Forum: oxidative stress status
Urinary aldehydes as indicators of lipid peroxidation in vivo

https://doi.org/10.1016/S0891-5849(00)00367-1Get rights and content

Abstract

The excretion of malondialdehyde (MDA), lipophilic aldehydes and related carbonyl compounds in rat and human urine was investigated. MDA was found to be excreted mainly in the form of two adducts with lysine, indicating that its predominant reaction in vivo is with the lysine residues of proteins. Adducts with the phospholipid bases serine and ethanolamine and the nucleic acid bases guanine and deoxyguanosine also were found. Except for the adduct with deoxyguanosine (dG-MDA), the excretion of these compounds increased with peroxidative stress imposed in the form of vitamin E deficiency or the administration of iron or carbon tetrachloride. Marked differences in the concentration of dG-MDA in different tissues were correlated with their content of fatty acids having three or more double bonds, the putative source of MDA. Fourteen nonpolar and eleven polar lipophilic aldehydes and other carbonyl compounds were identified as their 2,4-diphenylhydrazine derivatives in rat urine. The excretion of five nonpolar and nine polar compounds was increased under conditions of peroxidative stress. The profile of lipophilic aldehydes obtained for human urine resembled that for rat urine. Except for a reported 4-hydroxynon-2-enal conjugate with mercapturic acid, the conjugated forms of the lipophilic aldehydes excreted in urine remain unidentified. Aldehyde excretion is influenced by numerous factors that affect the formation of lipid peroxides in vivo such as energy status, physical activity and environmental temperature, as well as by wide variations in the intake of peroxides in the diet. Consequently, urinalysis for aldehydic products of lipid peroxidation is an unreliable indicator of the general state of peroxidative stress in vivo.

Introduction

In 1991 [1] it was stated that “one of the greatest needs in the field of free radical biology is the development of reliable methods for measuring oxidative stress status (OSS) in humans.” A decade later such methods have yet to appear. Also, the clinical value that any such measurement would have in the absence of information on the oxidizing species involved and the metabolic site and nature of its reactions with cell constituents is problematic. Most proposed methods have been limited to the measurement of lipid peroxidative stress in animals using analysis of blood, urine, or solid tissues. Methane and pentane exhalation respond to increased in vivo lipid peroxidation in human subjects fed intravenously [2], but the method of measurement requires specialized equipment and is subject to analytical losses as well as to interference by intestinal bacteria.

Urinalysis for products of in vivo lipid peroxidation is conceptually the most practical means of acquiring information on the reactions of lipid free radicals with cellular compounds in human subjects. Animals fed peroxide-free diets have been shown to excrete increased amounts of malondialdehyde (MDA) [3] and lipophilic aldehydes [4] after vitamin E depletion or the administration of chemical oxidants but such conditions have little relevance to humans. Urinalysis has yielded valuable information on the products of lipid peroxidative stress in both animals and humans, but it does not include stress exerted by radicals of molecular oxygen except as they may initiate lipid peroxidation. This paper contains an account of the results of this urinalysis.

The effect of oxidative stress in the form of iron nitrilotriacetate or carbon tetrachloride administration and of vitamin E deficiency on the excretion of MDA adducts by rats fed an MDA-free diet is shown in Fig. 1 [5]. Total MDA and the proportion excreted in the free form increased with the severity of oxidative stress. MDA is excreted mainly in the form of adducts with lysine and its N-acetylated derivative, indicating that its predominant reaction in vivo is with the ϵ-amino groups of the lysine residues of proteins. Smaller amounts of adducts with the phospholipid bases serine and ethanolamine and the nucleic acid bases guanine and deoxyguanosine reflect reactions with these compounds. Two additional minor metabolites have been detected and one has been tentatively identified as an adduct with taurine. Analogous MDA adducts were found in the urine of humans consuming a normal diet (Fig. 2) [5].

