J Am Soc Nephrol 9: 2082-2088, 1998

Decrease of Serum Paraoxonase Activity in Chronic Renal Failure

THIERRY F. DANTOINE,* JEAN DEBORD,t JEAN-PIERRE CHARMES,* LOUIS MERLE,t PIERRE MARQUET,t GERARD LACHATRE,t and CLAUDE LEROUXROBERT* *Nephrology Unit and Department of Pharmacology-Toxicology, Dupuytren University Hospital, Limoges, France.

Abstract. Paraoxonase is an esterase that hydrolyzes organo- peritoneab dialysis, and 132 renal transplant patients. Patients phosphate compounds. The enzyme is associated with HDL were compared with two groups of aged-matched control sub-

and could protect LDL against peroxidation, which suggests a jects (total number = 195). Especially with 4-nitrophenyl possible involvement of paraoxonase in the antiatherogenic acetate, paraoxonase activity was lower in patients with some properties of HDL. Paraoxonase activity has been shown to be degree of renal insufficiency (chronic renal failure [P < 0.05], low in patients with myocardial infarction, diabetes mellitus, or chronic hemodialysis [P < b01, chronic peritoneab dialysis familial hypercholesterobemia. Because cardiovascular disease [P < l0]) than in control subjects. In transplant patients, is the main cause of death in chronic renal failure, serum paraoxonase activity was not found to be different from that in paraoxonase activity was measured by spectrophotometry us- control subjects. The decrease of paraoxonase activity and thus ing three synthetic substrates (phenyl acetate, paraoxon, and the reduction of its antiatherogenic properties in renal failure 4-nitrophenyl acetate) in 305 patients with kidney disease, could be an essential factor of premature vascular aging, es- including 47 patients with non-end-stage chronic renal failure, peciably when dialysis is used. Renal transplantation seems to 104 patients treated with hemodialysis, 22 patients treated with restore paraoxonase activity.

Cardiovascular disease is the main cause of death in chronic renal groups have studied the HDL-linked enzyme paraoxonase/ failure (CRF) patients (accounting for approximately 50% of arylesterase ( 12, 13). Paraoxonase was first known for its pro- cases; references 1 and 2). The relative risk of myocardial infarc- tective activity against the toxicity of organophosphorus com- tion is five times greater in patients on renal replacement therapy pounds (by hydrolyzing the toxic metabolite of parathion: than in the general population (3). No clear explanation of this paraoxon) and its ability to hydrolyze several aromatic carbox- “accelerated” atherosclerosis has been found. Several factors, ylic acid esters (such as phenyl acetate and 4-nitrophenyl including increased prevalence of hypertension, diabetes melbitus, acetate) (14). More recently, evidence has emerged linking and lipid abnormalities, may contribute (4,5). After renal trans- human serum paraoxonase to atherosclerosis: It has been plantation, atherosclerosis remains a major cause of mortality, and shown that in vitro purified paraoxonase decreased LDL lipid dyslipidemias are also frequently observed (6). Recently, several peroxidation (15) and that purified paraoxonase significantly studies have shown a higher susceptibility of LDL to oxidation in reduced the ability of mildly oxidized LDL to induce mono- CRF patients, suggesting that the protection of lipoproteins cyte-endothelial interactions (16). Thus, a protective role of against oxidation might be impaired (7-9). paraoxonase against atherosclerosis has been suggested, as Because the oxidation hypothesis of atherosclerosis seems to well as an explanation of the protective role of HDL lipopro- be currently established (oxidized LDL promoting a number of teins against LDL oxidation. Other studies support this hypoth- processes leading to the formation of atherosclerotic plaques in esis by showing that a decreased activity was noticed in high the arterial wall: enhancement of macrophage uptake, cytotox- cardiovascular risk patients with diabetes mellitus, heterozy- icity toward endothelial cells, immunogenicity, and production gous familial hypercholesterolemia, or myocardial infarction of oxysterols, cytokines, and growth factors) (10,1 1), several (1 7, 18). In a previous study, we showed that paraoxonase activity decreased with age (19). Caucasian populations display a triphasic distribution of serum Received August 12, 1997. Accepted May 4, 1998. paraoxonase activity toward paraoxon: Three phenotypes and These data were presented in part at the 29th Annual Meeting of the American genotypes (AA [low activity], AB [intermediate activity], and BB Society of Nephrology, New Orleans, LA, November 3-6, 1996, and received [high activity]) have been defined (20-22). Genetic studies have a blue ribbon. associated that variation with two paraoxonase isozymes whose Correspondence to Dr. Jean-Pierre Charmes, Service de Nephrologie, Chu Limoges, 2 Avenue Martin Luther King, 87042 Limoges Cedex, France. difference is related to a single amino acid (glutarnine or arginine 191) (23,24). A second polymorphism of the paraoxonase gene in 1046-6673/0901 1-2082$03.00/0 Journal of the American Society of Nephrology position 54 is known; it has been shown to affect the concentration Copyright © 1998 by the American Society of Nephrology of paraoxonase, but not its specific activity (24-26). Recently, J Am Soc Nephrol 9: 2082-2088. 1998 Paraoxonase in Chronic Renal Failure 2083

