The present invention relates to methods and uses for determining whether an individual has or has an increased risk of developing an autism spectrum disorder

The listing or discussion of a prior-published document in this specification should not necessarily be taken as an acknowledgement that the -document is part of the state of the art or is common general knowledge Autism spectrum disorder (ASD) consists of a continuum or spectrum of complex neurodevelopmental disorders with a serious lifelong impact on individuals from all ethnic and socioeconomic backgrounds (Minschew, 2007 The Summer Institute of Neurodevelopmental Disorders UC Davis M I N D Institute, Sacramento, California, August 2-3, Pessah, 2006 Understanding immunological and neurobiology susceptibilities contributing to autism risk UC Davis MIND Institute Conference, Sacramento, California, November 2-3, Grandjean & Landigan, 2006 www thelancet com Published online November 8, 2006 DOI 10 1016/S0140-6736 (06) 69665-7, London & Etzel, 2000 Environmental Health Perspectives Supplements 108) Both genetic and environmental factors appear to contribute to the development of ASD (Herbert, 2006 Clinical implications of environmental toxicology for children's neurodevelopment in autism UC Davis MIND Institute Conference, Sacramento, California, November 2-3, Hertz-Picciotto, Croen, Hansen, Jones, Van de Water, & Pessah, 2006 Environmental Health Perspectives 114, from www ehponline org, London & Etzel, 2000 Environmental Health Perspectives Supplements 108) ASD includes five disorders Autistic Disorder, Pervasive developmental disorder not otherwise specified (PDD, NOS), Asperger's Disorder, Retts Disorder, and Childhood Disintegrative Disorder Common manifestations include challenges in communication, imaginative play, and socializing with others Preoccupation with unusual interests and a restricted/repetitive pattern of interests and/or behaviors are typical Routines, inflexibility, and difficulty with new situations characterize many individuals Intellectual disabilities are common

The prevalence of ASD has increased over the years from a prevalence of 4 in 10000 children before 1980 [2] to 53 in 10000 in 2006 [12] but this varies regionally and with ethnicity Some areas have much higher incidences of ASD Current diagnosis of ASD is subjective and depends on observations of a cluster of behaviours and fulfillment of several criteria set out in the Diagnostic and Statistical Manual of Mental Disorders 4 ^th edition (DSM-IV) for the diagnosis of ASD by trained clinicians At present, there are no reliable biochemical- or genetic-screening tests and in some cases such as late onset autism, normal development might switch to delay acquisition of new skills adding to the difficulty for diagnosing. Thus, there is a pressing need for a reliable diagnostic tool for ASD that is both sensitive and reliable given that early diagnosis generally leads to timely interventions and optimises outcomes.

Using NMR-based metabolic profiling, the inventors have identified clear metabolic differences between autistic and normal children and have realised that such changes may be used as early diagnostic markers of ASD.

Accordingly, a first aspect of the invention provides a method for determining whether an individual has, or has an increased risk of developing, autism spectrum disorder (ASD), the method comprising assessing any one or more of a mammalian-microbial co- metabolite, alanine, glycine, creatinine, glutamate, taurine, formate, inosine, succinate, acetate, an Λ/-acetyl glycoprotein, or a vitamin B metabolite such as Λ/-methyl-4- pyridone-3-carboxamide (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND), in a sample taken from an individual, wherein the method does not comprise assessing only hippurate and/or 4-hydroxyhippurate, and wherein the method does not comprise assessing Λ/-methyl-2-pyridone-5-carboxamide alone.

Thus, the invention provides a method for determining whether an individual has, or has an increased risk of developing, autism spectrum disorder (ASD), the method comprising assessing any one or more of a mammalian-microbial co-metabolite, alanine, glycine, creatinine, glutamate, taurine, formate, inosine, succinate, acetate, an Λ/-acetyl glycoprotein, or a vitamin B metabolite such as Λ/-methyl-4-pyridone-3-carboxamide (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND), in a sample taken from an individual, wherein the method does not comprise assessing any of: only hippurate; only 4-hydroxyhippurate; only Λ/-methyl-2-pyridone-5-carboxamide; only the combination of hippurate and 4-hydroxyhippurate; only the combination of hippurate and Λ/-methyl-2-pyridone-5-carboxamide; only the combination of 4-hydroxyhippurate and N- methyl-2-pyridone-5-carboxamide; or only the combination of hippurate, 4- hydroxyhippurate and Λ/-methyl-2-pyridone-5-carboxamide.

By a 'mammalian-microbial co-metabolite', we include the meaning of a metabolite which is the product of a mammal metabolising a microbial metabolite. Thus, any metabolite produced by gut bacteria and co-metabolised by a mammalian host may be a mammalian-microbial co-metabolite. For example, Clostridia produce para-cresol, which is co-metabolised in a mammal to para-cresol sulfate. The mammalian-microbial co- metabolite may be any metabolite produced by Clostridia, which metabolite has been co- metabolised by a mammalian host. Metabolites produced by gut bacteria undergo Phase 1 and Phase 2 metabolism by the mammal. Phase 1 metabolism adds extra functionality (eg additional OH groups) and Phase 2 metabolism forms chemically bonded conjugates to these functional groups, which may increase water solubility or aid excretion. Thus, it is appreciated that the mammalian-microbial co-metabolite may be any metabolite produced by gut bacteria that has undergone Phase 1 and 2 metabolism. For example, the mammalian-microbial co- metabolite may be a cresol metabolite including any metabolite that is the product of cresol being metabolised in a mammalian host (eg by Phase 1 and/or Phase 2 metabolism) such as p-tolyl sulfate, 4-methylenecyclohexa-2,5-dienone, A- hydroxybenzoate, hippurate and 4-hydroxyhippurate.

Common mammalian-microbial co-metabolites include dimethylamine (DMA), hippurate, phenylacetylglutamine (PAG) and para-cresol sulfate (PCS). Thus, when determining whether an individual has, or has an increased risk of developing, ASD, it is preferred if any one, two or all three of dimethylamine (DMA), phenylacetylglutamine (PAG) and para-cresol sulfate (PCS) are assessed.

By a 'vitamin B metabolite' we include any B vitamin (eg vitamin B3) and its associated metabolites (eg any metabolite on the same metabolic pathway). For example, a vitamin B metabolite may be a metabolite within 8, 7, 6, 5, 4, 3, 2 or 1 reaction steps of the B vitamin. Particular examples of vitamin B3 metabolites include Λ/-methyl-4-pyridone-3- carboxamide (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND), all of which the inventors have found to be perturbed in autistic children relative to normal children. Thus, when determining whether an individual has, or has an increased risk of developing, ASD, it is preferred if any one, two or all three of Λ/-methyl-4-pyridone-3- carboxamide (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND) are assessed.

