- Email: [email protected]
On the physiological role(s) of the paraoxonases
Chemico-Biological Interactions 119 – 120 (1999) 379 – 388
On the physiological role(s) of the paraoxonases Bert N. La Du a,*, Michael Aviram b, Scott Billecke a, Mohamad Navab c, Sergio Primo-Parmo a, Robert C. Sorenson a, Theodore J. Standiford d a
Departments of Anesthesiology and Pharmacology, Uni6ersity of Michigan Medical School, Ann Arbor, MI 48109, USA b Lipid Research Laboratory, Rambam Medical Center and the Bruce Rappaport Faculty of Medicine, Technion and the Rappaport Family Institute for Research in the Medical Sciences, Haifa 31096, Israel c Di6ision of Cardiology, Department of Medicine, School of Medicine, Uni6ersity of California, Los Angeles, CA 90095, USA d Department of Medicine, Uni6ersity of Michigan Medical School, Ann Arbor, MI 48109, USA
Abstract In recent years several lines of evidence have indicated that serum paraoxonase (PON1), and perhaps other mammalian paraoxonases, act as important guardians against cellular damage from toxic agents, such as organophosphates, oxidized lipids in the plasma low density lipoproteins (LDL), and against bacterial endotoxins. For some of these protective activities but not all, PON1 requires calcium ion. The catalyzed chemical reactions generally seem to be hydrolytic, but for some types of protection this may not be so. Several other metals have very high affinity for PON1 and may displace calcium. Replacement or substitution of calcium by other metals could extend the range of catalytic properties and the substrate specificity of the paraoxonases, as it does for the mammalian DFPases. Although this Third International Meeting on Esterases Reacting with Organophosphorus Compounds focuses on the organophosphatase activities of paraoxonase and related enzymes, it is important to also briefly review some of the current directions in several laboratories searching for additional functions of the paraoxonases to extend our understanding of the properties of this family of enzymes which now seem to have both physiological and toxicological importance. © 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Bacterial endotoxin; HDL; LDL; Paraoxonases; PON1 mutants; PON evolution; PON polymorphisms
* Corresponding author. Tel.: +1-734-7636429; fax: +1-734-7649332. E-mail address: [email protected] (B.N. La Du) 0009-2797/99/$ - see front matter © 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 0 0 9 - 2 7 9 7 ( 9 9 ) 0 0 0 4 9 - 6
380
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
1. Introduction: search for natural substrates for the paraoxonases Several lines of evidence in recent years have suggested that serum paraoxonase/arylesterase (PON1) and closely related enzymes may have physiological functions as well as the ability to hydrolyze toxic organophosphate compounds. The first line of evidence is recent genetic information. There has been a long standing assumption about the origin of paraoxonase family; it was presumed that the paraoxonases arose as an offshoot from the serine esterases, the carboxyesterases [1]. Structural analyses of the respective genes and proteins; however, have shown no homology between the arylesterases and the carboxyesterases, even though they share a few substrates, such as phenyl acetate. There is no structural homology supporting the claim that they share some common evolutionary heritage or were derived from a common ancestral esterase. The PON family in mammals (mice and humans, at least) have three distinct PON genes, located very close to each other on the same chromosome (chromosome 7 in man, chromosome 6 in the mouse) [2]. Presumably gene duplication has occurred during evolutionary development to cause this redundancy, and it suggests that this family of enzymes may contribute to some important physiological functions, quite independently from the unusual ability of PON1 to hydrolyze toxic organophosphates, carbamates and aromatic esters. The limited nucleotide sequence information we have, to date, has been translated into the predicted amino acid sequences of PON proteins from C. elegans to human beings and summarized in Fig. 1. A logical evolutionary pattern of the PON family of proteins can be constructed which brings out distinctive characteristics that make it a completely different family than any other. It is unrelated to the cholinesterases, the bacterial phosphotriesterases or the mammalian DFPases sequenced until now (Fig. 2). PON2 appears to be an older family member than PON1, and the latter seems to be a relatively recent arrival, compared to PONs 2 and 3. Secondly, several lines of evidence are emerging that PON1 (at least) protects the mammalian host to some degree against atherosclerosis. PON1 is a key component of the HDL (high density lipoprotein) complex, and PON1 is able to inactivate toxic products resulting from the oxidation of lipid components of the blood low density lipoproteins. PON1 also protects against the toxicity of bacterial endotoxins, lipopolysaccharides (LPS). These endotoxins can produce fatal endotoxemia and shock in the course of Gram-negative infections. Recent studies with knockout mice lacking PON1 in the Lusis laboratory at UCLA have confirmed the importance of this enzyme in protection against organophosphate exposure and lipoprotein oxidation [3]. Our working hypothesis is that the PON family of enzymes are important endogenous protective proteins against several types of acute and chronic toxicities associated with cellular oxidative damage.
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
381
2. Protection against organophosphates The best known protective function of PON1 has been its ability to hydrolyze organophosphate nerve agents and insecticides. Unfortunately, the efficiencies of the mammalian paraoxonases are quite limited in both their affinities and their turnover numbers with these substrates. Thus, acute exposure to organophosphate
Fig. 1. Deduced amino acid sequences of PON proteins from several sources. Dashes represent residues identical to those present in human PON1 Q. The numbering of the residues start with the amino acid residue of the secreted mature protein. The locations of the two polymorphic sites in human PON1 are indicated with crosses at residues 55 (M/L) and 192 (R/Q). The asterisk indicates position 106 which is only present in mammalian PONs 1, but absent in PONs 2 and 3. The black dot indicates a polymorphic site in human PON2 which can be either cysteine or serine. Blank spaces represent gaps which were created to obtain the best alignment between the various PON sequences. Modified from Primo-Parmo et al. [2].