two groups associated a higher coronary risk to the AB and BB was determined so that the final absorbance did not exceed I . Initial phenotypes (27,28), whereas three other groups found no associ- rates were computed by linear regression and expressed in U/mb. The ation between paraoxonase gene polymorphism and coronary within-day and day-to-day coefficients of variation were <5% (32). heart disease (29-31). Because cardiovascular disease is the main cause of death in Assays CRF, we measured paraoxonase activity in that population in Paraoxonase activity was measured with three synthetic esters: paraoxon (20), phenyl acetate (20), and 4-nitrophenyl acetate (33). which atherosclerosis incidence is high and premature. The With paraoxon, 500 pl of 2/5 prediluted serum was added to 2 ml relation between CRF and atherosclerosis is not well known. of glycine buffer (50 mM, pH 10.05, 25#{176}C)containing 1.0 M NaC1, 1 .0 mM CaCb2, and 1 .25 mM paraoxon (0, O-diethyl-O-4-nitrophe- Materials and Methods nylphosphate). The rate of generation of 4-nitrophenol was deter-

Patients and Control Subjects mined at 412 nm (#{128}42 = 17,000 M’ . cm at pH 10.05). In The control population (n) consisted of 195 (120 women, 75 men; addition, the hydrolysis of paraoxon was followed over 15 mm in a mean age: 50.5 ± 24.5 yr [range, 17 to 105]) apparently healthy small group of 20 subjects (five of each category) with a measurement subjects who either attended a routine health check or were staff of the wavelength set at 450 nm (#{128}4s() 3500 M I . cm I to avoid Dupuytren Hospital. Ninety-seven subjects were under age 65 and excessive absorbances. were not taking medication; 98 were older and were taking less than With phenyl acetate, 500 pA of 1/100 prediluted serum was added three medications. None had renal disease, familial hyperchobesterol- to 2 ml of Tris-HC1 buffer (9 mM, pH 8.00, 25#{176}C)containing 0.9 mM emia, diabetes, or myocardial infarction. The 305 CRF patients stud- CaCl2 and 1.25 niM phenyl acetate. The rate of generation of phenol ied ( 1 32 women. 173 men) were either in- or outpatients of Dupuytren was determined at 270 nm (#{128}270 1306 M ‘. cm ‘). Nephrology Unit. There were 47 nondialyzed, non-end-stage renal With 4-nitrophenyl acetate, 250 M1 of 1/20 prediluted serum was disease (NESRD) patients (20 women, 27 men; mean age: 62.6 ± added to 2 ml of Tris-HC1 buffer (25 mM, pH 7.4, 25#{176}C)containing 17.6 yr [17 to 86 yrl; creatinine clearance: 22.1 ± 15.5 mI/mm per 1 .0 mM CaCI2, 2.5% methanol, and 0.625 mM 4-nitrophenyl acetate. 1.73 m2 [range, 6 to 76]), 104 hemodialysis (HD) patients (40 women, The rate of generation of 4-nitrophenol was determined at 402 nm

64 men; mean age: 58.7 ± 15.2 yr [20 to 83 yr]), 22 peritoneal dialysis (#{128}402 14,000 M’ . cm’ at pH 7.4). (PD) patients (10 women, 12 men; mean age: 63.0 ± 18.7 yr [23 to With each substrate, the kinetic curves were linear for at least I 85 yr]), and 132 renal transplant (RT) patients (62 women, 70 men; mm. There was no evidence of pre-steady-state kinetics. mean age: 45.6 ± 12 yr [16 to 68 yr]; creatinine clearance: 48.4 ± 17.9 mb/mm per 1.73 m2 [range, 1 1 to 140]). All control subjects and Statistical Analyses patients were unrelated and Caucasian. Treatments were recorded. The S-Plus software (34) was used. The statistical distribution of The study protocol was approved by the local ethics committee. enzyme activity was studied by the Kolmogorov-Smirnov test. Case-

