Contrasting Roles of E2F2 and E2F3 in Cardiac Neovascularization

Insufficient neovascularization, characterized by poor endothelial cell (EC) growth, contributes to the pathogenesis of ischemic heart disease and limits cardiac tissue preservation and regeneration. The E2F family of transcription factors are critical regulators of the genes responsible for cell-cycle progression and growth; however, the specific roles of individual E2Fs in ECs are not well understood. Here we investigated the roles of E2F2 and E2F3 in EC growth, angiogenesis, and their functional impact on myocardial infarction (MI). An endothelial-specific E2F3-deficient mouse strain VE-Cre; E2F3fl/fl was generated, and MI was surgically induced in VE-Cre; E2F3fl/fl and E2F2-null (E2F2 KO) mice and their wild-type (WT) littermates, VE-Cre; E2F3+/+ and E2F2 WT, respectively. The cardiac function, infarct size, and vascular density were significantly better in E2F2 KO mice and significantly worse in VE-Cre; E2F3fl/fl mice than in their WT littermates. The loss of E2F2 expression was associated with an increase in the proliferation of ECs both in vivo and in vitro, while the loss of E2F3 expression led to declines in EC proliferation. Thus, E2F3 promotes while E2F2 suppresses ischemic cardiac repair through corresponding changes in EC proliferation; and differential targeting of specific E2F members may provide a novel strategy for therapeutic angiogenesis of ischemic heart disease.


Introduction
Ischemic heart disease (IHD) represents one of the largest epidemics facing the aging population. Insufficient neovascularization, characterized by poor vessel growth and survival, contributes to the pathogenesis of IHD and limits cardiac tissue preservation and regeneration; thus enhancement of cardiac neovascularization is a therapeutic goal.
Following tissue ischemia, vascular injury recruits endothelial cells (ECs) required for neovascularization to restore blood perfusion [1][2][3][4][5]. Neovascularization is a complex sequence of events involving coordinated vascular cell proliferation, migration and tube formation [6][7][8][9][10]. Accumulating evidence indicates that neovascularization is critically dependent on the appropriate regulation of EC cell cycle [11,12]; a deficient endothelial proliferative response to ischemia can result in tissues necrosis [13]. Additionally, ECs can contribute to cardiac repair by mediating favorable cell:cell interactions and secreting paracrine factors that protect the function and survival of cardiomyocytes [14].
The E2F family of transcription factors, with eight members identified, play a central role in regulating the expression of genes responsible for cell-cycle control and provide an ideal target for therapeutic modulation of vascular growth [15][16][17][18][19][20]. However, the PREVENT trial, targeting E2F activity in the vascular smooth muscle cells in the autologous vein grafts of CABG surgery to prevent graft overgrowth and failure, generated negative results presumably due to the non-specific inhibition of both activating and repressive E2F species [21,22]; therefore, it is imperative to elucidate the specific function of individual E2F members in the vascular biology [20]. The E2F1, 2 and 3, in particular, are considered ''transactivators'' that activate transcription of target genes for DNA replication and G1/S transition, thus cell proliferation [23][24][25][26]. However, we have recently found that E2F1 inhibits ischemic angiogenesis by suppressing the expression of angiogenic factors, VEGF and PlGF [27]. The functions of E2F2 and E2F3 in ECs during ischemic disease are largely unknown.
In this study, we sought to investigate the roles of E2F2 and E2F3 in EC proliferation and their functional impact on myocardial infarction (MI). We found that E2F3 is essential for EC growth, neovascularization, and preservation of cardiac function, while E2F2 plays a contrasting role. Thus, individual E2Ffactors, rather than E2F family as a whole, may provide more specific molecular targets for therapeutic neovascularization of IHD.

Mice
The E2F3 fl/fl and E2F2 +/2 mice were obtained from Dr. Gustavo Leone's lab (Ohio State University) [23]. VE-cadherin-Cre (VE-Cre) mice were purchased from The Jackson Laboratory (Bar Harbor, ME). VE-Cre; E2F3 fl/fl mice and littermate VE-Cre; E2F3 +/+ mice were generated by crossing VE-Cre; E2F3 fl/+ male and E2F3 fl/+ female parents. E2F2 2/2 (E2F2 KO) and littermate E2F2 +/+ (E2F2 WT) mice were generated by crossing E2F2 +/2 parents as previously described [28]. All these mice were on C57BL/6 background. Mouse genotypes were determined via polymerase chain reaction (PCR) with tail DNA. All the animal work presented in this report was approved by the Institutional Animal Care and Use Committee of Northwestern University and performed in the barrier facilities of the Center for Comparative Medicine of the university.

