Anticoagulant therapy for osteonecrosis associated with heritable hypofibrinolysis and thrombophilia

Osteonecrosis develops as the end-result of reduced blood flow to the femoral head. We postulate that venous thrombosis leads to increased intraosseus venous pressure, reduced arterial flow and hypoxic bone death. Hypofibrinolysis (reduced ability to lyse thrombi) and thrombophilia (increased tendency to form thrombi) appear to play an important role in osteonecrosis. If coagulation disorders cause osteonecrosis, then anticoagulation might ameliorate osteonecrosis. In subjects with coagulation disorders and osteonecrosis of the hip, provided that anticoagulant therapy is started before irreversible segmental collapse of the head of the femur, osteonecrosis may be arrested or, speculatively, sometimes reversed. This has the potential of preventing femoral head collapse which usually leads to total hip replacement.

Metformin reduces weight, centripetal obesity, insulin, leptin, and low-density lipoprotein cholesterol in nondiabetic, morbidly obese subjects with body mass index greater than 30

We studied 31 nondiabetic, habitually (> or =5 years) morbidly obese subjects (mean +/- SD body mass index [BMI] 43 +/- 8.7, median 43). Our specific aim was to determine whether metformin (2.55 g/d for 28 weeks) would ameliorate morbid obesity and reduce centripetal obesity; lipid and lipoprotein cholesterol, insulin, and leptin levels; and plasminogen activator inhibitor activity (PAI-Fx), risk factors for coronary heart disease (CHD). The patients were instructed to continue their prestudy dietary and exercise regimens without change. After 2 baseline visits 1 week apart, the 27 women and 4 men began receiving metformin, 2.55 g/d, which was continued for 28 weeks with follow-up visits at study weeks 5, 13, 21, and 29. Daily food intake was recorded by patients for 7 days before visits then reviewed with a dietitian. Kilocalories per day and per week were calculated. At each visit, fasting blood was obtained for measurement of lipid profile, insulin, leptin, and PAI-Fx. The mean +/- SD kilocalories consumed per day, 1,951 +/- 661 at entry, fell by week 29 to 1,719 +/- 493 (P =.014) but did not differ at weeks 5, 13, and 21 from that at week 29 (P >.2). Weight fell from 258 +/- 62 pounds at entry to 245 +/- 54 pounds at week 29 (P =.0001). Girth was reduced from 51.8 +/- 6.2 to 49.2 +/- 4.5 inches (P =.0001). Waist circumference fell from 44.0 +/- 6.4 inches to 41.3 +/- 5.9 (P =.0001). The waist/hip ratio fell from 0.85 +/- 0.09 to 0.84 +/- 0.09 (P =.04). Fasting serum insulin, 28 +/- 15 microU/mL at entry, fell to 21 +/- 11 microU/mL at week 29 (P =.0001), and leptin fell from 79 +/- 33 ng/mL to 55 +/- 27 ng/mL (P =.0001). On metformin, there were linear trends in decrements in weight, girth, waist circumference, waist/hip ratio, insulin, and leptin throughout the study period (P <.007). Low-density lipoprotein (LDL) cholesterol, 126 +/- 34 mg/dL at study entry, fell to 112 +/- 43 mg/dL at week 29 (P =.001), with a linear trend toward decreasing levels throughout (P =.036). By stepwise linear regression, the higher the entry weight, the larger the reduction in weight on metformin therapy (partial R(2) = 31%, P =.001). The greater the reduction in kilocalories consumed per day, the greater the decrease in weight on metformin therapy (partial R(2) = 15%, P =.011). The higher the waist/hip ratio at entry, the greater its reduction on metformin therapy (partial R(2) = 11%, P =.004). The higher the entry serum leptin, the greater its reduction on metformin therapy (partial R(2) = 29%, P =.002). The greater the reduction in insulin on metformin, the greater the reduction in leptin (partial R(2) = 8%, P =.03). The higher the entry PAI-Fx, the greater the reduction in PAI-Fx on metformin (partial R(2) = 43%, P =.0001). Metformin safely and effectively reduces CHD risk factors (weight, fasting insulin, leptin, LDL cholesterol, centripetal obesity) in morbidly obese, nondiabetic subjects with BMI > 30, probably by virtue of its insulin-sensitizing action.

Hypofibrinolysis, thrombophilia, osteonecrosis

In the context of additional characterization of the pathoetiologic associations of heritable hypofibrinolysis and thrombophilia with osteonecrosis of the hip, the authors assessed 15 women and 21 men at entry to a 12-week treatment study of the amelioration of Ficat Stages I or II osteonecrosis by low molecular weight heparin (Enoxaparin). All 36 patients had osteonecrosis of the hip; four patients had unifocal osteonecrosis, 25 patients had two joints affected, five had three affected joints, and two had four affected joints. In 11 of 15 women (73%), hyperestrogenemia of pregnancy (20%) or exogenous estrogen supplementation (53%) were associated with the development of osteonecrosis. Five gene mutations affecting coagulation and nine serologic coagulation tests were studied. Compared with control subjects, patients were more likely to have heterozygosity and homozygosity for the hypofibrinolytic 4G polymorphism of the plasminogen activator inhibitor-1 gene. Moreover, the plasminogen activator inhibitor-1 gene product, plasminogen activator inhibitor activity, the major determinant of hypofibrinolysis, was 10 times more likely to be high (> 21.1 U/mL) in patients than in control subjects (31% versus 3%), with a median of 15.7 versus 6.3 U/mL. Compared with controls, patients were more likely to have the thrombophilic methylenetetrahydrofolate reductase gene mutation. In addition, the thrombophilic methylenetetrahydrofolate reductase gene product, homocysteine, was four times more likely to be high (> 13.5 umol/L) in patients than in control subjects (20% versus 5%), with a median of 9.1 versus 7 umol/L. Twenty-three percent of patients had low levels (< 65%) of the thrombophilic free protein S versus 3% of control subjects. Patients were more likely than control subjects to have hypofibrinolytic high lipoprotein (a) (> or = 35 mg/dL), 33% versus 13%. Median lipoprotein (a) was higher in patients than in control subjects, 15 versus 5 mg/dL. Heritable hypofibrinolysis and thrombophilia, often augmented in women by hyperestrogenemia, seem to be major pathoetiologies of osteonecrosis. If the association between coagulation disorders and osteonecrosis reflects cause and effect, as postulated, then anticoagulation with Enoxaparin should be a promising therapy for patients with osteonecrosis.

