- Email: [email protected]
Serum VEGF—As a prognostic factor of atherosclerosis
Atherosclerosis 194 (2007) 182–188
Serum VEGF—As a prognostic factor of atherosclerosis Kuriko Kimura a , Teruto Hashiguchi b,∗ , Takahisa Deguchi a , Shuji Horinouchi a , Tadashi Uto a , Hiroko Oku a , Shiro Setoyama c , Ikuro Maruyama b , Mitsuhiro Osame a , Kimiyoshi Arimura a a
Department of Neurology and Geriatrics, Graduate School of Medical and Dental Sciences, Kagoshima University, Japan b Laboratory and Vascular Medicine, Graduate School of Medical and Dental Sciences, Kagoshima University, 8-35-1 Sakuragaoka, Kagoshima 890-8520, Japan c Kagoshima Prefectural Comprehensive Health Center, Kagoshima, Japan Received 8 October 2005; received in revised form 5 June 2006; accepted 24 July 2006 Available online 1 December 2006
Abstract Vascular endothelial growth factor (VEGF) has been noted in the pathogenesis of atherosclerosis. To examine the usefulness of the serum concentration of VEGF as an index of atherosclerosis, we analyzed the serum VEGF concentrations in 443 adults who underwent a medical checkup. The mean serum VEGF concentration of men (229 ± 147 pg/ml) was significantly higher than that of women (182 ± 112 pg/ml). The platelet count showed a slight correlation with the serum VEGF concentration in both genders (men R = 0.287, women R = 0.296), corresponding with the results of experiments that platelets are the major source of VEGF in circulating peripheral blood. In men, the serum VEGF concentrations correlated with platelet counts, body fat percentages, leukocyte counts, and HDL-cholesterol concentrations (negative correlation). In the multiple regression analysis performed for men’s serum VEGF concentrations, the decision coefficient (R2) was maximized (R2 = 0.173) when the leukocyte count, the body fat percentage, and the HDL-cholesterol concentration were taken into account besides the platelet count. Male smokers’ serum VEGF concentrations were higher than non-smokers’. Smoking in men significantly affected the sex difference in the serum VEGF concentration, leukocyte count, and HDL-cholesterol concentration. We concluded that the serum VEGF concentration might be closely related to atherosclerosis accelerating factor, especially in men. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: VEGF; Vascular endothelial growth factor; Atherosclerosis; Pathogenesis; Prognosis; Men; Women; Serum; Plasma; Platelet; Platelets; Fat; HDL; Cholesterol; Leukocyte; Smoking; Smoker; Estrogen; Menstruation; Menopause; C-reactive protein; CRP; hsCRP
1. Introduction It has been thought that atherosclerosis is a result of chronic inflammation of blood vessels. The failure of the restoration mechanism leads to chronic inflammation and thus gradually develops atherosclerosis [1,2]. At present, there are some methods which detect atherosclerosis in the early stage, such as the concentration of serum C-reactive protein [3,4], ultrasonographic measurement of intima-media thickness (IMT) of the carotid arteries [5], and measurement of pulse wave velocity (PWV) [6]. The vascular endothe∗
Corresponding author. Tel.: +81 99 275 5437; fax: +81 99 275 2629. E-mail address: [email protected] (T. Hashiguchi).
0021-9150/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2006.07.025
lial growth factor (VEGF) is a multifunctional cytokine which increases vascular permeability and is mitogenic for endothelial cells, acting early and at most points of the angiogenic cascade [7,8]. The role of VEGF in the molecular mechanism of atherosclerotic lesions remains controversial. Some studies have reported that VEGF inhibits thickening of the media by promoting regeneration of vascular endothelial cells and improving endothelial function. Others have indicated that VEGF promotes atherosclerosis via promotion of monocyte chemotaxis and plaque neovascularization [9]. Furthermore, several studies have investigated the association between the plasma/serum concentrations of VEGF and atherosclerosis-associated disorders such as diabetes and hypertension [10–15]. However, no study has
K. Kimura et al. / Atherosclerosis 194 (2007) 182–188
clarified differences between plasma VEGF and serum VEGF. We emphasize that the serum VEGF but not plasma VEGF level in a population including healthy adults is closely associated with risk factors for atherosclerosis, and that VEGF in circulating peripheral blood is derived from platelets.
2. Materials and methods 2.1. Subjects The subjects of this study were 443 continuous outpatients who attended the general health center of Kagoshima prefecture. They underwent a health check for the prevention of life-style related diseases from the period of October 2000 to November 2000 and July 2001 to September 2001. We obtained informed consent in writing from all the people enrolled. The study population consisted of 304 males and 139 females (average age, 49.31 ± 8.81 and 50.68 ± 8.40 years, respectively, p = 0.1339). Ten female and 137 male subjects were smokers. The age range was from 31 to 75 years. As for the details of their health condition, 48 people were diagnosed with hypertension (41 of whom received drug medication), 19 people had hyperlipidemia (8 of whom received medication), 12 people had diabetes mellitus (7 of whom received medication), and 17 people had gout (9 of whom received medication). There were three post-cancer patients, all of whom had undergone resection therapy, and five post-cerebrovascular disease patients (all of whom received medication). Four women had undergone uterus resection operations because of uterus myoma or endometriosis. 2.2. Physical examination Height, body weight, body mass index, body fat percentage, and systolic and diastolic blood pressure were examined. The smoking habit, the Blinkmann Index, the alcohol drinking habit and daily exercise habit were also investigated. 2.3. Laboratory tests and imaging analyses Fasting venous blood was obtained by a traumatic venepuncture into Vacutainers with EDTA-2Na or into tubes free of anticoagulant (for serum collection). The erythrocyte, leukocyte, and platelet count were routinely analyzed. HbA1c was measured by liquid chromatography. Serum aspartate aminotransferase (AST), alanine aminotransferase (ALT), ␥-glutamyl transferase (GTP), total cholesterol, HDLcholesterol, LDL-cholesterol, triglyceride, and amylase were analyzed by using a commercially available kit. A sufficient amount of serum remained for the measurement of serum C-reactive protein (CRP) in 324 subjects (237 males and 87 females), and CRP was measured using a commercially
183
