Vein graft disease in a knockout mouse model of hyperhomocysteinaemia

Summary

A major reason for vein graft failure after coronary artery bypass grafting is neointimal hyperplasia and thrombosis. Elevated serum levels of homocysteine (Hcy) are associated with higher incidence of cardiovascular disease, but homocysteine levels also tend to increase during the first weeks or months after cardiac surgery. To investigate this further, C57BL/6J mice (WT) and cystathionine‐beta‐synthase heterozygous knockout mice (CBS+/−), a mouse model for hyperhomocysteinaemia, underwent interposition of the vena cava of donor mice into the carotid artery of recipient mice. Two experimental groups were examined: 20 mice of each group underwent bypass surgery (group 1: WT donor and WT recipient; group 2: CBS+/− donor and CBS+/− recipient). After 4 weeks, the veins were harvested, dehydrated, paraffin‐embedded, stained and analysed by histomorphology and immunohistochemistry. Additionally, serum Hcy levels in CBS knockout animals and in WT animals before and after bypass surgery were measured. At 4 weeks postoperatively, group 2 mice showed a higher percentage of thrombosis compared to controls, a threefold increase in neointima formation, higher general vascularization, a lower percentage of elastic fibres with shortage and fragmentation in the neointima, a lower percentage of acid mucopolysaccharides in the neointima and a more intense fibrosis in the neointima and media. In conclusion, hyperhomocysteinaemic cystathionine‐beta‐synthase knockout mice can play an important role in the study of mechanisms of vein graft failure. But further in vitro and in vivo studies are necessary to answer the question whether or not homocysteine itself or a related metabolic factor is the key aetiologic agent for accelerated vein graft disease.

Coronary artery disease is one of the most common causes of death in western countries, responsible for an enormous economic burden. Coronary artery bypass grafting (CABG) is effective in relieving the symptoms of angina and prolongs survival. However, atherosclerotic changes in saphenous vein grafts (SVGs) after CABG are emerging as the major determinant of long‐term vein graft viability.

Saphenous vein graft disease is composed of three processes: thrombosis, intimal hyperplasia and atherosclerosis. In short, surgically induced endothelial cell damage leads to neointimal proliferation, caused by migration of activated smooth muscle cells from the medial to the intimal layer of venous vascular walls and transformation from a contractile to a proliferative phenotype. These processes are mediated by growth factors and cytokines released from platelets, endothelial cells and macrophages.

Additionally, synthesis and deposition of extracellular matrix by activated smooth muscle cells result in a progressive increase in intimal fibrosis (Soyombo et al. 1995).

Several factors including smoking, high cholesterol levels and diabetes mellitus have been identified as risk factors for SVG disease. But also hyperhomocysteinaemia (HHcy) is widely regarded as an independent risk factor for cardiovascular disease, atherosclerosis, arterial and venous thrombosis, myocardial infarction and stroke.

A number of cardiovascular risk factors have been linked to HHcy, including male gender, advancing age, smoking, high blood pressure, elevated cholesterol and lack of exercise. Furthermore, rare genetic disorders, as well as hormonal abnormalities, such as thyroid deficiency and decreased oestrogen levels, and the use of methotrexate, cyclosporine, metformin, nitrous oxide, l‐dopa, carbamazepine, phenytoin and the lipid‐lowering agents colestipol and niacin can also raise homocysteine levels; but the most common cause of elevated homocysteine levels is deficiency of folate, vitamin B6 or vitamin B12.

Homocysteine (Hcy), a product resulting from the metabolism of the amino acid methionine – a non‐essential, non‐protein‐forming, sulphur‐containing amino acid, produced from the metabolism of dietary methionine – was first reported by Butz and du Vigneaud (1932). The biochemical finding of homocystinuria was first described in children in 1962 (Carson & Neill 1962; Gerritsen et al. 1962).

Free homocysteine and homocysteine thiolactone (Hcy‐th) are highly reactive molecules, forming non‐covalent and covalent bonds with a variety of molecules, leading to loss of electrical charge, conformational changes or even precipitation of the protein, resulting in loss or degradation of the biological function of multiple enzymes, receptors, growth factors and structural proteins (Gao et al. 1996; Krumdieck & Prince 2000).

Severe HHcy (>100 μmol/l) results from rare homozygous genetic defects such as deficiency of cystathionine‐beta‐synthase (CBS), of methylenetetrahydrofolate reductase (MTHFR) or of enzymes involved in methyl‐B₁₂ synthesis and homocysteine methylation.

