å¡ ORIGINAL
ARTICLE
å¡
Abnormalities in Platelets and Vascular Endothelial Cells Induced by Glycated Lipoproteins Haruo Takeda, Tomohiko Yano, Hideki Kishikawa, Takao Miyata, Moritsugu Shinohara, Eiji Yamaguchi, Shozo Kobori, Jian-Lin Fan*, Osamu Tokunaga* and Motoaki Shichiri We studied the effects of glycated lipoproteins of low- and high-density (LDL and HDL) on platelets and vascular endothelial cells. After pretreatment for 5 minutes at 37°C, the thrombin induced synthesis of thromboxane B2 in washed platelets was significantly increased by glycated LDL as compared with native LDL (198.9± 16.2 vs 90.3±29.4ng/109 platelets, n=8, p < 0.01). Platelet aggregation was also increased by glycated LDL as compared with native LDL. After treatment with platelet-rich plasma for 5 hours at 37°C, these values were suppressed by native HDL vs the control (buffer), but not by glycated HDL. Abnormalities in the release of 6-keto prostaglandin Fla and lactate dehydrogenase from vascular endothelial cells were also induced by glycated LDL and/or HDL. These observations suggest that abnormalities induced in platelets and vascular endothelial cells by glycated lipoproteins may (Internal Medicine role 31: 746-751, 1992) play an important in the development of atherosclerosis in patients with diabetes mellitus. Key words: platelet aggregation, thromboxanes, prostacyclin, glycation, low-density lipoproteins , high-density lipoproteins
Introduction
sclerosis. Recently, Watanabe et al (14) reported that
platelet aggregation and the synthesis of thromboxane The role of glycated proteins the initiation and incu A2 (TXA2) are enhanced by in glycated LDL after development of diabetic macroand micro-angiopathy bation for a short period. It is generally accepted that has received considerable attention (1, 2). Glycated native LDL may and HDL affect the metabolism prosta lipoproteins play an important role in theofdevelopglandins (PGs) in the platelets and endothelium during ment of macroangiopathy, which increases the morbidity incubation for long (15-17) as well as for short periods and mortality of cardiovascular disease in diabetic patients (18).InThe study wasof designed to evaluate the (3). fact,present increased levels serum glycated lowdensity (LDL) and lipoeffects lipoproteins of glycated LDL and(4-6) HDL onhigh-density human platelets proteins (HDL) (7, 8) have been demonstrated in diabetic and cultured human umbilical vein endothelial cells patients. It hasincubation also been for reported glycated LDL is in vitro after short that and long periods. poorly recognized by LDL receptors and scavenger Materials and Methods receptors in several types of cells (9-12) which may slow its catabolism (ll, 12). In contrast to glycated LDL, Subjects Witztumhuman et al (13) reported thatobtained the catabolism of Washed platelets were from 8 healthy glycated HDL is accelerated, while its uptake by macrosubjects, 5 males and 3 females, aged 25-40 years. None phagesisnot. had received any medication for at least 2 weeks before There little evidence to indicate that glycated lipoblood issampling. proteins directly affect the functions of macrophages, platelets, and the vascular endothelial or smooth muscle Reagents cells, which arearachidonic important in thewere pathogenesis atheroThrombin and acid purchasedoffrom
From the Department of Metabolic Medicine, Kumamoto University Medical School, Kumamoto and *the Department of Pathology, Sa Medical School, Saga Received for publication August 5, 1991; Accepted for publication February 26, 1992 Reprint requests should be addressed to Dr. Haruo Takeda, the Department of Metabolic Medicine, Kumamoto University Medical Scho 1-1-1 Honjo, Kumamoto 860, Japan 746 Internal Medicine Vol. 31, No. 6 (June 1992)
Glycated
Lipoproteins