382
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
Fig. 2. Evolutionary relationship between the PONs based upon the number of amino acid differences predicted from the DNA analyses of the respective genes. Derived from data in Primo-Parmo et al. [2].
agents gives little opportunity for such enzymes to afford much protection. The paraoxonases are more effective against low-level, chronic exposure, where their catalytic (rather than stoichiometric) inactivation of the toxic agent gives them an advantage. If the oxime derivatives of organophosphates prove to also be substrates for the mammalian PONs, this would further extend their usefulness, and the combination of oximes and paraoxonase treatment for organophosphate exposure would delay aging and allow additional time and opportunity for enzymatic detoxification to take place. The history of the development of organophosphates as nerve gases and insecticides is relatively recent, primarily since the Second World War. However, the history of the organophosphates as natural products must be extremely old. For example, the organophosphate neurotoxins produced by the freshwater cyanobacterium (blue-green algae), Anabaena flos-aquae, such as anatoxin-a(s), are very potent anti-cholinesterases. The LD50 was estimated by Mahmood and Carmichael to be about 50 mg/kg in mice given the compound i.p. [4]. Whether such environmental toxic chemicals could have been responsible for selection for survival of organisms having PON-like enzymes is not known, but these very toxic organophosphates found in nature suggest that possibility. Dr Carmichael kindly tested our purified human type Q PON1 with the isolated anatoxin-a(s) and found it not to be hydrolyzed, but there remain many other potential substrates of this type that should be tested. Protection against organophosphates will depend not only on the level of the enzyme in blood and tissues but on the particular isozymes present as well. The B-type, now called the R-isozyme, is several times more efficient than the A-type (Q-isozyme) in hydrolyzing paraoxon, but most organophosphates are hydrolyzed appreciably better by the Q- than the R-isozyme (Fig. 3). Thus, both the level and the type of PON1 must be taken into consideration in evaluating the protective role of PON1 against such compounds that may be substrates catalytically inactivated by the enzyme.
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
383
3. Protection of low density lipoproteins from oxidative modification Oxidation of serum low density lipoproteins (LDL) is an important early step in the development of atherosclerosis. The oxidized products are scavenged by macrophages which eventually transform into foam cells, filled with cholesterol esters. They eventually become fatty streaks in the endothelium, and finally are seen as atheromatous plaques. Mackness [5] was the first to suggest that serum paraoxonase may be able to protect against the initial stage of this process, the oxidation of the LDL phospholipids, and his laboratory as well as several others have undertaken basic research in this important direction. Of course, the lack of protection of the HDL fraction obtained from the recently reported PON1 knockout mice, compared with HDL from the wild type mice [3] leaves little doubt that PON1 is the major component of HDL responsible for the previously demonstrated protective effects of HDL on LDL oxidation. Minimally oxidized LDL phospholipids adversely affect endothelial cells, unless this process is prevented by the
Fig. 3. Representation of the human PON1 structure illustrating the two polymorphic sites at positions 55 and 192, the internal disulfide bond between cysteine residues 42 and 353, the free cysteine at position 284, the location of potential sugar chains, and the hydrophobic retained leader sequence at the amino terminal end.
384
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
addition of either HDL or purified PON1 [6]. These authors proposed that this protection is due to the ability of PON1 to hydrolyze certain oxidized phospholipids generated by the oxidation of LDL phospholipids. Our laboratory has studied the protection by human serum PON1 against LDL phospholipid oxidation using either copper ions to catalyze the oxidation, or azobis-amidinopropane.HCl (AAPH), a free radial generator. Both simple in vitro systems [7] show human serum PON1 to protect against LDL lipoprotein oxidation. However, the requirements for this protection with the enzyme are not exactly the same as those required for hydrolysis of paraoxon or phenyl acetate [8]. For example, the free sulfhydryl residue at position 284 is not required for the hydrolysis reactions mentioned above [9], but it is essential for its protective actions against oxidized LDL [8]. The type R-isozyme is also less effective than the Q type isozyme in the protection systems [8]. Mackness et al. using a similar in vitro assay with HDL fractions representing the three phenotypes, also found this difference in the isozymes [10]. During the protection experiments against LDL oxidation by PON1, the enzyme is gradually inactivated to some degree, and this inactivation seems to be at least in part related to modifications of the essential 284 sulfhydryl group. Some reducing agents, such as flavonoids and quercetin, help to preserve the PON’s protective activity and prevent inactivation [11]. It is clear that the protective function of PON1 can not be predicted exactly or is it directly related to the hydrolytic activity of the enzyme with either paraoxon or phenyl acetate. Furthermore, even though the active centers for these two catalytic activities are very similar, they are not identical. Thus, the protective activity is not accurately predicted by PON1’s hydrolytic activity, as it may be oxidative or peroxidative, rather than hydrolytic in nature. Relating the hydrolytic and protective activities is also difficult because we still do not have convenient quantitative methods to measure the protective effects of PON1. Such an assay must take into account the gradual inactivation of the enzyme during incubation, the effect of stabilizing factors, the effects of metals, sulfhydryl reagents and reducing agents, etc. Dispute these limitations, it is clear that PON1 is effective at very low concentrations in protecting cells from oxidative damage from oxidized LDL lipoproteins, and this function of PON1 may represent an important physiological function of this protein.
