Você está na página 1de 7

Special issue: review

Received: 13 February 2013, Accepted: 5 March 2013 Published online in Wiley Online Library: 22 April 2013

(wileyonlinelibrary.com) DOI 10.1002/bmc.2911

Chromatographic methods in the study

of autism
Ewa Żurawicz, Joanna Kałużna-Czaplińska* and Jacek Rynkowski
ABSTRACT: Research into biomarkers of autism is a new means of medical intervention in this disease. Chromatographic
techniques, especially coupled with mass spectrometry, are widely used in determination of biomarkers and assessment of
effectiveness of autism therapy owing to their sensitivity and selectivity. Among the chromatographic techniques gas
chromatography and liquid chromatography, especially high-performance liquid chromatography, have found application
in clinical trials. The high-performance liquid chromatography technique allows an analysis of liquid samples with a wide
range of molecules, small and large, providing an opportunity to perform advanced assays within a short time frame.
Gas chromatography with the appropriate preparation of samples (gaseous and liquid) and a selection of analysis conditions
enables the separation of thermally stable, volatile and non-volatile organic substances in short runtimes. The chromatographic
techniques that are currently used in metabolic studies in autism are designed to identify abnormalities in three areas: the
metabolism of neurotransmitters, nutritional and metabolic status and manifestations of oxidative stress. This review presents
a necessary theoretical introduction and examples of applications of chromatographic studies of disorder markers in autism.
Copyright © 2013 John Wiley & Sons, Ltd.

Keywords: autism; chromatography; HPLC; GC-MS; biomarker

Introduction allows an analysis of various substances. Clinical analysis is based

on the fact that metabolic changes associated with specific dis-
According to the Diagnostic and Statistical Manual of Mental orders can cause changes in the biochemical profile of a specific
Disorders, autistic disorder is a pervasive developmental disorder body fluid (Fig. 1), so when the disease is suspected, specific
and is characterized by social, verbal and nonverbal communi- parameters are tested in certain body fluids (e.g. serum, plasma,
cation abnormalities. Manifestations of autism typically occur whole blood, and urine; Lindon et al., 2007). Blood, coming into
before the age of 36 months (American Psychiatric Association, contact with all organs of the body, is a source of various
2000). The report presented by the Autism and Developmental biomarkers, therefore it has the longest track record for use in
Disabilities Monitoring Network for 2008 presents the average modern diagnostic procedures. It is possible to see the whole
prevalence of autism as 11.3 per 1000 (1 in 88) children aged spectrum of different biomarkers, but this makes the biomarkers
8, and the diagnosis is more frequent in males than females, found in blood less specific and generally they may not be con-
with ratios from 3.7:1.0 to 7.2:1.0 (Centers for Disease Control fined to emanating from a single organ. Urine and cerebral spi-
and Prevention, 2012). The etiology of autism remains unknown, nal fluid (CSF) are very amenable to biomarker analysis. CSF is
although interactions of genes and environmental factors have quite specific to the central nervous system (CNS). CSF samples
been suggested (Hertz-Picciotto et al., 2006). are taken by means of a lumbar puncture (LP), which may not
The use of chromatographic techniques and metabolomics in be possible for all patients with a CNS disease, and it is generally
the study of autism is a result of the broad spectrum of this dis- not applicable. Urine seems a more suitable biofluid than blood
ease, which goes far beyond the purely developmental disorder. or CFS because it can be obtained in large quantities by non-
A number of different disorders associated with autism have invasive sampling. Urine is also more organ-specific than blood.
been found, with metabolic disorders taking an important place It can be obtained in large volumes (deciliter to liter amounts)
(Table 1). The study of abnormal accumulation or deficit of and can be collected over a period of time, even offering the
specific metabolites in a defined pathway can provide clues into
relevant candidate genes and/or environmental exposure in
autism. Such knowledge can also be helpful in targeted inter-
vention strategies to restore metabolic balance and potentially * Correspondence to: Joanna Kałużna-Czaplińska, Institute of General and
improve the symptoms of autism (Melnyk et al., 2012). A more Ecological Chemistry, Department of Chemistry, Lodz University of Tech-
nology, Zeromskiego 116, 90–924 Lodz, Poland. E-mail: jkaluzna@p.lodz.pl
complete understanding of metabolic conditions is possible
owing to the studies of biomarkers. According to International Institute of General and Ecological Chemistry, Department of Chemistry,
Union of Pure and Applied Chemistry nomenclature, a bio- Lodz University of Technology, Zeromskiego 116, 90-924 Lodz, Poland
marker is an indicator signaling an event or condition in a bio-
logical system or sample and giving a measure of exposure, Abbreviations used: 5-HIAA, 5-hydroxyindoleacetic acid; AA, amino acids;
CFS, cerebral spinal fluid; CNS, central nervous system; DA, D-arabinitol;
effect or susceptibility (Duffus, 1993). Because of the possibility IAG, indolyl-3-acryloylglycine; LA, L-arabinitol; LP, lumbar puncture; PUFA,
of using a large number of metabolites as biomarkers, in clinical plasma polyunsaturated fatty acid; SAH, S-adenosylhomocysteine; SAM,

analysis it is extremely important to adapt a technique that S-adenosylmethionine; SLOS, Smith–Lemli–Opitz Syndrome.

Biomed. Chromatogr. 2013; 27: 1273–1279 Copyright © 2013 John Wiley & Sons, Ltd.
E. Żurawicz et al.

