Hematopoietic Stem Cells Differentiate into Vascular Cells that Participate in the Pathogenesis of Atherosclerosis

Masataka Sata
University of Tokyo Graduate School of Medicine, Tokyo, Japan

Circulating smooth muscle progenitor cells may contribute to neointima formation, hypothesized these investigators. They first performed a heterotopic cardiac transplantation between wild type and LacZ mice that express the marker gene LacZ in all tissues. Ninety percent of the neointimal cells were LacZ-positive originating from the recipient mice when a wild-type heart was transplanted in a LacZ mouse, and conversely LacZ negative cells for neointimal formation were observed when the LacZ heart was transfected into the wild-type mouse. The medial cells remained in the media. Recipient-derived LacZ positive cells express various smooth muscle markers, including myosin heavy chain, b-Caldesmon, Calponin, and actin.

To investigate the potential source of recipient cells that may contribute to graft vasculopathy, they performed bone marrow transplantation from GFP mice to wild-type mice. Wild-type hearts transfected into GFP-BMT mice, resulted in accumulation of bone marrow-derived cells (BMDC), expressing smooth muscle actin, along graft coronary arteries.

The contribution of bone marrow cells to vascular remodeling after mechanical injury was studied by inserting a large wire into a femoral artery of the LacZ-BMT mouse, leading to endothelial denudation and marked enlargement of the lumen with rapid onset of medial smooth muscle cell apoptosis. After 4 weeks, the injured artery developed neointima hyperplasia; 60% of neointima and 40% of media were composed of LacZ-positive cells originating from bone marrow (Figure 1).

A time course study showed that in the absence of injury, no BMDC were observed at the artery. At 1 week after injury, BMDC accumulated on the injured artery. At 4 weeks, BMDC contributed to neointima formation and re-endothelialization.

The potential contribution of BMDC to atherosclerosis was then studied. Transplantation was performed of BMDC from GFP mice to ApoE knockout mice, which developed severe hyperlipidemia and atherosclerosis. Injected cells settled at the bone marrow of the vertebra. When mice developed atherosclerosis, injected cells also accumulated at the aorta, where severe atherosclerosis developed. On cross-section, GFP-positive bone marrow cells expressed smooth muscle actin. (Figure 2).

In LacZ to ApoE knockout GFP-BMT, about 50% of smooth muscle actin cells were derived from bone marrow. Media smooth muscle cells were negative for LacZ, indicating a specificity for this method. Anti-LacZ immunogold labeling detected BMDC with muscle fiber, which may be called smooth muscle-like cells. But in wild type mice, they found medial smooth muscle cells with a large body size and containing many secretory organelles.

These results suggest that BMDC can give rise to vascular progenitor cells that form as injured artery and differentiate and proliferate and thus contribute to vascular remodeling and neointima formation. However, another group reported that BMDC do not contribute to neointima formation in a model of aortic transplantation. Hence, the origin of neointima cells is diverse.

Thus, this group compared the contribution of BMDC using 3 different models of neointimal hyperplasia. Inserting a large wire to cause severe injury resulted in wire-mediated expansion. Ligation of the common carotid artery resulted in neointima containing a few BMDC. The placement of a cuff around the right femoral artery induced neointimal hyperplasia, after which many BMDC were observed at the site of injury. However, most of the cells were inflammatory cells in adventitia. Surprisingly, no BMDC were found in the neointima.

To determine the mechanism by which the contribution differs, they investigated the vascular changes induced by wire, cuff, and ligation injury. After wire injury, most of the medial cells are killed by apoptosis. In contrast, the cuff or ligation model induced mild injury and smooth muscle medial cells remained relatively intact. In the cuff model, inflammatory cells enter, smooth muscle cells or other advential cells migrate and proliferate and cause neointima formation. For a previously severe injury, such as wire, most of the cells are killed and BMDC are recruited to repair the vasculature.

The degree to which BMDC contribute to the pathogenesis of human disease is unknown. However, human pathology visualizes neovascalurization and microhemmorhage in human atherosclerosis. Other investigators have documented intima medial surface damage and accumulation of fresh blood cells inside the atheroma, suggesting those blood cells may contribute to the remodeling of advanced atherosclerosis.