4. Protection against toxicity from bacterial endotoxins For some years, the plasma high density lipoprotein complex (HDL) has been known to be protective against endotoxin toxicity [12]. The HDL complex appears to provide some degree of protection against the development of endotoxemia during Gram-negative bacterial infections. In some unknown way, the HDL can prevent the lipoprotein polysaccharide (LPS) derived from the surface of the bacteria from interacting with a specific binding protein (CD14) on the surface of certain macrophage cells that otherwise can initiate the release of a cascade of cytokines, including tumor necrosis factor-a (TNF-a), interleukin 1, and interleukin
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
385
Fig. 4. The survival of mice injected i.p. with 600 mg/animal of bacterial endotoxin (LPS), and 50 mg of purified human PON1 A (Type Q), i.p., either 2 h before the LPS (+), 2 h after LPS (%), or LPS plus vehicle alone (.)..
6. Collectively, these cytokines produce the various symptoms of septic shock that may lead to liver and kidney failure and, quite frequently, death. Evidence that a single protein within the HDL fraction of human serum can inactivate bacterial lipopolysaccharide (LPS) and thereby prevent endotoxemic symptoms was published over 20 years ago [13]. However, only recently has it been possible to obtain conclusive evidence that the protein component of HDL responsible for these protective effects is PON1. Partial purification and characterization of the LPS inactivator by the earlier workers [13] enabled them to demonstrate that the LPS inactivation process was clearly enzymatic, rather than being an immunological inactivation, and that the enzyme involved was not a serine esterase. In collaboration with Dr Standiford we have found that the i.p. injection of purified human PON1 (type Q) 2 h before an intraperitoneal injection of 600 mg of LPS to adult mice resulted in the survival of 60% of the animals. In contrast, giving the PON1 about 2 h after the LPS reduced the survival rate to about 30%. Without the PON1 injection all the mice died from the LPS treatment (Fig. 4). These and other experiments have convinced us that PON1 is able to protect cells from LPS and prevent, or greatly reduce the release of cytokines. It will be important to find out whether the level and type of PON1 make a difference in the resistance or sensitivity of different individuals to endotoxins, and how these characteristics of PON1 are correlated with the sensitivity and outcome following the endotoxin-producing bacterial infections.
386
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
5. Other functions of the paraoxonases Dr Martin Chalfie in the Department of Biology at Columbia University has determined that there are two or three, possibly more, PON-like genes in C. elegans. One of these was also independently sequenced in our laboratory (Fig. 1). According to Dr Chalfie (personal communication), one of the PONs is clearly involved in the complicated response to gentle touch. These interesting findings raise even wider possibilities in respect to possible physiological functions of some members of the PON gene family [14]. However, it should be mentioned that squidulin, the calcium-binding protein found in the squid optic nerve and giant axon, is structurally related to calmodulin, and is definitely not a member of the PON family [15]. Drugs other than organophosphates may also be substrates of the PON enzymes although there is little published about such possibilities, to date. A recent report shows PON1 to be the important enzyme for activating the antibiotic prodrug prulifloxacin to the active antibiotic [16]. There have also been a few verbal reports suggesting that some other drugs require PON1 for either their activation from a prodrug form or their inactivation. Presumably, more of these reports will be published soon.
6. Common features in these protective roles The protection of endothelial and other cells against oxidative damage due to oxidized LDL lipoproteins, and the in vivo protection by PON1 against endotoxic damage by LPS suggest that PON1 has a stabilizing property for cellular membranes that undergo either acute or a chronic exposure to oxidative agents and free radicals. Both of the latter challenge the ability of the membranes to maintain their selective permeability functions and osmotic integrity. Both types of protection need to have PON1 present at an early stage of the exposure; adding PON1 later, after the damage has taken place, is generally not effective. Protection, rather than reversal of the damage, may be a serious limitation for any therapeutic possibilities with PON1. Neither of the above types of protection seems to be due to a hydrolytic function of the enzyme. Although catalytic, there is some gradual loss of enzymatic activity with incubation, as if the enzyme has been damaged during the protection process. These features of the protective process with PON1 must be studied in further detail to learn the mechanism of the protective function, and to find whether PON1 can be stabilized by other agents.