control difference in continuous variables was studied by t test or Methods Mann-Whitney test. The x test was used to compare phenotypic Serum paraoxonase activity was measured by spectrophotometry distributions. using three synthetic substrates: paraoxon (diethyl-4-nitrophenyl Two methods were used to compare the enzyme activities of the phosphate), phenyl acetate, and 4-nitrophenyl acetate. patients and the control subjects. The first method involved the The hydrolysis of paraoxon by the B isozyme was shown to be selection, within the control population, of a subgroup whose age stimulated by a molar solution of NaCl, whereas the A isozyme was matched the patients’ ages (i.e. , no significant difference between the not. The hydrolysis of phenyl acetate was the same with both mean ages of the control and patient groups). This analysis needed one isozymes. This property allowed Eckerson et al. (20) to calculate the control subgroup (n = 156) for the NESRD, HD, and PD patients phenotype of a given subject by the following ratio: (whose ages were similar), and another one (n 185) for the RT patients (whose age was significantly lower). A second method used with the whole set of data (patients and R - U(paraoxon) x 1O - U(phenyl acetate) control subjects, without previous selection) involved a multiple re- gression analysis, taking into account the influence of age, sex, and (Enzyme activities are expressed in international units (U) per renal disease on enzyme activity, as well as the interaction between milliliter of serum: 1 U corresponds to the quantity of enzyme that age and renal disease. The regression model was expressed as: hydrolyses 1 mol of substrate per minute, at the given pH and log(Activity) a0 + a1 X Sex + a2 X Age + a3 X NESRD + a4 temperature.) This ratio compares the hydrolysis of paraoxon in the XHD+aXPD+a6XRT+b3XAgeXNESRD+b4XAgeX presence of I M NaCI to the hydrolysis of phenyl acetate without HD + b5 X Age X PD + b6 X Age X RT (with Sex = 1 for male NaCI and usually identifies three groups in Caucasian healthy popu- and 0 for female, NESRD = 1 for NESRD patients or 0, HD 1 for lations. These groups match the three phenotypes: AA, AB, or BB. hemodialysis patients or 0, PD = 1 for peritoneab dialysis patients or This “dual substrate method” correlates well with results obtained 0, and RT = 1 for renal transplant patients or 0). This model was used from genetic studies (25). in the total population (500 subjects), including control subjects (for Venous blood was sampled (5 ml in a Vacutainer tube without whom NESRD HD PD RT 0). additive) after an overnight fast, and serum was assayed on the same day or stored at -80#{176}C until the assay. All reagents used in the Results different assays were supplied by Sigma (Paris, France). Analyses were performed at 25#{176}Cwith an ultraviolet/visible spectrophotometer Figure 1 shows an example of kinetic curves obtained with (UV 1 205 Shimadzu, Courtaboeuf, France) equipped with a Peltier paraoxon in four NESRD patients. There was no evidence of thermostated unit. Absorbance was measured automatically every pre-steady-state kinetics (i.e. , exponential decrease of hydro- 0.5 5 or every second during 1 mm. The amount of serum in the assay bysis rate toward a constant steady-state value), even when the 2084 Journal of the American Society of Nephrology J Am Soc Nephrol 9: 2082-2088, 1998

Absorbance

0.8

0.6

0.4

02

0 0 20 30 40 50 60

Time (s)

Figure 1. Kinetic curves of paraoxon hydrolysis in four patients with non-end-stage renal disease. Lines were fitted by linear regression.

hydrolysis was followed for 15 mm. The enzyme activities higher and 4-nitrophenyl acetate hydrolysis about 10-fold observed in our subjects are summarized in Tables 1 and 2. The higher. This was observed in every group of subjects. The hydrolysis rates of paraoxon exhibited wide variations in the Kobmogorov-Smirnov test did not show significant departure control population as in CRF patients. With phenyl acetate and from the log-normal distribution in each group. 4-nitrophenyl acetate, the variations (in both control and CRF The dual substrate method showed a trimodal repartition in groups) were high but smaller than with paraoxon. Wide dis- our Caucasian control population (Figure 2), as in the patho- persion was also noticed within the different phenotype groups logic CRF populations with a ratio R < 1 .6 for homozygous (data not shown). Paraoxonase activity showed very different AA, 1 .6 R < 4 for heterozygous AB, and R 4 for levels with the substrate used: When compared to the activity homozygous BB subjects. According to the y test, the appar- with paraoxon, phenyl acetate hydrolysis was almost 500-fold ent phenotype distributions were not significantly different in

Table 1. Paraoxonase activity (U/mb) assayed with the substrates 4-nitrophenyl acetate, paraoxon, and phenyl acetate in the serum of age-matched control subjects, patients with non-end-stage chronic renal disease, hemodialysis, and peritoneal dialysi?

Subjects 4-Nitrophenyl Acetate Paraoxon Phenyl Acetate

Control subjects (n 156) 1.64 ± 0.66 0.20 ± 0.14 94 ± 32 [0.45; 4.78] [0.003; 0.804] [37; 211] 1.18 0.09 70 NESRD (n 47) 1 .42 ± 060b 0. 1 9 ± 0. 16 95 ± 37 [0.63; 3.78] [0.02; 0.98] [42; 197] 1.27 0.15 84

HD (n = 104) 1.07 ± 0.37c 0.14 ± 0.07c 74 ± 20C [0.1 1; 2.28] [0.01; 0.34] [25; 128] 1.0 0.13 71