It is appreciated that more than one of a mammalian-microbial co-metabolite, alanine, glycine, creatinine, glutamate, taurine, formate, inosine, succinate, acetate, an Λ/-acetyl glycoprotein, or a vitamin B metabolite such as Λ/-methyl-4-pyridone-3-carboxamide

(4PY), A/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND) may be assessed. Further, and as mentioned above, more than one mammalian-microbial co- metabolite and/or more than one vitamin B metabolite may be assessed.

Thus, any combination of two, three, four, five, six, seven, eight, nine, ten, eleven, twelve, thirteen, fourteen, fifteen, sixteen or all seventeen of dimethylamine (DMA), hippurate, phenylacetylglutamine (PAG), para-cresol sulfate (PCS), alanine, glycine, creatinine, glutamate, taurine, formate, inosine, succinate, acetate, an Λ/-acetyl glycoprotein, Λ/-methyl-4-pyridone-3-carboxamide (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND) may be assessed.

In one embodiment at least two, three, four, five, six, seven, eight, nine, ten, eleven, twelve, thirteen, fourteen, fifteen or sixteen of dimethylamine (DMA), hippurate, phenylacetylglutamine (PAG), para-cresol sulfate (PCS), alanine, glycine, creatinine, glutamate, taurine, formate, inosine, succinate, acetate, an Λ/-acetyl glycoprotein, N- methyl-4-pyridone-3-carboxamide (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND) are assessed.

Conveniently, the assessing comprises comparing the amount of any one or more of a mammalian-microbial co-metabolite, alanine, glycine, creatinine, glutamate, taurine, formate, inosine, succinate, acetate, an Λ/-acetyl glycoprotein, or a vitamin B metabolite such as Λ/-methyl-4-pyridone-3-carboxarnide (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND), in the sample to the amount of the respective any one or more of a mammalian-microbial co-metabolite, alanine, glycine, creatinine, glutamate, taurine, formate, inosine, succinate, acetate, an Λ/-acetyl glycoprotein, or a vitamin B metabolite such as Λ/-methyl-4-pyridone-3-carboxamide (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND), in a standard sample obtained from one or more individuals who are known to have, or have a known risk of developing, ASD.

Any suitable method may be used to assess a given metabolite in a sample taken from an individual, and it is appreciated that more than one method may be employed. For example, the metabolite may be assessed using NMR spectroscopy and/or mass spectrometry (eg by multiple ion monitoring). In this case, the sample may be fractionated prior to analysis, for example by liquid chromatography. The use of ¹ H NMR to assess urinary metabolites is described in Example 1 , and the measurement of metabolites using NMR or mass spectrometry is discussed in US patent no. 7,373,256 and in WO 03/107270. Additionally or alternatively, the metabolite may be assessed by making use of a binding partner that binds selectively to the metabolite, or by making use of an enzyme linked assay (eg enzyme linked immunosorbent assay, (ELISA) or an assay in which the metabolite is converted (either directly or indirectly) into a molecule which can readily be detected) that can be used to quantify the metabolite. The sample taken from the individual may be any suitable sample. For example, the sample may be any of urine, blood, blood plasma, blood serum, saliva, sweat, tears, breath or breath condensate. Preferably, the sample is a urine sample.

The individual may be a human or mammalian individual, such as a horse, dog, pig, cow, sheep, rat, mouse, guinea pig or primate. Preferably, the individual is a human individual.

Given that ASD is a brain development disorder, in one embodiment, the individual is a pre-pubescent individual. By a pre-pubescent individual we include the meaning of a neonate up to the age at which the brain is fully developed. For example, the human individual may be aged 18 years or less, 17 years or less, 16 years or less, 15 years or less, 14 years or less, 13 years or less, 12 years or less, 11 years or less, 10 years or less, 9 years or less, 8 years or less, 7 years or less, 6 years or less, 5 years or less, 4 years or less, 3 years or less, 2 years or less or 1 year or less.

A second aspect of the invention provides a use of a means for assessing any one or more of a mammalian-microbial co-metabolite, alanine, glycine, creatinine, glutamate, taurine, formate, inosine, succinate, acetate, an Λ/-acetyl glycoprotein or a vitamin B metabolite such as Λ/-methyl-4-pyridone-3-carboxamide (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND), in a sample taken from an individual in determining whether an individual has, or has an increased risk of developing, ASD, wherein the use does not comprise the use of means for assessing only hippurate and/or 4-hydroxyhippurate, and wherein the use does not comprise the use of means for assessing Λ/-methyl-2-pyridone-5-carboxamide alone.

Thus, the invention provides a use of a means for assessing any one or more of a mammalian-microbial co-metabolite, alanine, glycine, creatinine, glutamate, taurine, formate, inosine, succinate, acetate, an Λ/-acetyl glycoprotein or a vitamin B metabolite such as Λ/-methyl-4-pyridone-3-carboxamide (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND) ₁ in a sample taken from an individual in determining whether an individual has, or has an increased risk of developing, ASD, wherein the use does not comprise the use of means for assessing any of: only hippurate; only 4- hydroxyhippurate; only Λ/-methyl-2-pyridone-5-carboxamide; only the combination of hippurate and 4-hydroxyhippurate; only the combination of hippurate and Λ/-methyl-2- pyridone-5-carboxamide; only the combination of 4-hydroxyhippurate and Λ/-methyl-2- pyridone-5-carboxamide; or only the combination of hippurate, 4-hydroxyhippurate and Λ/-methyl-2-pyridone-5-carboxamide.

Any suitable means may be used to assess one or more given metabolites in a sample taken from an individual. For example, the means may be a NMR spectrometer or a mass spectrometer arranged to detect one or more given metabolites in a sample. Example 1 describes the use of an NMR spectrometer to detect and quantify urinary metabolites in a sample. Similarly, a mass spectrometer may be used in multiple ion monitoring mode, such that only certain ion fragments are entered into the instrument and detected by the mass spectrometer. Thus, if assessing a particular metabolite, the mass spectrometer can be programmed to detect only those ion fragments derived from that particular metabolite. It is appreciated that to confirm NMR and mass spectrometry assignments, spectra may also be compared to known reference standards of the particular metabolites in question.