available kit (Dade Behring Marburg GmbH, Germany). For each datum, the correlation with the serum VEGF concentration was examined. Imaging analysis including chest X-ray, barium-contrast study of the stomach, abdominal ultrasonography were also performed. No person with on-going or latent malignant disease was included in this study. 2.4. Measurement of serum VEGF Serum VEGF concentration was measured in duplicate using a commercial ELISA kit (R&D Systems, Minneapolis, MN). The assay system used is sensitive to 15.6 pg/ml VEGF and does not cross-react with platelet-derived growth factor (PDGF) or other homologous cytokines. Optical density at 450 nm was measured on the ImmunoMini NJ-2300 (Nippon Inter Med, Tokyo, Japan) and VEGF concentration was determined by linear regression from a standard curve using the VEGF supplied with the kit as standard. 2.5. Statistical analyses Data are presented as means ± S.D. Differences between data sets were evaluated by analysis of variance (ANOVA) and the Scheff´e test was used as a post hoc test. Paired data were compared by paired t-test. A simple linear regression analysis was used to evaluate the correlation between the two parameters. A multiple regression analysis was performed to seek independent relationships between serum VEGF and four possible markers of pathogenic processes. All statistical analyses were performed with the StatView 5.0 software program (Abacus, Berkeley, CA) for Windows; p < 0.05 was considered statistically significant. 2.6. Quantitative analysis of serum VEGF derived from platelets Serum and plasma from 13 control health subjects were obtained on the same day. Platelet-rich plasma (PRP) and platelet-poor plasma (PPP) were freshly prepared as described previously [16]. Cell-DYN (Dainabot, Tokyo, Japan) was used for hematological counts and to assess the degree of contamination of PRP or PPP by other cells. Thrombin and calcium activation of PRP and PPP was carried out by adding human thrombin (Sigma Chemical Co., St. Louis, MO)/calcium chloride solution to a final concentration of 5 U/ml/10 mM to ensure maximal clotting. After standing at room temperature for 30 min, samples were centrifuged at 750 × g at 4 ◦ C for 20 min and the supernatants were removed for measurement of VEGF concentration. VEGF released from a single platelet and platelet-derived VEGF in serum (pg/ml) were obtained as described previously [16].
184
K. Kimura et al. / Atherosclerosis 194 (2007) 182–188
3. Results 3.1. Difference of the serum VEGF concentration by sex The average serum VEGF concentration of all subjects was 214 ± 139 pg/ml. The maximum value was 830 pg/ml, and the minimum value was less than the measurement sensitivity. According to the t-test, the men’s serum VEGF concentrations are significantly higher than the women’s (p = 0.0008). The average serum VEGF concentration of men is 229 ± 147 pg/ml, and of women, 182 ± 112 pg/ml (Fig. 1). However, the significant difference disappeared when male (n = 167) and female (n = 129) non-smokers were compared (Fig. 1, 209 ± 144 and 183 ± 113 pg/ml, respectively, p = 0.0951). 3.2. Serum VEGF and platelet count showed a positive correlation We classified men and women into five age groups (30s, 40s, 50s, 60s, and 70s) and compared the serum VEGF concentrations, but there was no significant difference between groups. Though the correlation is weak, platelet counts are directly correlated with serum VEGF concentrations both in men and women (men: r = 0.287, p < 0.0001; women: r = 0.296, p = 0.0029) (Fig. 2a and c). 3.3. Four factors which influence the serum VEGF In the case of men, from the tendency of the point distribution on the charts, leukocyte count (r = 0.236, p < 0.0001), the body fat percentage (r = 0.204, p = 0.0004), and HDLcholesterol (r = −0.183, p = 0.0013) showed weak correlation with the serum VEGF concentration (Fig. 3a–c). In women’s data, there was no item that showed correlation except for the platelet count. Multiple regression analysis was performed on the men’s data, and the decision coefficient R2 was maximized (R2 = 0.173, r = 0.416) when leukocyte count, body fat percentage, and HDL-cholesterol con-
Fig. 1. Difference of serum VEGF concentration between men and women.
centration were taken into account besides the platelet counts. 3.4. Serum VEGF and smoking habit in men Men’s serum VEGF concentrations were compared between smokers and non-smokers. The smokers’ average serum VEGF concentration was 255 ± 148 pg/ml and nonsmokers’ was 209 ± 144 pg/ml (p = 0.0063). No significant differences were noted in the platelet count or body fat percentage between male smokers (n = 137) and non-smokers (n = 167) (Figs. 2b and 3e), but the leukocyte count and HDLcholesterol concentration were significantly different (Fig. 3d and f). The Blinkmann index, drinking habit and exercise habit were unrelated to the serum VEGF concentration. 3.5. Serum VEGF and women’s menstruation period Women were classified into the following four groups and serum VEGF concentration compared. The first group consisted of women who have regular menses (n = 56, VEGF 179 ± 111 pg/ml), the second group, irregular menses (n = 11, VEGF 167 ± 93 pg/ml), the third group consisted of menopausal women (n = 60, VEGF 180 ± 111 pg/ml), and the fourth group consisted of those who had undergone uterus
Fig. 2. (a and c) Correlation between serum VEGF concentration and platelet number in peripheral blood (a, men; c, women). (b) Difference of platelet number in the peripheral blood between smokers and non-smokers in men.
K. Kimura et al. / Atherosclerosis 194 (2007) 182–188
185
Fig. 3. (a–c) Correlation between serum VEGF concentration and leukocyte count in the peripheral blood (a), body fat percentage (b), and the concentration of HDL-cholesterol (c). (d–f) Difference of leukocyte count (d), body fat percentage (e), and the concentration of HDL-cholesterol (f) between smokers and non-smokers in men.
resection because of myoma or endometriosis (n = 4, VEGF 299 ± 187 pg/ml). Although the fourth uterus resection group showed high serum VEGF concentrations, with two of the four cases presenting over 400 pg/ml, there were no significant differences between groups. 3.6. Serum C-reactive protein and VEGF In men, the serum CRP concentration was significantly higher in smokers (0.094 ± 0.114 mg/dl) than in non-smokers (0.059 ± 0.084 mg/dl) (p = 0.0063). The serum CRP concentration was weakly correlated with the serum VEGF concentration (Fig. 4a, r = 0.181, p = 0.0052). The serum CRP concentration was also correlated with the leukocyte count (Fig. 4b, r = 0.318, p < 0.0001) and HDL-cholesterol concentration (Fig. 4c, r = −0.142, p = 0.0290), but not with the platelet count or body fat percentage (data not shown).