PD (n = 22) 1 . 12 ± 0.48’ 0. 1 3 ± 0.08” 95 ± 25 [0.53; 2.40] [0.01; 0.32] [54; 164] 0.98 0.12 92

a Comparisons were performed with the t test. Results are given as mean ± SD, [minimum; maximum], and median. NESRD, non-end- stage renal disease; HD, hemodialysis; PD, peritoneal dialysis. “P < 0.05. C p < 0.001. J Am Soc Nephrol 9: 2082-2088, 1998 Paraoxonase in Chronic Renal Failure 2085

Table 2. Paraoxonase activity (U/mb) assayed with the substrates 4-nitrophenyl acetate, paraoxon, and phenyl acetate in the serum of age-matched control subjects and renal transplant patients’

Subjects 4-Nitrophenyl Acetate Paraoxon Phenyl Acetate

Control subjects (n = 185) 1.89 ± 0.92 0.22 ± 0.17 100 ± 33 [0.45; 6.18] [0.016; 0.84] [37; 252] 1.67 0.17 97

RT (n = 132) 1.80 ± 0.68 0.20 ± 0.12 106 ± 25 [0.33; 4.16] [0.03; 0.69] [34; 203] 1.66 0.15 106

a Comparisons were performed with the t test. Results are given as mean ± SD, [minimum; maximum], and median. No significant differences were found between patients and controls. RT, renal transplant.

Paraoxon (U/mL) doubtful, because we observed a significant increase of the 1 activating effect of NaCl in the HD and TR groups (but not in the NESRD patients) (35). Creatinine clearances of NESRD patients were significantly 0.8 U lower than creatinine clearances of RT patients (P < 0.0001). U . No significant correlation between paraoxonase activity and 0.6 creatinine clearance was found. uI I Comparison of patients with age-matched control groups I .. showed that the mean paraoxonase activity was diminished in U. : 0.4 . . NESRD, HD, and PD patients (Table 1 ). The highest modifi- U C: I cation was noticed in HD patients, whatever the substrate. In : U PD patients, the decrease was found mainly with paraoxon and 0.2 . 4-nitrophenyl acetate. In NESRD patients, a decrease was .‘: observed only with 4-nitrophenyl acetate, and was smaller than 0 #{149},..,.__- in HD or PD patients. The activity in RT patients did not differ 0 100 200 300 significantly from the control subjects (Table 2). It is interest- Phenyl acetate (U/mL) ing to note that, among the three substrates, 4-nitrophenyl

Figure 2. Control subject phenotypes, determined by the dual sub- acetate was the only one that showed a modified activity in strate method. #{149},A; 0, AB; #{149},B. most patient groups. Multiple regression analysis (Table 3) confirmed these re- suits and, in addition, showed that paraoxonase activity de- control subjects (AA: 46%; AB: 42%; BB: 12%), in nontrans- creased with age. This phenomenon was observed with all plant CRF patients (AA: 49.7%; AB: 49.3%; BB: 6.4%), and in three substrates, and its magnitude was similar in control RT patients (AA: 49.2%; AB: 49.3%; BB: 6.4%). However, subjects, NESRD, and PD patients (due to the absence of the calculated distributions in the pathologic groups may be significant interaction terms for these groups in the equation).

Table 3. Paraoxonase activity (U/mb) according to different substrates and factors’

Coefficient Factor 4-Nitrophenyl Acetate Paraoxon Phenyl Acetate

20856h a0 Gender 0.3903L a, Age o0040h 0oo22t) a3 NESRD NS NS a4 HD as PD NS b4 Age X HD 00043b 0.0043c NS r2 0.57 0.32 0.47

a Log (Activity) = a0 + a1 X Sex + a2 X Age + a3 X NESRD + a4 X HD + a5 X PD + a6 X RT + b1 x Age X NESRD + h4

x Age X HD + b5 X Age X PD + b6 x Age X RT. Abbreviations as in Tables 1 and 2. bp < 0.001.

C p < 0.05.

dp < 0.01. 2086 Journal of the American Society of Nephrology J Am Soc Nephrol 9: 2082-2088, 1998