Other means of assessing one or more given metabolites in a sample taken from an individual include an enzyme linked assay (eg ELISA or an assay in which the metabolite is converted (either directly or indirectly) into a molecule which can be readily detected) or an agent that binds to a metabolite. For example, the metabolite may be a substrate in an enzymatic reaction, such that an enzyme linked assay can be used to quantify the metabolite. Enzyme linked assays are standard practice in the art and typically involve colorimetric, fluorescent, or chemiluminescent detection. An agent that binds to a metabolite may be used to assess the metabolite by measuring the binding between the agent and the metabolite. Conveniently, the agent is detectably labelled so that the presence of the metabolite can readily be detected. Examples of labels include peptide labels, chemical labels, fluorescent labels or radio labels.

Preferences for assessing any one or more of a mammalian-microbial co-metabolite, alanine, glycine, creatinine, glutamate, taurine, formate, inosine, succinate, acetate, an Λ/-acetyl glycoprotein, or a vitamin B metabolite such as Λ/-methyl-4-pyridone-3- carboxamide (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND) in a sample taken from an individual, for the mammalian-microbial co-metabolite, for the sample and for the individual, for this and all subsequent aspects of the invention are defined above with respect to the first aspect of the invention. A third aspect of the invention provides a method for classifying an individual as being one who has, or has an increased risk of developing, a pathological condition, the method comprising assessing a mammalian-microbial co-metabolite in a sample taken from the individual, wherein when assessing para cresol sulphate (PCS) alone the individual is not classified according to whether or not the individual has multiple sclerosis, wherein when assessing only hippurate and/or 4-hydroxyhιppurate the individual is not classified according to whether or not the individual has ASD, and wherein when assessing Λ/-methyl-2-pyπdone-5-carboxamιde alone, the individual is not classified according to whether or not the individual has ASD

Thus, the invention provides a method for classifying an individual as being one who has, or has an increased risk of developing, a pathological condition, the method comprising assessing a mammalian-microbial co-metabolite in a sample taken from the individual, wherein when assessing para cresol sulphate (PCS) alone the individual is not classified according to whether or not the individual has multiple sclerosis, and

wherein when assessing any of only hippurate, only 4-hydroxyhιppurate, only N- methyl-2-pyrιdone-5-carboxamιde, only the combination of hippurate and 4- hydroxyhippurate, only the combination of hippurate and Λ/-methyl-2-pyrιdone-5- carboxamide, only the combination of 4-hydroxyhιppurate and Λ/-methyl-2-pyrιdone-5- carboxamide, or only the combination of hippurate, 4-hydroxyhιppurate and Λ/-methyl-2- pyrιdone-5-carboxamιde, the individual is not classified according to whether or not the individual has ASD Preferably, the individual is classified as one who has, or has an increased risk of developing, ASD However, the individual may also be classified as one who has, or has an increased risk of developing, any of the following pathological conditions motor neuron disease, cirrhosis of the liver and migraine headaches, which are conditions related to sulfate deficiencies, Parkinson's disease, a neurological condition such as Schizophrenia, Huntington's disease and Alzheimer's disease, idiosyncratic drug toxicity, diabetes, an immunological-based condition such as Lupus, an inflammatory condition, atherosclerosis, rheumatoid arthritis, pre-eclampsia, post-operative sepsis, gut dysbiosis, cystic fibrosis, obesity, diabetes, muscular dystrophy, and cancer Preferences for assessing the mammalian-microbial co-metabolite include assessing any one or two or three or all four of dimethylamine (DMA), hippurate, phenylacetylglutamine (PAG), para-cresol sulfate (PCS) Conveniently, assessing a mammalian-microbial co-metabolite in a sample taken from the individual to classify that individual as one who has or has an increased risk of developing a pathological condition, comprises comparing the amount of a mammalian- microbial co-metabolite in the sample to the amount of a mammalian-microbial co- metabolite in a standard sample obtained from one or more individuals who are known to have, or have a known risk of developing a pathological condition. For example, to classify an individual as being one who has, or has an increased risk of developing, ASD, the amount of a mammalian-microbial co-metabolite in the sample taken from the individual may be compared to the amount of a mammalian-microbial co-metabolite in a standard sample obtained from one or more individuals who are known to have, or have a known risk of developing, ASD.

It is appreciated that the individual may be classified as being one who has, or has an increased risk of developing, any pathological condition for which assessing a mammalian-microbial co-metabolite is predictive. Thus, any pathological condition whose class representation variables are correlated with a mammalian-microbial co- metabolite may be characterised by assessing a mammalian-microbial co-metabolite. Methods for identifying such pathological conditions are standard practice in the art and are described in, for example, US 7,373,256 and WO 03/107270.

It is also appreciated that by assessing a mammalian-microbial co-metabolite in a sample taken from the individual, the individual may be classified as having any one or more particular pathological conditions.

A fourth aspect of the invention provides a use of a means for assessing a mammalian- microbial co-metabolite in a sample taken from an individual, in classifying an individual as being one who has, or has an increased risk of developing, a pathological condition, wherein when assessing para cresol sulphate (PCS) alone the individual is not classified according to whether or not the individual has multiple sclerosis, wherein when assessing only hippurate and/or 4-hydroxyhippurate the individual is not classified according to whether or not the individual has ASD, and wherein when assessing N- methyl-2-pyridone-5-carboxamide alone, the individual is not classified according to whether or not the individual has ASD. Thus, the invention provides a use of a means for assessing a mammalian-microbial co- metabolite in a sample taken from an individual, in classifying an individual as being one who has, or has an increased risk of developing, a pathological condition,

wherein when assessing para cresol sulphate (PCS) alone the individual is not classified according to whether or not the individual has multiple sclerosis, and

wherein when assessing any of only hippurate; only 4-hydroxyhippurate; only N- methyl-2-pyridone-5-carboxamide; only the combination of hippurate and 4- hydroxyhippurate; only the combination of hippurate and Λ/-methyl-2-pyridone-5- carboxamide; only the combination of 4-hydroxyhippurate and Λ/-methyl-2-pyridone-5- carboxamide; or only the combination of hippurate, 4-hydroxyhippurate and Λ/-methyl-2- pyridone-5-carboxamide, the individual is not classified according to whether or not the individual has ASD.

Preferences for classifying the individual are as described above with respect to the third aspect of the invention.

In any aspect of the invention, in addition to assessing any one or more of a mammalian- microbial co-metabolite, alanine, glycine, creatinine, glutamate, taurine, formate, inosine, succinate, acetate, Λ/-acetyl glycoprotein, or a vitamin B metabolite such as Λ/-methyl-4- pyridone-3-carboxamide (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND) in a sample taken from the individual, the invention may comprise assessing at least one further biological parameter. The at least one further biological parameter may be assessed either in the same sample or in a different sample taken from the individual. Further biological parameters that may be assessed include any one or more of genetic information, proteomic information, and metabolomic information. Thus, the at least one further biological parameter may be any of the sequence of a particular region of a chromosome, the status of a specific protein or the status of a particular metabolite or a combination thereof. Additionally or alternatively, the at least one further biological parameter may be a phenotypic trait of an individual such as behaviour.