3.7. Platelet-derived VEGF in serum for healthy control subjects Serum levels of VEGF were significantly higher than the corresponding plasma level as previously reported [16–18]. We analyzed VEGF release in PRP (Fig. 5a and b, Table 1) from 13 healthy control subjects. The mean VEGF concentrations of serum/plasma/PPP were 146 ± 74, 12 ± 9 and 16 ± 14 pg/ml, respectively (Table 1). Clotting of PRP with thrombin/CaCl2 resulted in a significantly higher level of VEGF compared to plasma and PPP level. A significant correlation was found between VEGF in the directly prepared serum and platelet-derived VEGF in serum (r = 0.726, p = 0.0006, Fig. 5a). We also calculated the amount of VEGF released from a single platelet (Table 1). The amount of VEGF released from a single platelet varied from 0.02 to 0.77 (×10−6 pg, 0.25 ± 0.20) and also significantly corre-
Fig. 4. Correlation between serum C-reactive protein concentration and serum VEGF concentration (a), leukocyte count (b), and HDL-cholesterol (c) in men.
186
K. Kimura et al. / Atherosclerosis 194 (2007) 182–188
Table 1 Platelet-derived VEGF in serum for healthy control subjects
PPP, platelet-poor plasma; PRP, platelet-rich plasma; th/Ca2+ , thrombin/CaCl2 ; VEGF released from a single platelet was obtained using the equation: (PRP+th/Ca2+ − Plasma)/Platelet count in PRP.
Fig. 5. Correlation between serum VEGF concentration and platelet-derived VEGF in serum (a) and amount of VEGF released from a single platelet (b).
lated with the directly prepared serum VEGF (r = 0.766, p = 0.0023, Fig. 5b).
4. Discussion Atherosclerosis has been recognized as a disease condition in which low-grade, chronic inflammation causes structure change of the vascular wall [1,2]. In the analysis of data of 443 people who underwent a health check, there was a significant difference in the serum VEGF concentrations between men and women (p = 0.0008; Fig. 1). A multiple regression analysis was performed for men’s serum VEGF concen-
trations, and the decision coefficient (R2) was maximized (R2 = 0.173) when leukocyte count, body fat percentage, and HDL-cholesterol concentration (negative correlation) were taken into account besides the platelet count. As those factors are all related to atherosclerosis or inflammation, we speculate that serum VEGF concentration will be a candidate marker for atherosclerosis development. In laboratory animals, it has been demonstrated that introduction of the VEGF gene led to therapeutic neovascularization, improving tissue ischemia, and that it inhibited thickening of the media after injury via regeneration of vascular endothelial cells and functional improvement. However, the efficacy of VEGF was not suggested in a prospective clinical study of neovascularization therapy using VEGF or in a clinical study targeting restenosis after coronary intervention. A study has reported that up-regulation of VEGF and its receptors, flt-1/Flk-1, is involved in proliferation of intraplaque microvessels in advanced atherosclerotic foci in humans [9]. Other studies have indicated that VEGF promotes atherosclerotic plaque formation in mice [19], and that VEGF promotes atherosclerosis via infiltration of macrophages and mobilization of myelocytes in rabbits [20]. Also oxidized low density lipoproteins reportedly upregulated the VEGF expression in human macrophages [21]. Based on the results of our study, VEGF may be a cytokine that promotes arteriosclerosis in humans. Moreover, the male smokers’ serum VEGF concentrations tended to be higher than the non-smokers’. Smoking habit is widely known as a strong promoting factor of atherosclerosis [22] and it also increased the serum VEGF concentration. Our study revealed that smoking significantly affected the sex difference in the serum VEGF concentration and the leukocyte count and HDL-cholesterol concentration in men (Figs. 1 and 3d and f). In the women’s data, platelet count was the only item that was correlated with serum VEGF concentration. Two of four people who had undergone uterus resection presented over 400 pg/ml of serum VEGF concentration. Women’s mean serum VEGF was below 200 pg/ml, so they presented over twice the mean. As dosing with estrogen was reported to decrease the serum VEGF [10], it may be modified by factors like estrogen. More accumulation of women’s data is needed. The well-known inflammatory marker for atherosclerosis, serum CRP concentration, was higher in male smokers than in non-smokers (p = 0.0063), and correlated with the serum VEGF concentration, leuko-
K. Kimura et al. / Atherosclerosis 194 (2007) 182–188
cyte count, and HDL-cholesterol concentration (Fig. 4), but not with the body fat percentage or platelet count. The serum VEGF concentration may be a marker to predict the development of arteriosclerosis earlier than serum CRP concentration. Platelet count showed a slight correlation with serum VEGF concentration in both genders. The cytoplasmic presence of VEGF within bone marrow megakaryocytes has been demonstrated [23,24]. Furthermore, the platelets are the major source of VEGF in the circulating blood in the state of malignancy [16–18]. We here demonstrated that even in healthy control subjects, platelets were the major source of VEGF in circulating peripheral blood (Table 1; Fig. 5a). VEGF released from a single platelet was calculated at 0.25 ± 0.20 (×10−6 pg), which corresponded well with our previous report, 0.4 ± 0.3 (×10−6 pg, n = 5) [16]. Interestingly, amount of VEGF released from a single platelet had a significant correlation with serum VEGF concentration (Fig. 5b). Taken together, serum VEGF concentration could be influenced not only by platelet count but also by platelet VEGF content. We believe that VEGF generated locally in an autocrine or paracrine manner may be immediately scavenged from the circulating blood, being trapped by its specific receptors, Flt-1 and KDR. We have observed that in the state of disseminated intravascular coagulation (DIC), the situation when platelets are almost activated or aggregated, both serum and plasma VEGF became less than the measurement sensitivity [25]. Furthermore, the serum VEGF was the better practical indicator than the plasma VEGF [25]. We would like to emphasize that the VEGF in the circulating blood should be measured by sera. Because platelets are easily activated by stimulations like collection of blood or centrifugation, plasma must include “leaking VEGF” from platelets. Measuring serum VEGF is valid and reliable since platelets are certainly activated and VEGF is thoroughly removed. The results of this study suggest that the serum VEGF concentration is correlated with risk factors for atherosclerosis even in a healthy state. The serum VEGF concentration may be a new prognostic factor for atherosclerosis.