7. Structural requirements for these activities It is important to find out whether these various protective functions of PON1 make use of the same structural components of the enzyme protein, and whether
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
387
the different protective functions can be predicted quantitatively by the use of simple in vitro tests, such as the rate PON1 hydrolyzes paraoxon, phenyl acetate, or some other model substrate. We doubt that all of these functions are hydrolytic, or that they depend upon the same functional groups of the enzyme. The replacement of calcium by zinc and other metals may also indicate how specific the metal requirements are for these different protective functions. Fortunately, we have available not only the two common human PON1 isozymes, and the PON1 homologues from the serum of other mammalian species, but also a large number of expressed mutant forms of the human PON1s to explore the important structureactivity relationships. For example, Dr Robert Sorenson from our laboratory is presenting a report at this Third International Meeting on Esterases Reacting with Organophosphorus Compounds [17] (this issue) on a mutant of PON1 in which he substituted two alanines at positions 20 and 21 of the hydrophobic leader sequence which allows the leader sequence to be cleaved. Surprisingly, this mutant enzyme does not localize in the HDL fraction when added to normal serum, as occurs when purified wild type human serum PON1 is added to serum. It is clear that the retained hydrophobic leader sequence of PON1 is very critical in explaining why the wild type PON1 is found almost exclusively in the HDL fraction. In 1989, Mackness raised important questions about what the physiological role of the A-esterases might be [5]. Now, about 10 years later, we can say that we have started to uncover some very exciting clues, but a proper answer to his question is just starting to emerge. References [1] K.B. Augustinsson, The evolution of esterases, in: N. van Thoai, J. Roche (Eds.), Homologous Enzymes and Biochemical Evolution, Gordon and Breach, New York, 1968, pp. 299 – 311. [2] S.L. Primo-Parmo, R.C. Sorenson, J. Teiber, B.N. La Du, The human serum paraoxonase/ arylesterase gene (PON1) is one member of a multigene family, Genomics 33 (1996) 498 – 507. [3] D.M. Shih, L. Gu, Y.-R. Xia, M. Navab, W.-F. Li, S. Hama, L.W. Castellani, C.E. Furlong, L.G. Costa, A.M. Fogelman, A.J. Lusis, Mice lacking serum paraoxonase are susceptible to organophosphate toxicity and atherosclerosis, Nature 394 (1998) 284 – 287. [4] N.A. Mahmood, W.W. Carmichael, The pharmacology of anatoxin-a(s), neurotoxin produced by the freshwater cyanobacterium Anabaena flos-aquae NRC 525 – 17, Toxicon. 24 (1986) 425 – 434. [5] M.I. Mackness, Commentary A-esterases: enzymes looking for a role?, Biochem. Pharmacol. 38 (1989) 385–390. [6] A.D. Watson, J.A. Berliner, S.Y. Hama, B.N. La Du, K.F. Faull, A.M. Fogelman, M. Navab, Protective effect of HDL associated paraoxonase-inhibition of the biological activity of minimally oxidized low density lipoprotein, J. Clin. Invest. 96 (1995) 2882 – 2891. [7] M. Aviram, M. Rosenblat, C.L. Bisgaier, R.S. Newton, S.L. Primo-Parmo, B.N. La Du, Paraoxonase inhibits high-density lipoprotein oxidation and preserves its functions. A possible peroxidative role for paraoxonase, J. Clin. Invest. 101 (1998) 1581 – 1590. [8] M. Aviram, S. Billecke, R. Sorenson, C. Bisgaier, R. Newton, M. Rosenblat, J. Erogul, C. Hsu, C. Dunlop, B. La Du, Paraoxonase active site required against LDL oxidation involves its free sulfhydryl group and is different than that irequired for its arylesterase/paraoxonase activities: selective action for human paraoxonase allozymes Q and R, Arterioscler. Thromb. Vasc. Biol. 18 (1998) 1617–1624.
388
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
[9] R.C. Sorenson, S.L. Primo-Parmo, C-L. Kuo, S. Adkins, O. Lockridge, B.N. La Du, Reconsideration of the catalytic center and mechanism of mammalian paraoxonase/arylesterase, Proc. Natl. Acad. Sci. USA 92 (1995) 7187–7191. [10] M.I. Mackness, S. Arrol, B. Mackness, P.N. Durrington, Alloenzymes of paraoxonase and effectiveness of high-density lipoproteins in protecting low-density lipoprotein against lipid peroxidation, Lancet 349 (1997) 851–852. [11] M. Aviram, M. Rosenblat, S. Billecke, J. Erogul, R. Sorenson, C.L.Bisgaier, R. Newton, B. La Du, Human serum paraoxonase (PON1) inactivation by oxidized low density lipoprotein: Role for PON’s free sulfhydryl group and effect of antioxidants. Free Radic. Biol. Med. in press. [12] D.M. Levine, T.S. Parker, T.M. Donnelly, A. Walsh, A.L. Rubin, In vivo protection against endotoxin by plasma high density lipoprotein, Proc. Nat. Acad. Sci. USA 90 (1993) 12040 – 12044. [13] K.J. Johnson, P.A. Ward, S. Gorainick, M.J. Osborn, Isolation from human serum of an inactivator of bacterial lipopolysaccharide, Am. J. Pathol. 88 (1977) 559 – 574. [14] G. Gu, G.A. Caldwell, M. Chalfie, Genetic interactions affecting touch sensitivity in Caenorhabditis elegans, Proc. Nat. Acad. Sci. USA 93 (1996) 6577 – 6582. [15] J.F. Head, S. Spielberg, B. Kaminer, Two low-molecular-weight Ca2-binding proteins isolated from squid optic lobe by phenothiazine-sepharose affinity chromatography, Biochem. J. 209 (1983) 797–802. [16] K. Tougou, A. Nakamura, S. Watanabe, Y. Okuyama, A. Morino, Paraoxonase has a major role in the hydrolysis of prulifloxacin (NM441), a prodrug of a new antibacterial agent, Drug Metab. Dispo. 26 (1998) 355–359. [17] R.C. Sorenson, M. Aviram, C.L. Bisgaier, S. Billecke, C. Hsu, B.N. La Du, Properties of the retained N-terminal hydrophobic leader sequence in human serum paraoxonase/arylesterase. Chem.-Biol. Interact. 119–120 (1999) 243 – 249.