This group also detected apoptosis of endothelial cells and smooth muscle cells in advanced atherosclerosis. So they investigated the contribution of BMDC to the progression and remodeling of established atherosclerosis. They replaced bone marrow using very old ApoE knockout mice with advanced atherosclerosis. At 3 months, many BMDC were seen in atheroma, mostly macrophages or T cells. However, they also identified BMDC and luminal endothelial cells, indicating those cells contributed to vascular remodeling. Notably, they found the BMDC also contributed to the neovascularization inside the atheroma. BMDC were seen at the site of calcification, which expressed osteoblast-related proteins. Thus, they speculated that circulating progenitor cells may play a role in the initiation and progression of remodeling and the calcification of advanced atheroscleroisis. BMDC may participate in cell turnover and repair after apoptotic cell death.

Consistent with their results, another group reported that BMDC are present in human coronary atherosclerosis. That group analyzed patients with leukemia who received gender mis-matched bone marrow. At 90 days, BMDC were seen in coronary arteries. This phenomenon was more prominent when the artery was diseased.

Sata and colleagues are now developing therapeutic strategies targeting smooth muscle progenitor cells (SMPC). Possible targets include mobilization, circulation, homing, differentiation and proliferation of SMPC (Figure 3). The delivery of a high concentration of a drug at the site of SMPC accumulation may be beneficial.

An Increase in Plasma Ox-LDL Levels during Coronary Artery Bypass Grafting are Associated with Plaque Injuries of the Aorta

Shoichi Ehara
Osaka City University Graduate School of Medicine, Osaka, Japan

Recently, oxidized LDL (ox-LDL) is considered to play a key role in the pathogenesis of the inflammatory process in atherosclerotic lesions.

This group recently developed a sensitive method to measure the plasma levels of ox-LDL. Blood samples were collected in a tube containing EDTA, and LDL fractions were obtained from the samples by sequential ultracentrifugation for 2 days. Standard ox-LDL measurement was prepared by incubation of LDL from healthy volunteers with CuSO₄ at 37 degrees for 3 hours. Diluted LDL fractions were added to microtiter wells precoated with an anti-ox-LDL antibody (DLH3). After extensive washing, the remaining ox-LDL was detected with a sheep anti-human ApoB antibody and an alkaline phosphatase-conjugated anti-sheep Ig G antibody. The measurement of plasma levels of ox-LDL used a standard curve (ng/5 Äæg LDL protein).

The characteristics of the ox-LDL method developed by this group include 1) the use of LDL fraction from the blood plasma, but does not use whole blood plasma, 2) minimizes potential interference with other plasma constituents, such as ox-VLDL, anti-oxidized LDL autoantibodies, and anti-phospholipid antibodies, and 3) takes 4 days to perform each measurement, but is superior to other methods in obtaining a sensitive and accurate detection of ox-LDL. A continuum of degrees of oxidation exists. Oxidized phospholipids are a prominent component of minimally modified LDL and fully ox-LDL.

This group previously reported that elevated levels of ox-LDL show a positive relationship with the severity of acute coronary syndromes (ACS). Patients with acute myocardial infarction had significantly higher levels of ox-LDL than in patients with unstable angina, stable angina, or control patients (Figure 1). Macrophage levels are also elevated. Their findings suggest a pivotal role for ox-LDL in the genesis of coronary plaque instability and the development of ACS.

They investigated the plasma levels of ox-LDL in patients who had undergone coronary artery bypass grafting to determine whether there is a relationship between plasma ox-LDL levels and the release of aortic atheromatous debris due to aortic clamp manipulation in CABG patients.

Plasma ox-LDL levels were measured in 3 groups: 1) CABG group of 27 patients who had cardiopulmonary bypass and aortic clamp manipulation, 2) Off-pump CABG group of 6 patients, without aortic clamp manipulation, and 3) non-CABG cardiac surgery group of 6 patients who had atrial septal defect (ASD) or mitral valve prolapse and aortic clamp manipulation as a control group. In the non-CABG group, although aortic clamp manipulation was performed, no atherosclerosis of the ascending aorta was seen. Both the CABG and Off-pump surgery group had mass aortic atherosclerosis.