Interaction of estrogen replacement therapy with the thrombophilic 20210 G/A prothrombin gene mutation for atherothrombotic vascular disease: a cross-sectional study of 275 hyperlipidemic women

In a consecutive case series, cross-sectional study of 275 women referred for therapy of hyperlipidemia, (75 [27%] on estrogen replacement therapy [ERT]), our specific aim was to determine whether ERT-mediated thrombophilia and heterozygosity for the thrombophilic 20210 G/A prothrombin gene mutation interacted as risk factors for atherothrombotic cardiovascular disease (ATCVD). Of the 275 women, 100 (36%) had ATCVD; 10 (3.6%) were heterozygous for the 20210 G/A prothrombin gene mutation. In women without the 20210 G/A prothrombin gene mutation, 15 of 71 (21%) on ERT had ATCVD versus 78 of 194 (40%) not on ERT (X(2) = 8.31, P =.004). By stepwise logistic regression, in 261 women with ATCVD risk factor data, positive explanatory variables for ATCVD included the 20210 G/A prothrombin mutation (risk odds ratio, 5.8; 95% confidence intervals [CI], 1.4 to 30.2; P =.021) and a 20210 G/A prothrombin gene mutation*ERT interaction term (risk odds ratio, 2.70; 95% CI, 1.4 to 5.4; P =.004). ATCVD events were more likely in 2 subgroups of women (ERT minus [-] and 20210 G/A prothrombin gene mutation -) or (ERT plus [+] and 20210 G/A prothrombin gene mutation +), P =.004. Other positive explanatory variables for ATCVD events included age (P =.004), triglycerides (P =.012), lipoprotein (a) (P =.03), and homocysteine (P =.032). ERT may be protective against ATCVD when the thrombophilic 20210 G/A prothrombin gene mutation is absent, whereas the 20210 G/A prothrombin gene mutation may increase risk for ATCVD, particularly in the presence of ERT. We suggest that the 20210 G/A prothrombin gene mutation be measured in all women on ERT or before beginning ERT to identify those heterozygous for the thrombophilic prothrombin gene mutation (4%) in whom ERT is contraindicated because of increased risk for ATCVD and thromboembolism, and a second, much larger group of women without the 20210 G/A prothrombin gene mutation (96%) in whom ERT may possibly reduce risk for ATCVD.

Interaction of heritable and estrogen-induced thrombophilia: possible etiologies for ischemic optic neuropathy and ischemic stroke

Our specific aim was to assess how thrombophilic exogenous estrogens interacted with heritable thrombophilias leading to non-arteritic ischemic optic neuropathy (NAION) and ischemic stroke. Coagulation measures were performed in a 74 year old patient and her immediate family. The proband had a 47 year history of 9 previous thrombotic episodes, and developed unilateral NAION 4 years after starting estrogen replacement therapy (ERT). The proband was heterozygous for two thrombophilic gene mutations (G20210A prothrombin gene, platelet glycoprotein IIIa P1A1/A2 polymorphism), and homozygous for the C677T mutation in the methylenetetrahydrofolate reductase (MTHFR) gene. Of 238 normal controls, none had these 3 gene mutations together. The proband's mother and brother had deep venous thrombosis (DVT). The proband's brother, sister, nephew, daughter, and two granddaughters were homozygous for the C677T MTHFR mutation. The proband's brother was heterozygous for the G20210A prothrombin gene mutation. The proband's niece was heterozygous for the G20210A prothrombin gene mutation, homozygous for the C677T MTHFR mutation, homozygous for the hypofibrinolytic 4G polymorphism of the plasminogen activator inhibitor-1 (PAI-1) gene, and heterozygous for the platelet glycoprotein IIIa P1A1/A2 polymorphism. Of 238 normal controls, none had the niece's combination of 4 gene mutations. When ERT-mediated thrombophilia was superimposed on the proband's heritable thrombophilias, unilateral ischemic optic neuropathy developed, her tenth thrombotic event over a 5 decade period. When estrogen-progestin oral contraceptives were given to the proband's niece, she had an ischemic stroke at age 22. Exogenous estrogen-mediated thrombophilia superimposed on heritable thrombophilia and hypofibrinolysis is associated with arterial and venous thrombi, and appears to be a preventable, and potentially reversible etiology for ischemic optic neuropathy and ischemic stroke.

Genetic hypofibrinolysis in complicated pregnancies

OBJECTIVE

To assess the hypofibrinolytic 4G/4G mutation of the plasminogen activator inhibitor (PAI-1) gene as a possible factor contributing to severe preeclampsia, abruptio placentae, fetal growth restriction, and stillbirth.

METHODS

We compared 94 women from a previous report who had obstetric complications to 95 controls with normal pregnancies matched for ethnic background and age. We collected blood and extracted DNA after delivery. All subjects had been tested for thrombophilic mutations factor V Leiden, C677T mutation in the methylenetetrahydrofolate reductase gene, and the G20210A mutation in the prothrombin gene. In the present study we tested for the hypofibrinolytic 4G/4G mutation in the PAI-1 gene.

RESULTS

Women who had obstetric complications were more likely than controls to be 4G/4G homozygotes, 32% (30 of 94) women versus 19% (18 of 95) controls, odds ratio (OR) and 95% confidence intervals (CI) 2.0 (1.02, 3.9). Mutations in the PAI-1 gene were independently associated with obstetric complications (OR 1.56, 95% CI 1.005, 2.43). Heterozygosity for the factor V Leiden mutation was more common in the 30 women who had PAI-1 4G/4G than in the 18 4G/4G controls (33% versus 0%, Fisher P =.008). Seventy-six percent of women had some form of thrombophilia or hypofibrinolysis compared with 37% of controls (Fisher P <.001).

CONCLUSIONS

Women with severe preeclampsia, abruptio placentae, fetal growth restriction, and stillbirth had increased incidence of the hypofibrinolytic 4G/4G mutation of the PAI-1 gene that is frequently associated with the thrombophilic factor V Leiden mutation, further predisposing them to thrombosis.

