Nutrition and biomarkers in psychiatry : research on micronutrient deficiencies in schizophrenia, the role of the intestine in the hyperserotonemia of autism, and a method for non-hypothesis driven discovery of biomarkers in urine

摘要

This thesis describes the study of markers of nutrition and intestinal motility in mental disorders with a focus on schizophrenia and autism, and the development, evaluation and application of a biomarker discovery method for urine. The aim of the thesis is to investigate the role of long-chain polyunsaturated fatty acids (LCPUFA), B-vitamins and platelet (PLT) serotonin (5-HT) in schizophrenia and autism. The thesis proposes also that biomarker research in psychiatric disease is of great relevance and describes a biomarker discovery method in urine using a non-hypothesis driven ‘-omics’-like approach. The thesis ends by summarizing its contents and putting biomarker research in psychiatric disease and its implications in a broader perspective.In the Introduction the complex etiology and potential role of non-hypothesis driven biomarker research in psychiatric disease is reviewed, with an accent on schizophrenia and autism. The enormous economic and psychosocial global burden of mental disorders is described as well as their epidemiology, clinical presentation and classification/diagnosis. Hypothesized etiological factors are discussed to create a framework in which biomarkers and the research thereof can be positioned. Furthermore, advances in the field of biomarker research in psychiatry are discussed in the context of epigenetics, proteomics and metabolomics.The first part (Part I) of this thesis describes a study of LCPUFA and B-vitamins in schizophrenia, of PLT 5-HT and intestinal permeability in autism, and of the value of PLT 5-HT as marker of intestinal motility in newborns. The chapters, in which these studies are described, are preceded by a review (Chapter 1), which gives an overview of the role of LCPUFA and folate in the etiology and severity of psychiatric diseases such as depression, bipolar disorders, schizophrenia and autism. Pregnancy complications and folate-substrated carbon-1 metabolism are considered and their possible epigenetic effect on the etiology of mental disorders is described. Other nutritional factors, such as LCPUFA, that are important for brain development, physico-chemical properties of membranes, signal transduction and DNA-transcription, and that have been used in supplementation trials, are suggested to be important factors in the origin and severity of schizophrenia. In Chapter 2 we describe the results from a study concerning the essential fatty acid (EFA) and functional B-vitamin status in patients with schizophrenia. Aberrant EFA-status and increased homocysteine (Hcy; a marker of functional B-vitamin deficiency), have been reported before in subgroups of patients with schizophrenia. We describe the characteristics of large subgroups with marginal to severe deficiencies of LCPUFA and B-vitamins, notably folate and vitamin B12. Deficiencies proved easily correctable in the most severely deficient patients upon supplementation with ω3 fatty acids and B-vitamins. Chapter 3 attempts to integrate and link previously reported findings of increased intestinal permeability and increased PLT 5-HT levels in subgroups of children with pervasive developmental disorders (PDD). Platelet 5-HT and intestinal permeability were assessed in children with PDD in Curaçao. Differential urinary excretion of inert sugars after ingestion of a sugar solution was used as marker of intestinal permeability. In Chapter 4 we examined the potential of PLT 5-HT as marker of intestinal motility. For this we studied whole blood and PLT 5-HT in mothers (normal motility) and their newborns (developing intestinal motility) at birth. The course of PLT 5-HT in relation to changes in feeding mode (i.e. parenteral/enteral) was investigated in a small group of preterm born infants to see whether PLT 5-HT was responsive to changes in intestinal motility. Chapters 2, 3 and 4 describe hypothesis-driven research in mental disorders. However, the advance of knowledge about mental disorders is slow and it is likely to benefit from complementation by information that is generated through non-hypothesis driven research with state-of-the art techniques that profile proteins (i.e. proteomics) and metabolites (i.e. metabolomics). These ‘-omics’ techniques are likely to deliver a multitude of candidate diagnostic and prognostic markers as well as therapeutic targets, compared to hypothesis driven research.Part II is thus devoted to the development, evaluation and application of such a non-hypothesis driven method. In Chapter 5 we describe the comparative analysis of low molecular weight urinary components using LC-MS and subsequent multivariate statistical analysis of the processed LC-MS data. This chapter deals with the development, evaluation and preliminary application of the method to proteinuria in humans. The potential and pitfalls of the method are contemplated upon. Chapter 6 describes an advanced proof-of-principle of the method through the comparison of urinary profiles from pregnant and non-pregnant females using the improved methodology described in Chapter 5. The methodology is significantly optimized with respect to data processing and multivariate statistical analysis. More focus is put on the selection of discriminatory peaks.---------------------------------------------------------------Éénderde van de patiënten met schizofrenie blijkt matige tot ernstige tekorten aan bepaalde B-vitaminen en ω3- en ω6-vetzuren te hebben. Deze tekorten zijn eenvoudig op te heffen met voedingsupplementen (o.a. foliumzuur en visolie). Een lage status van deze micronutriënten speelt waarschijnlijk een rol in het ontstaan en de ernst van diverse psychiatrische ziekten en het ontstaan van hart- en vaatziekten. De behandelende artsen vermoedden het bestaan van deze tekorten niet.Ongeveer een kwart van de patiënten met autisme heeft een verhoogd serotoninegehalte in bloedplaatjes (hyperserotonemie: “biomarker”). Serotonine is een neurotransmitter in onze hersenen en darmen. Autisten zouden ook vaker maagdarmstoornissen hebben. We vonden geen relatie tussen hyperserotonemie (26%) en de darmdoorlaatbaarheid (in 0% verhoogd) in kinderen met autisme. Het idee dat een verhoogde darmmotiliteit hyperserotonemie veroorzaakt, werd ondersteund door een twee maal hoger bloedplaatjesserotonine van moeders (actieve darm) t.o.v. hun pasgeboren baby’s (inactieve darm). Tevens bleek starten en staken van enterale voeding in pasgeborenen gecorreleerd aan respectievelijk stijgingen en dalingen van hun bloedplaatjesserotonine. Meer onderzoek naar het maagdarmstelsel (onze “second brain”) bij autisme is gewenst.Niet-hypothesegedreven onderzoek als aanvulling op het veelal hypothesegedreven onderzoek kan helpen bij het vinden van diagnostische en therapeutische biomarkers voor psychiatrische stoornissen. De hiertoe opgezette methode onderzoekt urinemonsters met vloeistofchromatografie-massaspectrometrie, waarna de 3D-data multivariaat geanalyseerd worden om biomarkers te ontdekken. Experimenten met urinemonsters van patiënten met eiwit in hun urine (vanwege een nierziekte) en controles, en van zwangere en niet-zwangere vrouwen, lieten duidelijke groepsverschillen zien (“proof-of-principle”). Dit geeft hoop dat de ontwikkelde methode in de toekomst gebruikt kan worden voor het opsporen van biomarkers bij psychiatrische stoornissen en andere ziektes.