Dogma Disputed
IN PRAISE OF PEROXIDATION
T. L. DORMANDY
Department of Chemical Pathology, Whittington Hospital, London N19 5NF
Summary Free-radical-mediated lipid peroxidation has become closely associated with destructive biochemical processes and, more recently, with disease. Its potential survival value may be overlooked.

BACKGROUND

PEROXIDATION has had a bad press. In the 1940s E. H. Farmer and his team at the British Rubber Producers Association laboratories identified it as the key to the rancidification of fats and oils.1 The discovery put free radicals on the map. It also hatched two notions which are fast becoming articles of faith. The first is that free-radical attack on polyunsaturated lipids means lipid peroxidation; and the second is that lipid peroxidation is bad. Both notions are questionable.

In Farmer's autocatalytic scheme a hydrogen atom is abstracted from a carbon atom in the unsaturated lipid molecule, converting the molecule into a free radical. This lipid free radical immediately interacts with oxygen to become a lipid-peroxy free radical. This in turn wrenches an electron off a previously stable lipid molecule, generating a lipid peroxide with a paired electron complement and a second lipid free radical. The latter interacts with a second molecule of oxygen, and so on. Lipid peroxides tend to break down and it is their fragments that endow rancid food with its characteristic smell, taste, consistence, and toxicity.

Three propositions are at the heart of this scheme. The first is that molecular species with an unpaired electron (free radicals) can exist, however fleetingly. The second is the extraordinary affinity of lipid free radicals for molecular oxygen. The third is the aggressive reactivity of lipid-peroxy free radicals. All three propositions were heretical forty years ago. All three are now established wisdom. But established wisdom should always be questioned-and never more so than when it is being transposed from a non-biological to a biological setting.

A decade before Farmer embarked on his pursuit of rancidity A. E. Gillam and colleagues observed that stored fats and oils develop an absorption peak in the ultraviolet. 2 In most naturally occurring polyunsaturated lipids the double bonds are separated by at least two single bonds ( – C = C – C – C = C – ). Farmer and his group suggested that when a single electron is abstracted from a carbon atom between two double bonds the double bonds close up and become separated by only one single bond ( – C = C – C = C –). The ultraviolet absorption peak is the function of this "diene-conjugated" bind sequence. 3 Thus the first step in lipid autoxidation is not only the conversion of a lipid molecule into a lipid free radical but also the conversion of a non-diene-conjugated into a diene-conjugated species. The dieneconjugation, moreover, is retained by the peroxides and by most of their secondary products. In simple lipid systems its characteristic absorption peak is one of the most sensitive and easily measured indicators of the imminence of rancidification.

DIENE CONJUGATION IN BIOCHEMISTRY

Free-radical processes and products (plastics and polymers) went on to transform evervday life but for many years they made little impression on biochemistry. Free radicals were difficult enough to detect in simple chemical systems. Any suggestions that they might also be involved in biological processes rested largely on the evidence of "markers" –i.e., relatively stable molecules whose origin could be ascribed to free-radical-mediated reactions. Could be, but did not have to be. Often it was a matter of faith. Did polyunsaturated lipids, essential components of all cells, undergo rancidification? It seemed unlikely. Yet the characteristic diene conjugation absorption peak was not only detected in lipid extracts from tissues but a significant increase could also be demonstrated under experimental conditions known to initiate free-radical reactions.' The method was eventually applied to the study of disease, and the findings broadly supported the hunch that free radicals play a part in such diverse conditions as rheumatoid arthritis, toxaemia of pregnancy, alcoholism, and endotoxic shock.' 10 But diene conjugation was invariably interpreted as lipid peroxidation and this interpretation never quite made sense. The specific measurement of lipid peroxides is not impossible and in fresh serum and tissues their concentration is very low. Diene-conjugated lipids, on the other hand, are present in relatively large amounts. Doubts were not stilled by the non-specificity of the spectroscopic technique: theoretically the diene-conjugated bond sequence could be part of a wide variety of molecules. The explanation that emerged was unexpected and aspects of it are still controversial.

More than 90% of diene conjugation in fresh human material was shown to be due to a single fatty acid or, more precisely, a single fatty acid residue. Contradicting the assumption that diene conjugation means lipid peroxidation, this fatty acid is not a peroxide or a peroxidation product: it is a simple isomer of linoleic acid with the double bonds shifted from the 9 and 12 to the 9 and 11 positions. 11,12 The steric configuration of the double bonds is cis-cis in linoleic acid and cis-trans in the isomer. How does this fit into Farmer's scheme of free-radical--mediated peroxidation?

When linoleic acid is exposed to free-radical activity in the presence of oxygen peroxides form, as predicted. But when albumin is added diene-conjugated non-peroxide isomers appear.13 Protein, in other words, can initiate alternative pathways to peroxidation which depend either on the reduction of the carbon-centred diene-conjugated lipid free radical before it could interact with oxygen or on the reversal of this interaction.

But the origin of diene conjugation in living tissues is even more complex. Hughes and others have described the enzymic conversion of linoleic acid to diene-conjugated non-peroxide isomers by anaerobic microorganisms in the rumen of cattle. 14 The conversion probably accounts for the high concentration of diene conjugates in milk. Other microorganisms that can produce diene-conjugated nonperoxide isomers are constituents of the normal flora of the mouth, nose, and vagina. 15

The fact that non-peroxide products account for the bulk of diene conjugation in fresh human material does not mean that they are produced in preference to peroxidation. However, peroxidised cells perish and the peroxides themselves are unstable. By contrast, cells containing large amounts of non-peroxide diene conjugates survive. Do these products reflect or represent a block to "normal" peroxidation?