However, in HD patients, with 4-nitrophenyb acetate and para- The reduced activity measured in patients did not seem to be oxon, the interaction term (b4, Table 3) nearly compensated the related to phenotype differences (i.e., no excess of AA phenotypes coefficient of age (a,), thus making the influence of age in the patient population). In HD patients, the decrease in enzyme negligible in this category of patients, with these substrates. activity was detected with all three substrates, including phenyl acetate, which is not sensitive to the phenotype. In addition, this study did not demonstrate a significant difference between the Discussion phenotype distribution in the patient and control groups, or be- Paraoxonase activity in CRF was investigated with three tween these groups and literature data (38,39). It is not possible to substrates: phenyl acetate, paraoxon, and 4-nitrophenyl acetate. completely dismiss a methodobogic bias, because it has been The kinetic curves were always linear in our experimental shown that the stimulation of paraoxonase by NaCI, which is the conditions. Some authors (36,37) reported a pre-steady-state basis of the phenotyping method used here, is more important in phase with paraoxon in some samples, but their conditions HD and RT patients, but not in NESRD patients (37). The direct (substrate and enzyme concentrations, pH, solvent) differed determination of the genotype was not available to us. However, widely from ours. It has been suggested that the reaction of it seems justified to admit that, at least for NESRD patients, the paraoxon with serum cholinesterase could explain the initial decrease in enzyme activities is not related to the phenotype. In “burst” of 4-nitrophenol observed in some experiments (36). addition, the recent results of Hasselwander et a!. (40) have shown Because paraoxon is an irreversible inhibitor of cholinesterase, that the amount of enzyme protein is not modified in patients with the amount of 4-nitrophenol released by this reaction cannot CRY, so that the observed alterations represent a decrease in the exceed the molar concentration of the enzyme, which is ap- specific activity of the enzyme. These last results also seem to rule proximately 50 nM in the average patient (0. Lockridge, out the possible implication of a second polymorphism (Met/Leu) personal communication). It is very unlikely that this reaction in position 54, since this polymorphism has a strong effect on could have influenced our results, because the serum was enzyme concentration (26). We were not able to check these diluted 2.5 times and the amount of 4-nitrophenol released was findings in our patients, however, because neither the determina- in the range 10 to 100 j.tM. tion of the genotype nor the immunoassay of the enzyme was As shown previously in other high cardiovascular risk pop- available to us. ulations (17, 18), this study showed that paraoxonase activity In an Italian population, Schiavon et al. also found that was lower in patients suffering from CRF than in control paraoxonase activity (toward paraoxon) was significantly subjects, especially when HD or PD was required. In transplant lower (P < 0.01) in 60 HD patients than in 64 healthy subjects patients, paraoxonase activity was not different from that in (41). Mean paraoxonase activity was approximately I .5 times control subjects. higher in our French groups than in the Italian groups. Higher The decrease of paraoxonase activity was marked and sig- activities in our group could be explained by different assay nificant in dialysis patients, leading to very bow activities conditions (buffers, pH, calcium concentrations) or by un- whatever the substrate. In NESRD patients, results were less known environmental or population differences. But paraoxo- significant with paraoxon than with 4-nitrophenyl acetate. nase activity variations between non-HD and HD patients in Even if the latter substrate was never used when the relation- each population were about the same (decrease ratio: 0.636 in ship between paraoxonase activity and atherosclerosis was our study versus 0.620 in Schiavon’s study), which corrobo- investigated, 4-nitrophenyl acetate has already been used in rates the notion that dialysis or uremia probably induces several studies for measuring paraoxonase activity and has changes in the enzyme even if the studied population or the been shown to be rather specific for this enzyme; it also has paraoxonase activity assay conditions change. Schiavon et a!. been shown to be a “discriminating” substrate (like paraoxon did not explore NESRD or PD patients. but unlike phenyl acetate), useful to phenotype paraoxonase In CRF patients, an increased effect of oxidative stress on (33). For the first time, this study suggests that 4-nitrophenyl LDL lipoproteins in vitro has been reported (7-9), as it had acetate may be very sensitive for detecting variations in para- been noticed in patients with familial hypercholesterolemia oxonase activity, perhaps because it provides higher activities (42), diabetes mebbitus (43), or coronary heart disease (44). within narrower ranges and thus probably diminishes analytical This work sustains the relationship between bow paraoxonase errors. Moreover, 4-nitrophenyl acetate is a nontoxic ester, activity and diseases with low antioxidant defense and exces- unlike paraoxon, and thus would be more suitable for routine sive lipid peroxidation. Paraoxonase may play its protective determinations. role in atherogenesis by hydrolyzing some products of lipid As we noticed previously (19), paraoxonase activity de- peroxidation and consequently by limiting LDL oxidation and creased with age in control subjects. Considering the suspected foam cell synthesis (16). This allows us to hypothesize that protective role of paraoxonase in atherosclerosis, it is interest- during