Sigma Chemical Co. (St. Louis, MO, USA), prostacyclin (PGI2) sodium salt from Funakoshi Pharmaceutical Co. (Tokyo, Japan), RPMI 1640 and Ham's F10 medium from Nissui Pharmaceutical Co. (Tokyo, Japan), fetal calf serum (FCS) from Sanko Junyaku Co. (Tokyo, Japan), and endothelial cell growth supplement (ECGS) from Cosmo-Bio Co. (Tokyo, Japan). Prepacked Glycogelá" B columns (Affinity Paká") containing 1 ml of aminophenylboronate gel were obtained from Pierce Co. (Rock ford, IL, USA), and lactate dehydrogenase (LDH) UV test from Boehringer- Mannheim-Yamanouchi (Tokyo, Japan). Other reagents were of analytical grade. Preparation of glycated lipoproteins LDL and HDL were prepared from pooled normo lipidemic human serum by density gradient ultracentri fugation (19). After the dialysis of these proteins against phosphate-buffered saline (PBS, pH 7.4) containing 0.05% EDTA, they were redialyzed against PBS over night. Glycation of these samples was performed as described by Steinbrecher et al (ll), with a slight modi fication. Briefly, the LDL or HDL solution was incubated with 80mmol/l glucose at 37°C for 7 days. These samples were dialyzed against either modified Tyrode's buffer for the study with platelets (20) or RPMI 1640 or Ham's F10 medium for the study with endothelial cells (21). Native protein control samples were prepared using the same solutions without glucose. The glycation rate of the lipoproteins glycated in vitro was determined as pre viously described (6), according to the method of Jack et al (5) using an aminophenylboronate column. The lipid peroxide content of these lipoproteins was assessed by the thiobarbituric acid (TBA) method (19) as a para meter of TBA-reacting substances (TBARS), and by the methylene blue-hemoglobin method (20) using Determinerá" LPO (Kyowa Medix Co. Ltd., Tokyo, Short-term Japan) as that incubation of lipid of hydroperoxides. the lipoproteinIncreases with platelets in the Washed human platelets from citrated venous blood negative charge of these glycated lipoproteins were also that had been obtained 8 healthy subjects assessed by agarose gelfrom electrophoresis (21). were prepared by centrifugation as described elsewhere (22). Platelet pellets were suspended in modified Tyrode's buffer (pH 7.4) containing lmmol/1 CaCl2 (20-25 x 104/mm3) and incubated with native or glycated LDL or HDL (0.5mg prot/ml) at 37°C for 5min. This was then stimulated by either arachidonic acid (50/imol/l) or near-threshold concentrations of thrombin (16.0 ± 1.3mU/ml, n=8) at 37°C for 5min as described pre viously (22). After monitoring the aggregation of plate lets photometrically, the TXB2 content of the supernatant was determined by radioimmunoassay. To examine the stimulation of thrombin-induced platelet aggregation and Internal Medicine synthesis Vol. 31, No. 6 (June 1992) thromboxane by glycated LDLs with various
and Vascular Cells extents of glycation, glycated LDLs with different extents of glycation were prepared by mixing the glycated LDL with known extent of glycation and native LDL. The washed platelet suspension was pretreated with same amount of the glycated LDLs with various extents of glycation in modified Tyrode's buffer (0.5mg/ml) for 5 min at 37°C. This suspension was then further stimu lated by incubation with thrombin (16mU/ml) for 5 min at 37°C. Long-term incubation of the lipoproteins with platelets Platelet-rich plasma, which was obtained as an inter mediate sample for preparation of the washed platelets, was pretreated with or without various types of lipo proteins in modified Tyrode buffer (0.5mg/ml) for 5 hours at 37°C. It was then stimulated by near-threshold concentrations of thrombin (116.3 ± 14.4mU/ml, n = 8) or arachidonic acid (10-100/imol/l) for 5min at 37°C before and after the preparation of washed platelets from the platelet-rich plasma. Platelet aggregation and the release of TXB2 from the samples were measured as described for the short-term incubation study. Short-term incubation of the lipoproteins with endothelial cells Endothelial cells were separated from human umbilical cord vein by incubation with dispase at 37°C for 90min and cultured in RPMI 1640 medium with 20% FCS containing lOO U/ml penicillin, 100 /zg/ml streptomycin, 5U/ml heparin and 5/ig/ml ECGS as described else where (23). Sub-confluent