Polycystic ovary syndrome, infertility, familial thrombophilia, familial hypofibrinolysis, recurrent loss of in vitro fertilized embryos, and miscarriage

OBJECTIVE

To study reversible determinants of infertility and recurrent loss of transferred embryos after failure of 7 of 10 embryo transfers, 1 live birth, and 2 miscarriages.

DESIGN

Measures of thrombophilia, hypofibrinolysis, reproductive hormones, and androgenic steroids before and after metformin therapy.

SETTING

Outpatient clinical research center.

PATIENT(S)

A 32-year-old amenorrheic, infertile woman with polycystic ovary syndrome (PCOS) who had 7 of 10 embryo transfers fail, 1 premature live birth, and 2 miscarriages at 8 and 17 weeks.

INTERVENTION(S)

Metformin (2.55 g/d) was given to ameliorate the endocrinopathy of PCOS.

MAIN OUTCOME MEASURE(S)

Coagulation, insulin, reproductive hormones, and androgenic steroids.

RESULT(S)

The propositus had thrombophilia (familial protein S deficiency [free protein S 32%; normal >/=65%]). She also had familial hypofibrinolysis with 4G4G polymorphism of the plasminogen activator inhibitor (PAI-1) gene and high PAI-1 activity (PAI-Fx), 42.5 U/mL, normal <21.1. Polycystic ovary syndrome was characterized by amenorrhea, polycystic ovaries, high fasting serum insulin (39 microU/mL, normal <20), androstenedione (763 ng/dL, normal <250), and testosterone (229 ng/dL, normal <83). After she received metformin for 4 months, PAI-Fx normalized (12.4 U/mL), as did insulin (12 microU/mL), androstenedione (185 ng/dL), and testosterone (39 ng/dL); weight fell from 109 to 91.3 kg (16%).

CONCLUSION(S)

Metformin reversed the endocrinopathy of PCOS. Familial thrombophilia and hypofibrinolysis may lead to thrombosis-mediated uteroplacental vascular insufficiency, failure to achieve pregnancy after embryo transfer, and miscarriage.

The 4G/4G polymorphism of the hypofibrinolytic plasminogen activator inhibitor type 1 gene: an independent risk factor for serious pregnancy complications

The specific aim of the current study of 133 women with at least 1 pregnancy and measures of hypofibrinolytic and thrombophilic gene mutations was to determine retrospectively whether the mutations were associated with adverse pregnancy outcomes including prematurity, miscarriage, stillbirth, intrauterine growth retardation (IUGR), eclampsia, and abruptio placentae. Four gene mutations (factor V Leiden, methylenetetrahydrofolate reductase [MTHFR], prothrombin, and 4G/5G polymorphism of the plasminogen activator inhibitor type 1 [PAI-1] gene) were assessed by polymerase chain reaction (PCR). One hundred twenty-two women were genotyped for all 4 genes and divided into gene mutation (n = 68) and non-gene (n = 54) groups. The gene mutation group included those with at least 1 thrombophilic mutation (heterozygous for factor V Leiden, heterozygous for prothrombin, and homozygous for MTHFR), or hypofibrinolysis with homozygosity for the 4G polymorphism of the PAI-1 gene. The non-gene mutation group included those with no mutation for all 4 genes (wild-type normal) or who were wild-type normal for the prothrombin and factor V Leiden mutations and heterozygous for MTHFR and/or 4G/5G for the PAI-1 gene, neither heterozygosity associated with coagulation abnormalities. The 68 women with gene mutations, versus 54 in the non-gene mutation group, has more prematurity (10% v 4%, chi2 = 5.4, P = .021), more IUGR (3% v 0%, P = .035), and more total complications of pregnancy (37% v 21%, chi2 = 11.6, P = .001). The number of pregnancies (P = .0001) and 4G/4G polymorphism of the PAI-1 gene (P = .029) were positively associated with complications of pregnancy by stepwise logistic regression when the age, number of pregnancies, and all 4 gene mutations were the explanatory variables. Heritable hypofibrinolysis, mediated by 4G/4G homozygosity for the PAI-1 gene, is an independent significant, potentially reversible risk factor for pregnancy complications, probably acting through thrombotic induction of placental insufficiency.

Plasminogen activator inhibitor activity: an independent risk factor for the high miscarriage rate during pregnancy in women with polycystic ovary syndrome

In 41 women with at least one pregnancy drawn from a group of 149 (108 never-pregnant) women with polycystic ovary syndrome (PCOS), our specific aim was to determine whether hypofibrinolysis mediated by high plasminogen activator inhibitor activity (PAI-Fx) is an independent risk factor for miscarriage. The 41 women had 77 total pregnancies with 34 miscarriages (44%) and 42 live births (55%). There were 12 women with at least one pregnancy, at least one miscarriage, and no live births (16 pregnancies and 16 miscarriages). There were 15 women with at least one pregnancy, no miscarriages, and at least one live birth (25 pregnancies and 28 live births). Of 12 women with only miscarriages and no live births, 67% had PAI-Fx greater than 16.4 U/mL (normals' 95th percentile), versus 29% of 15 women with no miscarriages and all live births (chi2 = 3.8, P = .052). By stepwise logistic regression, the number of pregnancies (P = .0001) and PAI-Fx (P = .016) were significant positive explanatory variables for the number of miscarriages. Age, 4G/5G polymorphisms of the PAI gene, factor V Leiden, methylenetetrahydrofolate reductase (MTHFR) gene mutations, androstenedione, testosterone, sex hormone-binding globulin, the Quetelet index, and fasting serum insulin and glucose were not significant variables in the logistic regression model. In a separate stepwise logistic regression, three nonoverlapping groups of women (12 with > or = 1 pregnancy, > or = 1 miscarriage, and 0 live births, 10 with > or = 1 pregnancy, > or = 1 miscarriage, and > or = 1 live births, and 15 with > or = 1 pregnancy, 0 miscarriages, and > or = 1 live births) were the dependent variables. PAI-Fx was positively associated (P = .05) with the group with the worst pregnancy outcome (> or = 1 pregnancy, > or = 1 miscarriage, and 0 live births). The 41 women with PCOS and at least one pregnancy were more likely than healthy normal controls to have heterozygosity and homozygosity for the 4G/5G polymorphism of the PAI-1 gene (P = .028), but did not differ from normals for factor V Leiden (P > .10) or MTHFR (P > .09) mutations. PAI-Fx is a predominant independent significant positive reversible risk factor for miscarriage in women with PCOS.