.
On the physiological role(s) of the paraoxonases Bert N. La Du a,*, Michael Aviram b, Scott Billecke a, Mohamad Navab c, Sergio Primo-Parmo a, Robert C. Sorenson a, Theodore J. Standiford d a
Departments of Anesthesiology and Pharmacology, Uni6ersity of Michigan Medical School, Ann Arbor, MI 48109, USA b Lipid Research Laboratory, Rambam Medical Center and the Bruce Rappaport Faculty of Medicine, Technion and the Rappaport Family Institute for Research in the Medical Sciences, Haifa 31096, Israel c Di6ision of Cardiology, Department of Medicine, School of Medicine, Uni6ersity of California, Los Angeles, CA 90095, USA d Department of Medicine, Uni6ersity of Michigan Medical School, Ann Arbor, MI 48109, USA
Abstract In recent years several lines of evidence have indicated that serum paraoxonase (PON1), and perhaps other mammalian paraoxonases, act as important guardians against cellular damage from toxic agents, such as organophosphates, oxidized lipids in the plasma low density lipoproteins (LDL), and against bacterial endotoxins. For some of these protective activities but not all, PON1 requires calcium ion. The catalyzed chemical reactions generally seem to be hydrolytic, but for some types of protection this may not be so. Several other metals have very high affinity for PON1 and may displace calcium. Replacement or substitution of calcium by other metals could extend the range of catalytic properties and the substrate specificity of the paraoxonases, as it does for the mammalian DFPases. Although this Third International Meeting on Esterases Reacting with Organophosphorus Compounds focuses on the organophosphatase activities of paraoxonase and related enzymes, it is important to also briefly review some of the current directions in several laboratories searching for additional functions of the paraoxonases to extend our understanding of the properties of this family of enzymes which now seem to have both physiological and toxicological importance. © 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Bacterial endotoxin; HDL; LDL; Paraoxonases; PON1 mutants; PON evolution; PON polymorphisms
* Corresponding author. Tel.: +1-734-7636429; fax: +1-734-7649332. E-mail address: [email protected] (B.N. La Du) 0009-2797/99/$ - see front matter © 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 0 0 9 - 2 7 9 7 ( 9 9 ) 0 0 0 4 9 - 6
380
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
1. Introduction: search for natural substrates for the paraoxonases Several lines of evidence in recent years have suggested that serum paraoxonase/arylesterase (PON1) and closely related enzymes may have physiological functions as well as the ability to hydrolyze toxic organophosphate compounds. The first line of evidence is recent genetic information. There has been a long standing assumption about the origin of paraoxonase family; it was presumed that the paraoxonases arose as an offshoot from the serine esterases, the carboxyesterases [1]. Structural analyses of the respective genes and proteins; however, have shown no homology between the arylesterases and the carboxyesterases, even though they share a few substrates, such as phenyl acetate. There is no structural homology supporting the claim that they share some common evolutionary heritage or were derived from a common ancestral esterase. The PON family in mammals (mice and humans, at least) have three distinct PON genes, located very close to each other on the same chromosome (chromosome 7 in man, chromosome 6 in the mouse) [2]. Presumably gene duplication has occurred during evolutionary development to cause this redundancy, and it suggests that this family of enzymes may contribute to some important physiological functions, quite independently from the unusual ability of PON1 to hydrolyze toxic organophosphates, carbamates and aromatic esters. The limited nucleotide sequence information we have, to date, has been translated into the predicted amino acid sequences of PON proteins from C. elegans to human beings and summarized in Fig. 1. A logical evolutionary pattern of the PON family of proteins can be constructed which brings out distinctive characteristics that make it a completely different family than any other. It is unrelated to the cholinesterases, the bacterial phosphotriesterases or the mammalian DFPases sequenced until now (Fig. 2). PON2 appears to be an older family member than PON1, and the latter seems to be a relatively recent arrival, compared to PONs 2 and 3. Secondly, several lines of evidence are emerging that PON1 (at least) protects the mammalian host to some degree against atherosclerosis. PON1 is a key component of the HDL (high density lipoprotein) complex, and PON1 is able to inactivate toxic products resulting from the oxidation of lipid components of the blood low density lipoproteins. PON1 also protects against the toxicity of bacterial endotoxins, lipopolysaccharides (LPS). These endotoxins can produce fatal endotoxemia and shock in the course of Gram-negative infections. Recent studies with knockout mice lacking PON1 in the Lusis laboratory at UCLA have confirmed the importance of this enzyme in protection against organophosphate exposure and lipoprotein oxidation [3]. Our working hypothesis is that the PON family of enzymes are important endogenous protective proteins against several types of acute and chronic toxicities associated with cellular oxidative damage.
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
381
2. Protection against organophosphates The best known protective function of PON1 has been its ability to hydrolyze organophosphate nerve agents and insecticides. Unfortunately, the efficiencies of the mammalian paraoxonases are quite limited in both their affinities and their turnover numbers with these substrates. Thus, acute exposure to organophosphate
Fig. 1. Deduced amino acid sequences of PON proteins from several sources. Dashes represent residues identical to those present in human PON1 Q. The numbering of the residues start with the amino acid residue of the secreted mature protein. The locations of the two polymorphic sites in human PON1 are indicated with crosses at residues 55 (M/L) and 192 (R/Q). The asterisk indicates position 106 which is only present in mammalian PONs 1, but absent in PONs 2 and 3. The black dot indicates a polymorphic site in human PON2 which can be either cysteine or serine. Blank spaces represent gaps which were created to obtain the best alignment between the various PON sequences. Modified from Primo-Parmo et al. [2].