mono-layer cells at passage 2-4 (50-60 x 104/well) were incubated in Cell-Wellsá" having 6 wells (Corning, NY, USA) with various lipo proteins in RPMI 1640 medium (0.2mg/ml) for 5 min at 37°C, then stimulated by arachidonic acid (50 /imol/1) for 5min at 37°C. The 6-keto prostaglandin Fla (PGF^) content of the supernatant was determined by radio immunoassay (22). of the lipoproteins with endothelial Long-term incubation cells Cultured endothelial cells were pretreated with or without various types of LDL in RPMI 1640 or Ham's F1Q medium (0.2mg/ml) for 1 to 24 hours at 37°C in a CO2 incubator. The time-dependent release of 6-keto PGFlar content from the endothelium was determined as described above. The LDH content of the super natant released from the endothelium was assayed photometrically by the LDH UV Test to examine cellular integrity (24). Statistical analysis All results were expressed as mean ± SE. Statistical analysis was performed with the paired Student's t-test, with p < 0.05 accepted as statistically significant. 747
Takeda et al (data not shown). We then examined the differences in the rate of stimulation of thrombin-induced platelet aggregation and the synthesis of TXB2 produced by glycated LDLs 60 with various extents of glycation (Fig. 2). Dose-response curves were obtained for platelet aggregation and for TXB2 synthesis, with the half-maximal effect occurring at around 10% of the glycation. The percentage of glycated LDL prepared from native LDL in vitro was 63.4± 3.3% (n = 6) in the present study. With respect to glycated HDL, the in vitro value was 2.82 ± 0.93% (n=6). Long-term effect of glycated lipoproteins on platelets Fig. 1. Short-term effects of glycated LDL on thrombin The effects of glycated lipoproteins on thrombin induced platelet aggregation (A) and the release of thromboxane B2 (B) in washed platelets. Values are mean ± SE (n = 8). Open induced platelet aggregation and TXB2 synthesis in columns indicate the values obtained without LDL, hatched platelet-rich plasma were studied after pretreatment for columns those with native LDL, and dotted columns those with 5 hours at 37°C (Fig. 3). While both values were signifi glycated LDL. cantly suppressed by native HDL to less than half that found with the control (buffer), they were increased by glycated HDL as compared to the LDL by control. On the other hand, both values werenative increased native or glycated LDL as compared with the control values.
Glycation
of LDL (%)
Fig. 2. Stimulation of thrombin-induced platelet aggregation and thromboxane synthesis by glycated LDLs with various extents of glycation. Open circles indicate the % increase of thromboxane B2 synthesis when compared with the native LDL control, while closed circles represent those of platelet aggregation. Each point represents the mean value of 3 duplicate determinations. Fig. 3. Long-term effects of glycated HDL on thrombin induced platelet aggregation (A) and the release of thromboxane Results B2 (B) in platelet-rich plasma. Values are mean± SE (n= 8). Open columns indicate the values obtained without HDL, hatched columns those with native HDL, and dotted columns those with Short-term effect of glycated lipoproteins on platelets glycated HDL. The effects of glycated lipoproteins on thrombin induced platelet aggregation and TXB2 synthesis in Table 1. Short-term Effects of Glycated Lipoproteins on the washed platelets were studied following pretreatment for Arachidonic Acid-induced Synthesis of Prostacyclin in 5 min at 37°C (Fig. 1). Platelet aggregation was slightly Vascular Endothelial Cells but significantly suppressed by native LDL as compared to LDL Addition None HDL the control (buffer), but the synthesis of TXB2 was not. (0.2 mg/ml) (0.2 mg/ml) Glycaton ( control) Both values were elevated to more than two-fold follow ing treatment with glycated LDL as compared with (-) 21.4±1.8 12.6± 1.9* 13.6± 1.7* treatment with native LDL. Neither platelet aggregation (+) 21.4±1.8 8.8±0.9** 10.9± 1.3** p value# nor TXB2 synthesis was significantly affected by HDL,