Effect of exogenous estrogen on atherothrombotic vascular disease risk related to the presence or absence of the factor V Leiden mutation (resistance to activated protein C)

Estrogen replacement therapy (ERT), which produces acquired resistance to activated protein C when superimposed on heritable resistance to activated protein C (the mutant Factor V Leiden trait), may promote venous and arterial thrombosis. In a cross-sectional study of 423 women referred for hyperlipidemic therapy (93 of whom [22%] were on ERT), our specific aim was to determine whether ERT and heterozygosity for the Factor V Leiden mutation and/or resistance to activated protein C interacted as risk factors for atherothrombosis. Of the 423 women, 168 (40%) had atherothrombosis, 19 (4%) were heterozygous for Factor V Leiden mutation or had resistance to activated protein C <2 (Factor V Leiden mutation+), and 404 were wild-type normal for the Factor V gene and/or had resistance to activated protein C > or =2 (Factor V Leiden mutation-). By stepwise logistic regression, positive explanatory variables for atherothrombosis included hypertension (p = 0.002), age (p = 0.003), relatives with atherothrombosis (p = 0.002), anticardiolipin antibody immunoglobulin-M (p = 0.02), and a Factor V Leiden mutation*ERT interaction term where atherothrombosis events were more likely in 2 subgroups of women (ERT- and Factor V Leiden mutation-) or (ERT+ and Factor V Leiden mutation+) (p = 0.02). High-density lipoprotein cholesterol was inversely associated with atherothrombosis (p = 0.004). In a separate logistic regression model for the 213 women with a polymerase chain reaction measurement of the Factor V gene, ERT was protective (p = 0.008); the Factor V Leiden mutation was positively associated with atherothrombosis (p = 0.05). The atherothrombosis odds ratio risk for ERT (yes vs no) was 0.36 (95% confidence intervals [CI] 0.16 to 0.74, p = 0.007). The atherothrombosis risk odds ratio in women heterozygous for the Factor V Leiden mutation (vs normal) was 2.00 (95% CI 1.02 to 4.22, p = 0.05). ERT may be protective against atherothrombosis when the Factor V Leiden mutation is absent, whereas the Factor V Leiden mutation may increase risk for atherothrombosis, particularly in the presence of ERT. We suggest that the Factor V Leiden mutation be measured in all women on ERT or before beginning ERT to identify those heterozygous for the Factor V Leiden mutation (4%), in whom ERT is relatively or absolutely contraindicated because of increased risk for atherothrombosis and thromboembolism. A second, much larger group of women will also be identified without the factor V Leiden mutation (96%), in whom ERT may reduce the risk for atherothrombosis.

The plasminogen activator inhibitor-1 gene, hypofibrinolysis, and osteonecrosis

In 59 patients with osteonecrosis of the hip, four genes associated with thrombophilia or hypofibrinolysis along with coagulation tests were studied to determine the pathoetiologic associations of heritable coagulation disorders with osteonecrosis. Patients did not differ from healthy control subjects for the thrombophilic Factor V Leiden, prothrombin, or methylenetetrahydrofolate reductase mutations. The plasminogen activator inhibitor-1 gene was shifted toward homozygosity for the 4G polymorphism; 41% of patients with osteonecrosis were homozygous for the 4G/4G polymorphism versus 20% of 40 healthy control subjects. The gene product of the 4G polymorphism, hypofibrinolytic plasminogen activator inhibitor activity, was higher in patients than in control subjects (median 19.2 versus 6.3 U/mL); 61% of patients had high plasminogen activator inhibitor activity (> or = 16.4 U/mL) versus 5% of control subjects. Stimulated tissue plasminogen activator activity (inhibited by plasminogen activator inhibitor activity) was lower in patients than in control subjects (3.10 versus 5.98 IU/mL); 31% of patients had low stimulated tissue plasminogen activator activity (< 2.28 IU/mL) versus 3% of control subjects. Heritable hypofibrinolysis conferred by the 4G/4G mutation of the plasminogen activator inhibitor-1 gene seems to be a major pathoetiology of primary osteonecrosis.

Heritable thrombophilia and hypofibrinolysis. Possible causes of retinal vein occlusion

OBJECTIVE

To determine whether heritable thrombophilia and hypofibrinolysis were risk factors for retinal vein occlusion.

DESIGN

Measures of thrombophilia (increased likelihood of thrombus formation) included anticardiolipin antibodies (IgG and IgM), the lupus anticoagulant (including dilute Russell viper venom clotting time), antigenic proteins C and S, and homocysteine. Polymerase chain reaction assays were performed for 3 thrombophilic gene mutations (factor V Leiden, methylenetetra-hydrofolate reductase, and prothrombin gene). Measures of hypofibrinolysis (reduced ability to lyse thrombi) included lipoprotein Lp(a), plasminogen activator inhibitor activity, and polymerase chain reaction analysis of the hypofibrinolytic 4G/5G polymorphism of the PAI1 gene. These coagulation measures were performed in 17 patients with retinal vein occlusions with comparison with serologic coagulation measures and polymerase chain reaction assays in 40 and 234 healthy normal volunteers as controls, respectively.