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
Acknowledgements We would like to express our special thanks to the members of the Kagoshima Prefectural Comprehensive Health Center for their kind cooperation.
References [1] Ross R, Glomset JA. The pathogenesis of atherosclerosis (first of two parts). N Engl J Med 1976;295(7):369–77. [2] Ross R, Glomset JA. The pathogenesis of atherosclerosis (second of two parts). N Engl J Med 1976;295(8):420–5. [3] Ridker PM, Buring JE, Shih J, Matias M, Hennekens CH. Prospective study of C-reactive protein and the risk of future cardiovascular
[19]
[20]
[21]
[22]
[23]
187
events among apparently healthy women. Circulation 1998;98(8): 731–3. Carlson CS, Aldred SF, Lee PK, et al. Polymorphisms within the Creactive protein (CRP) promoter region are associated with plasma CRP levels. Am J Hum Genet 2005;77(1):64–77. de Maat MP, Bladbjerg EM, Drivsholm T, et al. Inflammation, thrombosis and atherosclerosis: results of the Glostrup study. J Thromb Haemost 2003;1(5):950–7. Kobayashi K, Akishita M, Yu W, et al. Interrelationship between noninvasive measurements of atherosclerosis: flow-mediated dilation of brachial artery, carotid intima-media thickness and pulse wave velocity. Atherosclerosis 2004;173(1):13–8. Leung DW, Cachianes G, Kuang WJ, Goeddel DV, Ferrara N. Vascular endothelial growth factor is a secreted angiogenic mitogen. Science 1989;246(4935):1306–9. Keck PJ, Hauser SD, Krivi G, et al. Vascular permeability factor, an endothelial cell mitogen related to PDGF. Science 1989;246(4935):1309–12. Inoue M, Itoh H, Ueda M, et al. Vascular endothelial growth factor (VEGF) expression in human coronary atherosclerotic lesions: possible pathophysiological significance of VEGF in progression of atherosclerosis. Circulation 1998;98(20):2108–16. Sumino H, Nakamura T, Ichikawa S, et al. Serum level of vascular endothelial growth factor is decreased by hormone replacement therapy in postmenopausal women without hypercholesterolemia. Atherosclerosis 2000;148(1):189–95. Matsui K, Yoshioka T, Murakami Y, et al. Serum concentrations of vascular endothelial growth factor and monocyte-colony stimulating factor in peripheral arterial disease. Circ J 2003;67(8):660–2. Heliopoulos I, Papazoglou D, Piperidou H, et al. Carotid plaque echomorphology and serum vascular endothelial growth factor levels. Eur Neurol 2004;51(2):104–8. Alber HF, Frick M, Dulak J, et al. Vascular endothelial growth factor (VEGF) plasma concentrations in coronary artery disease. Heart 2005;91(3):365–6. Tsai WC, Li YH, Huang YY, et al. Plasma vascular endothelial growth factor as a marker for early vascular damage in hypertension. Clin Sci (Lond) 2005;109(1):39–43. Lim HS, Lip GY, Blann AD. Angiopoietin-1 and angiopoietin2 in diabetes mellitus: relationship to VEGF, glycaemic control, endothelial damage/dysfunction and atherosclerosis. Atherosclerosis 2005;180(1):113–8. Hashiguchi T, Arimura K, Matsumuro K, et al. Highly concentrated vascular endothelial growth factor in platelets in Crow–Fukase syndrome. Muscle Nerve 2000;23(7):1051–6. Banks RE, Forbes MA, Kinsey SE, et al. Release of the angiogenic cytokine vascular endothelial growth factor (VEGF) from platelets: significance for VEGF measurements and cancer biology. Br J Cancer 1998;77(6):956–64. Gunsilius E, Petzer A, Stockhammer G, et al. Thrombocytes are the major source for soluble vascular endothelial growth factor in peripheral blood. Oncology 2000;58(2):169–74. Celletti FL, Waugh JM, Amabile PG, et al. Vascular endothelial growth factor enhances atherosclerotic plaque progression. Nat Med 2001;7(4):425–9. Ohtani K, Egashira K, Hiasa K, et al. Blockade of vascular endothelial growth factor suppresses experimental restenosis after intraluminal injury by inhibiting recruitment of monocyte lineage cells. Circulation 2004;110(16):2444–52. Salomonsson L, Pettersson S, Englund MC, Wiklund O, Ohlsson BG. Post-transcriptional regulation of VEGF expression by oxidised LDL in human macrophages. Eur J Clin Invest 2002;32(10):767–74. Pryor WA, Stone K. Oxidants in cigarette smoke. Radicals, hydrogen peroxide, peroxynitrate, and peroxynitrite. Ann NY Acad Sci 1993;686:12–27 (discussion 8). M¨ohle R, Green D, Moore MA, Nachman RL, Rafii S. Constitutive production and thrombin-induced release of vascular endothelial growth
188
K. Kimura et al. / Atherosclerosis 194 (2007) 182–188
factor by human megakaryocytes and platelets. Proc Natl Acad Sci USA 1997;94(2):663–8. [24] Arisato T, Hashiguchi T, Sarker KP, et al. Highly accumulated platelet vascular endothelial growth factor in coagulant thrombotic region. J Thromb Haemost 2003;1(12):2589–93.
[25] Tokashiki T, Hashiguchi T, Arimura K, et al. Predictive value of serial platelet count and VEGF determination for the management of DIC in the Crow–Fukase (POEMS) syndrome. Intern Med 2003;42(12):1240–3.