ARTICLE
å¡
Abnormalities in Platelets and Vascular Endothelial Cells Induced by Glycated Lipoproteins Haruo Takeda, Tomohiko Yano, Hideki Kishikawa, Takao Miyata, Moritsugu Shinohara, Eiji Yamaguchi, Shozo Kobori, Jian-Lin Fan*, Osamu Tokunaga* and Motoaki Shichiri We studied the effects of glycated lipoproteins of low- and high-density (LDL and HDL) on platelets and vascular endothelial cells. After pretreatment for 5 minutes at 37°C, the thrombin induced synthesis of thromboxane B2 in washed platelets was significantly increased by glycated LDL as compared with native LDL (198.9± 16.2 vs 90.3±29.4ng/109 platelets, n=8, p < 0.01). Platelet aggregation was also increased by glycated LDL as compared with native LDL. After treatment with platelet-rich plasma for 5 hours at 37°C, these values were suppressed by native HDL vs the control (buffer), but not by glycated HDL. Abnormalities in the release of 6-keto prostaglandin Fla and lactate dehydrogenase from vascular endothelial cells were also induced by glycated LDL and/or HDL. These observations suggest that abnormalities induced in platelets and vascular endothelial cells by glycated lipoproteins may (Internal Medicine role 31: 746-751, 1992) play an important in the development of atherosclerosis in patients with diabetes mellitus. Key words: platelet aggregation, thromboxanes, prostacyclin, glycation, low-density lipoproteins , high-density lipoproteins
Introduction
sclerosis. Recently, Watanabe et al (14) reported that
platelet aggregation and the synthesis of thromboxane The role of glycated proteins the initiation and incu A2 (TXA2) are enhanced by in glycated LDL after development of diabetic macroand micro-angiopathy bation for a short period. It is generally accepted that has received considerable attention (1, 2). Glycated native LDL may and HDL affect the metabolism prosta lipoproteins play an important role in theofdevelopglandins (PGs) in the platelets and endothelium during ment of macroangiopathy, which increases the morbidity incubation for long (15-17) as well as for short periods and mortality of cardiovascular disease in diabetic patients (18).InThe study wasof designed to evaluate the (3). fact,present increased levels serum glycated lowdensity (LDL) and lipoeffects lipoproteins of glycated LDL and(4-6) HDL onhigh-density human platelets proteins (HDL) (7, 8) have been demonstrated in diabetic and cultured human umbilical vein endothelial cells patients. It hasincubation also been for reported glycated LDL is in vitro after short that and long periods. poorly recognized by LDL receptors and scavenger Materials and Methods receptors in several types of cells (9-12) which may slow its catabolism (ll, 12). In contrast to glycated LDL, Subjects Witztumhuman et al (13) reported thatobtained the catabolism of Washed platelets were from 8 healthy glycated HDL is accelerated, while its uptake by macrosubjects, 5 males and 3 females, aged 25-40 years. None phagesisnot. had received any medication for at least 2 weeks before There little evidence to indicate that glycated lipoblood issampling. proteins directly affect the functions of macrophages, platelets, and the vascular endothelial or smooth muscle Reagents cells, which arearachidonic important in thewere pathogenesis atheroThrombin and acid purchasedoffrom
From the Department of Metabolic Medicine, Kumamoto University Medical School, Kumamoto and *the Department of Pathology, Sa Medical School, Saga Received for publication August 5, 1991; Accepted for publication February 26, 1992 Reprint requests should be addressed to Dr. Haruo Takeda, the Department of Metabolic Medicine, Kumamoto University Medical Scho 1-1-1 Honjo, Kumamoto 860, Japan 746 Internal Medicine Vol. 31, No. 6 (June 1992)
Glycated
Lipoproteins