RESULTS

Of 14 patients with retinal vein occlusion with measures of dilute Russell viper venom clotting time, a thrombophilic antiphospholipid antibody, 6 (43%) had abnormal results (> 38.8 seconds) compared with 1 (3%) of 30 controls (P = .002). Of 17 patients with vein occlusion, 3 (18%) were heterozygous for the thrombophilic factor V Leiden G1691A mutation compared with 7 (3%) of 233 controls (P = .02). Of 17 patients with vein occlusion, 2 (12%) had normal alleles (5G/5G) for the plasminogen activator inhibitor gene promoter; the other 15 (88%) were heterozygous or homozygous for the 4G polymorphism, which is associated with hypofibrinolysis. Of 234 controls, 85 (36.3%) had the 5G/5G allele; 149 (63.7%) were heterozygous or homozygous for the 4G polymorphism (P = .03). Patients with vein occlusion were more likely to have high levels of the major determinant of hypofibrinolysis, plasminogen activator inhibitor activity. These levels were high (> 22 U/L) in 6 (38%) of 16 patients with vein occlusion compared with 1 (2%) of 40 controls (chi 2 = 12.8; P = .001). Patients with vein occlusion were more likely (8/16 [50%]) to have high levels of hypofibrinolytic Lp(a) (> 35 mg/dL) than controls (5/40 [13%]; chi 2 = 9; P = .003). The median Lp(a) level in patients with vein occlusion who had the 4G/4G genotype was 62 mg/dL compared with 5.3 mg/dL in controls with the 4G/4G genotype (P = .05).

CONCLUSION

Thrombophilia and hypofibrinolysis are possible causes of retinal vein occlusion.

Myocardial infarction in a 35-year-old man with homocysteinemia, high plasminogen activator inhibitor activity, and resistance to activated protein C

Our specific aim was to examine the interface between risk factors for atherosclerosis, thrombosis, and hypofibrinolysis in a previously healthy 35-year-old male who had sustained a recent myocardial infarction. By angiography, the right, left main, and left anterior descending coronary arteries were smooth-walled, widely patent, and free of significant obstruction; the circumflex exhibited total, probably thrombotic occlusion of the distal large second marginal branch. The patient was found to have prothrombotic high homocysteine (46.4 mumol/L), prothrombotic resistance to activated protein C (ratio, 1.47), and hypofibrinolytic high plasminogen activator inhibitor (PAI-Fx) activity (54 U/mL). He was homozygous for the 677C-->T; A-->V mutation in the methylenetetrahydrofolate reductase (MTHFR) gene causing homocysteinemia, heterozygous for the mutant factor V Leiden gene causing resistance to activated protein C, and heterozygous for the 4G/5G polymorphism in the PAI-1 promoter gene causing high PAI-Fx. Other major risk factors for coronary artery disease included previously undiagnosed adult-onset diabetes, high triglycerides (291 mg/dL), and low high-density lipoprotein (HDL) cholesterol (26 mg/dL). The patient's prothrombotic status (homocysteinemia and resistance to activated protein C) and hypofibrinolysis (high PAI-Fx) apparently facilitated occlusive coronary artery thrombus formation and retention. Prothrombotic factors and hypofibrinolysis appear to play important pathogenetic roles in premature myocardial infarction. In patients with severe premature coronary artery disease, we suggest that interactions between prothrombotic factors, hypofibrinolysis, and hyperlipidemia-atherosclerosis be regularly evaluated, since such interactions may have ramifications for the outcome of short- and long-term secondary prevention. Moreover, in patients with heritable prothrombotic factors or hypofibrinolysis, it should be important to optimize lipid and lipoprotein cholesterol levels with the goal of stabilizing coronary plaques to reduce the likelihood of plaque rupture and thrombosis.

Prospective 10-year evaluation of hypobetalipoproteinemia in a cohort of 772 firefighters and cross-sectional evaluation of hypocholesterolemia in 1,479 men in the National Health and Nutrition Examination Survey I

Our specific aim in a 10-year prospective study of 772 Cincinnati firemen (predominantly aged 26 to 46 years) was to determine the prevalence, attributes, and etiology of persistent hypobetalipoproteinemia, defined by entry low-density lipoprotein cholesterol (LDLC) less than 75 mg/dL. A second specific aim was to cross-sectionally assess hypocholesterolemia (defined by total serum cholesterol [TC] < 130 mg/dL) in 1,314 white and 165 black men aged 26 to 46 years in the National Health and Nutrition Examination Survey (NHANES I). The 141 black and 631 white firemen had 4,973 person-years of follow-up time (median, 7.1 yr/man). Of 772 men, 44 (5.7%) had entry LDL levels less than 75 mg/dL; they had a mean follow-up time of 7.3 yr/man. Of these 44 men, there were 12 (1.8% of the cohort) with entry LDLC less than 75 mg/dL, and at least 67% of their follow-up LDLC levels were less than 75. Their mean entry TC and LDLC levels were low (130 and 58 mg/dL), mean triglyceride (TG) was low (63 mg/dL), and mean high-density lipoprotein cholesterol (HDLC) was high (60 mg/dL), LDLC remained at less than 75 mg/dL in 81% of their follow-up samples. Their mean entry and follow-up cholesterol and LDLC did not differ (P > .1, 130 v 133 mg/dL and 58 v 63 mg/dL). Compared with 32 men with entry LDLC less than 75 mg/dL but with less than 87% of follow-up LDLC less than 75 mg/dL, the 12 men with persistently low LDLC had lower mean Quetelet indices and diastolic blood pressure at entry (2.36 v 2.58, P = .056; 73 v 80 mm Hg, P = .03) and on follow-up study (2.45 v 2.69, P = .04; 72 v 79 mm Hg, P = .05). Of 12 men with persistently low LDLC, two had truncated apolipoprotein (apo) B (familial hypobetalipoproteinemia, two had the apo E genotype 2/3, and two had acquired hypobetalipoproteinemia that antedated mortality from melanoma by 9 years and from alcoholism by 2 years. Comparable to white and black firemen aged 26 to 46 years, 2.9% and 3.6% of whom had entry serum TC less than 130 mg/dL, of 1,314 white and 165 black men in the NHANES I study (aged 26 to 46), 1.8% and 3.6% had hypocholesterolemia (entry TC < 130 mg/dL). Daily mean calorie, fat, and protein intake (grams per day) did not differ (P > .05) in men with entry TC less than 130 mg/dL compared with those with TC 130 to 230 or greater than 230 mg/dL. Hypocholesterolemia in white and black men in NHANES I could not be attributed to hypocaloric intake or to protein, fat, or carbohydrate undernutrition. There appear to be racial differences in the prevalence of hypocholesterolemia. Blacks comprised 18% of the firemen's cohort but 42% of those with persistent hypobetalipoproteinemia; among NHANES I subjects, 3.6% of blacks were hypocholesterolemic versus 1.8% of whites. Unless persistent hypobetalipoproteinemia reflects an underlying disease, alcoholism, etc., it is often heritable, and may be associated with a reduced likelihood of coronary heart disease (CHD) and with increased longevity.