Serum VEGF—As a prognostic factor of atherosclerosis Kuriko Kimura a , Teruto Hashiguchi b,∗ , Takahisa Deguchi a , Shuji Horinouchi a , Tadashi Uto a , Hiroko Oku a , Shiro Setoyama c , Ikuro Maruyama b , Mitsuhiro Osame a , Kimiyoshi Arimura a a
Department of Neurology and Geriatrics, Graduate School of Medical and Dental Sciences, Kagoshima University, Japan b Laboratory and Vascular Medicine, Graduate School of Medical and Dental Sciences, Kagoshima University, 8-35-1 Sakuragaoka, Kagoshima 890-8520, Japan c Kagoshima Prefectural Comprehensive Health Center, Kagoshima, Japan Received 8 October 2005; received in revised form 5 June 2006; accepted 24 July 2006 Available online 1 December 2006
Abstract Vascular endothelial growth factor (VEGF) has been noted in the pathogenesis of atherosclerosis. To examine the usefulness of the serum concentration of VEGF as an index of atherosclerosis, we analyzed the serum VEGF concentrations in 443 adults who underwent a medical checkup. The mean serum VEGF concentration of men (229 ± 147 pg/ml) was significantly higher than that of women (182 ± 112 pg/ml). The platelet count showed a slight correlation with the serum VEGF concentration in both genders (men R = 0.287, women R = 0.296), corresponding with the results of experiments that platelets are the major source of VEGF in circulating peripheral blood. In men, the serum VEGF concentrations correlated with platelet counts, body fat percentages, leukocyte counts, and HDL-cholesterol concentrations (negative correlation). In the multiple regression analysis performed for men’s serum VEGF concentrations, the decision coefficient (R2) was maximized (R2 = 0.173) when the leukocyte count, the body fat percentage, and the HDL-cholesterol concentration were taken into account besides the platelet count. Male smokers’ serum VEGF concentrations were higher than non-smokers’. Smoking in men significantly affected the sex difference in the serum VEGF concentration, leukocyte count, and HDL-cholesterol concentration. We concluded that the serum VEGF concentration might be closely related to atherosclerosis accelerating factor, especially in men. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: VEGF; Vascular endothelial growth factor; Atherosclerosis; Pathogenesis; Prognosis; Men; Women; Serum; Plasma; Platelet; Platelets; Fat; HDL; Cholesterol; Leukocyte; Smoking; Smoker; Estrogen; Menstruation; Menopause; C-reactive protein; CRP; hsCRP
1. Introduction It has been thought that atherosclerosis is a result of chronic inflammation of blood vessels. The failure of the restoration mechanism leads to chronic inflammation and thus gradually develops atherosclerosis [1,2]. At present, there are some methods which detect atherosclerosis in the early stage, such as the concentration of serum C-reactive protein [3,4], ultrasonographic measurement of intima-media thickness (IMT) of the carotid arteries [5], and measurement of pulse wave velocity (PWV) [6]. The vascular endothe∗
Corresponding author. Tel.: +81 99 275 5437; fax: +81 99 275 2629. E-mail address: [email protected] (T. Hashiguchi).
0021-9150/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2006.07.025
lial growth factor (VEGF) is a multifunctional cytokine which increases vascular permeability and is mitogenic for endothelial cells, acting early and at most points of the angiogenic cascade [7,8]. The role of VEGF in the molecular mechanism of atherosclerotic lesions remains controversial. Some studies have reported that VEGF inhibits thickening of the media by promoting regeneration of vascular endothelial cells and improving endothelial function. Others have indicated that VEGF promotes atherosclerosis via promotion of monocyte chemotaxis and plaque neovascularization [9]. Furthermore, several studies have investigated the association between the plasma/serum concentrations of VEGF and atherosclerosis-associated disorders such as diabetes and hypertension [10–15]. However, no study has
K. Kimura et al. / Atherosclerosis 194 (2007) 182–188
clarified differences between plasma VEGF and serum VEGF. We emphasize that the serum VEGF but not plasma VEGF level in a population including healthy adults is closely associated with risk factors for atherosclerosis, and that VEGF in circulating peripheral blood is derived from platelets.
2. Materials and methods 2.1. Subjects The subjects of this study were 443 continuous outpatients who attended the general health center of Kagoshima prefecture. They underwent a health check for the prevention of life-style related diseases from the period of October 2000 to November 2000 and July 2001 to September 2001. We obtained informed consent in writing from all the people enrolled. The study population consisted of 304 males and 139 females (average age, 49.31 ± 8.81 and 50.68 ± 8.40 years, respectively, p = 0.1339). Ten female and 137 male subjects were smokers. The age range was from 31 to 75 years. As for the details of their health condition, 48 people were diagnosed with hypertension (41 of whom received drug medication), 19 people had hyperlipidemia (8 of whom received medication), 12 people had diabetes mellitus (7 of whom received medication), and 17 people had gout (9 of whom received medication). There were three post-cancer patients, all of whom had undergone resection therapy, and five post-cerebrovascular disease patients (all of whom received medication). Four women had undergone uterus resection operations because of uterus myoma or endometriosis. 2.2. Physical examination Height, body weight, body mass index, body fat percentage, and systolic and diastolic blood pressure were examined. The smoking habit, the Blinkmann Index, the alcohol drinking habit and daily exercise habit were also investigated. 2.3. Laboratory tests and imaging analyses Fasting venous blood was obtained by a traumatic venepuncture into Vacutainers with EDTA-2Na or into tubes free of anticoagulant (for serum collection). The erythrocyte, leukocyte, and platelet count were routinely analyzed. HbA1c was measured by liquid chromatography. Serum aspartate aminotransferase (AST), alanine aminotransferase (ALT), ␥-glutamyl transferase (GTP), total cholesterol, HDLcholesterol, LDL-cholesterol, triglyceride, and amylase were analyzed by using a commercially available kit. A sufficient amount of serum remained for the measurement of serum C-reactive protein (CRP) in 324 subjects (237 males and 87 females), and CRP was measured using a commercially
183