Sigma Chemical Co. (St. Louis, MO, USA), prostacyclin (PGI2) sodium salt from Funakoshi Pharmaceutical Co. (Tokyo, Japan), RPMI 1640 and Ham's F10 medium from Nissui Pharmaceutical Co. (Tokyo, Japan), fetal calf serum (FCS) from Sanko Junyaku Co. (Tokyo, Japan), and endothelial cell growth supplement (ECGS) from Cosmo-Bio Co. (Tokyo, Japan). Prepacked Glycogelá" B columns (Affinity Paká") containing 1 ml of aminophenylboronate gel were obtained from Pierce Co. (Rock ford, IL, USA), and lactate dehydrogenase (LDH) UV test from Boehringer- Mannheim-Yamanouchi (Tokyo, Japan). Other reagents were of analytical grade. Preparation of glycated lipoproteins LDL and HDL were prepared from pooled normo lipidemic human serum by density gradient ultracentri fugation (19). After the dialysis of these proteins against phosphate-buffered saline (PBS, pH 7.4) containing 0.05% EDTA, they were redialyzed against PBS over night. Glycation of these samples was performed as described by Steinbrecher et al (ll), with a slight modi fication. Briefly, the LDL or HDL solution was incubated with 80mmol/l glucose at 37°C for 7 days. These samples were dialyzed against either modified Tyrode's buffer for the study with platelets (20) or RPMI 1640 or Ham's F10 medium for the study with endothelial cells (21). Native protein control samples were prepared using the same solutions without glucose. The glycation rate of the lipoproteins glycated in vitro was determined as pre viously described (6), according to the method of Jack et al (5) using an aminophenylboronate column. The lipid peroxide content of these lipoproteins was assessed by the thiobarbituric acid (TBA) method (19) as a para meter of TBA-reacting substances (TBARS), and by the methylene blue-hemoglobin method (20) using Determinerá" LPO (Kyowa Medix Co. Ltd., Tokyo, Short-term Japan) as that incubation of lipid of hydroperoxides. the lipoproteinIncreases with platelets in the Washed human platelets from citrated venous blood negative charge of these glycated lipoproteins were also that had been obtained 8 healthy subjects assessed by agarose gelfrom electrophoresis (21). were prepared by centrifugation as described elsewhere (22). Platelet pellets were suspended in modified Tyrode's buffer (pH 7.4) containing lmmol/1 CaCl2 (20-25 x 104/mm3) and incubated with native or glycated LDL or HDL (0.5mg prot/ml) at 37°C for 5min. This was then stimulated by either arachidonic acid (50/imol/l) or near-threshold concentrations of thrombin (16.0 ± 1.3mU/ml, n=8) at 37°C for 5min as described pre viously (22). After monitoring the aggregation of plate lets photometrically, the TXB2 content of the supernatant was determined by radioimmunoassay. To examine the stimulation of thrombin-induced platelet aggregation and Internal Medicine synthesis Vol. 31, No. 6 (June 1992) thromboxane by glycated LDLs with various
and Vascular Cells extents of glycation, glycated LDLs with different extents of glycation were prepared by mixing the glycated LDL with known extent of glycation and native LDL. The washed platelet suspension was pretreated with same amount of the glycated LDLs with various extents of glycation in modified Tyrode's buffer (0.5mg/ml) for 5 min at 37°C. This suspension was then further stimu lated by incubation with thrombin (16mU/ml) for 5 min at 37°C. Long-term incubation of the lipoproteins with platelets Platelet-rich plasma, which was obtained as an inter mediate sample for preparation of the washed platelets, was pretreated with or without various types of lipo proteins in modified Tyrode buffer (0.5mg/ml) for 5 hours at 37°C. It was then stimulated by near-threshold concentrations of thrombin (116.3 ± 14.4mU/ml, n = 8) or arachidonic acid (10-100/imol/l) for 5min at 37°C before and after the preparation of washed platelets from the platelet-rich plasma. Platelet aggregation and the release of TXB2 from the samples were measured as described for the short-term incubation study. Short-term incubation of the lipoproteins with endothelial cells Endothelial cells were separated from human umbilical cord vein by incubation with dispase at 37°C for 90min and cultured in RPMI 1640 medium with 20% FCS containing lOO U/ml penicillin, 100 /zg/ml streptomycin, 5U/ml heparin and 5/ig/ml ECGS as described else where (23). Sub-confluent