Several CBS mutations are known: the most frequent are 833TC and 919GA, located in exon 8; and 1224‐2AC, which causes the deletion of the entirety of exon 12.

This deficiency in CBS is characterized by arteriosclerosis, thromboembolic complications, skeletal abnormalities, ectopia lentis, mental retardation and a predisposition to schizophrenia and epilepsy (Robert et al. 2003).

An association between elevated blood levels of homocysteine (Hcy) and cardiovascular disease was first described by McCully (1969), who observed premature atherothrombosis of the peripheral, coronary and cerebral vasculature in children with homocystinuria.

Seven years later, Wilcken and Wilcken (1976) showed that the concentration of homocysteine–cysteine mixed disulphide after a methionine load was slightly higher in patients with coronary heart disease than in controls.

Homocysteine causes endothelial cell desquamation, inhibits DNA synthesis in vascular endothelial cells and is responsible for growth arrest at the G1 phase of cell cycle (Starkebaum & Harlan 1986; Dudman et al. 1999). Pathological concentrations of homocysteine also increase the interaction between neutrophils and endothelial cells. This results in neutrophil migration across the endothelium with concurrent damage and detachment of endothelial cells (Heinecke et al. 1987), smooth muscle cell proliferation and intimal thickening. It is suggested that smooth vascular cell proliferation may be directly stimulated by homocysteine or may be secondary to the mitogenic effect of endothelial and/or platelet‐derived growth factors (PDGFs) released by homocysteine‐induced endothelial cell damage.

Homocysteine induces mitogenesis in vascular smooth muscle cells by stimulating MAP kinase signal transduction pathway and by the induction of C‐fos and C‐myc genes (Upchurch et al. 1997; Dudman et al. 1999).

Furthermore, homocysteine stimulates platelet generation of thromboxane A2 which is a vasoconstrictor and pro‐aggregant. Hyperhomocysteinaemia causes activation of factor V and interferes with protein C activation and thrombomodulin expression, the bioavailability of prostacyclin and the expression of anti‐coagulant substance heparan sulphate. Platelet activation due to HHcy results in thrombosis and smooth muscle proliferation and is therefore important in the pathogenesis of atherothrombosis (Ross 1986; Poddar et al. 2001).

Homocysteine also plays a role in activating matrix metalloproteinases via oxidative stress and has also been shown to utilize the extracellular signal‐related kinases (ERK) pathway (Vacek et al. 2015).

Additionally, homocysteine reduces the bioavailability of nitric oxide (NO)² and is thought to cause deterioration of the elastic structure of the arterial wall through alteration in metalloproteinase activity. It may also increase vascular rigidity by augmenting breakdown of elastin in vascular cells (Hiltunen & Ylä‐Herttuala 2003).

The purpose of this study was to investigate vein graft changes 4 weeks after bypass surgery in a heterozygous knockout mouse model of HHcy in comparison with a control group.

Materials and methods

Mice and vein grafting

C57BL/6J mice and cystathionine‐beta‐synthase heterozygous knockout mice were purchased from Harlan‐Winkelmann (Borchen, Germany). They were maintained at 24°C and received food and water ad libitum.

The operation was performed as described previously (Zou et al. 1998). In brief, the inferior vena cava was harvested from a donor mouse. In the recipient mouse, the right common artery was exposed, ligated and divided and the two stumps were everted over a nylon cuff. The vein from the donor mouse was then interposed between the carotid artery cuffs (Figures 1 and 2).

Figure 1.

(a + b) Bypass surgery with the interposed vein graft into the carotid artery (scale bar = 1 cm).

Figure 2.

(a) Native vein of a C57BL/6J mouse, (b) 4 weeks after bypass surgery with increased vessel wall thickness (haematoxylin–eosin, original magnification ×10).

Tissue preparation

Two groups were performed: group 1 – WT donor and WT recipient; and group 2 – CBS+/− donor and CBS+/− recipient.

Twenty mice of each group underwent surgical procedure and were sacrificed 4 weeks postoperatively. The interposed vein segments were cut out at the cuff ends, embedded in mouse liver segments, fixed with 4% phosphate‐buffered formaldehyde for 24 h and then dehydrated and wax‐embedded.