14C-bovine serum albumin that has been exposed to MDA in aqueous solution is subject to an increased rate of proteolysis in vitro by a macroprotease isolated from human erythrocytes by the procedure of Davies and Goldberg [6] (Fig. 3) and a protease isolated from rat liver mitochondria by the procedure of Marcillat et al. [7]. The magnitude of these increases is similar to that observed for proteins exposed to hydroxyl radicals [6], [7], indicating that these enzymes hydrolyze proteins damaged not only directly by oxygen radicals but by MDA formed as a result of their reaction with polyunsaturated lipids.

Excretion of the lipophilic products of in vivo lipid peroxidation by fasting rats is illustrated in Fig. 4, Fig. 5 [8]. The sensitivity of the method of measurement was ≤ 0.5 ng. Fourteen nonpolar and 11 polar lipophilic aldehydes and related carbonyl compounds were isolated as their 2,4-dinitrophenylhydrazine derivatives by HPLC. Nine polar and three polar compounds were identified. The excretion of polar compounds was much greater than that of nonpolar compounds. Except for a conjugate of 4-hydroxynon-2-enal with mercapturic acid, the conjugated forms of the lipophilic products of in vivo lipid peroxidation excreted in urine have not been identified.

The effect of vitamin E deficiency induced in rats by feeding a diet high in polyunsaturated fatty acids from distilled corn oil and cod liver oil on the excretion of lipophilic products of in vivo lipid peroxidation is shown in Fig. 6, Fig. 7 [4]. The excretion of five polar and six nonpolar compounds was significantly increased (p ≤ .05) in vitamin E deficiency. There was no significant difference in their excretion by animals fed a normal concentration of vitamin E (NE) or ten times this concentration (HE). To the extent that excretion of the lipophilic products of lipid peroxidation in vivo reflects the peroxidative status of rats, it appears that a high intake of vitamin E has no advantage over a normal intake.

The profile of polar and nonpolar aldehydes found in the urine of humans consuming free-choice diets Fig. 8, Fig. 9 was similar to that found in rats, but their concentrations in human urine were much lower [8]. This and a similar finding with respect to MDA excretion are attributable to a species difference in basal metabolic rate [9]. The profile of MDA adducts in the urine of rats fed a stock diet of natural ingredients and that of rats fed a purified diet free of peroxidizable fat, though much lower, are both dominated by adducts with lysine, indicating a primary association of MDA with proteins in both the diet and the tissues. Fasting humans and rats excrete analogous adducts of MDA. The effect of fasting and a peroxide-free diet on excretion of lipophilic products of lipid peroxidation by humans has not been determined, nor has the identity of the conjugates.

Rats fed so-called “purified diets” excrete markedly higher amounts of MDA and lipophilic metabolites than fasting animals [4], [10]. Vegetable oils used as a source of essential polyunsaturated fatty acids in these diets and inorganic salts used as a source of essential trace minerals provide a fertile milieu for lipid peroxidation. Short-term replacement of corn oil with hydrogenated coconut oil results in a marked decrease in MDA excretion, whereas prolonged fasting results in an increase [10]. The increase in the level of free fatty acids in blood serum induced by administration of the lipolytic hormones epinephrine and adrenocorticotropic hormone [10] and the lipolysis induced by strenuous exercise [11] are accompanied by an increase in MDA excretion. It is possible that the increase in MDA excretion caused by exercise is due to enhanced initiation of lipid peroxidation by hydroxyl radicals associated with a higher rate of oxygen inhalation, but this does not explain the MDA response to the administration of lipolytic hormones. Adipose tissue is low in PUFA, but these acids may be more susceptible to peroxidation when mobilized in the free state. It is also possible that the increase in MDA excretion is simply due to mobilization of preformed fatty acid peroxides. In any event, normal fluctuations in lipolysis undermine the usefulness of urinalysis for products of lipid peroxidation as a method of assessing the steady state peroxidative stress status of human subjects.