In one embodiment, the at least one further biological parameter comprises the status of a gene that encodes factors related to sulfation or sulfation deficiency. Analysis of the status of said gene factor applies to any attributes inherent in a gene, for instance its nucleotide sequence, chromosomal position, or copy number. Variations of these factors may expose differences from a sequence normal to a population, specifically a single nucleotide polymorphism (SNPs), multiple site polymorphisms or other mutations, differences in normal gene copy number (copy number variations, CNVs), and variance in epigenetic regulation, including factors regulating expression such as DNA methylation. In a preferred embodiment, the analysis of the gene measures an SNP that relates to a defect in the gene-encoded factor. In one embodiment, said at least one further biological parameter to be assessed is a sulfate transporter. Sulfate transporters are critical to regulation of free sulfate available to cells. Defects in said sulfate transporters are correlated with decreased salvage of dietary or systemic sulfate necessary for detoxification and other processes. The sulfate transporters measured in this invention may be any of the group comprising the renal transporters NaS1, sat-1, or CFEX, the intestinal transporters NaS1 , Diastrophic Dysplasia Sulfate Transporter (DTDST), or Down Regulated in Adenoma transporter (DRA), and the mucosal sodium sulfate symporter.

In another embodiment, said at least one further biological parameter to be assessed is a transferase detoxification enzyme, specifically a sulfotransferase or glutathione transferase. In an embodiment the at least one further biological parameter is the identification of a polymorphism in a phenol sulfotransferase, either SULT1A1 , or

SULT1A2. In an alternate embodiment, the at least one further biological parameter is a polymorphism that confers hyper-activity to a competing sulfotransferase or glutathione transferase (by competing transferase it is meant a transferase enzyme that conducts sulfation reactions on metabolites or xenobiotics other than the phenolic compounds of interest of this invention). Examples of said competing sulfotransferases are those for androsterones (SULT2A1), estrogens (SULT1 E1), dopaminergics (SULT1A3), cholesterol (SULT2B1), and tyrosyl protein sulfotransferases (TS-PSTs), among others. Specific glutathione-s-transferases (GSTs) for polymorphism analysis include GST P1 haplotype, and GST MuI

In another embodiment, the at least one further biological parameter to be assessed is an enzyme involved in the generation of sulfur-containing compounds that can be used as substrates for Phase Il detoxification reactions. Said enzymes include cysteine dioxygenase, cystathione B-synthase, methionine synthetase, and Organic Anion Transporter B (OATPB).

In another embodiment, the at least one further biological parameter to be assessed is a sulfur-containing metabolite or metabolite ratio indicative of a subject's sulfation capacity.

Said metabolites and metabolite ratios may comprise sulfate, sulfite, thiosulfate, thiocyanate, transulfurated androgens, plasma glutathione (GSH)/ glutathione disulfide (GSSG) ratio, cysteine, taurine, sulfate, free sulfate, GSSG, or paracetamol sulfate/paracetamol glucuronide ratio.

The at least one further biological parameter may be known to be discriminatory for a particular pathological condition (eg ASD), such that it would be beneficial to assess at least said one further biological parameter in addition to any one or more of a mammalian-microbial co-metabolite, alanine, glycine, creatinine, glutamate, taurine, formate, inosine, succinate, acetate, an Λ/-acetyl glycoprotein, or a vitamin B metabolite such as Λ/-methyl-4-pyridone-3-carboxamide (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND). For example, where the individual is classified as being one who has, or has an increased risk of developing, ASD, the at least one further biological parameter may be a further marker for ASD such as a diagnostic biomarker of ASD or any characteristic trait of an individual with ASD. In one embodiment, the at least one further biological parameter to be assessed is a gene known in the art to be associated with ASD such as a neuronal cell adhesion/synapse formation marker (PCDH10, CDH10, CDH9, NRXN1 , CNTN4), a copy number variation on NLGN1 , ASTN2, a factor in the ubiquitin pathway displaying copy number variations (for example, UB3A, PARK2, RFWD2, FBXO40), dopamine B- hydroxylase, or MET kinase. Alternatively, the at least one further biological parameter may include genes correlated to pathologies associated with autism spectrum disorders, such as Rett syndrome (genes MECP2 and CDKL5), Angelman syndrome (genes SLC9A6 and UBE3A), and Fragile X (gene FMR4). In one embodiment, the at least one further biological parameter to be assessed is at least one further metabolite that is assessed in the same or a different sample.

Samples containing said at least one further biological parameter may be collected in blood (blood, cells, plasma or serum), tissue homogenates, fecal samples, urine samples, hair samples, interstitial fluid, cerebrospinal fluid, synovial fluid, or saliva samples that allow for the preservation of genetic information. The preservation of these samples is conducted by common methods in the art, including the use of gentle homogenization and collection, and frozen storage until use. The detection and measurement of genetic information is accomplished by any of the techniques common in the art, including RNA or DNA microarrays, fluorescence hybridizations such as fluorescence in situ hybridization (FISH), or genomic sequencing techniques. For example, the at least one further biological parameter to be assessed may be a further mammalian-microbial co-metabolite or a further mammalian metabolite Thus, the method may comprise assessing more than one mammalian-microbial co-metabolite or mammalian metabolite in a sample taken from the individual Assessing more than one mammalian-microbial co-metabolite or mammalian metabolite may provide for more powerful discriminatory models for classifying individuals, than when only one such metabolite is assessed Methods for generating models to classify individuals based on multiple variables that may be used in the context of the present invention, include those described in US patent no 7,373,256 and WO 03/107270

It is appreciated that the assessment of any one or more of a mammalian-microbial co- metabolite, alanine, glycine, creatinine, glutamate, taurine, formate, inosine, succinate, acetate, an Λ/-acetyl glycoprotein, or a vitamin B metabolite such as Λ/-methyl-4- pyrιdone-3-carboxamιde (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND), may comprise assessing the ratio of any one or more of a mammalian- microbial co-metabolite, alanine, glycine, creatinine, glutamate, taurine, formate, inosine, succinate, acetate, an Λ/-acetyl glycoprotein, or a vitamin B metabolite such as Λ/-methyl- 4-pyrιdone-3-carboxamιde (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND) with at least one other metabolite, such as any of creatinine, creatine, glycine, hippurate, NMNA, NMND, PAG, succinate and taurine