Histomorphology and immunohistochemistry

Sections of 2 μm thickness were routinely stained by haematoxylin and eosin, elastica staining for verifying elastic fibres, Gordon and Sweet's staining for visualizing reticulin fibres, alcian blue staining for examination of acid mucopolysaccharides and Masson's trichrome staining for the assessment of vessel wall fibrosis.

For immunohistochemistry, the following rabbit polyclonal antibodies (all Abcam, Cambridge, UK) were used: smooth muscle actin antibody (1:50, ab5694), anti‐Ki‐67 antibody (1:1000, clone SP6; ab16667), metalloproteinase‐2 (MMP2) antibody (1:50, ab37150) and metalloproteinase‐9 (MMP9) antibody (1:100, ab38898). Smooth muscle actin and MMP9 antibody needed microwave antigen retrieval with 10 mM citrate buffer (pH 6.0).

For assessment of the neointimal and media thickness and luminal area, pictures were taken under light microscope at a 10‐fold magnification. Measurements were performed using a computerized image analysis system (imageJ 1.32j software for Java; National Institutes of Health, Bethesda, MD, USA). For achieving reproducible results, the cross sections of the veins were divided into four quadrants and four measurements in each quadrant were performed. The median value of all measurements was regarded as representative for the neointimal and media thickness.

The occurrence of elastic fibres, reticulin fibres, fibrosis, acid mucopolysaccharides and the re‐endothelialization was given as percentage of the area. The number of capillaries in intima, media and adventitia was counted in 10 high‐power fields (HPF) at 40‐fold magnification. For inflammation grading, a semiquantitative score system was used: (i) mild inflammation/inflammation in 1/3 of the circumference of the vessel wall; (ii) moderate inflammation/inflammation in 2/3 of the circumference; and (iii) intense inflammation/inflammation throughout the whole circumference.

The number of positively stained vascular smooth muscle cells was counted in HPF at 40‐fold magnification. The number of positive cells was extrapolated to 1 mm².

Measurement of serum Hcy levels

Mouse plasma of WT animals and CBS‐deficient mice was collected before and after bypass surgery. Serum Hcy levels were measured by mass spectrometry.

Statistical analysis

All data are expressed as mean ± standard error of the mean (SEM). The spss for Windows statistical software package (spss 22.0; Microsoft, Redmond, WA, USA) was used for analysis. The Mann–Whitney U‐test was used to compare histomorphological measurements of both groups (group 1 – CBS+/− donor and CBS+/− recipient; group 2 – WT donor and WT recipient).

The bivariate correlation between continuous variables was calculated using Spearman's rho correlation coefficient. P‐values < 0.05 were considered statistically significant.

Ethical Approval

All animal procedures were approved by the institutional ethics committee at the Medical University Innsbruck and were performed according to protocols approved by the Austrian Ministry of Science in accordance with §8 of the law on animal experiments. All animals were treated according to the Guide for the Use and Care of Laboratory Animals, published by the National Institutes of Health (NIH Publications no. 86‐23, revised 1985).

Results

Twenty mice of each group (group 1 – WT donor and WT recipient; group 2 – CBS+/− donor and CBS+/− recipient) underwent bypass grafting and were sacrificed 4 weeks after surgery (results in Table 1).

Table 1.

Characteristics of the different groups after bypass surgery

	GROUP 1 (n = 20)	GROUP 2 (n = 20)	P‐value
Intimal thickness (μm)	82.15 ± 64.46	245.59 ± 75.72	<0.001
Media thickness (μm)	155.08 ± 51.61	124.28 ± 40.22	n.s.
Luminal area (mm2)	2.17 ± 1.20	1.98 ± 1.69	n.s.
Re‐endothelialization (%)	87.50 ± 11.39	80.00 ± 18.37	n.s.
Vascularization (n/10 high‐power fields)
Intima	0 ± 0	0 ± 0	n.s.
Media	7.93 ± 4.36	20.00 ± 6.20	<0.001
Adventitia	25.79 ± 7.74	29.20 ± 13.35	n.s.
Vessels in general	33.71 ± 7.25	49.20 ± 16.74	<0.01
Inflammation grading (score)
Intima	1.36 ± 0.63	1.20 ± 0.45	n.s.
Media	2.36 ± 0.63	2.00 ± 1.00	n.s.
Adventitia	1.64 ± 0.50	2.00 ± 0.71	n.s.
Elastic fibres in the intima (%)	46.67 ± 25.82	38.00 ± 16.43	0.03
Elastic fibres in the media (%)	11.67 ± 4.08	8.80 ± 10.28	n.s.
Acid mucopolysaccharides in the intima (%)	37.14 ± 4.88	20.00 ± 23.45	0.03
Fibrosis in the intima (%)	21.00 ± 0	49.13 ± 5.48	<0.01
Fibrosis in the media (%)	26.67 ± 19.66	39.00 ± 10.95	0.03
Fibrosis in the adventitia (%)	33.33 ± 17.51	22.00 ± 21.68	n.s.
SMA+ cells in the neointima (mm2)	450 ± 130	630 ± 350	n.s.