382
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
Fig. 2. Evolutionary relationship between the PONs based upon the number of amino acid differences predicted from the DNA analyses of the respective genes. Derived from data in Primo-Parmo et al. [2].
agents gives little opportunity for such enzymes to afford much protection. The paraoxonases are more effective against low-level, chronic exposure, where their catalytic (rather than stoichiometric) inactivation of the toxic agent gives them an advantage. If the oxime derivatives of organophosphates prove to also be substrates for the mammalian PONs, this would further extend their usefulness, and the combination of oximes and paraoxonase treatment for organophosphate exposure would delay aging and allow additional time and opportunity for enzymatic detoxification to take place. The history of the development of organophosphates as nerve gases and insecticides is relatively recent, primarily since the Second World War. However, the history of the organophosphates as natural products must be extremely old. For example, the organophosphate neurotoxins produced by the freshwater cyanobacterium (blue-green algae), Anabaena flos-aquae, such as anatoxin-a(s), are very potent anti-cholinesterases. The LD50 was estimated by Mahmood and Carmichael to be about 50 mg/kg in mice given the compound i.p. [4]. Whether such environmental toxic chemicals could have been responsible for selection for survival of organisms having PON-like enzymes is not known, but these very toxic organophosphates found in nature suggest that possibility. Dr Carmichael kindly tested our purified human type Q PON1 with the isolated anatoxin-a(s) and found it not to be hydrolyzed, but there remain many other potential substrates of this type that should be tested. Protection against organophosphates will depend not only on the level of the enzyme in blood and tissues but on the particular isozymes present as well. The B-type, now called the R-isozyme, is several times more efficient than the A-type (Q-isozyme) in hydrolyzing paraoxon, but most organophosphates are hydrolyzed appreciably better by the Q- than the R-isozyme (Fig. 3). Thus, both the level and the type of PON1 must be taken into consideration in evaluating the protective role of PON1 against such compounds that may be substrates catalytically inactivated by the enzyme.
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
383
3. Protection of low density lipoproteins from oxidative modification Oxidation of serum low density lipoproteins (LDL) is an important early step in the development of atherosclerosis. The oxidized products are scavenged by macrophages which eventually transform into foam cells, filled with cholesterol esters. They eventually become fatty streaks in the endothelium, and finally are seen as atheromatous plaques. Mackness [5] was the first to suggest that serum paraoxonase may be able to protect against the initial stage of this process, the oxidation of the LDL phospholipids, and his laboratory as well as several others have undertaken basic research in this important direction. Of course, the lack of protection of the HDL fraction obtained from the recently reported PON1 knockout mice, compared with HDL from the wild type mice [3] leaves little doubt that PON1 is the major component of HDL responsible for the previously demonstrated protective effects of HDL on LDL oxidation. Minimally oxidized LDL phospholipids adversely affect endothelial cells, unless this process is prevented by the
Fig. 3. Representation of the human PON1 structure illustrating the two polymorphic sites at positions 55 and 192, the internal disulfide bond between cysteine residues 42 and 353, the free cysteine at position 284, the location of potential sugar chains, and the hydrophobic retained leader sequence at the amino terminal end.
384
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
addition of either HDL or purified PON1 [6]. These authors proposed that this protection is due to the ability of PON1 to hydrolyze certain oxidized phospholipids generated by the oxidation of LDL phospholipids. Our laboratory has studied the protection by human serum PON1 against LDL phospholipid oxidation using either copper ions to catalyze the oxidation, or azobis-amidinopropane.HCl (AAPH), a free radial generator. Both simple in vitro systems [7] show human serum PON1 to protect against LDL lipoprotein oxidation. However, the requirements for this protection with the enzyme are not exactly the same as those required for hydrolysis of paraoxon or phenyl acetate [8]. For example, the free sulfhydryl residue at position 284 is not required for the hydrolysis reactions mentioned above [9], but it is essential for its protective actions against oxidized LDL [8]. The type R-isozyme is also less effective than the Q type isozyme in the protection systems [8]. Mackness et al. using a similar in vitro assay with HDL fractions representing the three phenotypes, also found this difference in the isozymes [10]. During the protection experiments against LDL oxidation by PON1, the enzyme is gradually inactivated to some degree, and this inactivation seems to be at least in part related to modifications of the essential 284 sulfhydryl group. Some reducing agents, such as flavonoids and quercetin, help to preserve the PON’s protective activity and prevent inactivation [11]. It is clear that the protective function of PON1 can not be predicted exactly or is it directly related to the hydrolytic activity of the enzyme with either paraoxon or phenyl acetate. Furthermore, even though the active centers for these two catalytic activities are very similar, they are not identical. Thus, the protective activity is not accurately predicted by PON1’s hydrolytic activity, as it may be oxidative or peroxidative, rather than hydrolytic in nature. Relating the hydrolytic and protective activities is also difficult because we still do not have convenient quantitative methods to measure the protective effects of PON1. Such an assay must take into account the gradual inactivation of the enzyme during incubation, the effect of stabilizing factors, the effects of metals, sulfhydryl reagents and reducing agents, etc. Dispute these limitations, it is clear that PON1 is effective at very low concentrations in protecting cells from oxidative damage from oxidized LDL lipoproteins, and this function of PON1 may represent an important physiological function of this protein.
