5th Congress Autism-Europe
Articulos / Proceeding
Autism-Spain

ROLE OF DIET IN CAUSE & TREATMENT OF AUTISTIC SYNDROMES:
Biochemical and Physiological Evidence

Michael L.G. Gardner*, A-M. Knivsberg, and K-L. Reicheit

*Department of Biomedical Sciences, University of Bradford, BRADFORD BD7 1DP,
U.K and Department of Paediatric Rescareh, The National Hospital, OSLO, N-0027, Norway.

Introduction

Dietary components are frequently blarned for playing some role in the causation of various mental or behavioural disorders, including autism and schizophrenia. However, in the past, the evidence has often been only anecdotal or subjective: adequately designed trials conducted in a double-blind manner with appropriate control subjects (ideally sex- and age-matched) are comparatively rare. Hence, a truly valid role for diet in the aetiology or treatment of these conditions has never Been proven, and this general concept frequently receives scepticism among medical and scientific professionals. Furthermore, knowledge of biochemical and/or physiological mechanisms that could explain in rigorous scientific terms a link between diet and autism or schizophrenia has been lacking. This too has made it difficult to take seriously any hypothesis involving such a link.

Quite often in the past, the view has been expressed that intake of particular dietary proteins notably gluten, the major protein in many cereals such as what - can be associated with autism and (at least in some cases) schizophrenia; also, that a constructive treatment is simply to remove the offending protein from the diet. Such a view requires to be accompanied by evidence that some 'toxic' component exists in the gluten and that it enters the body in adequate amounts to elicit harmful consequences. Until recently, it was believed that dietary proteins were digested wholly to free amino acids in the gastrointestinal tract (especially, in the lumen of the small intestine and in the enterocytes linning it) before entering the circulation. Hence, if this were true, the actual biochemical moities entering the body during protein assimilation would be the same for all ingested proteins. The hypothesis linking specific proteins with disease would have to be rejected. New knowledge shows that this view is incorrect.

This paper therefore reviews the evidence that it is physiologically feasible for biologically active peptides arising during digestion of a protein-containing meal to be absorbed and, thereafter, to exert important "pharmacological" effects. These effects would, it is argued, be likely to influence the central nervous system and its development. Hence, with this new knowledge, there is a sound rationale for (a) measuring the peptide content of body fluids in patients and control subjects, and (b) excluding specific proteins (particularly gluten and casein) from the patients' diet. Gluten and casein are particularly implicated because their structures are known to contain the amino acid sequences of the endorphins (endogenous morphine-like compounds) - see below.

Peptide Content of Body Fluids

Although major roles as horrnones and neurotransmitters have long been accepted for peptides, remarkably little is known about "general" peptides in tissues and in the circulation. There is some tacit assumption that peptides are only an intermediate between free amino acids and proteins, and that they have no biochemical or physiological significance apart from those as horrnones or neuromodulators which are elicited at very low concentrations. The possibility that peptides might be a significant form for transport of amino-N between organs has been neglected until recently (Grimble & Backwefl, 1996).

Only a few workers have attempted to measure the peptide content of blood. This is largely because the possibility that such molecules might be of pathological or physiological relevance has often been overlooked, but also because there are quite serious methodological difficulties in measuring reliably peptides in body fluids (Gardner, 1996). Such studies as have been undertaken indicate that at least 7% (and possibly up to75%) of the non-protein amino-N (excluding urea) in plasma could be in the form of small peptides. One of the difficulties in obtaining a reliable estimate arises because blood contains a high activity of peptidase enzymes that hydrolyse peptides during the blood collection and during preparation for analysis unless very stringent precautions are taken. Urine is a much easier body fluid for biochemical analysis of peptides, and most workers would agree with our estimate that about 55 to 65% of the amino-N (excluding urea) in urine from healthy subjects is in the form of small peptides. These peptides must have arisen from other organs or tissues: hence, it seems inescapable that peptides are a quantitatively important part of the body's amino-N pool, and that it is vital to consider and investigate their role(s) in normal metabolism and in pathological circumstances.

The source of the peptides in blood (and, subsequently, in urine) is largely unknown. However, the evidence of Noguchi et al (1982) very strongly suggests that the main source of these is catabolism of tissue proteins (not just collagen, as has been assumed), and that a smaller contribution arises from dietary intake. The size of the latter contribution is still unclear, but the strong indications are that it is large enough and consistent enough to wan-ant fuller investigation: this is, of course, central to the hypothesis that peptides arising from dietary proteins play a role in the pathophysiology of mental and behavioural disorders.

Evidence for Absorption of Intact Peptides

The evidence that significant amounts of small peptides, also small amounts of much large molecules including some proteins, can cross the healthy small intestine of Man and many other species has been discussed in detail elsewhere (Gardner, 1994), and so only a few examples are provided here. It is important to stress that we still are unable to estimate reliably what fraction of the protein in a protein meal may be absorbed as intact peptide or intact protein. However, this question is less critical than may at first be supposed because the subsequent biological effect elicited is a compound result of both potency and quantity. Hence, only minute amounts of a highly potent molecule need to be absorbed in order to produce potentially dramatic consequences - absorption of the Botulinum toxin protein is a good example.