Identification of an MDA adduct with deoxyguanosine (dG-MDA) in human and rat urine [9] suggested the possibility that the excretion of this compound might serve as a specific marker for in vivo lipid peroxidation. However, in contrast to the increase in MDA excretion seen in vitamin E deficiency and following the administration of chemical oxidants, these instruments of oxidative stress had no effect on the excretion of dG-MDA by rats or its prevalence in liver DNA [12]. Feeding oils containing a wide range of fatty acids (coconut, olive, linseed, safflower, evening primrose and salmon) produced no significant differences in the dG-MDA content of rat liver nuclei [13]. DNA appears to be shielded from lipid peroxidative damage, possibly by a nuclear barrier to oxidants or conservation of vitamin E in the nuclear membrane.

Despite this evidence that DNA is resistant to oxidative stress, there are major differences between rat tissues in the concentration of dG-MDA in DNA (62 pmol/100 μg in brain vs. 10 pmol in liver and 2 pmol in kidney) [12]. Furthermore, its frequency in DNA increases with age. The high concentration of dG-MDA in brain DNA suggests a relationship with the prevalence of highly unsaturated n-3 fatty acids in this organ. Adding 2% cod liver oil to a 10% distilled corn oil diet as a source of n-3 fatty acids resulted in a significant increase in the dG-MDA content of rat liver DNA [12]. Humans fed a diet high in PUFA as opposed to one high in monounsaturated fatty acids have been found to have a higher concentration of dG-MDA in their white blood cells [14]. The factors responsible for the differences between tissues in the concentration of dG-MDA in DNA and for the increase in its prevalence with age are prime objectives for future research into the effect of lipid peroxidative stress on human health, with particular reference to afflictions such as Alzheimer’s disease.

The mutagenicity of free MDA [15] raised the question as to whether its presence in the diet presents a risk to human health. This concern has been mitigated by the finding that digestion of foods in vitro with proteolytic enzymes releases MDA mainly in the form of an adduct with lysine [16]. This compound, which is nonmutagenic, arises from a reaction between MDA released in the peroxidation of food lipids and the ϵ-amino groups of the lysine residues of food proteins, from which it is released in the course of protein digestion. Under the acidic conditions of the thiobarbituric acid method for the determination of MDA, its bound forms are hydrolyzed to give a spuriously high value for the concentration of free MDA [16].

Section snippets

Discussion

Halliwell [18] among others has discussed the difficulties involved in distinguishing between lipid peroxidation as a cause and as a result of disease. The list of clinical ailments reportedly associated with an increased concentration of MDA in the blood, urine, or tissues covers most of the major metabolic diseases (atherosclerosis, diabetes, congestive heart failure, cancer, arthritis, cold injury, alcoholic liver disease). This list could be extended to include perinatal hypoxia;

Acknowledgements

This research was supported by the Natural Sciences and Engineering Research Council of Canada, the Minnesota Agricultural Experiment Station, the U.S Agency for International Development Contract No. 608-0160, and the North Dakota Agricultural Experiment Station.

References (22)

  • S. Agarwal et al.

    Identification of a deoxyguanosine-malondialdehyde adduct in rat and human urine

    Lipids

    (1994)
  • Cited by (0)

    1

    Harold H. Draper received a Ph.D. in nutritional biochemistry from the University of Illinois in 1952 and subsequently pursued research on the function and metabolism of vitamin E at this institution. In 1975, he joined the Department of Nutritional Sciences at the University of Guelph, Ontario, where he conducted research on the metabolism and toxicity of the hydrophilic products of lipid peroxidation with an emphasis on malondialdehyde.

    2

    A. Saari Csallany was awarded an Sc.D. in food chemistry by the Technical University of Budapest for work conducted at the University of Illinois on the metabolism of vitamin E. In 1973 she joined the Department of Food Science and Nutrition at the University of Minnesota where she is Professor of Food Chemistry and Nutritional Biochemistry. Her research has been focused on identification of the lipophilic products of lipid peroxidation formed in vivo.

    3

    Mary Hadley received a Ph.D. in nutritional biochemistry from the University of Guelph. She is currently Associate Professor in the Department of Food and Nutrition at North Dakota State University.

    View full text