4. Protection against toxicity from bacterial endotoxins For some years, the plasma high density lipoprotein complex (HDL) has been known to be protective against endotoxin toxicity [12]. The HDL complex appears to provide some degree of protection against the development of endotoxemia during Gram-negative bacterial infections. In some unknown way, the HDL can prevent the lipoprotein polysaccharide (LPS) derived from the surface of the bacteria from interacting with a specific binding protein (CD14) on the surface of certain macrophage cells that otherwise can initiate the release of a cascade of cytokines, including tumor necrosis factor-a (TNF-a), interleukin 1, and interleukin
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
385
Fig. 4. The survival of mice injected i.p. with 600 mg/animal of bacterial endotoxin (LPS), and 50 mg of purified human PON1 A (Type Q), i.p., either 2 h before the LPS (+), 2 h after LPS (%), or LPS plus vehicle alone (.)..
6. Collectively, these cytokines produce the various symptoms of septic shock that may lead to liver and kidney failure and, quite frequently, death. Evidence that a single protein within the HDL fraction of human serum can inactivate bacterial lipopolysaccharide (LPS) and thereby prevent endotoxemic symptoms was published over 20 years ago [13]. However, only recently has it been possible to obtain conclusive evidence that the protein component of HDL responsible for these protective effects is PON1. Partial purification and characterization of the LPS inactivator by the earlier workers [13] enabled them to demonstrate that the LPS inactivation process was clearly enzymatic, rather than being an immunological inactivation, and that the enzyme involved was not a serine esterase. In collaboration with Dr Standiford we have found that the i.p. injection of purified human PON1 (type Q) 2 h before an intraperitoneal injection of 600 mg of LPS to adult mice resulted in the survival of 60% of the animals. In contrast, giving the PON1 about 2 h after the LPS reduced the survival rate to about 30%. Without the PON1 injection all the mice died from the LPS treatment (Fig. 4). These and other experiments have convinced us that PON1 is able to protect cells from LPS and prevent, or greatly reduce the release of cytokines. It will be important to find out whether the level and type of PON1 make a difference in the resistance or sensitivity of different individuals to endotoxins, and how these characteristics of PON1 are correlated with the sensitivity and outcome following the endotoxin-producing bacterial infections.
386
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
5. Other functions of the paraoxonases Dr Martin Chalfie in the Department of Biology at Columbia University has determined that there are two or three, possibly more, PON-like genes in C. elegans. One of these was also independently sequenced in our laboratory (Fig. 1). According to Dr Chalfie (personal communication), one of the PONs is clearly involved in the complicated response to gentle touch. These interesting findings raise even wider possibilities in respect to possible physiological functions of some members of the PON gene family [14]. However, it should be mentioned that squidulin, the calcium-binding protein found in the squid optic nerve and giant axon, is structurally related to calmodulin, and is definitely not a member of the PON family [15]. Drugs other than organophosphates may also be substrates of the PON enzymes although there is little published about such possibilities, to date. A recent report shows PON1 to be the important enzyme for activating the antibiotic prodrug prulifloxacin to the active antibiotic [16]. There have also been a few verbal reports suggesting that some other drugs require PON1 for either their activation from a prodrug form or their inactivation. Presumably, more of these reports will be published soon.
6. Common features in these protective roles The protection of endothelial and other cells against oxidative damage due to oxidized LDL lipoproteins, and the in vivo protection by PON1 against endotoxic damage by LPS suggest that PON1 has a stabilizing property for cellular membranes that undergo either acute or a chronic exposure to oxidative agents and free radicals. Both of the latter challenge the ability of the membranes to maintain their selective permeability functions and osmotic integrity. Both types of protection need to have PON1 present at an early stage of the exposure; adding PON1 later, after the damage has taken place, is generally not effective. Protection, rather than reversal of the damage, may be a serious limitation for any therapeutic possibilities with PON1. Neither of the above types of protection seems to be due to a hydrolytic function of the enzyme. Although catalytic, there is some gradual loss of enzymatic activity with incubation, as if the enzyme has been damaged during the protection process. These features of the protective process with PON1 must be studied in further detail to learn the mechanism of the protective function, and to find whether PON1 can be stabilized by other agents.