Group 1: WT donor and WT recipient; Group 2: CBS+/− donor and CBS+/− recipient. All data are expressed as mean ± SEM.

In general, we observed a higher rate of thrombosis in CBS knockout mice than in controls; 10 of 40 transplanted vein grafts (25%) developed thrombosis at 4 weeks postoperatively [two veins in group 1 (10%) and eight veins in group 2 (40%); Figure 3a]. Although HHcy is implicated as a hazardous prothrombotic trigger, acute thrombosis in vein grafts is also caused by surgical trauma, mechanical stress and vessel wall cell damage or loss, occurring with even the most careful handling (Motwani & Topol 1998; Torsney et al. 2004). But nevertheless, rapidly accumulating evidence links elevated homocysteine levels to thrombosis via several mechanisms such as increased tissue factor expression, attenuated anti‐coagulant processes, enhanced platelet reactivity, increased thrombin generation, augmented factor V activity, impaired fibrinolytic potential, and vascular injury, including endothelial dysfunction. Molecular mechanisms underlying prothrombotic actions of homocysteine are incompletely understood and involve oxidative stress, DNA hypomethylation and proinflammatory effects.

Figure 3.

(a) Venous bypass graft of a cystathionine‐beta‐synthase heterozygous knockout mouse (group 2) 4 weeks after surgery with thrombus formation (haematoxylin–eosin, original magnification ×10). (b) Group 1 vein graft (control) with calcification in the intima (haematoxylin–eosin, original magnification ×10, white arrow). (c + d) Group 2 vein graft with ossification and bone marrow formation (haematoxylin–eosin; c – original magnification ×10; d – original magnification ×40, black arrows).

Two veins of group 1 (10%) showed coarse calcification in the neointima (Figure 3b). Furthermore, two vein grafts of group 2 (10%) showed metaplastic ossification in the neointima (Figure 3c,d), one with bone marrow formation and the other one revealing additional metaplastic hyaline cartilage. On the one hand, ossification is a rather unusual finding in the case of HHcy, which is generally associated with several skeletal malformations, osteopenia and osteoporosis, but on the other hand, calcifications and metaplastic ossifications are common findings in vein graft atheroma and atherosclerotic lesions, first described by Benivieni in 1507 and Fallopius in 1575 and also noticed by Virchow in 1863.

Vein grafts of group 2 mice developed an increased total vessel wall thickness and a significant neointimal hyperplasia compared to ungrafted native venae cavae and controls (245.59 ± 75.72 μm vs. 82.15 ± 64.46 μm in group 1, P < 0.001; Figure 4a,b), which is in accordance with other studies (Dudman et al. 1999; Tan et al. 2014).

Figure 4.

Vein grafts 4 weeks after bypass surgery: group 1 vein graft (WT donor and WT recipient) on the left side and group 2 vein graft (CBS+/− donor and CBS+/− recipient) on the right side. (a + b) Haematoxylin–eosin (black line – neointima; white line – media), (c + d) elastica staining verifying elastic fibres, (e + f) Masson's trichrome staining visualizing intima and media fibrosis, (g + h) Gordon and Sweet's staining showing the dense reticulin network and (i + j) immunohistochemistry with anti‐SMA antibody (original magnification ×20).

Media thickness was slightly higher in group 1, but the difference was not significant (155.08 ± 51.61 μm in group 1 vs. 124.28 ± 40.22 μm in group 2, P = n.s.).

Group 2 mice also revealed significantly increased total numbers of capillaries in the vein graft wall and an increased capillary density in the media compared to group 1 animals [20.00 ± 6.20 vessels vs. 7.93 ± 4.36 vessels (P < 0.001)].