It is also appreciated that it may be necessary to express the amount of any one or more of a mammalian-microbial co-metabolite, alanine, glycine, creatinine, glutamate, taurine, formate, inosine, succinate, acetate, an Λ/-acetyl glycoprotein, or a vitamin B metabolite such as Λ/-methyl-4-pyπdone-3-carboxamιde (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND), as a ratio with at least one other metabolite, for example to account for bulk mass differences between samples In this way, the amount of any one or more of a mammalian-microbial co-metabolite, alanine, glycine, creatinine, glutamate, taurine, formate, inosine, succinate, acetate, an Λ/-acetyl glycoprotein, or a vitamin B metabolite such as Λ/-methyl-4-pyrιdone-3-carboxamιde (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND), is normalised Preferably, the amount of any one or more of a mammalian-microbial co-metabolite, alanine, glycine, creatinine, glutamate, taurine, formate, inosine, succinate, acetate, an Λ/-acetyl glycoprotein, or a vitamin B metabolite such as Λ/-methyl-4-pyrιdone-3-carboxamιde (4PY), Λ/-methyl nicotinic acid (NMNA) or Λ/-methyl nicotinamide (NMND) is expressed as a ratio with at least one other metabolite, typically in the same sample, whose excretion is relatively stable (millimoles/day/kg body weight) For example, the concentration of creatinine in urine is relatively constant and so the amount of bacterial metabolite conjugated to a sulfur containing moiety may be expressed as a ratio with creatinine. Other ways of normalising to account for changes in dilution of samples that may be used in the present invention include lyophilisation and reconstitution into a specific volume, or for spectroscopic data, normalising to the total sum of the residual spectra.

In another embodiment, the at least one further biological parameter to be assessed is at least one protein assessed in the same or different sample. For example, interleukin-13 or cysteine deoxygenase may be assessed in a sample taken from the individual.

The data produced from carrying out the methods of the invention may conveniently be recorded on a data carrier. Thus, the invention includes a method of recording data on whether an individual has, or has an increased risk of developing, autism spectrum disorder (ASD) by using any of the methods of the first aspect of the invention and recording the results on a data carrier. Similarly, the invention includes a method of recording data on whether an individual is one who has, or has an increased risk of developing, a pathological condition by using any of the methods of the third aspect of the invention and recording the results on a data carrier. Typically, the data are recorded in an electronic form and the data carrier may be a computer, a disk drive, a memory stick, a CD or DVD or floppy disk or the like.

Information recorded on the data carrier may include the name, date of birth, age, sex, height and body mass index of the individual or individuals.

The invention will now be described in more details with the aid of the following Figures and Examples. Figure 1. (A) Urine 600-MHz ¹ H NMR spectroscopy median spectra of control, sibling and autistic individuals. Key: 4PY ^* , tentatively assigned as N-methyl-4-pyridone-3- carboxamide; DMA, dimethylamine; DMG, dimethylglycine; NMNA, N-methyl nicotinic acid; NMND, N-methyl nicotinamide; TMA, trimethylamine; TMAO, trimethylamine N- oxide; PAG, phenylacetylglutamine; PCS, p-cresol sulfate;

Figure 2. (A) PCA scores plot of the first 2 components from the normalized UV-scaled NMR data and (B) the corresponding PLS-DA cross-validated scores plot for all three group. PLS-DA cross-validated scores plot of pairwise comparison between (C) controls versus siblings and (D) controls versus autistics.

Figure 3. O-PLS-DA coefficients plot showing differences in urinary profiles between (A) controls and autistics, and (B) controls and siblings.

Example 1: Urinary metabolic phenotvpes of autistic and non-autistic siblings and age-matched normal children Summary

Autism is an early-onset developmental disorder with severe life long impact on behavior and social functioning. Here, the urinary metabolic profile of individuals diagnosed with autism (n=42), together with their non-autistic siblings (n=36) and age-matched healthy volunteers (n=34) have been characterized for the first time using NMR spectroscopy and pattern recognition methods. Clear metabolic phenotype (metabotype) differences were observed between autistic and controls, which were associated with perturbation in urinary mammalian-microbial co-metabolites including dimethylamine, hippurate and phenyacetylglutamine (PAG).

Levels of urinary amino acids alanine and glycine were also changed in autism indicating perturbation in amino acid metabolism. Additional alterations in nicotinic acid metabolism was also observed with autistic individuals showing changes in urinary N- methyl-4-pyridone-3-carboxamide (4PY), Λ/-methyl nicotinic acid (NMNA), and Λ/-methyl nicotinamide (NMND). This study demonstrates the potential of NMR-based metabolic profiling as a top-down systems biology driver for investigating the mechanistic basis of autism and aids identification of metabolites, which could be used as diagnostic biomarkers of autism. Introduction

Autism spectrum disorders (ASD) represent a series of related highly complex socio- psychological and neurodevelopmental problems with associated metabolic and gastrointestinal abnormalities of poorly-defined etiology. ASD typically develop during the first 3 years of life and are characterized by a myriad of deficits in language/communication skills, and socially detached as well as repetitive and stereotypic behaviours. [1 ,2] The etiopathology of ASD is multi-factorial and has been linked to genetic abnormalities such as fragile X syndrome, inborn errors of metabolism but there are many possible, but largely ill-defined, triggers including infectious agents and environmental toxins. Various abnormalities in immunological function and gastrointestinal disturbances are also involved in ASD though their etiological significance is unknown. [3-6] Recent studies have suggested the involvement of gastrointestinal microflora in gut dysfunction of ASD individuals. [7] The gastrointestinal microbiota has been implicated because repeated multiple courses of antibiotic therapies are common in ASD individuals and the stability and composition of their microbiota are uncertain. [8] Repeated antimicrobial treatment has been shown to create an environment that is favorable for colonization of toxin-producing microflora through the disruption of protective commensal microflora. [7, 9] Abnormalities of the gut microbiota including Clostridium sp. a group of bacteria with a wide range of species . variation in host hostility have also been noted in autistic children. Many species of Clostridia can produce powerful neurotoxins, and these have been implicated in ASDs with certain Clostridial species being found to be specific to autistic individuals (not present in normal children). [10] The faecal flora of ASD individuals have been found to contain higher numbers of Clostridium histolyticum group of bacteria as compared to healthy children and their non-autistic siblings were found to have intermediate levels of the same group of bacteria indicating the importance of gastrointestinal microbiota compositions in ASD. [8] Individuals with ASD has also been shown to have abnormal sulfur metabolism. [11] Waring et al showed that individuals with autism have lower levels of plasma sulfate but higher levels of urinary sulfate as compared to normal individuals, indicating that autistic individuals suffer from a deficiency in detoxification involving sulfation as evidenced by their inability to sulfate paracetamol[11]. We have recently shown that even in normal individuals high levels of microbial 4-cresol can compromise sulfation of Tylenol (paracetamol) at therapeutic doses, and further that N-acetyl cysteinyl and cysteinyl adducts of paracetamol are lower in urines of people with high levels of pre-dose 4-cresol sulfate suggesting that a generalised depletion of sulfur metabolism (including the glutathione pathway) and a consequent inability to detoxify reactive intermediates.