available kit (Dade Behring Marburg GmbH, Germany). For each datum, the correlation with the serum VEGF concentration was examined. Imaging analysis including chest X-ray, barium-contrast study of the stomach, abdominal ultrasonography were also performed. No person with on-going or latent malignant disease was included in this study. 2.4. Measurement of serum VEGF Serum VEGF concentration was measured in duplicate using a commercial ELISA kit (R&D Systems, Minneapolis, MN). The assay system used is sensitive to 15.6 pg/ml VEGF and does not cross-react with platelet-derived growth factor (PDGF) or other homologous cytokines. Optical density at 450 nm was measured on the ImmunoMini NJ-2300 (Nippon Inter Med, Tokyo, Japan) and VEGF concentration was determined by linear regression from a standard curve using the VEGF supplied with the kit as standard. 2.5. Statistical analyses Data are presented as means ± S.D. Differences between data sets were evaluated by analysis of variance (ANOVA) and the Scheff´e test was used as a post hoc test. Paired data were compared by paired t-test. A simple linear regression analysis was used to evaluate the correlation between the two parameters. A multiple regression analysis was performed to seek independent relationships between serum VEGF and four possible markers of pathogenic processes. All statistical analyses were performed with the StatView 5.0 software program (Abacus, Berkeley, CA) for Windows; p < 0.05 was considered statistically significant. 2.6. Quantitative analysis of serum VEGF derived from platelets Serum and plasma from 13 control health subjects were obtained on the same day. Platelet-rich plasma (PRP) and platelet-poor plasma (PPP) were freshly prepared as described previously [16]. Cell-DYN (Dainabot, Tokyo, Japan) was used for hematological counts and to assess the degree of contamination of PRP or PPP by other cells. Thrombin and calcium activation of PRP and PPP was carried out by adding human thrombin (Sigma Chemical Co., St. Louis, MO)/calcium chloride solution to a final concentration of 5 U/ml/10 mM to ensure maximal clotting. After standing at room temperature for 30 min, samples were centrifuged at 750 × g at 4 ◦ C for 20 min and the supernatants were removed for measurement of VEGF concentration. VEGF released from a single platelet and platelet-derived VEGF in serum (pg/ml) were obtained as described previously [16].
184
K. Kimura et al. / Atherosclerosis 194 (2007) 182–188
3. Results 3.1. Difference of the serum VEGF concentration by sex The average serum VEGF concentration of all subjects was 214 ± 139 pg/ml. The maximum value was 830 pg/ml, and the minimum value was less than the measurement sensitivity. According to the t-test, the men’s serum VEGF concentrations are significantly higher than the women’s (p = 0.0008). The average serum VEGF concentration of men is 229 ± 147 pg/ml, and of women, 182 ± 112 pg/ml (Fig. 1). However, the significant difference disappeared when male (n = 167) and female (n = 129) non-smokers were compared (Fig. 1, 209 ± 144 and 183 ± 113 pg/ml, respectively, p = 0.0951). 3.2. Serum VEGF and platelet count showed a positive correlation We classified men and women into five age groups (30s, 40s, 50s, 60s, and 70s) and compared the serum VEGF concentrations, but there was no significant difference between groups. Though the correlation is weak, platelet counts are directly correlated with serum VEGF concentrations both in men and women (men: r = 0.287, p < 0.0001; women: r = 0.296, p = 0.0029) (Fig. 2a and c). 3.3. Four factors which influence the serum VEGF In the case of men, from the tendency of the point distribution on the charts, leukocyte count (r = 0.236, p < 0.0001), the body fat percentage (r = 0.204, p = 0.0004), and HDLcholesterol (r = −0.183, p = 0.0013) showed weak correlation with the serum VEGF concentration (Fig. 3a–c). In women’s data, there was no item that showed correlation except for the platelet count. Multiple regression analysis was performed on the men’s data, and the decision coefficient R2 was maximized (R2 = 0.173, r = 0.416) when leukocyte count, body fat percentage, and HDL-cholesterol con-
Fig. 1. Difference of serum VEGF concentration between men and women.
centration were taken into account besides the platelet counts. 3.4. Serum VEGF and smoking habit in men Men’s serum VEGF concentrations were compared between smokers and non-smokers. The smokers’ average serum VEGF concentration was 255 ± 148 pg/ml and nonsmokers’ was 209 ± 144 pg/ml (p = 0.0063). No significant differences were noted in the platelet count or body fat percentage between male smokers (n = 137) and non-smokers (n = 167) (Figs. 2b and 3e), but the leukocyte count and HDLcholesterol concentration were significantly different (Fig. 3d and f). The Blinkmann index, drinking habit and exercise habit were unrelated to the serum VEGF concentration. 3.5. Serum VEGF and women’s menstruation period Women were classified into the following four groups and serum VEGF concentration compared. The first group consisted of women who have regular menses (n = 56, VEGF 179 ± 111 pg/ml), the second group, irregular menses (n = 11, VEGF 167 ± 93 pg/ml), the third group consisted of menopausal women (n = 60, VEGF 180 ± 111 pg/ml), and the fourth group consisted of those who had undergone uterus
Fig. 2. (a and c) Correlation between serum VEGF concentration and platelet number in peripheral blood (a, men; c, women). (b) Difference of platelet number in the peripheral blood between smokers and non-smokers in men.
K. Kimura et al. / Atherosclerosis 194 (2007) 182–188
185
Fig. 3. (a–c) Correlation between serum VEGF concentration and leukocyte count in the peripheral blood (a), body fat percentage (b), and the concentration of HDL-cholesterol (c). (d–f) Difference of leukocyte count (d), body fat percentage (e), and the concentration of HDL-cholesterol (f) between smokers and non-smokers in men.
resection because of myoma or endometriosis (n = 4, VEGF 299 ± 187 pg/ml). Although the fourth uterus resection group showed high serum VEGF concentrations, with two of the four cases presenting over 400 pg/ml, there were no significant differences between groups. 3.6. Serum C-reactive protein and VEGF In men, the serum CRP concentration was significantly higher in smokers (0.094 ± 0.114 mg/dl) than in non-smokers (0.059 ± 0.084 mg/dl) (p = 0.0063). The serum CRP concentration was weakly correlated with the serum VEGF concentration (Fig. 4a, r = 0.181, p = 0.0052). The serum CRP concentration was also correlated with the leukocyte count (Fig. 4b, r = 0.318, p < 0.0001) and HDL-cholesterol concentration (Fig. 4c, r = −0.142, p = 0.0290), but not with the platelet count or body fat percentage (data not shown).