Although there are contradictory results concerning the angiogenesis in HHcy, one may speculate that the higher number of vessels is due to an increased expression of vascular endothelial growth factor (VEGF) (Roybal et al. 2004).

Reports from the literature vary on the correlation between VEGF and HHcy from positive in some studies to negative or no correlation in other studies in relation to ischaemia and other disease conditions (Maeda et al. 2003; Bosch‐Marcé et al. 2005; Lee et al. 2007; Yan et al. 2010; Sen et al. 2012). These reports indicate that homocysteine could be a potential regulator of VEGF levels and angiogenesis.

It has already been demonstrated that HHcy can induce endothelial dysfunction in part by promoting oxidant stress, as illustrated in cellular studies (Upchurch et al. 1997; Heydrick et al. 2004) and in various animal models (Eberhardt et al. 2000; Tan et al. 2006; Dayal & Lentz 2007, 2008). These results are in accordance with our study, which showed a lower re‐endothelialization 4 weeks after surgery in group 2 (80% of the circumference vs. 87% in group 1), although these differences reached no significance. Endothelial cells were flat and inconspicuous in both groups without swelling or vacuolization.

Group 2 animals also exhibited a slight luminal narrowing (with a luminal area of 1.98 ± 1.69 mm²) compared to the controls (2.17 ± 1.20 mm², P = n.s.).

In contrast, native veins of C57BL/6J mice before bypass grafting showed a luminal area of 0.35 ± 0.19 mm²; native veins of CBS knockout mice showed a luminal area of 0.36 ± 0.13 mm².

Although HHcy is reported to enhance vascular inflammation (Hofmann et al. 2001) and promotes inflammatory myocyte generation in murine mouse models (Zhang et al. 2009; Fang et al. 2014), we observed no significant differences concerning the inflammation score, the inflammation grading and the composition of the inflammation cells in the vein graft wall. In both groups, the inflammation was more intense in the media and adventitia and less distinctive in the intima.

Generally, the inflammation infiltrate consisted of neutrophil granulocytes, lymphocytes, macrophages, plasma cells and some histiocytes. Some veins showed additionally foreign body reaction in media or adventitia with foreign body granulomas and phagocytosis of suture material.

Surprisingly, immunohistochemical staining of the vein grafts with MMP2 and MMP9 was negative in both groups, although homocysteine has been shown to activate MMPs via an increase in oxidative stress and to act as a signalling molecule on receptors such as the peroxisome proliferator‐activated receptor‐γ and N‐methyl‐d‐aspartate receptor.

Homocysteine is able to modulate the activity of matrix metalloproteinases (MMPs) and their inhibitors leading to elastin degradation and accumulation of collagen within the intima and media in a dose‐dependent manner. In a study by Southgate et al. (1999), MMP2 was found in the neointima 1–4 weeks postoperatively, and MMP9 was localized in the neointima and media starting from day 2 (and increased until day 28) after bypass grafting. Obviously, matrix degradation is a time‐dependent process which occurs in the early days and weeks after bypass surgery and is already completed 4 weeks postoperatively. Our study shows that in this advanced state of vein graft disease, fibrosis is the predominant process.

Due to the metalloproteinase‐related elastolysis, group 2 animals revealed a significant loss of elastic fibres in the neointimal area (38.00 ± 16.43%) in comparison with group 1 (46.67 ± 25.82%, P = 0.03).

Elastic fibres in group 2 were short and fragmented, and also the internal elastic lamina consisted of short, fragmented elastic fibres in contrast to group 1 (Figure 4c,d).

Group 2 mice developed a significantly higher percentage of intimal fibrosis (49.13 ± 5.48%) compared to group 1 (21.00 ± 0%, P < 0.01) and also a significantly increased media fibrosis (39.00 ± 10.95% vs. 26.67 ± 19.66%, P = 0.03; Figure 4e,f). Intimal fibrosis was continuous throughout the whole circumference without intimal plaque formation.

Furthermore, the percentage of acid mucopolysaccharides in the intima was significantly higher in group 1 (37.14 ± 4.88%) compared to group 2 (20.00 ± 23.45%, P = 0.03), and acid mucopolysaccharides were completely absent in the media and adventitia of both groups.

The reticulin fibre network could be recognized in all three layers (intima, media and adventitia) in the vein graft wall of both groups with no significant difference, although it seemed less intense in the adventitia (Figure 4g,h).