Surgical MI Model and Echocardiographic Assessments of Left Ventricular (LV) Function
MI was induced in 8 week-old male VE-Cre; E2F3 fl/fl , VE-Cre; E2F3 +/+ , E2F2 KO, and E2F2 WT mice by permanent ligation in the middle of the left anterior descending (LAD) coronary artery as described previously [29,30]. Mice were anesthetized by inhaling Isoflurane TM delivered at 2-4% throughout the surgical procedure, and were injected subcutaneously with Metacam (1 mg/kg) as analgesic immediately after the surgery and then daily for the next 2 to 3 days. Trans-thoracic 2-dimensional echocardiographic measurements were performed before MI (baseline) and at 7, 14 and 28 days post-MI by using a commercially available high resolution system (VEVO 770 TM , VisualSonics Inc., Toronto, Canada) equipped with a 30-MHz transducer. M-mode tracings were used to measure LV wall thickness, end-systolic diameter (LVESD), and end diastolic diameter (LVEDD). Systolic and diastolic LV areas were determined by M-mode in long-axis configuration and fractional shortening (FS) was measured at the mid-ventricular level. The LV chamber volumes in diastole and systole were derived from their respective measured 2D areas using a LV volume algorithm within the Vevo770 echo software. Cardiac ejection fraction (EF) was determined offline by the equation: EF = (Diastolic Volume -Systolic Volume/Diastolic Volume) x100.

Histological Assessments of Infarct Size, Vascular Density, and in vivo EC Proliferation
Fourteen and 28 days after surgically induced MI, the vasculature was labeled by injecting 50 mL BS Lectin I (Vector laboratories, Inc., Burlingame, CA) into the tail vein, and mice were euthanized 10 min later by CO 2 inhalation (primary method) and cervical dislocation (secondary method). A portion of animals also received Bromodeoxyuridine (BrdU) (30 mg/kg IP) for 48 h (Q12H64) before euthanasia, which permits identification of proliferating cells. Cardiac tissues were fixed in 4% paraformaldehyde for 4 h, incubated overnight in 30% sucrose, embedded in OCT compound (Sakura Finetek U.S.A., Inc., Torrance, CA), snap-frozen in liquid nitrogen, and cut into 5 mm sections. Serial cryosectioning was performed starting at 1 mm below the suture (used to ligate the LAD) moving toward the apex, with three consecutive sections per 1 mm to allow for quantitative pathohistological analysis at each level. To evaluate infarct area, the Masson Trichrome elastic tissue staining was performed as described previously [30,31]. Infarct size was reported as the ratio of the length of fibrotic area to the length of the LV inner

Isolation of Mouse Primary Cardiac ECs and Adenoviral Vector Transduction
Primary ECs were isolated from mouse heart tissues as described previously [32]. Briefly, tissues were minced and digested with collagenase and dispase, then the mixture was passed through a 100-mm cell strainer to obtain single cell suspensions. Cell debris were removed via density centrifugation with Histopaque-1.083 (Sigma), then the ECs were immunostained with CD31-PE, FACS sorted to $95% purity, and cultured in EBM-2 medium (Lonza, Walkersville, MD). The cells were used before passage 5. The ECs were infected by adenovirus-Cre (Vector BioLabs, Philadelphia, PA) by following manufacturer's instructions.

EdU Incorporation Assay
Proliferation of primary ECs was measured with a commercially available Click-iT TM EdU (5-ethynyl-29-deoxyuridine) Flow Cytometry Assay kit (Invitrogen) by following the manufacturer's instructions. Prior to the addition of EdU, subconfluent ECs were synchronized by incubation in EBM-2 medium with 0.1% FBS for 24 h and stimulated to proliferate by addition of EBM-2 supplemented with 5% FBS.

Statistical Analysis
All values are expressed as mean 6 SEM. Comparison between two means was performed with an unpaired Student's t test, whereas ANOVA with Fisher's protected least significant differences and Bonferroni-Dunn post hoc analysis were used for comparisons of more than two means.