Liver and intestinal fatty acid binding proteins in control and TGF beta 1 gene targeted deficient mice

The effect of transforming growth factor beta-1 (TGF beta 1) expression on fatty acid binding proteins was examined in control and two strains of gene targeted TGF beta 1-deficient mice. Homozygous TGF beta 1-deficient 129 x CF-1, expressing multifocal inflammatory syndrome, had 25% less liver fatty acid binding protein (L-FABP) when compared to control mice. The decrease in L-FABP expression was not due to multifocal inflammatory syndrome since homozygous TGF beta 1-deficient/immunodeficient C3H mice on a SCID background had 36% lower liver L-FABP than controls. This effect was developmentally related and specific to liver, but not the proximal intestine, where L-FABP is also expressed. Finally, the proximal intestine also expresses intestinal-FABP (I-FABP) which decreased 3-fold in the TGF beta 1-deficient/immunodeficient C3H mice only. Thus, TGF beta 1 appears to regulate the expression of L-FABP and I-FABP in the liver and the proximal intestine, respectively.

Effects of chronic ethanol consumption on sterol transfer proteins in mouse brain

Although lipids are essential to brain function, almost nothing is known of lipid transfer proteins in the brain. Early reports indicates cross-reactivity of brain proteins with antisera against two native liver sterol transfer proteins, sterol carrier protein-2 (SCP-2) and the liver form of fatty acid-binding protein (L-FABP). Herein, polyclonal antibodies raised against the recombinant liver sterol transfer proteins SCP-2 and L-FABP were used to identify the lipid transfer proteins in the brains of alcohol-treated and control mice. L-FABP was not detectable in brain of either control or chronic ethanol-treated mice. In contrast, SCP-2 not only was present, but its level was significantly (p < 0.05) increased 23 and 50%, respectively, in brain homogenates and synaptosomes of mice exposed to alcohol. To determine whether antibodies against the recombinant liver SCP-2 reflected true levels of SCP-2 in brain, the cDNA sequence for brain SCP-2 was isolated from a brain cDNA library. The mouse brain SCP-2 sequence was 99.99% identical to the mouse liver SCP-2 sequence. The translated sequence differed by only one amino acid, and the replacement was conservative. Thus, unlike the fatty acid binding proteins, the SCP-2 moieties of brain and liver are essentially identical. Polyclonal antibodies against acyl-CoA binding protein, a lipid-binding protein that does not bind or transfer sterol, showed that increased levels of brain SCP-2 with chronic ethanol consumption did not represent a general increase in content of all lipid transfer proteins. Changes in the amount of SCP-2 may contribute to membrane tolerance to ethanol.

Reverse cholesterol transport with acute exercise

Physical activity has been shown to be inversely related to coronary heart disease (CHD). The role of high density lipoprotein (HDL) particles in the process of reverse cholesterol transport may be a link between exercise and the prevention of CHD. The aim of the present study was to evaluate the effects of acute exercise on cholesterol efflux (C-EF) from human monocyte derived macrophages overloaded with cholesterol and subsequently incubated with HDL fractions isolated from plasma. Ten males; five sedentary (NR) and five runners (R) exercised 30 min on a cycle ergometer at 60% of maximum oxygen consumption. HDL-C was higher in R when compared with NR (49.2 +/- 2.6 vs 36.8 +/- 4.6 mg.dl-1; P < 0.05). Plasma lipid profiles did not differ between groups and were unchanged with exercise. C-EF was higher to HDL obtained from NR compared with R before exercise (1.05 +/- 0.17 vs 0.59 +/- 0.09 microgram/mg protein, P < 0.05). Acute exercise increased HDL's ability to act as an acceptor of cellular cholesterol in R, whereas it decreased in NR. These preliminary studies suggest that functional changes in HDL fractions may differ in NR and R, and appear to be influenced by acute exercise.

Structure of the rat pancreatic cholesterol esterase gene

The gene encoding the rat pancreatic cholesterol esterase has been isolated and characterized. Analysis of overlapping genomic clones showed that the cholesterol esterase gene spans approximately 8 kb, containing 11 exons interrupted by 10 introns. The exons ranged in size from 83 to 201 bp except for the last exon, which was 548 bp in length. A TAAATA sequence was present at -31 nucleotides from the transcriptional initiation site. A putative pancreas-specific enhancer sequence was found at -90 bp upstream from the CAP site. Although cholesterol esterase shares three domains of similarity with cholinesterase and acetylcholinesterase, these domains were found to be localized in distinct exons of the cholesterol esterase gene. The organization of the cholesterol esterase gene suggests its divergent evolution with other members of the serine esterase gene family.

Site-specific mutagenesis of an essential histidine residue in pancreatic cholesterol esterase

The histidine residue essential for the catalytic activity of pancreatic cholesterol esterase (carboxylester lipase) has been identified in this study using sequence comparison and site-specific mutagenesis techniques. In the first approach, comparison of the primary structure of rat pancreatic cholesterol esterase with that of acetylcholinesterase and cholinesterase revealed two conserved histidine residues located at positions 420 and 435. The sequence in the region around histidine 420 is quite different between the three enzymes. However, histidine 435 is located in a 22-amino acid domain that is 47% homologous with other serine esterases. Based on this sequence homology, it was hypothesized that histidine 435 is the histidine residue essential for catalytic activity of cholesterol esterase. The role of His435 in the catalytic activity of pancreatic cholesterol esterase was then studied by the site-specific mutagenesis technique. Substitution of the histidine in position 435 with glutamine, arginine, alanine, serine, or aspartic acid abolished the ability of cholesterol esterase to hydrolyze p-nitrophenyl butyrate and cholesterol [14C]oleate. In contrast, mutagenesis of the histidine residue at position 420 to glutamine had no effect on cholesterol esterase enzyme activity. The results of this study strongly suggested that histidine 435 may be a component of the catalytic triad of pancreatic cholesterol esterase.