The diagnosis of ASD has improved tremendously over the years from a prevalence of 4 in 10 000 children before 1980 [2] to 53 in 10 000 in 2006 [12] but this varies regionally and with ethnicity some areas have much higher incidences of ASD. However, it is not clear whether the increase is due to higher prevalence of the disorder, improved awareness by clinicians, improved early detection/diagnosis or a combination of all the above. Current diagnosis of ASD is subjective and depends on observations of a cluster of behaviors and fulfillment of several criteria set out in the Diagnostic and Statistical Manual of Mental Disorders 4th edition (DSM-IV) for the diagnosis of ASD by trained clinician. [13] At present, there are no reliable biochemical- or genetic-screening tests and in some cases such as late onset autism, normal development might switch to delay in acquisition of new skills adding to the difficulty for diagnosing ASD. Thus, there is pressing need for a reliable diagnostic tool for ASD that is both sensitive and reliable since early diagnosis usually leads to timely interventions and optimise outcomes.

Metabonomic technologies offer the possibility of measuring the metabolic end points (metabolic profiles) that are determined by host genetic and environmental factors, and enable possible identification of metabolites that could be indicative of ASD. [14] These high throughput metabolic profiling methods using high resolution spectroscopic platforms (NMR and/or mass spectrometry (MS)) with subsequent multivariate statistical analyses is a well-established strategy for differential metabolic pathway profiling and disease diagnosis.[15-19] The aim of the current study is to use nuclear magnetic resonance (NMR) spectroscopy-based metabonomics strategy to metabolically profile individuals with autism and a healthy comparison group in population-based samples to characterize and understand the metabolic phenotype and pathophysiology of ASD individuals respectively as well as to identify urinary metabolites, which can be used to study and assist diagnosis of childhood ASDs.

Methods

Sample Collection.

The urine samples used in this study were obtained from University of South Australia and Swiss Tropical Institute, and complied with Australian and Swiss local and national regulations on ethics. Each individual and their family gave informed consent for the study to take place. Autistic patients were diagnosed according to the DSM-IV criteria for diagnosis of autism or Asperger syndrome.[13]

Sample Preparation.

Urine samples were prepared by mixing 400 μl of urine with 220 μl of a phosphate buffer (90% D ₂ O, 1 mM 3-trimethylsilyl-1-[2,2,3,3- ² H4] propionate (TSP), and 3 mM sodium azide; pH 7.4) and left to stand for 10 min. The samples were centrifuged at 11000 g for 10 min and 600 μl of the supematants were then transferred into 5 mm (outer diameter) NMR tubes.

¹ H NMR Spectroscopy.

Spectra were obtained on a Bruker DRX600 spectrometer (Bruker Biospin; Rheinstetten, Germany) at 600.13 MHz (ambient probe temperature 27°C). A standard 1 -dimensional (1 D) pulse sequence was used [recycle delay (RD)-90°-f ₁ -90 ^o -f _m -90°-acquire free induction decay (FID)]. The water signal was suppressed by irradiation during RD of 2 s, and mixing time (t _m ) of 150 ms. t^ was set to 3 μs and the 90° pulse length was adjusted to ~10 μs. For each sample, a total of 128 transients were accumulated into -32,000 data points using a spectral width of 20 ppm. Prior to Fourier transformation, all FIDs were multiplied by an exponential function equivalent to a line broadening of 0.3 Hz. The assignment of the peaks to specific metabolites was based on 2-dimensional (2D) ¹ H- ¹ H correlation spectroscopy (COSY) and ¹ H- ¹ H total correlation spectroscopy (TOCSY) NMR, published literature, [20,21] and statistical total correlation spectroscopy (STOCSY). [22]

Data Processing and Analysis.

¹ H NMR spectra of urine samples were manually phased and baseline corrected using XwinNMR 3.5 (Bruker Biospin; Rheinstetten, Germany). The ¹ H NMR spectra were referenced to the TSP resonance at δ 0.0. The spectra were digitized using a MATLAB (version 7, The Mathworks Inc.; Natwick, MA, USA) script developed in-house. The regions containing the water and urea resonances were removed from each spectrum to eliminate baseline effects of imperfect water saturation. For each spectrum, recursive peak alignment algorithm [23] was applied to minimize spectral peak shift due to residual pH differences within samples prior to normalization to the total sum of the residual spectrum and pattern recognition analyses.

Principal component analysis (PCA) was applied to the unit variance (UV)-scaled spectral data to reveal intrinsic illness-related patterns within the data. Projection to latent structure discriminant analysis (PLS-DA) with UV-scaled spectral data was also carried out in order to improve classification of the different groups of individuals as well as to identify changes in urinary metabolites that are unique to a particular group. Permutation testing (200 permutations) was performed on the PLS-DA models to ensure statistical model validity. Inter-individual variation can confound data interpretation, particularly in multivariate data of high dimensionality. Therefore, orthogonal-projection to latent structure discriminant analysis (O-PLS-DA) [24] was performed on UV-scaled spectral data in a MATLAB environment to optimally model class differences and to systematically identify metabolites contributing to the differences between autistic, siblings and control groups. The O-PLS-DA method decomposes the variation in X (NMR data) into 3 parts; the first being the variation in X related to Y (the class variable), and the last 2 containing the specific systemic variation in X and residual, respectively. The contribution of each metabolite to sample classification is interpreted using the O-PLS coefficients with back-scaling transformation. [25] Here, the colors projected onto the spectrum indicate the correlation of the metabolites discriminating autistic and siblings from the corresponding controls. Red indicates a high correlation and blue denotes no correlation with sample class. The direction and magnitude of the signals relate to the covariation of the metabolites with the classes in the model. A coefficient of 0.22, corresponding to 5% significance level, was used as a cutoff value to select variables that had a significant correlation with class and the model predictive performance (robustness) was evaluated using a 7-fold cross validation method. [25]

Results Urinary ¹ H NMR spectroscopic profiles of the cohort.