3.7. Platelet-derived VEGF in serum for healthy control subjects Serum levels of VEGF were significantly higher than the corresponding plasma level as previously reported [16–18]. We analyzed VEGF release in PRP (Fig. 5a and b, Table 1) from 13 healthy control subjects. The mean VEGF concentrations of serum/plasma/PPP were 146 ± 74, 12 ± 9 and 16 ± 14 pg/ml, respectively (Table 1). Clotting of PRP with thrombin/CaCl2 resulted in a significantly higher level of VEGF compared to plasma and PPP level. A significant correlation was found between VEGF in the directly prepared serum and platelet-derived VEGF in serum (r = 0.726, p = 0.0006, Fig. 5a). We also calculated the amount of VEGF released from a single platelet (Table 1). The amount of VEGF released from a single platelet varied from 0.02 to 0.77 (×10−6 pg, 0.25 ± 0.20) and also significantly corre-
Fig. 4. Correlation between serum C-reactive protein concentration and serum VEGF concentration (a), leukocyte count (b), and HDL-cholesterol (c) in men.
186
K. Kimura et al. / Atherosclerosis 194 (2007) 182–188
Table 1 Platelet-derived VEGF in serum for healthy control subjects
PPP, platelet-poor plasma; PRP, platelet-rich plasma; th/Ca2+ , thrombin/CaCl2 ; VEGF released from a single platelet was obtained using the equation: (PRP+th/Ca2+ − Plasma)/Platelet count in PRP.
Fig. 5. Correlation between serum VEGF concentration and platelet-derived VEGF in serum (a) and amount of VEGF released from a single platelet (b).
lated with the directly prepared serum VEGF (r = 0.766, p = 0.0023, Fig. 5b).
4. Discussion Atherosclerosis has been recognized as a disease condition in which low-grade, chronic inflammation causes structure change of the vascular wall [1,2]. In the analysis of data of 443 people who underwent a health check, there was a significant difference in the serum VEGF concentrations between men and women (p = 0.0008; Fig. 1). A multiple regression analysis was performed for men’s serum VEGF concen-
trations, and the decision coefficient (R2) was maximized (R2 = 0.173) when leukocyte count, body fat percentage, and HDL-cholesterol concentration (negative correlation) were taken into account besides the platelet count. As those factors are all related to atherosclerosis or inflammation, we speculate that serum VEGF concentration will be a candidate marker for atherosclerosis development. In laboratory animals, it has been demonstrated that introduction of the VEGF gene led to therapeutic neovascularization, improving tissue ischemia, and that it inhibited thickening of the media after injury via regeneration of vascular endothelial cells and functional improvement. However, the efficacy of VEGF was not suggested in a prospective clinical study of neovascularization therapy using VEGF or in a clinical study targeting restenosis after coronary intervention. A study has reported that up-regulation of VEGF and its receptors, flt-1/Flk-1, is involved in proliferation of intraplaque microvessels in advanced atherosclerotic foci in humans [9]. Other studies have indicated that VEGF promotes atherosclerotic plaque formation in mice [19], and that VEGF promotes atherosclerosis via infiltration of macrophages and mobilization of myelocytes in rabbits [20]. Also oxidized low density lipoproteins reportedly upregulated the VEGF expression in human macrophages [21]. Based on the results of our study, VEGF may be a cytokine that promotes arteriosclerosis in humans. Moreover, the male smokers’ serum VEGF concentrations tended to be higher than the non-smokers’. Smoking habit is widely known as a strong promoting factor of atherosclerosis [22] and it also increased the serum VEGF concentration. Our study revealed that smoking significantly affected the sex difference in the serum VEGF concentration and the leukocyte count and HDL-cholesterol concentration in men (Figs. 1 and 3d and f). In the women’s data, platelet count was the only item that was correlated with serum VEGF concentration. Two of four people who had undergone uterus resection presented over 400 pg/ml of serum VEGF concentration. Women’s mean serum VEGF was below 200 pg/ml, so they presented over twice the mean. As dosing with estrogen was reported to decrease the serum VEGF [10], it may be modified by factors like estrogen. More accumulation of women’s data is needed. The well-known inflammatory marker for atherosclerosis, serum CRP concentration, was higher in male smokers than in non-smokers (p = 0.0063), and correlated with the serum VEGF concentration, leuko-
K. Kimura et al. / Atherosclerosis 194 (2007) 182–188
cyte count, and HDL-cholesterol concentration (Fig. 4), but not with the body fat percentage or platelet count. The serum VEGF concentration may be a marker to predict the development of arteriosclerosis earlier than serum CRP concentration. Platelet count showed a slight correlation with serum VEGF concentration in both genders. The cytoplasmic presence of VEGF within bone marrow megakaryocytes has been demonstrated [23,24]. Furthermore, the platelets are the major source of VEGF in the circulating blood in the state of malignancy [16–18]. We here demonstrated that even in healthy control subjects, platelets were the major source of VEGF in circulating peripheral blood (Table 1; Fig. 5a). VEGF released from a single platelet was calculated at 0.25 ± 0.20 (×10−6 pg), which corresponded well with our previous report, 0.4 ± 0.3 (×10−6 pg, n = 5) [16]. Interestingly, amount of VEGF released from a single platelet had a significant correlation with serum VEGF concentration (Fig. 5b). Taken together, serum VEGF concentration could be influenced not only by platelet count but also by platelet VEGF content. We believe that VEGF generated locally in an autocrine or paracrine manner may be immediately scavenged from the circulating blood, being trapped by its specific receptors, Flt-1 and KDR. We have observed that in the state of disseminated intravascular coagulation (DIC), the situation when platelets are almost activated or aggregated, both serum and plasma VEGF became less than the measurement sensitivity [25]. Furthermore, the serum VEGF was the better practical indicator than the plasma VEGF [25]. We would like to emphasize that the VEGF in the circulating blood should be measured by sera. Because platelets are easily activated by stimulations like collection of blood or centrifugation, plasma must include “leaking VEGF” from platelets. Measuring serum VEGF is valid and reliable since platelets are certainly activated and VEGF is thoroughly removed. The results of this study suggest that the serum VEGF concentration is correlated with risk factors for atherosclerosis even in a healthy state. The serum VEGF concentration may be a new prognostic factor for atherosclerosis.