mono-layer cells at passage 2-4 (50-60 x 104/well) were incubated in Cell-Wellsá" having 6 wells (Corning, NY, USA) with various lipo proteins in RPMI 1640 medium (0.2mg/ml) for 5 min at 37°C, then stimulated by arachidonic acid (50 /imol/1) for 5min at 37°C. The 6-keto prostaglandin Fla (PGF^) content of the supernatant was determined by radio immunoassay (22). of the lipoproteins with endothelial Long-term incubation cells Cultured endothelial cells were pretreated with or without various types of LDL in RPMI 1640 or Ham's F1Q medium (0.2mg/ml) for 1 to 24 hours at 37°C in a CO2 incubator. The time-dependent release of 6-keto PGFlar content from the endothelium was determined as described above. The LDH content of the super natant released from the endothelium was assayed photometrically by the LDH UV Test to examine cellular integrity (24). Statistical analysis All results were expressed as mean ± SE. Statistical analysis was performed with the paired Student's t-test, with p < 0.05 accepted as statistically significant. 747
Takeda et al (data not shown). We then examined the differences in the rate of stimulation of thrombin-induced platelet aggregation and the synthesis of TXB2 produced by glycated LDLs 60 with various extents of glycation (Fig. 2). Dose-response curves were obtained for platelet aggregation and for TXB2 synthesis, with the half-maximal effect occurring at around 10% of the glycation. The percentage of glycated LDL prepared from native LDL in vitro was 63.4± 3.3% (n = 6) in the present study. With respect to glycated HDL, the in vitro value was 2.82 ± 0.93% (n=6). Long-term effect of glycated lipoproteins on platelets Fig. 1. Short-term effects of glycated LDL on thrombin The effects of glycated lipoproteins on thrombin induced platelet aggregation (A) and the release of thromboxane B2 (B) in washed platelets. Values are mean ± SE (n = 8). Open induced platelet aggregation and TXB2 synthesis in columns indicate the values obtained without LDL, hatched platelet-rich plasma were studied after pretreatment for columns those with native LDL, and dotted columns those with 5 hours at 37°C (Fig. 3). While both values were signifi glycated LDL. cantly suppressed by native HDL to less than half that found with the control (buffer), they were increased by glycated HDL as compared to the LDL by control. On the other hand, both values werenative increased native or glycated LDL as compared with the control values.
Glycation
of LDL (%)
Fig. 2. Stimulation of thrombin-induced platelet aggregation and thromboxane synthesis by glycated LDLs with various extents of glycation. Open circles indicate the % increase of thromboxane B2 synthesis when compared with the native LDL control, while closed circles represent those of platelet aggregation. Each point represents the mean value of 3 duplicate determinations. Fig. 3. Long-term effects of glycated HDL on thrombin induced platelet aggregation (A) and the release of thromboxane Results B2 (B) in platelet-rich plasma. Values are mean± SE (n= 8). Open columns indicate the values obtained without HDL, hatched columns those with native HDL, and dotted columns those with Short-term effect of glycated lipoproteins on platelets glycated HDL. The effects of glycated lipoproteins on thrombin induced platelet aggregation and TXB2 synthesis in Table 1. Short-term Effects of Glycated Lipoproteins on the washed platelets were studied following pretreatment for Arachidonic Acid-induced Synthesis of Prostacyclin in 5 min at 37°C (Fig. 1). Platelet aggregation was slightly Vascular Endothelial Cells but significantly suppressed by native LDL as compared to LDL Addition None HDL the control (buffer), but the synthesis of TXB2 was not. (0.2 mg/ml) (0.2 mg/ml) Glycaton ( control) Both values were elevated to more than two-fold follow ing treatment with glycated LDL as compared with (-) 21.4±1.8 12.6± 1.9* 13.6± 1.7* treatment with native LDL. Neither platelet aggregation (+) 21.4±1.8 8.8±0.9** 10.9± 1.3** p value# nor TXB2 synthesis was significantly affected by HDL,