Detection of P-glycoprotein with JSB-1 monoclonal antibody in B-5 fixed and paraffin-embedded cell lines and tissues

We analyzed the expression of P-glycoprotein (Pgp) by immunohistochemistry using JSB-1 monoclonal antibody (MAb) on paraffin-embedded sections of the multi-drug resistant (MDR) (CHrC5 and CEM-VLB), and sensitive (AuxB1 and CEM) cell lines, and also in normal kidney, colon, adrenal and in kidney and colon carcinomas. After comparing the sensitivity of three different immunohistochemical techniques the peroxidase-antiperoxidase method was found to be the best. We then tested six different fixation methods. The MDR cell lines and human tissues demonstrated the strongest staining with B-5 fixative. Both MDR cell lines, but not the tissues fixed in 1% paraformaldehyde and Zamboni's fixative demonstrated weak staining. No immuno- reactivity could be detected in MDR cell lines and tissues fixed in 10% buffered or nonbuffered formalin or by the AMeX method of tissue processing. The present study clearly shows that the type of fixative is critical for the preservation of Pgp epitope recognized by JSB-1 MAb, and that B-5 fixative is expected to be equally applicable for the detection of Pgp in normal and neoplastic tissues.

Identification of the active site serine in pancreatic cholesterol esterase by chemical modification and site-specific mutagenesis

Chemical modification and site-specific mutagenesis approaches were used in this study to identify the active site serine residue of pancreatic cholesterol esterase. In the first approach, purified porcine pancreatic cholesterol esterase was covalently modified by incubation with [3H]diisopropylfluorophosphate (DFP). The radiolabeled cholesterol esterase was digested with CNBr, and the peptides were separated by high performance liquid chromatography. A single 3H-containing peptide was obtained for sequence determination. The results revealed the binding of DFP to a serine residue within the serine esterase homologous domain of the protein. Furthermore, the DFP-labeled serine was shown to correspond to serine residue 194 of rat cholesterol esterase (Kissel, J. A., Fontaine, R. N., Turck, C. W., Brockman, H. L., and Hui, D. Y. (1989) Biochim. Biophys. Acta 1006, 227-236). The codon for serine 194 in rat cholesterol esterase cDNA was then mutagenized to ACT or GCT to yield mutagenized cholesterol esterase with either threonine or alanine, instead of serine, at position 194. Expression of the mutagenized cDNA in COS-1 cells demonstrated that substitution of serine 194 with threonine or alanine abolished enzyme activity in hydrolyzing the water-soluble substrate, p-nitrophenyl butyrate, and the lipid substrates cholesteryl [14C]oleate and [14C] lysophosphatidylcholine. These studies definitively identified serine 194 in the catalytic site of pancreatic cholesterol esterase.

Molecular cloning and expression of cDNA for rat pancreatic cholesterol esterase

A full-length cDNA complementary to the rat pancreatic cholesterol esterase mRNA was isolated by screening a rat pancreatic cDNA expression library in lambda gt11 vector with antibodies against the porcine pancreatic cholesterol esterase. The isolated cholesterol esterase cDNA is 2050 bp in length and contains an open reading frame coding for a protein of 612 amino acids. A 20-amino acid hydrophobic leader sequence is predicted, based on the position of the first ATG initiation codon upstream from the sequenced amino terminus of the isolated cholesterol esterase. The cholesterol esterase cDNA was subcloned into a mammalian expression vector, pSVL, for transfection studies. Expression of the cDNA in COS cells resulted in the production of bile salt-stimulated cholesterol esterase. Comparison of the cholesterol esterase cDNA sequence with other proteins revealed that the pancreatic cholesterol esterase is identical to rat pancreatic lysophospholipase. The primary structure of cholesterol esterase displayed no significant homology with other lipases, although the putative lipid interfacial recognition site of G-X-S-X-G is present in the cholesterol esterase sequence. However, the cholesterol esterase sequence revealed a 63-amino-acid domain which is highly homologous to the active site domain of other serine esterases. These data suggest that cholesterol esterase may be a member of the serine esterase supergene family. Analysis of the cholesterol esterase structure also revealed a repetitive sequence enriched with Pro, Asp, Glu, Ser, and Thr residues at the C-terminal end of the protein. This sequence is reminiscent of the PEST-rich sequences in short-lived proteins, suggesting that cholesterol esterase may have a short half-life in vivo. Northern blot hybridization showed that the bile salt-stimulated cholesterol esterase mRNA is present in liver suggesting that this protein may also be synthesized by liver cells.

A fluorescence and radiolabel study of sterol exchange between membranes

The fluorescent sterols delta 5,7,9(11),22-ergostatetraen-3 beta-ol (dehydroergosterol) and delta 5,7,9,(11)-cholestatrien-3 beta-ol (cholestatrienol) as well as [1,2-3H]cholesterol were utilized as cholesterol analogues to examine spontaneous exchange of sterol between 1-palmitoyl-2-oleoylphosphatidylcholine (POPC) small unilamellar vesicles (SUV). Exchange of fluorescent sterols was monitored at 24 degrees C by release from self-quenching of polarization from the time of mixing without separation of donor and acceptor vesicles. The polarization curve for 35 mol% sterol in POPC best fitted a two-exponential function, with a fast-exchange rate constant k1 = 0.0217 min-1, 1t1/2 = 32 min, size pool 1 = 12%, and a slow rate constant k2 = 2.91.10(-3) min-1, 2t1/2 = 238 min, size pool 2 = 88%. In addition to the above two exchangeable pools of sterol, the data were consistent with the presence of a slowly or nonexchangeable pool, 42% of total sterol, that was highly dependent on sterol content. These results were confirmed by simultaneous monitoring of [1,2-3H]cholesterol radioactivity and dehydroergosterol fluorescence intensity after separation of donor and acceptor vesicles by ion-exchange column chromatography. Thus, dehydroergosterol or cholestatrienol exchange as measured by fluorescence parameters (polarization and/or intensity) provides two new methods to follow cholesterol spontaneous exchange. These methods allow resolution and quantitation of a shorter exchange t1/2 near 30 min previously not reported. Thus, the cholesterol desorption rate from membranes may be faster than previously believed. In addition, the presence of a slowly non-exchangeable pool was confirmed.