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
Acknowledgements We would like to express our special thanks to the members of the Kagoshima Prefectural Comprehensive Health Center for their kind cooperation.
References [1] Ross R, Glomset JA. The pathogenesis of atherosclerosis (first of two parts). N Engl J Med 1976;295(7):369–77. [2] Ross R, Glomset JA. The pathogenesis of atherosclerosis (second of two parts). N Engl J Med 1976;295(8):420–5. [3] Ridker PM, Buring JE, Shih J, Matias M, Hennekens CH. Prospective study of C-reactive protein and the risk of future cardiovascular
[19]
[20]
[21]
[22]
[23]
187
events among apparently healthy women. Circulation 1998;98(8): 731–3. Carlson CS, Aldred SF, Lee PK, et al. Polymorphisms within the Creactive protein (CRP) promoter region are associated with plasma CRP levels. Am J Hum Genet 2005;77(1):64–77. de Maat MP, Bladbjerg EM, Drivsholm T, et al. Inflammation, thrombosis and atherosclerosis: results of the Glostrup study. J Thromb Haemost 2003;1(5):950–7. Kobayashi K, Akishita M, Yu W, et al. Interrelationship between noninvasive measurements of atherosclerosis: flow-mediated dilation of brachial artery, carotid intima-media thickness and pulse wave velocity. Atherosclerosis 2004;173(1):13–8. Leung DW, Cachianes G, Kuang WJ, Goeddel DV, Ferrara N. Vascular endothelial growth factor is a secreted angiogenic mitogen. Science 1989;246(4935):1306–9. Keck PJ, Hauser SD, Krivi G, et al. Vascular permeability factor, an endothelial cell mitogen related to PDGF. Science 1989;246(4935):1309–12. Inoue M, Itoh H, Ueda M, et al. Vascular endothelial growth factor (VEGF) expression in human coronary atherosclerotic lesions: possible pathophysiological significance of VEGF in progression of atherosclerosis. Circulation 1998;98(20):2108–16. Sumino H, Nakamura T, Ichikawa S, et al. Serum level of vascular endothelial growth factor is decreased by hormone replacement therapy in postmenopausal women without hypercholesterolemia. Atherosclerosis 2000;148(1):189–95. Matsui K, Yoshioka T, Murakami Y, et al. Serum concentrations of vascular endothelial growth factor and monocyte-colony stimulating factor in peripheral arterial disease. Circ J 2003;67(8):660–2. Heliopoulos I, Papazoglou D, Piperidou H, et al. Carotid plaque echomorphology and serum vascular endothelial growth factor levels. Eur Neurol 2004;51(2):104–8. Alber HF, Frick M, Dulak J, et al. Vascular endothelial growth factor (VEGF) plasma concentrations in coronary artery disease. Heart 2005;91(3):365–6. Tsai WC, Li YH, Huang YY, et al. Plasma vascular endothelial growth factor as a marker for early vascular damage in hypertension. Clin Sci (Lond) 2005;109(1):39–43. Lim HS, Lip GY, Blann AD. Angiopoietin-1 and angiopoietin2 in diabetes mellitus: relationship to VEGF, glycaemic control, endothelial damage/dysfunction and atherosclerosis. Atherosclerosis 2005;180(1):113–8. Hashiguchi T, Arimura K, Matsumuro K, et al. Highly concentrated vascular endothelial growth factor in platelets in Crow–Fukase syndrome. Muscle Nerve 2000;23(7):1051–6. Banks RE, Forbes MA, Kinsey SE, et al. Release of the angiogenic cytokine vascular endothelial growth factor (VEGF) from platelets: significance for VEGF measurements and cancer biology. Br J Cancer 1998;77(6):956–64. Gunsilius E, Petzer A, Stockhammer G, et al. Thrombocytes are the major source for soluble vascular endothelial growth factor in peripheral blood. Oncology 2000;58(2):169–74. Celletti FL, Waugh JM, Amabile PG, et al. Vascular endothelial growth factor enhances atherosclerotic plaque progression. Nat Med 2001;7(4):425–9. Ohtani K, Egashira K, Hiasa K, et al. Blockade of vascular endothelial growth factor suppresses experimental restenosis after intraluminal injury by inhibiting recruitment of monocyte lineage cells. Circulation 2004;110(16):2444–52. Salomonsson L, Pettersson S, Englund MC, Wiklund O, Ohlsson BG. Post-transcriptional regulation of VEGF expression by oxidised LDL in human macrophages. Eur J Clin Invest 2002;32(10):767–74. Pryor WA, Stone K. Oxidants in cigarette smoke. Radicals, hydrogen peroxide, peroxynitrate, and peroxynitrite. Ann NY Acad Sci 1993;686:12–27 (discussion 8). M¨ohle R, Green D, Moore MA, Nachman RL, Rafii S. Constitutive production and thrombin-induced release of vascular endothelial growth
188
K. Kimura et al. / Atherosclerosis 194 (2007) 182–188
factor by human megakaryocytes and platelets. Proc Natl Acad Sci USA 1997;94(2):663–8. [24] Arisato T, Hashiguchi T, Sarker KP, et al. Highly accumulated platelet vascular endothelial growth factor in coagulant thrombotic region. J Thromb Haemost 2003;1(12):2589–93.
[25] Tokashiki T, Hashiguchi T, Arimura K, et al. Predictive value of serial platelet count and VEGF determination for the management of DIC in the Crow–Fukase (POEMS) syndrome. Intern Med 2003;42(12):1240–3.