Typical urinary 600 MHz ¹ H NMR median spectra of individuals from the 3 groups (Figure 1) consist of a wide range of low-molecular-weight metabolites of diverse chemical classes (typically <1 kDa) from both mammalian and associated gut-microbial metabolism. The urinary NMR spectra are dominated by dietary and microbial-derived methylamines (dimethylamine (DMA), trimethylamine Λ/-oxide (TMAO)) and phenolics such as hippurate and phenylacetylglutamine (PAG) (Figure 1). Additionally, mammalian metabolites such as citrate, succinate, creatinine, lactate, α-hydroxyisobutyrate and amino acids (alanine, glutamine and glycine) are also excreted. Visually, the urinary spectra were very similar with the autistic individuals showing subtle differences in urinary succinate, Λ/-methyl nicotinic acid (NMNA) and N-methyl nicotinamide (NMND) as compared to the controls (Figure 1). The NMR aromatic spectral region of both the ASDs and siblings are very similar to those of the controls. Metabolic differences within the cohort. To further examine the metabolic differences between the three groups of individuals, pattern recognition analyses were employed on the data set to extract useful metabolic information. PCA was carried out on the UV-scaled data to identify any inherent differences within the dataset. The resulting scores plot (Figure 2A) showed no differences between all three groups. However, by utilizing group information in PLS-DA analysis (Q ² (goodness of prediction): 23%; R ² (goodness of fit): 17%), clear differences can be observed in the three groups. The corresponding cross-validate PLS-DA scores plot (Figure 2B) showed clear separation between autistic individuals and the controls and partial separation between siblings and the controls. PLS-DA pair-wise comparison between controls and siblings (Figure 2C; Q ² : 36.2%, R ² : 16.8%), and between controls and autistics (Figure 2D; Q ² : 36.2%, R ² : 12.1%) showed clear separation between the two groups, with both siblings and autistics clearly separated from the controls in the cross-validate scores plot (Figure 2C and D) respectively on the first component. O-PLS-DA analyses were carried out on the full NMR data set consisting of 34 controls, 36 siblings and 42 autistic urine samples. Systematic comparisons of the autistics and normal control groups using O-PLS-DA generated a model with goodness of fit, R ² , of 15% and goodness of prediction, Q ² , of 30%. The corresponding coefficients plot (Figure 3A) indicated differences in the urinary metabolic profile between the two groups with the autistics showing higher levels of urinary acetate, DMA, Λ/-acetyl glycoproteins (NAG), glycine, succinate, alanine, taurine, formate, inosine, NMNA, NMND and Λ/-methyl-4-pyridone-3-carboxamide (4PY), whilst the healthy controls showing higher levels of urinary glutamate, hippurate and PAG. Comparison between siblings and the controls using O-PLS-DA (Q ² : 19%; R ² : 17%) showed some differences in the urinary metabolic profile (Figure 3B) although the changes observed were subtler as compared to the effects observed in the autistic versus control model. The coefficient plots (Figure 3B) showed that the siblings have higher levels of urinary creatinine, glycine and DMA whereas the controls have higher levels of urinary hippurate and PAG. Discussion

The results of this study have shown clear and significant differences in the urinary metabolic profiles between children with autism, their normal siblings and unrelated normal control. Pattern recognition of the urine NMR data indicated that children with autism and their normal siblings have lower levels of urinary hippurate and PAG, and higher levels of urinary glycine and DMA. More importantly, this study showed that children with autism are very different metabolically as compared to normal controls and pattern recognition analyses indicated changes in mammalian-microbial co-metabolites such as DMA, hippurate and PAG, vitamin B metabolism such as 4PY, NMNA and NMND, amino acids such as alanine and glycine, NAG, succinate and glutamate Gut microbial variations

Changes in urinary aromatics such as hippurate and phenylacetylglutamine suggested the involvement of gut microflora in autism, since their precursors, benzoic acid, phenylacetic acid and p-hydroxyphenylacetate, are produced by bacterial metabolism in the intestine [26-29] Hippurate is predominantly formed by hepatic glycine conjugation of gut microbial-derived benzoate, which is produced from plant phenolics [27] Protein and aromatic amino acids such as phenylalanine, tyrosine and tryptophan are also catabohzed by host gut microbiota to form PAG and PCS [28,29] Gut microbiota has been shown to facilitate host energy recovery from dietary sources by providing refined control mechanisms on energy recovery through catabolism of otherwise poorly digestible nutrients i e resistant starch [30] Changes in urinary hippurate and PAG have been previously linked to the effects of xenobiotics on the intestinal microfloral metabolism [31] Depletion of hippurate and PAG has been previously reported after ingestion of antibiotics such as vancomycin, [9] and gentamicin [32] In addition, gastrointestinal microflora analysis of fecal samples from children with late-onset autism has been shown to have higher counts as well as increased diversity of Clostridial species [8,10] and individuals with autism has been shown to have abnormalities in hippurate excretion [33] However, further metagenomic study on the faecal samples of these children will be required to confirm the gut microbiota species variation as postulated in this study The changes in urinary mammalian-microbial co-metabolites, observed in the current study, highlighted the potential involvement of gut microflora in autism and warrant further investigation of the effect of gut microbe metabolites on early brain function development Biochemical changes in autism

In the present study, urinary ammo acids alanine and glycine, were found to be elevated in autistic individuals as compared to the controls Several studies have shown that children with ASD often suffer from dysregulated amino acid metabolism [34,35] In addition, Rolf et al reported that platelet serotonin was significantly increased and amino acids aspartic acid, glutamine, glutamic acid and γ-aminobutyπc acid were significantly decreased in autistic individuals as compared to the controls [36] This is in agreement with our current finding where urinary glutamate was found to be lower in autistic group. The increase in urinary NAG in autistic group could indicate the involvement of autoimmunity in autism since NAG is a marker of immune response [37] and autoimmunity has been implicated in autistic children. [38] 4PY, NMNA and NMND, metabolites of nicotinic acid metabolism, were found to be increase in individuals with autism. These metabolites are end product of nicotinamide metabolism, which is derived from nicotinic acid (vitamin B3). Nicotinamide is the amide derivative of nicotinic acid, which is involved in the tryptophan-NAD pathway that supplies pyridine nucleotides to the liver. Nicotinamide is metabolized to NMND via nicotinamide N-methyltransferase and subsequently to 2PY and 4PY via the action of aldehyde oxidase. [39] In addition, both 4PY and NMNA has been suggested as potential markers of peroxisome proliferation. [40] The changes in urinary 4PY, NMNA and NMND observed in the current study suggested possible perturbation of nicotinic acid metabolism in autistic individuals. Conclusion

This work has demonstrated the potential of NMR-based metabonomics to aid understanding and has provided insights into the metabolic changes that are associated with autism. Our metabolic profiling approach indicated changes in gut microbiota metabolism and amino acid metabolism in children with autism. A panel of urinary metabolites such as hippurate, NMNA and glycine were found to change, and so are potential markers for autism.