Comparison of cholesterol transport in pulmonary, peritoneal, and blood-derived macrophages from normo- and hypercholesterolemic rabbits

The influx and efflux components of cholesterol transport were separately determined in pulmonary, peritoneal, and monocyte-derived macrophages from rabbits fed a diet containing either 4.5% fat or 4.5% fat plus 2% cholesterol. Both influx and efflux in pulmonary macrophages increased with increasing concentration of either normal or hypercholesterolemic serum in the medium. The mass of cholesterol entering the macrophages continued to increase beyond the mass of cholesterol effluxed, leading to an increase in cholesterol mass. Similar results were obtained with peritoneal macrophages. Cholesterol-enriched peritoneal macrophages in most cases had a net efflux of sterol when incubated with normocholesterolemic serum. Pulmonary and peritoneal macrophages from cholesterol-fed rabbits tended to have slower sterol influx and a slightly faster sterol efflux than pulmonary and peritoneal macrophages from control rabbits, but the combined effect of these mechanisms did not prevent these macrophages from accumulating sterol ester from hypercholesterolemic serum. Hypercholesterolemic rabbit serum was fractionated by heparin-Sepharose affinity chromatography into a beta-VLDL-deficient nonadsorbed fraction, which had very little effect on pulmonary macrophage sterol content, and an adsorbed beta-VLDL-containing fraction which promoted a large increase in macrophage sterol. As with unfractionated hypercholesterolemic serum, macrophages incubated with the adsorbed beta-VLDL-containing fraction accumulated large amounts of cellular sterol. Monocyte-macrophages cultured in vitro for 21 hr, in contrast to extravascular macrophages, closely regulated their cellular sterol, primarily by limiting the rate of sterol influx.

LM fibroblast plasma membrane subfractionation by affinity chromatography on con A-sepharose

Affinity chromatography was used to determine the heterogeneity and orientation of plasma membrane vesicles isolated from LM fibroblasts subjected to Dounce homogenization. Two plasma membrane subfractions were obtained by Con A-Sepharose affinity chromatography of LM fibroblast plasma membranes prepared by Dounce homogenization. The desmosterol-phospholipid molar ratio, the phospholipid composition, and the phospholipid fatty acid composition were almost identical between the two fractions. However, the lipid to protein ratio was almost 2-fold greater in the nonadherent fraction A. The binding of fluorescein-concanavalin A was the same in both fractions indicating a right-sided-out orientation of the vesicles. Similarly and asymmetric distribution of phosphatidylethanolamine in both membrane fractions was the same. In contrast, sialic acid content, 5'-nucleotidase activity, and (Na+ + K+)-ATPase activity were 47%, 3.7-fold, and 2.5-fold greater, respectively, in the nonadherent, lipid-rich fraction A. Structural properties of the two membrane fractions determined by fluorescence polarization and arrhenius plots of trans-parinaric acid fluorescence were similar. These results indicate that concanavalin-A affinity chromatography separates two membrane fractions differing in sialic acid content, lipid content, and enzyme profile but having the same right-side-out orientation.

Drug-induced surface membrane phospholipid composition in murine fibroblasts

The effects of drug on phospholipid composition of cell surface membranes are not well understood at this time. The effects of membrane-active drugs and membrane depolarization on the phospholipid composition were determined in murine LM fibroblasts. Receptor-aggregating drugs such as concanavalin A and cytoskeleton-disrupting agents such as colchicine, vinblastine, and cytochalasin B decreased phosphatidylserine content of the plasma membrane from 5.4 +/- 1.5% to as low as 1.4 +/- 0.2%. In addition, concanavalin A and colchicine increased the phosphatidylglycerol content from 6.9 +/- 1.6% to 13.1 +/- 0.7% and 10.6 +/-1.7%, respectively, while vinblastine and cytochalasin B had no effect. Pentobarbital decreased the content of phosphatidylinositol+ phosphatidylserine and of phosphatidylglycerol almost 2-fold. Propranolol, ethanol, and depolarization with 120 mM KCl had small or ne effects on plasma membrane phospholipid composition. None of the above drugs or treatments significantly altered the asymmetric distribution of phosphatidylethanolamine across the LM cell plasma membrane under the conditions tested. In addition, energy inhibitors that deplete the proton-motive force of the cell (NaN3 and KCN) and inhibitors of ATP synthesis such as NaAsO4 did not affect the asymmetric distribution of phosphatidylethanolamine. It is concluded that the mechanism of action of membrane-active drugs such as concanavalin A, vinblastine, colchicine and pentobarbital may involve alterations in plasma membrane composition. It also appears that microfilaments, microtubules, beta-adrenergic receptors, membrane fluidity, and membrane potential are not critical for the regulation of the asymmetric distribution of membrane phosphatidylethanolamine.

Plasma membrane aminophospholipid distribution in transformed murine fibroblasts

The possibility that the asymmetric distribution of aminophospholipids may be an intrinsic property of mammalian plasma membranes was examined in LM cells, a transformed murine fibroblast cell line. The cells were grown in suspension culture in a chemically defined medium without lipid, protein, or serum and then treated with 2,4,6-trinitrobenzenesulfonic acid (TNBS). A maximum of 4% of LM cell plasma membrane phosphatidylethanolamine and 5% of the phosphatidylserine was labelled with TNBS. Furthermore, long chain and unsaturated fatty acids were preferentially esterified to the non-derivatized phosphatidylethanolamine (inner monolayer) as compared to phosphatidylethanolamine derivatized with TNBS (outer monolayer). Isethionyl acetimidate, an alternative non-penetrating reagent, confirmed the results obtained with TNBS and provided supportive evidence for the highly asymmetric distribution of phosphatidylethanolamine; 6% of the phosphatidylethanolamine was labelled with isethionyl acetimidate. When the penetrating reagent methylacetimidate was used, more than 80% of the phosphatidylethanolamine was derivatized. Although the growing of the LM cells in 10% calf serum significantly increased plasma membrane phosphatidylcholine while decreasing phosphatidylethanolamine, calf serum had no significant effect on phosphatidylethanolamine or phosphatidylserine asymmetry.