The Ki‐67 proliferation index was the same in both groups (1% in 10 HPF) with only a few positive vascular smooth muscle cells in the intima, media and adventitia of the vessel wall, although it is hypothesized that HHcy promotes smooth vascular cell proliferation and migration (Motwani & Topol 1998; Carmody et al. 1999; Chen et al. 2000; Tan et al. 2014). Our results lead to the conclusion that maybe only smooth vascular cell migration and not proliferation is responsible for intimal hyperplasia – or proliferation is a time‐dependent process, which has already finished 4 weeks after bypass surgery.

Immunohistochemical staining with smooth muscle actin antibody exhibited distinct expression predominantly in the neointima and little expression in the media and adventitia of both groups. Group 2 had a higher number of SMA‐positive cells in the neointima (630 ± 350), but there were no significant differences between the two groups (Figure 4i,j).

Serum Hcy levels of CBS knockout animals before bypass surgery were 3.9 ± 1.6 μM/l. This was significantly higher than that in wild‐type animals (1.5 ± 0.6 μM/l (P = 0.01)), which is due to the deletion of the enzyme and the resulting accumulation of homocysteine. After bypass surgery, serum Hcy levels were increased (7.2 ± 1.2 μM/l in CBS knockout mice vs. 3.4 ± 0.8 μM/l in WT animals; P = 0.01). These results are in accordance with some studies, which have already shown that homocysteine levels tend to increase during the first weeks or months after cardiac surgery procedures in the vein graft disease group (Iwama et al. 1998, 2001), suggesting that an elevation of homocysteine levels is related to SVG disease after CABG. Additionally, HHcy is significantly associated with total and cardiovascular mortality after CABG (Girelli et al. 2006).

Discussion

Coronary artery bypass grafting has become widely accepted and established as an effective therapy for coronary artery disease. However, its long‐term efficacy is limited by SVG disease. Not only is HHcy an independent risk factor for cardiovascular disease, but also increased levels of homocysteine could be observed after coronary artery bypass surgery (Iwama et al. 1998, 2001).

Animal models as the cystathionine‐beta‐synthase knockout mouse, first reported by Watanabe et al. (1995), are useful and essential tools, providing a unique opportunity to study biochemical consequences of a defective cystathionine‐beta‐synthase enzyme.

Hyperhomocysteinaemia in mouse models can be produced through dietary modifications, genetic approaches or a combination of dietary and genetic interventions, or by pharmacological approaches. Genetic approaches include the disruption of the CBS gene, the MTHFR gene, the methionine synthase (MTR) or the methionine synthase reductase (MSR) gene (Chen et al. 2001; Swanson et al. 2001; Elmore et al. 2007).

Cystathionine‐β‐synthase (CBS) is a multidomain enzyme that is encoded by the CBS gene located on chromosome 21q22.3 and catalyses the first step of the transsulphuration pathway, from homocysteine to cystathionine.

In 1995, Watanabe et al. reported the generation of a knockout mouse completely lacking cystathionine synthase with a 40 times higher plasma level of homocysteine.

Homozygous knockout mice (CBS−/−) were born at the expected frequency from matings of heterozygotes, but they suffered from severe growth retardation and a majority of them died within 5 weeks after birth.

They developed early thrombotic events (51% peripheral veins, 32% cerebrovascular accidents, 11% peripheral arteries, 4% myocardial infarctions), abdominal aortic aneurysm, mental retardation, psychiatric disorders, seizures, marfanoid habitus, osteoporosis, ectopia lentis, hyperpigmentation, pancreatitis and spontaneous pneumothorax. They exhibit similar disease manifestations compared with human homozygous patients, but human patients with CBS deficiency do not exhibit phenotypes as severe as homozygous CBS‐deficient mice, because most human homozygous mutations maintain at least some level of CBS activity.

Human arterial wall changes in large, medium‐sized and small arteries in hyperhomocysteinaemic patients include fibrous intimal plaque formation; disruption of the internal elastic membrane with reduplication, fraying, thickening and discontinuity of elastic fibres; proliferation of intimal loose fibrous connective tissue; enlarged vacuolated endothelial cells; swelling and hyperplasia of the endothelial cells; moderate thickening of the media; narrowing of the lumen; disorganization and fibrosis of the media; explicable and loss of elastic tissue of the media (Robert et al. 2003).