Blood samples were obtained on admission, and then post-operatively at 6 hours, day 1, day 3, and day 5. Epiaortic ultrasonography of the ascending aorta was performed during surgery to evaluate the severity of aortic atherosclerosis. Based on the severity of atherosclerosis, the patients were divided into 2 groups: mild atherosclerosis with intimal thickness < 3 mm and severe atherosclerosis with intimal thickness > 3 mm.

The time course of ox-LDL levels during the peri-operative period in the 3 groups showed in the CABG group a marked and significant increase in plasma ox-LDL at 6 hours and throughout until day 5 (Figure 2). In contrast, off-pump and non-CABG groups did not have an increase in plasma ox-LDL levels. Moreover, plasma ox-LDL levels at 6 hours were significantly higher in the severe atherosclerosis group compared to the mild atherosclerosis group. In the severe atherosclerosis group, plasma ox-LDL levels showed a maximum increase at 6 hours post-op and a subsequent decrease from day 1 post-op onward. In contrast, in the mild atherosclerosis group, no increase in the plasma ox-LDL levels was seen.

These results support the hypothesis that ox-LDL and atheromatous debris present within severe atherosclerotic plaques of the ascending aorta may be released into the blood stream, most probably during aortic clamp manipulation during CABG.

A balance mechanism between oxidation and anti-oxidation in the blood in humans is strongly suggested by the increase in plasma ox-LDL levels at 6 hours post-op that subsequently decrease in the severe atherosclerosis group (Figure 3).  It has been suggested that anti-oxidants would be able to inhibit cell-induced oxidative modifications. Moreover, previous experimental studies have shown that ox-LDL injected intravenously disappears from plasma with a very short half-life, only minutes in the rat or rabbit. These data strongly suggest that a balance between oxidation and anti-oxidation may play a role to determine plasma levels of ox-LDL in humans.

Graft Arterial Disease: A Model for Investigating Inflammation and T-cell Mediated Immunity in Coronary Atherosclerosis

Mitsuaki Isobe
Tokyo Medical and Dental University , Tokyo, Japan

Graft arterial disease (GAD) is one of the most serious complications after heart transplantation. Autopsy findings include nearly total occlusion of coronary artery vessels. More than 50% of cardiac transplant recipients have some extent of GAD at 5 years after transplantation.

Chronic rejection cannot be prevented by conventional immunosuppressants like tacrolimus or cyclosporin A. Therapeutic doses of tacrolimus can prevent acute rejection but not chronic rejection. A very high dose of tacrolimus can reduce the extent of intimal hyperplasia to some extent. But if the immunological tolerance is induced to cardiac allograft the intimal hyperplasia or chronic rejection is nearly completely prevented. This is indirect evidence that T-cell mediated immunity is involved in this processes.

The thickening intima is comprised of smooth muscle cells (SMCs). The natural form of SMCs decreases in the thickened intima, but the synthetic form of immature SMCs is increased in the thickened intima. This proliferation of SMCs can be prevented to some extent by oral tranilast. The decrease in SMCs proliferation is associated with the induction of  p21 expression, which is a major cell cycle inhibitor genes.

Thus, this group thought that T cells could activate SMCs for proliferation. An in vitro study in which SMCs were taken from mice aorta and co-cultured with T cells was performed. SMCs were not activated by naïve T cells, but activated T cells from mice with rejecting cardiac allograft activated the SMCs proliferation. This proliferation is associated with induction of PDGF, basic FGF, VCAM-1, SMemb, and GAPDH.  SMCs proliferation was prevented with antisense Egr-1 transfected onto the SMCs prior to the co-culture, suggesting that activated cells can interact with SMCs and causes their proliferation.

T-cells require two signals for optimal activation. One signal is from the antigen via T-cell receptor, and the other is a co-stimulatory signal for T-cell activation. In addition to CD28, a few novel costimulatory molecules have been identified, which are thought to play a role in the interaction of T-cells and SMCs. CD28 is the major co-stimulatory molecule and studied for more than 10 years. Fetal protein and immunoglobin of CTLA4 binds to the ligand for CD28 and blocks its signaling.

The induction of CTLA4-Ig  is a strong agent for inducing immunological tolerance. But at the same time, chronic rejection is nearly completely suppressed (Figure 1).