Functional Recovery of the Infarcted Heart is Enhanced by the Loss of E2F2 Expression and Impaired by the Loss of Endothelial E2F3 Expression
We induced MI by surgical ligation of LAD coronary artery in VE-Cre; E2F3 fl/fl and E2F2 KO mice and their WT littermates, VE-Cre; E2F3 +/+ and E2F2 WT mice, respectively. E2F2 KO mice exhibited a greater EF and FS and a smaller LV systolic and diastolic volume as compared with E2F2 WT mice ( Figure 1). The significantly better cardiac function in E2F2 KO mice was observed as early as day 7 and persisted till day 28 post-MI. In contrast, VE-Cre; E2F3 fl/fl exhibited a worsened heart function as compared with VE-Cre; E2F3 +/+ mice at days 14 and 28 post-MI ( Figure 1). These results suggest that E2F3 improved while E2F2 impairs cardiac function in response to ischemic injury.

Infarct Size and Vessel Density are Improved by the Loss of E2F2 Expression and Worsened by the Loss of Endothelial E2F3 Expression
At day 28 post-MI, infarct size was significantly smaller in E2F2 KO and significantly larger in VE-Cre; E2F3 fl/fl mice, than in their littermates with WT levels of E2F2 and endothelial E2F3 expression ( Figure 2). The loss of E2F2 expression in E2F2 KO mice also led to improvements in vessel density, while the endothelial deletion of E2F3 in VE-Cre; E2F3 fl/fl mice was associated with a decline in vessel density ( Figure 3). Thus, E2F3 enhance while E2F2 suppresses cardiac neovascularization.

EC Proliferation is Increased by Declines in E2F2 Expression and Reduced by Declines in E2F3 Expression
Because E2F2 and E2F3 are cell cycle regulators, we assessed EC proliferation in the ischemic tissue (i.e., the infarct border zone) at day 14 post-MI with CD31 and BrdU double staining. The frequency of proliferating ECs was significantly higher in E2F2 KO mice and significant lower in VE-Cre; E2F3 fl/fl mice, than in their WT littermates ( Figure 4A-B).
To determine whether the contrasting roles of E2F2 and E2F3 in EC proliferation in vivo are cell autonomous, we isolated primary EC from the hearts of E2F2 KO, E2F2 WT, and E2F3 fl/fl mice and the E2F3 fl/fl ECs were subsequently transduced with a vector coding for GFP alone or with a vector coding for both GFP and Cre expression to knockout the expression of E2F3. Cells were starved to quiescence by serum deprivation and then re-stimulated by the addition of serum. The EdU incorporation assays showed that the proliferating (i.e., EdU+) ECs were more frequent in the E2F2 KO population and much less frequent in the E2F3 KO  Figure 4C). These results indicate that E2F2 suppresses and E2F3 enhances EC proliferation.

Discussion
E2F1, 2, and 3 are classified as a subgroup of ''activating'' E2Fs that promote cell proliferation [33][34][35][36][37]. However, here we found that EC growth, neovascularization, and cardiac function post-MI are improved by the loss of E2F2 expression and impaired by the loss of endothelial E2F3 expression, which suggest that E2F2 and E2F3 play contrasting roles. Data from this current study and the results we have recently reported on E2F1 [27], collectively, suggest diverse but specific roles of these E2F species in the regulation of vascular growth.
Our data clearly show that E2F3 is essential for EC proliferation and ischemic angiogenesis, which cannot be compensated by other ''activating'' E2Fs (i.e., E2F1 and 2). Although consistent with reports in the literature documenting a critical role of E2F3 for cell cycle progression, this role appears to be greater in ECs than in other cell types [25,38,39]. Whether this role of E2F3 is specific to ECs or E2F3 plays a similar role in other vascular cells such as VSMCs remains to be investigated.
The finding that E2F2 suppresses EC proliferation is somewhat unexpected. It is in contrast to the observations made in several other cell types, in which overexpression of E2F2 activates cell cycle progression [40]. While the molecular mechanism underlying the enhanced growth of E2F2 KO ECs is yet to be determined, other lab did report that E2F2 KO mice display hyperproliferation of immune cells [41]. Given the strong transactivity of E2F3 in ECs, it is tempting to speculate that in the absence of E2F2 (i.e., in E2F2 KO cells), the regulatory DNA elements normally bound by E2F2 may be occupied by E2F3 instead, thereby exerting a stronger transactivity for the expression of these genes. This hypothesis, however, remains to be tested in our future study.
In summary, our study revealed that E2F2 and E2F3, two members in the ''activating'' E2F subfamily exert contrasting roles in the regulation of EC growth and neovascularization and in the preservation of cardiac tissues from ischemic injury. Thus, individual E2Fs, rather than E2F family as a whole, may provide more specific targets for therapeutic angiogenesis of IHD.