uremia, loss of that protection could represent a very ing to note that aging is accompanied by this (albeit modest) important factor in atherosclerosis. In a previous study, we reduction in enzyme activity. Paraoxonase activity diminishing found that in HD patients, HDL prevented LDL oxidation but with time could be a non-negligible factor of increased athero- with a significantly reduced effect when compared with control sclerosis development in elderly people. Age also decreased subjects (8). We have completed this study by measuring paraoxonase activity in NESRD and PD patients but not in HD serum paraoxonase activity and HDL paraoxonase activity in patients. five HD patients and in five control subjects. Results showed J Am Soc Nephrol 9: 2082-2088, 1998 Paraoxonase in Chronic Renal Failure 2087 that both serum and HDL paraoxonase activity were signifi- portant factor in premature vascular aging, especially when cantly lower in the five HD patients (data to be published). dialysis is used. This could represent additional proof that a decrease in para- oxonase activity may lead to failure in the protection of LDL oxidation in renal failure. Acknowledgments In this cross-sectional study, we measured serum lipids in a We thank Mr. J. Comte for technical support and Mrs. C. Frugier for the preparation of the manuscript. sample of 32 healthy subjects and 40 CRF patients. Total cholesterol, triglyceride, and HDL cholesterol levels were not significantly different between patients and control subjects References (data not shown). Several studies that tried to show a relation- I. Gomez-Farices MA, McClellan W, Soucie JM, Mitch WE: A ship between paraoxonase activity and apobipoproteins, serum prospective comparison of methods for determining if cardiovas- cholesterol, or triglyceride levels gave somewhat contradictory cular disease is a predictor of mortality in dialysis patients. Am J results (45-47). Schiavon et a!. (41) found no significant Kidney Dis 23: 382-388, 1994 difference between lipid profiles (total cholesterol, triglycer- 2. Owen WF, Madore F, Brenner BM: An observational study of ides, HDL cholesterol, apo A-l, apo B levels, HDL/apo A-b cardiovascular characteristics of long-term end-stage renal dis- ratio) of HD patients and control subjects, but they showed that ease survivors. Am J Kidney Dis 28: 931-936, 1996 3. Bloembergen WE, Port FK, Mauger EA, Wolfe RA: Causes of the paraoxonase activitylHDL cholesterol ratio was signifi- death in dialysis patients: Racial and gender differences. Am J cantly lower in uremic patients than in control subjects without Soc Nephrol 5: 1231-1242, 1994 significant correlation between paraoxonase activity and HDL 4. Attman P0, Samuelsson 0, Alaupovic P: Lipoprotein metabo- cholesterol bevel or apo Al bevel. The lack of prevention of lism and renal failure. Am J Kidney Dis 21: 573-592. 1993 LDL oxidation by HDL could be explained by this qualitative 5. Kronenberg F, Konig P. Neyer U, Auinger M, Pribasnig A, Lang alteration of HDL in renal failure. U, Reitinger J, Pinter 0, Utermann 0, Dieplinger H: Multicenter In transplant patients, paraoxonase activity seems to be study of lipoprotein(a) and apolipoprotein(a) phenotypes in pa- restored to normal bevels. Paraoxonase activity was higher in tients with end-stage renal disease treated by hemodialysis or RT patients whose creatinine clearances were also significantly continuous ambulatory peritoneal dialysis. J Am Soc Nephrol 6: higher than creatinine clearances of the NESRD patients. Para- 110-120, 1995 oxonase metabolism is still unknown. The DNA coding for 6. Hilbrand LB, Demacker PNM, Hoitsma AJ, Stalenhoef AFH, Koene RAP: The effects of cyclosporine and prednisone on paraoxonase has been found in liver cells (14), but no study has serum lipid and (apo)lipoprotein levels in renal transplant recip- explored paraoxonase in kidneys. Decreased activity during ients. JAm Soc Nephrol 5: 2073-2081, 1995 CRF, especially when renal replacement therapy is required 7. Panzetta 0, Cominacini L, Garbin U, Fratta Pasini A, Gammaro (HD and PD), and restoration of this activity after renal trans- L, Bianco F, Davoli A, Campagnola M, De Santis A, Pastorino plantation suggest, for the first time, that the kidneys play an AM: Increased susceptibility of LDL to in vitro oxidation in important robe in paraoxonase metabolism. During the pre- patients on maintenance hemodialysis: Effects of fish oil and transplantation evaluation, the patients with the most severe vitamin E administration. Cli,i Nephrol 44: 303-309, 1995 and widespread atherosclerosis lesions were excluded. Conse- 8. Cristol JP, Dantoine T, Morena M, Vaussenat F, Carbonneau A, quently, a bias may have been introduced. However, in five Leger C, Descomps B, Canaud B: Protective effects of HDL patients in whom paraoxonase activity was measured both against oxidative stress is impaired in hemodialysis patients before and after transplantation, values during the HD period [Abstract]. Nephrol Dial Transplant 12: A156, 1997 9. Maggi E, Bellazzi R, Falaschi F, Frattoni A, Perani 0, Finardi 0, (which were low) did not differ from those in other HD Gazo A, Nai M, Romanini D, Bellomo 0: Enhanced LDL oxi- patients; the values doubled after transplantation and still did dation in uremic patients: An additional mechanism for acceler- not differ from those in other RT patients (preliminary work to ated atherosclerosis? Kidney tnt 45: 869-876, 1994 be published). 10. Steinberg D, Parthasarathy 5, Carew TE, Khoo IC, Witztum JL: During renal failure, we suggest that within the environment of Beyond cholesterol: Modifications of low-density lipoprotein the arterial wall, the accumulation of nitrogen-derived products that increase its atherogenicity. N Engi J Med 320: 9 15-924, could decrease paraoxonase activity by directly inhibiting the 1989