PEROXIDATION AS A THREAT

In chemistry peroxidation has come to mean the irreversible loss of precious fats and oils. In biochemistry it has become closely linked to cell death. At first the relation was seen purely as cause-and-effect: peroxidation led to structural damage and disintegration. We now know that the reverse is equally true: any kind of structural damage leads to lipid peroxidation. This applies not only to ionising and ultraviolet irradiation, which we know generate free radicals: it also applies to physical injury, chemical poisons, living pathogens, and almost certainly to the ageing process. 16

Over the past few years interest in free radicals has blossomed in clinical medicine. Some of the hypotheses advanced will repay careful neglect. Others will stand the test of time. But across the board peroxidation still conveys a doom-laden message. "Damage" has become disease, though whether peroxidation is linked to it as a cause or effect is still debated. 17, 18 How far is this malevolent image justified?

SURVIVAL VALUE OF PEROXIDATION

The survival of complex organisms depends on a staggeringly fast turnover of its constituent units. Only under exceptional circumstances does Nature favour repair: anything remotely shopsoiled or outworn is discarded and replaced. Theoretically this extravagance could be controlled from the production end, the rate of disposal of old cells being governed by the rate of production of new, ones. For reasons unknown this is not what happens. In reality it is the rate of wastage that determines the rate of replacement. This implies a "self-destruct programme built into every cell and a sensitive feedback. Moreover, since the feedback may have to operate over relatively large distances it must depend on chemical messengers. Two other requirements can be added. First, the self-destruct programme must be switched off in healthy cells. Second, it must be capable of being activated not only by the encoded life-span of cells but also by "accelerated aging" –i.e., any kind of injury or disease. No biochemical mechanism fits this role better than free-radical-mediated lipid peroxidation. Except in certain specific sites and under controlled conditions it is not part of normal metabolism; it can be activated in every cell by any form of damage; and it generates a host of specific reaction products.

A role in the self-destruct programme would give peroxidation immense survival value even in the normal organism. It might acquire special significance in protection against certain diseases. Pre-eminent among them is cancer.

SURVIVAL OF MALIGNANT CELLS

The recognition of chemical, physical, and viral carcinogenesis has made hardly any impact on the incidence and course of clinical cancer. Do these mechanisms matter? The more causes are revealed that lead to malignant change in individual cells the more compelling the argument that the vast majority of these cells are eliminated with extraordinary efficiency. The corollary is that most clinical cancers arise because elimination fails. Neither contention is new. What is relatively new is the elimination process that can be envisaged.

The weeding out of noxious and foreign cells and material has long been conceived as the function of the immune system. That immunological factors operate in cancer is not in doubt. But again: how important are they? The multiplicity of cancer aetiologies recalls the fog that surrounded the origin of tuberculosis before Koch. Bramaud's 1898 textbook lists fifteen "important" causes, including poverty, malnutrition, alcohol, smoking and lack of sunlight. What is remarkable about this list is that it is still valid. Yet once the tubercle bacillus was added the other items lost all their importance. If malignant change in individual cells is an almost continuous phenomenon the first line of their elimination must be a latent function of all cells. Conversely, many or most clinical cancers must reflect a failure of this function. Unfortunately, the identification of such a negative event is infinitely more daunting than the search for positive causes. It might be less daunting if we assumed that the mechanism destroying malignant cells is not basically different from the self-destruct mechanism activated by other forms of injury, damage, or disease.

A remarkable fact about malignant tissue is its greatly diminished peroxidisability." 22 Electron spin resonance spectroscopy, the reference measurement of free-radical studies, shows that the strong peroxy free-radical signal demonstrable in the normal human cervix becomes attenuated or disappears in cervical cancer." The increase in diene conjugation in some cancers seemed to contradict this finding; but diene conjugation can no longer be equated with lipid peroxidation. 24 The mechanism of the diminished peroxidisability remains obscure. In non-biological material peroxidation can be inhibited by antioxidants; and antioxidants, especially vitamin E, do seem to accumulate in malignant cells. 25 But this can only be part of the explanation. The same is true of relative anoxia which might be expected to suppress peroxidation in some tumours.

But lack of an explanation does not alter the fact that the transformation of normal into malignant tissue makes cells less peroxidisable. Or is this a crazy reversal of the true sequence? Is it failure of peroxidation-an essential component of the normal self-destruct mechanism-which allows cancer cells to survive and multiply? Perhaps radiotherapy and many forms of cancer chemotherapy, powerful promoters of lipid peroxidation, are effective within limits because they prop up a cytotoxic process which is, or should be, a built-in programme in every cell.

TREATING PEROXIDATION
To talk of agents which suppress peroxidation as "inhibitory" is correct but uninspiring. To describe them as ,,protective" conjures up a beguiling image of fire-eating dragons (free radicals) and knights in armour (antioxidants). Biochemists have been playing this game for many years without being seriously misled; but the disneyfied message is now carried on a torrent of promotional literature to doctors and the lay public. Fortunately most "protective" agents currently on offer (vitamins, enzymes, and cranky diets) are only financially crippling. But the packaging may tentially useful idea into unmerited disrepute.
There may be an analogy with anti-inflammatory therapy. Such treatment has its uses when inflammation is inappropriately severe or prolonged. But the normal inflammatory response is a sine-qua-non of organised life, and suppressing it carries grave risks. Similarly, excessive peroxidation and free-radical activity may need inhibiting. Inadequate peroxidation and free-radical activity may need boosting. Normal peroxidation and free-radical activity should be left alone.

REFERENCES (supplied but not included here due to OCR difficulties). 
Questions re. references? Email webmaster...