In the case of CD28, because acute and chronic rejection is suppressed, it is not possible to differentiate the mechanism of chronic rejection from acute rejection. They found that a novel molecule, iCOS, inducible co-stimulator, a member of the CD28 family, was expressed on activated T-cells and also SMCs. It appears that the iCOS expression in the thickened intima plays some role in the interaction between T cells and SMCs. If the iCOS pathway is blocked by either anti-iCOS antibody or iCOS-Ig the development of intimal hyperplasia is nearly completely blocked.

The same observation was made with another co-stimulatory molecule, HVEM (herpes virus entry mediator). The immunological role of HVEM is not clearly understood. When mice were treated with HVEM Ig, the intima hyperplasia is suppressed in association with a reduction in the cytokines interferon gamma, IL6, and IL4.

Another interesting molecule is PD-1 (programmed death-1). PD-1 knockout mice show dilated cardiomyopathy, probably due to autoimmune mechanisms. PD-1 is believed to transduce negative co-stimulatory signals to T-cells. PD-1 has two ligands, PD-L1 and PD-L2. These investigators used an PD-L1 antibody to block PD-L1 and found that PD-L1 is expressed on thickened intima of recipient mice. Blockade of PD-L1 signaling with the PD-L1 antibody promotes the neointima formation compared to controls. The same observation was made in a model of arterial injury. After wire injury of mice femoral artery, the neointima thickening was exaggerated with the PD-L1 antibody.

There are many players of inflammation, including ICAM-1, MCP-1, TNF-R, MMP, and NF-κB. Adhesion molecules work as key players of the inflammation by recruitment of white blood cells to the site of inflammation. These investigators used gene transfer technology to the donor heart. When antisense ICAM-1 was transfected just prior to transplantation, neointima formation was dramatically reduced in association with reduction in cytokines.

MMP is a key molecule required for tissue inflammation and tissue fibrosis. MMP expression is induced on the thickened intima of recipient mice in the chronic rejected heart. Also, in the monkey model, tissue inhibitor of MMP was induced in the outer area of the occluded vessel. These investigators tried to block the expression of MMP-2 by using a transfection of ribozymes to the heart allograft. MMP-2 ribozyme nearly completely blocked the neointima formation.

Cytokines also play an important role in inflammation. TNF receptor knockout mice (receptor 1, receptor 2, and double knockout) were studied and showed that neointima hyperplasia was blocked only when the double knockout mice were transplanted into the wild-type recipient. The TNF receptors on the donor heart are essential for the development of neointima hyperplasia.

They tried to block inflammation using NF-κB, a key transcription factor required for inflammation (Figure 2 ). When synthetic DNA was transfected into the cell, the DNA works as a decoy and blocks the production of inflammatory molecules such as cell adhesion molecules and cytokines. When the decoy is transfected into the recipient mice heart just prior to transplantation, neointimal formation is nearly completely blocked.  This reduction in chronic rejection is associated with the reduction of the expression of VCAM-1 and PDGF transcription.

An hypothesis for the mechanism of GAD developed by these investigators based on their data and that from other investigators and from clinical observations is shown in Figure 3.

Inflammation of coronary arteries plays a key role. Inflammation is caused by several factors: acute rejection causing inflammation of the coronary arteries;  cytomegalovirus infection causing inflammation directly or through acute rejection; ischemia and reperfusion injury during the cold ischemia for preservation of the donor heart; mechanical injury of the donor heart during transport to the recipient hospital; immunosuppressants, hypertension, hyperlipidemia, and diabetes cause endothelial damage that promotes inflammation. After the damage activation, the coronary arteries show intimal hyperplasia and subsequent remodeling of the coronary arteries.

GAD of the transplanted heart and the restenosis after PTCA can occur on the same molecular basis as inflammation. The stimulus that causes GAD is immune injury to the whole coronary arteries. Local mechanical injury is the stimulus after PTCA.

In conclusion, T cell-mediated immunity and inflammation are tightly involved in the pathogenesis of GAD. Interventions on T cell-mediated immunity through suppression of costimulatory pathways and inhibition of inflammation by gene therapy would be effective for the attenuation or prevention of GAD.