enzyme or by modifying its synthesis/secretion. The recent study 1 1 . Witztum iL: The oxidation hypothesis of atherosclerosis. Lancet of Hasselwander et a!. seems to confirm the former mechanism by 344: 793-795, 1994 showing that the protein concentration was not decreased in HD 12. Mackness Ml, Arrol 5, Abbott C, Durrington PN: Protection of patients compared with control subjects (40). low-density lipoprotein against oxidative modification by high- density lipoprotein-associated paraoxonase. Atherosclerosis 104: In conclusion, this preliminary epidemiologic study opens a 129-135, 1993 new research area in the understanding of atherogenesis mech- 13. Kelso GJ, Stuart WD, Richter Ri, Furlong CE, Jordan-Stack T, anisms in CRF. The involvement of paraoxonase in the exces- Harmony JAK: Apolipoprotein J is associated with paraoxonase sive LDL lipoperoxidation, noticed in uremia, and its physio- in human plasma. Biochemistry 33: 832-839, 1994 logic substrates need to be determined. This study presents a 14. Aldridge WN: Serum esterases 2. An enzyme hydrolyzing di- potential explanation of accelerated atherosclerosis in CRF. ethyl p-nitrophenyl phosphate (E 600) and its identity with the The decrease of paraoxonase activity and thus the decrease of A-esterase of mammalian sera. Bioche,n J 53: 1 17-124, 1953 its antiatherogenic properties in renal failure could be an im- 15. Mackness MI, Arrol 5, Durrington PN: Paraoxonase prevents 2088 Journal of the American Society of Nephrology I Am Soc Nephrol 9: 2082-2088, 1998