Proof of absorption of intact peptides has been provided by several different complimentary approaches, each with its own set of limitations. However, they all point to the same conclusion, namely that significant amounts of intact peptides can be absorbed. Our own work has demonstrated passage of peptides across perfused animal small intestine (e.g. Gardner, 1978), and up to 30% of the absorbed amino-N could be in the form of peptides though the fraction does depend markedly on the particular protein digest from which absorption is occurring. We have also studied absorption of the dipeptide carnosine (SS-alanyl-histidine) in healthy human subjects, and up to 14% of an oral dose can be absorbed and excreted in the urine in intact form (Gardner et al, 1991.

The experiments with carnosine show that there is substantial variation between individual subjects in the %-age recovery of intact carnosine in urine. The results show that this is due largely due to variability in the carnosinase enzyme activity in the subject's plasma, and they emphasise the importance of plasma hydrolase enzymes in clearing (or metabolising) absorbed peptides. Hence, any deficiency in these enzyrnes (e.g. any genetic absence of enzyme or an enzyme-variant with low activity) would lead to a longer biological half-life of absorbed peptides in the circulation: in turn, this would increase the chances of such peptides gaining access to the central nervous system and exerting effects there.

Biological and Pharmacological Activity of Peptides from Dietary Proteins

It is now clear that peptic digestion of gluten and casein in experimental conditions in vitro produces peptides with opioid activity, and some of these peptides (the gluten exorphins and casomorphins, respectively) have been sequenced (references provided by Gardner, 1994). It is less certain whether such peptides are actually produced during digestion in vivo and, if so, whether they are absorbed; also, if they are absorbed, it is not clear yet what role plasma peptidases (and those in other non-gastrointestinal tissues) play in their clearance. However, decreases in urinary peptide output following gluten and casein exclusion from the diet are strongly suggestive that exorphins and casomorphins are absorbed. Also, decreases in urinary peptide output that have been reported following administration with neuroleptics (which activate peptide hydrolase enzymes) are wholly consistent with the view that systemic peptidases play important roles in the physiological handling of peptides (Reichelt & Teiglandqerstad, 1995). It has been argued that the production of pharmacologically active peptides during digestion of dietary proteins must be regarded as "likely" rather than "rare" (Gardner, 1985)

Discussion

From the above, it is clear that a pathophysiological mechanism could exist for peptides from dietary proteins to interfere with the central nervous system and its development. This, in turn, leads to a rationale for evaluating the effect on patients of excluding gluten and casein from the diet, though such exclusion must be prolonged and it must be complete. As patients with coeliac disease are aware, many processed foods, as well as wheat products, contain gluten, and hence substantial effort is required to establish a genuinely gluten-free diet: firm commitment is required to adhere to it. Evaluation of the validity of this mechanism requires rigorous biochemical experimentation on a large enough population of patients, together with appropriate control subjects, with emphasis being given to objective quantitative measurements that will withstand sceptical scrutiny. More and better analytical methods are also urgently required (see Gardner, 1996). We are currently completing a double-blind multi-centre analysis of urinary peptides using several different analytical methodologies. If increased levels of peptides in autistic patients are confirmed, then detailed investigation of intestinal absorption, intestinal permeability, and systemic peptide hydrolase activities will be required. Further, it should be stressed that a single defect or abnormality is unlikely to account for the aetiologies of all patients. Experimenters need to expect to find sub-sets of the patient populations.

References

Gardner, M.L.G., (1978) Amino acid and peptide absorption from partial digests of proteins in isolated rat small intestine. J. Physiol. (London) 284, 83-104.

Gardner, M.L.G. (1985) Production of phannacologically active peptides from foods in the gut. In: Hunter, J.O. & Alun Jones, V. (eds) Food and the Gut. 121-134. Bailfiere Tindafi, London.

Gardner, M.L.G., Illingworth, K.M., Kefleher, J. & Wood, D. (1991). Intestinal absorption of the intact peptide carnosine in Man, and comparison with intestinal permeabdity to lactulose. Journal of Physiology, 439, 411-422.

Gardner, M.L.G. (1994) Absorption of intact proteins and peptides. In: L.R- Johnson (ed) Physiology of Gastrointestinal Tract, 3rd edn., pp. 1795-1820. Raven Press, New York.

Gardner, M.L.G. (1996) Transmucosal passage of intact peptides. In: Gíimble, G. & Backwefl, C. (eds) The Significance of Circulating Peptides in Mammalian Protein Metabolism. In Press, Portland Press, London.

Grimble, G. & BackweU, C. (1996) (eds) The significance of Circulating Peptides in Mammalian Protein Metabolism. In Press, Portland Press, London.

Noguchi,T.,Okiyama, A.,Naito,H. et al.(1982)Agricultural and Biological Chemistry, 46, 2821-2828.

Reichelt, K-L. & Teigland-Gjerstad, B. (1995) Decreased urinary peptide excretion in schizophrenic patients after neuroleptic treatment. Phychiatry Research 58, 171-176.


Mail Address

Michael L.G.Gardner, D.Sc. F.I.Biol., Reader in Physiological Biochemistry, Department of Biomedical Sciences, University of Bradfort, Bradfort BD7 1DP, England, U.K.