References

1. Angley M. (2007) Australian Family Physician Vol. 36, No. 9.

2. Johnson, C.P., Myers, S.M. (2007) Pediatrics 120, 1183-1215.

3. Kidd P. M. (2002) Alternative Medicine Review\/o\ 7, Number 4.

4. Chugani et al (1999) Prog. Neuro-Psycopharmacol & Biol Psychiat 23, 635-641. 5. Ashwood et al Journal of Clinical Immunology Vol. 23, No. 6, November 2003.

6. Levy et al (2007) Biol Psychiatry 61 : 492-497.

7. Bolte, E. R. (1998) Autism and Clostridium tetani. Med Hypotheses 51, 133-144.

8. Parracho et al (2005) Journal of Medical Microbiology 54, 987-991.

9. Yap IK, Li JV, Saric J, et al (2008) J. Proteome Res. 7, 3718-3728.

10. Finegold, S. M., Molitoris, D., Song, Y. & 20 other authors (2002) Gastrointestinal microflora studies in late-onset autism. Clin Infect Dis 35 (Suppl. 1), S6-S16. 1 1. Waring R. H., Klovrza L.V. (2000) "Sulphur Metabolism in Autism". Journal of Nutritional and Environmental Medicine 10, 25-32.

12. Gurney, J.G. et al (2006) Arch. Pediatr. Adolesc. Med. 160, 825-830.

13. American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th edn text revised ed. Washington DC: APA, 2000.

14. Nicholson JK ₁ Connelly J ₁ Lindon JC, Holmes E (2002) Nat Rev Drug Discov 1 :

153-161.

15. Nicholson JK, Lindon JC, Holmes E (1999) Xenobiotica 29: 1181-1189.

16. Yap IKS, Clayton TA, Tang H ₁ Everett JR, Hanton G, Provost JP, Le Net JL, Charuel C ₁ Lindon JC, Nicholson JK. (2006) J. Proteome Res. 5(10), 2675-2684.

17. Skordi E, Yap IKS, Claus SP, Martin FP, Cloarec O, Lindberg J, Schuppe- Koistinen I ₁ Holmes E, Nicholson JK. (2007) J. Proteome Res. 6(12), 4572-4581.

18. Martin FP, Wang Y, Sprenger N ₁ Yap IKS, Lundstedt T, Lek P, Rezzi S, Ramadan Z ₁ van Bladeren P, Fay LB, Kochhar S, Lindon JC, Holmes E, Nicholson JK. (2008) MoI. Syst. Biol. 4, 157.

19. Marchesi J. R. et al (2007) J. Proteome Res. 6(2), 546 -551.

20. Nicholson, J. K.; Foxall, P. J.; Spraul, M.; Farrant, R. D.; Lindon, J. C. (1995) Anal. Chem. 67, 793-811.

21. Fan, T. W. M. (1996) Prog. Nucl. Magn. Reson. Spectrosc. 28, 161-219.

22. Cloarec, O.; Dumas, M. E.; Craig, A.; Barton, R. H.; Trygg, J.;Hudson, J.;

Blancher, C; Gauguier, D.; Lindon, J. C; Holmes, E.; Nicholson, J. K. (2005) Anal. Chem. 77, 1282-1289.

23. Kirill A. Veselkov, John C. Lindon, Timothy M. D. Ebbels, Derek Crockford, Vladimir V. Volynkin, Elaine Holmes, David B. Davies and Jeremy K. Nicholson (2009) Anal. Chem. 81 , 56-66.

24. Trygg, J. (2002) J. Chemom. 16, 283-293.

25. Cloarec, O.; Dumas, M. E.; Trygg, J.; Craig, A.; Barton, R. H.; Lindon, J. C;

Nicholson, J. K.; Holmes, E. (2005) Anal. Chem. 77, 517-526.

26. Nicholson JK, Holmes E, Wilson ID (2005) Nat Rev. Microbiol. 3, 431-438.

27. Schwab AJ, Tao L, Yoshimura T et al (2001 ) Am. J. Physiol. Gastrointest. Liver Physiol. 280, G1124-G1136.

28. Diaz E ₁ Ferra ^' ndez A, Prieta MA, Garcia J (2001) Microbiol. MoI. Biol. Rev. 65,

523-569.

29. Selmer T, Andrei Pl Eur. (2001 ) J. Biochem. 268, 1363-1372.

30. Backhed F, Ding H, Wang T, et al (2004) Proc. Natl. Acad. Sci. U.S.A. 101 ,

15718-15723. Nicholls, A W , Mortishire-Smith, R J , Nicholson, J K (2003) Chem Res Toxicol 16, 1395-1404

Lenz, E M , Bright, J , Knight, R , Westwood, F R , Davies, D , Major, H , Wilson, I D (2005) Biomarkers 10, 173-187

Lis AW, Mclaughlin I, Mclaughlin RK, Lis EW, Stubbs EG (1976) CIm Chem 22, 1528-1532

Evans C, Dunstan HR, Rothkirch T, Roberts TK, Reichelt KL, Cosford R, Deed G, Ellis LB, Sparkes DL (2008) Nutr Neurosci 11 , 9-17

Aldred S, Moore KM ₁ Fitzgerald M, Waring RH (2003) J Autism Dev Disord 33, 93-97

Rolf LH, Haarmann FY, Grotemeyer KH, Kehrer H (1993) Acta Psychiatr Scand 87, 312-316

Stadnyk, A , Gauldie, J (1991 ) Immunol Today 12, A7-12

Mostafa GA, El-Sayed ZA, El-Aziz MM, El-Sayed MF (2008) J Child Neurol 23, 1413-1418

Wolf, H (1974) Scandinavian Journal of Clinical and Laboratory Investigation Supplementum 136, 1-186

Ringeissen, Stephanie, Connor, Susan C , Brown, H Roger, Sweatman, Brian C , Hodson, Mark P , Kenny, Steve P , Haworth, Richard I , Mcgill, Paul, Price, Mark A , Aylott, Mike C , Nunez, Derek J , Haselden, John N and Waterfield, Catherine

J (2003) Biomarkers 8, 3, 240-271