accumulation of lipoperoxides in low density lipoprotein. FEBS (HUMPONA) is not associated with the risk of coronary artery Leti 286: 152-154, 1991 disease in Finns. J Clin invest 98: 883-885, 1996 16. Watson AD, Berliner JA, Hama SY, La Du BN, Faull KF, 32. Debord J, Dantoine T, Suchaud K, Harel M, Verneuil B, Merle Fogelman AM, Navab M: Protective effect of high density li- L: Etalonnage non bin#{233}aired’un dosage enzymatique par la poprotein associated paraoxonase. J Clin Invest 96: 2882-2891, m#{233}thodedes moindres carr#{233}sg#{233}n#{233}ralis#{233}s:Applicationaux ester- I 995 ases du serum humain. Analusis 25: 293-297, 1997 17. McElveen J, Mackness MI, Colley CM, Peard T, Warner 5, 33. Haagen L, Brock A: A new automated method for phenotyping Walker CH: Distribution of paraoxon hydrolytic activity of pa- arylesterase (EC 3.1.1.2) based upon inhibition of enzymatic tients after myocardiab infarction. Clin Chem 32: 671-673, 1986 hydrolysis of 4-nitrophenylacetate by phenyl acetate. Eur J Clin 18. Mackness MI, Harty D, Bhatnagar D, Winocour PH, Arrol S, Chem Clin Biochem 30: 391-395, 1992 Ishola M, Durrington PN: Serum paraoxonase activity in familial 34. S-Plus: Statistical Sciences, Inc., version 3.2 for Windows. Se- hypercholesterolemia and insulin-dependent diabetes mellitus. attle, 1994 Atherosclerosis 86: 193-199, 1991 35. Dantoine T, Debord i, Charmes iP, Merle L, Leroux-Robert C: 19. Dantoine T, Debord J, Charmes J-P, Marquet P, Lachatre 0, Paraoxonase activity stimulation by salts is higher in chronic Merle L: Baisse de l’activit#{233}arylest#{233}raseavec l’#{226}ge:Incidence renal failure patients than in controls [Letter]. Nephrol Dial dans l’ath#{233}romatose? [Abstract] Proc Soc Fr Gerontol Paris 1: Transplant 13: 816, 1998 63, 1995 36. McElveen J, Mackness MI, Colley CM, Peard T, Warner S, 20. Eckerson HW, Romson J, Wyte C, La Du BN: The human serum Walker CH: Distribution of paraoxon hydrolytic activity in the paraoxonase polymorphism: Identification of phenotypes by serum of patients after myocardial infarction. Cliri Chem 32: their response to salts. Am J Hunz Genet 35: 214-227, 1983 671-673, 1986 21 . La Du BN, Eckerson HW: The polymorphic paraoxonase/aryles- 37. Vitarius iA, Sultatos LG: Kinetic mechanism of the detoxifica- terase isozymes ofhuman serum. Fed Proc 43: 2338-2341, 1984 tion of the organophosphate paraoxon by human serum A-ester- 22. La Du BN, Novais J: Human serum organophosphatase: Bio- ase. Drug Metab Dispos 22: 472-478, 1994 chemical characteristics and polymorphic inheritance. In: En- 38. Eckerson HW, Wyte CM, La Du BN: The human serum para- zymes Hvdrolvsing Organophosphorus Compounds, edited by oxonase/arylesterase polymorphism. Am J Hum Genet 35: 1 126- Reiner E, Aldridge WN, Hoskin FCG, Chichester, United King- 1138, 1983 dom, Ellis-Horwood, 1989, pp 41-52 39. Playfer JR. Eze LC, Bullen MF, Evans DAP: Genetic polymor- 23. La Du BN, Adkins 5, Kuo CL, Lipsig D: Studies on human phism and interethnic variability of plasma paraoxonase activity. serum paraoxonase/arylesterase. Chem-Biol interact 87: 25-34, J Med Genet 13: 337-342, 1976 I 993 40. Hasselwander 0, Young IS, Mackness MI, Arrol 5, Mc Master 24. Humbert R, Adler DA, Disteche CM, Hassett C, Omiecinski CJ, D, Fogarty DO, Nicholls DP, Maxwell AP: Paraoxonase specific Furlong CE: The molecular basis of the human serum paraoxo- activity is reduced in patients with chronic renal failure [Ab- nase activity polymorphism. Nat Genet 3: 73-76, 1993 stract]. J A,n Soc Nephrol 8: 6b8A, 1997 25. Adkins 5, Gan KN, Mody M, La Du BN: Molecular basis for the 4 1 . Schiavon R, De Fanti E, Giavarina D, Biasioli 5, Cavalcanti 0, polymorphic forms of human serum paraoxonase/arylesterase: Guidi 0: Serum paraoxonase is decreased in uremic patients. Glutamine or arginine at position 191, for respective A or B Clin Chim Acta 247: 71-80, 1996 allozymes. Am J Hum Genet 52: 598-608, 1993 42. Lavy A, Brook Gi, Dankner 26. Blatter MC, James RW, Dussoix P, Blanche H, Passa P, Froguel 0, Amotz AM, Aviram M: En- P, Ruiz J: Paraoxonase polymorphism Met-Leu 54 is associated hanced in vitro oxidation of plasma lipoprotein derived from with modified serum concentrations of the enzyme. J Clin Invest hypercholesterolemic patients. Metabolism 40: 794-799, 1991 99: 62-66, 1997 43. Babiy AV, Gebicky JM, Sullivan DR: Increased oxidizabibity of 27. Serrato M, Marian AJ: A variant of human paraoxonase/aryles- plasma lipoproteins in diabetic patients can be decreased by terase (HUMPONA) gene is a risk factor for coronary artery probucol therapy and is not due to glycation. Biochem Pharma- disease. J Cliii Invest 96: 3005-3008, 1995 cot 43: 995-1 000, 1992 28. Ruiz J, Blanche H, James RW, Garin M-C, Vaisse C, Charpentier 44. Regnstrom iN, Nilsson J, Tornavabl P. Landou C, Hamsten A: 0, Cohen N, Morabia A, Passa P, Froguel P: Gln-Argl92 poly- Susceptibility to low-density lipoprotein oxidation and coronary morphism of paraoxonase and coronary heart disease in type 2 atherosclerosis in man. Lancet 339: 1 183-1 186, 1992 diabetes. Lancet 346: 869-872, 1995 45. Blatter-Garin MC, Blanche H, Ruiz J, Passa P, Froguel P. James 29. Herrmann SM, Blanc H, Poirier 0, Arveiler D, Luc 0, Evans A, R: The relationship between paraoxonase polymorphism and Maques-Vidal P. Bard J-M, Cambien F: The Oln!Arg polymor- plasma lipids/lipoproteins in non-insulin dependent diabetes. phism of human paraoxonase (PON 192) is not related to myo- Atherosclerosis 1b5[Suppl]: 523, 1995 cardial infarction in the ECTIM study. Atherosclerosis 126: 46. Nevin DN, Zambon A, Furlong CE, Richter RJ, Humbert R, 299-303, 1996 Hokanson iE, Brunzell iD: Paraoxonase genotypes, lipoprotein 30. Suehiro T, Nakauchi Y, Yamamoto M, An K, Itoh H, Hamash- lipase activity and HDL. Arterioscler Thromb Vasc Biol 16: ige N, Hashimoto K: Paraoxonase gene polymorphism in Japa- 1243-1249, 1996 nese subjects with coronary heart disease. mt i C’ardiol 57: 47. Abbott CA, Mackness MI, Kumar 5, Boulton AJ, Durrington 69-73, 1996 PN: Serum paraoxonase activity, concentration and phenotype 31. Antikainen M, Murtom#{228}ki 5, Syvanne M, Pahlman R, Tah- distribution in diabetes mellitus and its relationship to serum vanainen E, Jauhiainen M, Heikki Frick M, Ehnholm C: The lipids and lipoproteins. Arterioscler Thromb Vasc Biol 15: 18 12- Gln-Argl9l polymorphism of the human paraoxonase gene 1818, 1995