7. Structural requirements for these activities It is important to find out whether these various protective functions of PON1 make use of the same structural components of the enzyme protein, and whether
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
387
the different protective functions can be predicted quantitatively by the use of simple in vitro tests, such as the rate PON1 hydrolyzes paraoxon, phenyl acetate, or some other model substrate. We doubt that all of these functions are hydrolytic, or that they depend upon the same functional groups of the enzyme. The replacement of calcium by zinc and other metals may also indicate how specific the metal requirements are for these different protective functions. Fortunately, we have available not only the two common human PON1 isozymes, and the PON1 homologues from the serum of other mammalian species, but also a large number of expressed mutant forms of the human PON1s to explore the important structureactivity relationships. For example, Dr Robert Sorenson from our laboratory is presenting a report at this Third International Meeting on Esterases Reacting with Organophosphorus Compounds [17] (this issue) on a mutant of PON1 in which he substituted two alanines at positions 20 and 21 of the hydrophobic leader sequence which allows the leader sequence to be cleaved. Surprisingly, this mutant enzyme does not localize in the HDL fraction when added to normal serum, as occurs when purified wild type human serum PON1 is added to serum. It is clear that the retained hydrophobic leader sequence of PON1 is very critical in explaining why the wild type PON1 is found almost exclusively in the HDL fraction. In 1989, Mackness raised important questions about what the physiological role of the A-esterases might be [5]. Now, about 10 years later, we can say that we have started to uncover some very exciting clues, but a proper answer to his question is just starting to emerge. References [1] K.B. Augustinsson, The evolution of esterases, in: N. van Thoai, J. Roche (Eds.), Homologous Enzymes and Biochemical Evolution, Gordon and Breach, New York, 1968, pp. 299 – 311. [2] S.L. Primo-Parmo, R.C. Sorenson, J. Teiber, B.N. La Du, The human serum paraoxonase/ arylesterase gene (PON1) is one member of a multigene family, Genomics 33 (1996) 498 – 507. [3] D.M. Shih, L. Gu, Y.-R. Xia, M. Navab, W.-F. Li, S. Hama, L.W. Castellani, C.E. Furlong, L.G. Costa, A.M. Fogelman, A.J. Lusis, Mice lacking serum paraoxonase are susceptible to organophosphate toxicity and atherosclerosis, Nature 394 (1998) 284 – 287. [4] N.A. Mahmood, W.W. Carmichael, The pharmacology of anatoxin-a(s), neurotoxin produced by the freshwater cyanobacterium Anabaena flos-aquae NRC 525 – 17, Toxicon. 24 (1986) 425 – 434. [5] M.I. Mackness, Commentary A-esterases: enzymes looking for a role?, Biochem. Pharmacol. 38 (1989) 385–390. [6] A.D. Watson, J.A. Berliner, S.Y. Hama, B.N. La Du, K.F. Faull, A.M. Fogelman, M. Navab, Protective effect of HDL associated paraoxonase-inhibition of the biological activity of minimally oxidized low density lipoprotein, J. Clin. Invest. 96 (1995) 2882 – 2891. [7] M. Aviram, M. Rosenblat, C.L. Bisgaier, R.S. Newton, S.L. Primo-Parmo, B.N. La Du, Paraoxonase inhibits high-density lipoprotein oxidation and preserves its functions. A possible peroxidative role for paraoxonase, J. Clin. Invest. 101 (1998) 1581 – 1590. [8] M. Aviram, S. Billecke, R. Sorenson, C. Bisgaier, R. Newton, M. Rosenblat, J. Erogul, C. Hsu, C. Dunlop, B. La Du, Paraoxonase active site required against LDL oxidation involves its free sulfhydryl group and is different than that irequired for its arylesterase/paraoxonase activities: selective action for human paraoxonase allozymes Q and R, Arterioscler. Thromb. Vasc. Biol. 18 (1998) 1617–1624.
388
B.N. La Du et al. / Chemico-Biological Interactions 119–120 (1999) 379–388
[9] R.C. Sorenson, S.L. Primo-Parmo, C-L. Kuo, S. Adkins, O. Lockridge, B.N. La Du, Reconsideration of the catalytic center and mechanism of mammalian paraoxonase/arylesterase, Proc. Natl. Acad. Sci. USA 92 (1995) 7187–7191. [10] M.I. Mackness, S. Arrol, B. Mackness, P.N. Durrington, Alloenzymes of paraoxonase and effectiveness of high-density lipoproteins in protecting low-density lipoprotein against lipid peroxidation, Lancet 349 (1997) 851–852. [11] M. Aviram, M. Rosenblat, S. Billecke, J. Erogul, R. Sorenson, C.L.Bisgaier, R. Newton, B. La Du, Human serum paraoxonase (PON1) inactivation by oxidized low density lipoprotein: Role for PON’s free sulfhydryl group and effect of antioxidants. Free Radic. Biol. Med. in press. [12] D.M. Levine, T.S. Parker, T.M. Donnelly, A. Walsh, A.L. Rubin, In vivo protection against endotoxin by plasma high density lipoprotein, Proc. Nat. Acad. Sci. USA 90 (1993) 12040 – 12044. [13] K.J. Johnson, P.A. Ward, S. Gorainick, M.J. Osborn, Isolation from human serum of an inactivator of bacterial lipopolysaccharide, Am. J. Pathol. 88 (1977) 559 – 574. [14] G. Gu, G.A. Caldwell, M. Chalfie, Genetic interactions affecting touch sensitivity in Caenorhabditis elegans, Proc. Nat. Acad. Sci. USA 93 (1996) 6577 – 6582. [15] J.F. Head, S. Spielberg, B. Kaminer, Two low-molecular-weight Ca2-binding proteins isolated from squid optic lobe by phenothiazine-sepharose affinity chromatography, Biochem. J. 209 (1983) 797–802. [16] K. Tougou, A. Nakamura, S. Watanabe, Y. Okuyama, A. Morino, Paraoxonase has a major role in the hydrolysis of prulifloxacin (NM441), a prodrug of a new antibacterial agent, Drug Metab. Dispo. 26 (1998) 355–359. [17] R.C. Sorenson, M. Aviram, C.L. Bisgaier, S. Billecke, C. Hsu, B.N. La Du, Properties of the retained N-terminal hydrophobic leader sequence in human serum paraoxonase/arylesterase. Chem.-Biol. Interact. 119–120 (1999) 243 – 249.
.