Beard and Bearden (2011) observed an increased vessel wall thickness, lower ratios of non‐distensible (collagen and basement membrane) to distensible (elastin and smooth muscle) components, an increased matrix fibre deposition and increased expression and activities of matrix metalloproteinases (MMP2 and MMP9), fragmented elastic fibres, an impaired angiogenesis and proliferation of endothelial cells, and a promotion of smooth muscle cells in the aorta of CBS−/− mice.

Although morphological changes in the aorta of CBS‐deficient mice are well described, saphenous venous wall changes after bypass surgery are rarely studied. With this study, we tried to add some information for a better understanding of the complex molecular mechanisms, but further in vivo and in vitro studies are required.

Vein graft disease after bypass surgery is a multifactorial disease, involving endothelial injury in the form of extensive venous distension, denudation of the endothelium itself and some vasospasm occurring during harvest (LoGerfo et al. 1983). After the vein is harvested, the initial injury causes a decrease in nitric oxide. Following the ischaemic period and after implantation, nitric oxide synthesis will increase due to the reperfusion. Re‐implantation leads to a release of multiple growth factors, and cytokines that cause the migration of vascular smooth muscle cells and formation of extracellular matrix into the intimal compartment of the vein graft. Once neutrophils are adherent, they initiate further endothelial damage and activation of the coagulation cascade which can lead to thrombosis (Shuhaiber et al. 2002).

Neointimal formation is stimulated by several factors including PDGF, transforming growth factor‐beta and epidermal growth factor, which cause subsequent invasion of the smooth muscle cells into the intimal layer (Shuhaiber et al. 2002).

Additionally, matrix metalloproteinases, the mediators of matrix deposition and degradation, are important factors for cell growth and proliferation, cell migration, organ development, reproduction and tissue remodelling.

High levels of homocysteine activate and damage endothelial cells by the generation of reactive oxygen species combined with the removal of EC‐protective antioxidant mechanisms such as NO and glutathione. Hyperhomocysteinaemia is associated with increased activation of MMP2 and MMP9, leading to elastolysis and enhanced stiffness of the vein graft wall. It also promotes platelet adhesion to endothelial cells and has been associated with higher levels of prothrombotic factors, for example β‐thromboglobulin, tissue plasminogen activator and factor VIIc. Furthermore, increased levels of homocysteine result in vascular smooth muscle cell migration, interleukin‐6 production by human saphenous vein endothelial cells, IL‐1β production by human peripheral blood monocytes and TNF‐alpha production by monocyte‐derived macrophages (Dalal et al. 2003).

Hyperhomocysteinaemia induces connective tissue growth factor expression in vascular smooth muscle cells (Liu et al. 2008), upregulates PDGF levels via DNA demethylation in endothelial cells, affects crosstalk between endothelial cells and vascular smooth muscle cells (VSMCs) and leads to VSMC activation (Zhang et al. 2012).

Taken together, all these findings highlight the complexity of molecular events that are involved in vein graft disease. After CABG, HHcy interferes at certain molecular levels leading to accelerated development of neointimal formation and vein graft thrombosis.

Conclusion

In conclusion, our hyperhomocysteinaemic knockout mouse model of vein graft disease showed a fourfold higher percentage of thrombosis, a threefold increased neointima formation, a higher general vascularization, a lower percentage of elastic fibres with shortage and fragmentation in neointima and media, a lower percentage of acid mucopolysaccharides in the intima and a more intense fibrosis in the neointima and media compared to the control group 4 weeks after bypass surgery.

We could not find any significant differences between both groups regarding the inflammatory infiltrate, the luminal area, the number of SMA‐positive cells, the expression of MMP2 and MMP9, the Ki‐67 proliferation index or the reticulin fibre network.

In summary, HHcy is responsible for structural alterations in the vessel wall and the higher stiffness of the vein graft due to increased fibrosis and neointimal formation.

But whether homocysteine itself or a related metabolic factor is the key aetiologic agent, however, still remains an open question.

A future challenge, with important clinical implications, will be to design experimental approaches to distinguish between direct and indirect vascular effects of HHcy.

So further in vitro and in vivo studies are needed for a better understanding of the complex interplay between the various risk factors, which will be very beneficial in devising lifestyle modifications and pharmacological interventions to deal with vein graft disease and vein graft thrombosis after coronary artery bypass surgery.

Conflict of Interest

The authors declare that they have no conflict of interest.

Funding source

This study was not funded by any source.