemble more closely the pathogenesis of atherosclerosis development in humans, with elevated levels of cholesterol and distribution being confined mainly to the LDL fraction [13]. Interestingly, we found no difference in the progression of atherosclerotic lesions of LDLR−/− study groups at early or late time points as compared with controls. We hypothesized that these data are consistent with differences in the participation of components of hemostasis in atherosclerosis development between the ApoE−/− and LDLR−/− models. In contrast to the ApoE−/− model, in which there was amelioration of atherosclerosis by LXR agonists without modifications in the lipid profile, the use of these compounds in the LDLR−/− model lowered the lipid profile, while also ameliorating atherosclerosis development. These differences suggest that the underlying mechanism of atherosclerosis in the LDLR−/− model is lipid-dependent. Therefore, the data on progression of atherosclerosis in the FVIIIo/LDLR−/− model and its lipid profile support the findings of atherosclerosis development in PWH [22]. Further studies on hemophilia B in both atherosclerosis-prone models may provide further evidence for the role of the intrinsic pathway in the development of atherosclerosis. Previous studies [16] on deficiencies of distinct proteins in the coagulant and fibrinolytic pathways have commonly been performed in ApoE−/− or LDLR−/− models, but rarely in both [15,23], and this raises the possibility that similar discrepancies as determined in this current study can be expected.

In conclusion, our data demonstrate that FVIIIo/LDLR−/− mice provide an attractive model for testing both the efficacy and safety of novel therapies for atherosclerosis. These animals may prove to be useful for assessing the safety of such therapies, which may increase the risk of bleeding that is not readily evident in hemostatically normal models, but is enhanced in FVIII-deficient models.


We are grateful to D. Rader and staff from his laboratory at the University of Pennsylvania for their assistance in setting up the atherosclerosis assessment models and helpful discussions. We are also grateful to H. H. Kazazian from the University of Pennsylvania for providing us with the murine model of hemophilia A.

Disclosure of Conflict of Interests

This work was support by grants from FAPESP (to J. M. Annichino-Bizzacchi). The other authors state that they have no conflict of interest.