Table 1. Conditions related to autism

Psychiatric, behavior and Metabolic disorders Genetic disorders

neurodevelopmental disorders
Attention-deficit hyperactivity disorder Abormalities of purine metabolism Fragile X syndrome (Zafeiriou
(Yates and LeCouteur, 2009) (Celani, 2003) et al., 2007)
Tourette syndrome/tic disorder Creatine deficiency (Zecavati and Tuberous sclerosis complex (Zafeiriou
(Yates and LeCouteur, 2009) Spence, 2009) et al., 2007)
Dyspraxia/developmental coordination Biotinidase deficiency (Zecavati and Down syndrome (Zafeiriou
disorder (Yates and LeCouteur, 2009) Spence, 2009) et al., 2007)
Dyslexia (Yates and LeCouteur, 2009) Succinic-semialdehyde dehydrogenase Angelman syndrome (Zafeiriou
deficiency (Zecavati and Spence, 2009) et al., 2007)
Obsessive–compulsive disorder Cerebral folate deficiency (Zecavati and Neurofibromatosis type 1 (Zafeiriou
(Yates and LeCouteur, 2009) Spence, 2009) et al., 2007; Celani, 2003)
Specific phobias (Yates and Smith–Lemli–Opitz syndrome (Zecavati Phenylketonuria (Zafeiriou et al.,
LeCouteur, 2009) and Spence, 2009) 2007; Celani, 2003)
Anxiety (Yates and LeCouteur, 2009) Infantile ceroid lipofuscinosis (Zecavati Aristaless Related Homeobox
and Spence, 2009) Syndrome (Zafeiriou et al., 2007)
Depression/mood disorder (Yates and Histidinemia (Zecavati and Spence, 2009) Charge, Goldenhar and Moebius
LeCouteur, 2009) syndromes (Zafeiriou et al., 2007)
Sleeping difficulties (Yates and Urea cycle defects (Zecavati and Chromosome 2q37 deletion
LeCouteur, 2009; Williams et al., 2004) Spence, 2009) syndrome (Zafeiriou et al., 2007)
Feeding difficulties (Yates and Gastrointestinal Candida overgrowth Cohen syndrome (Zafeiriou et al., 2007)
LeCouteur, 2009) (Shaw et al., 2010)
Anorexia nervosa (Zafeiriou et al., 2007) Mitochondrial disorders (Zafeiriou
et al., 2007)

ability to collect multiple samples within 24 h. Urine analysis the use of chromatography, which is the most powerful, flexible
allows a very detailed monitoring of the patient’s therapeutic and simple analytical technique in separation science, especially
response (Good et al., 2007). in separation of complex organic mixtures (Bruner, 1993).
A very important task of the clinical analysis involves separa- Through the use of chromatography, it is possible to carry out
tion of species in complex matrices, which are the body fluids a simultaneous separation and quantitative analysis of the
(Varcoe, 2001). The problem of separation can be solved through substance. Separation efficiency is related to the type of the
chromatographic column (hydrophobic/hydrophilic) and solvent
system and column pressure. An appropriate selection of chro-
matographic techniques for the analysis is dependent on the
nature and amount of sample, the objective of the separation
and the limitations of available time, equipment and expertise.
Simple economical and effective chromatographic techniques
such as paper chromatography and thin layer chromatography,
but also higher-sensitivity and more expensive high-performance
liquid chromatography (HPLC) and gas chromatography (GC) find
broad application in clinical investigations (Rao et al., 2009). Gas or
liquid chromatography helps in the analysis of various categories
of metabolites such as volatiles, steroids, steroids, organic acids
and amino acids as well as similar compound types (Agilent
Technologies I2, 2007) from physiological fluids (urine, serum,
amniotic fluid, cerebral spinal fluid, etc.). Providing selectivity and
sensitivity, which are necessary for clinical trials, chromatographic
techniques contribute to the success of the analytical process
(Vékey et al., 2007). The application of clinical research with
the use of chromatographic techniques with subsequent multi-
variate statistical analyses provides a well-established strategy
for differential metabolic pathway profiling and chemical con-
stituents that may serve as markers for the presence of disease
or its absence.
Gas chromatography is a powerful tool for separating volatile
and thermally stable compounds. It can also be suitable for
compounds of low volatility or those with the polar functional

Figure 1. The metabolic profile as reflection of the organism state. group after chemical derivatization. Derivatization performed

wileyonlinelibrary.com/journal/bmc Copyright © 2013 John Wiley & Sons, Ltd. Biomed. Chromatogr. 2013; 27: 1273–1279
Chromatographic methods in the study of autism

by alteration of functional groups provides increased sample (metabolic profiles), which are determined by host genetic
volatility, improved selectivity, chromatographic efficiency and and environmental factors. Figure 3 shows the stages of work
enhanced detection. Combination of GC with mass spectrome- in the study of metabolic profiles. Metabolic profiling, or
ter (GC-MS) is recommended as a technique for metabolic metabolomics, is the study of low-molecular weight metabo-
profiling, mainly owing to its high sensitivity and specificity lites and their intermediates found in biological fluid samples.
and access to extensive mass spectral databases. GC-MS instru- A metabolic profile illustrates the dynamics of the biological
mentation is available for many clinical laboratories, although system as a response to genetic modification (e.g. transgenic
GC-MS analysis is relatively expensive in terms of equipment, or viral) as well as physiological (e.g. gender), pathophysiological
personnel and time required (Kwong, 2002). (e.g. disease morbidity) and/or developmental stimuli (e.g. aging;
High-performance liquid chromatography (sometimes referred Clarke and Haselden, 2008). Metabolic profiles can provide
to as high-pressure liquid chromatography) is the chromato- information about mitochondrial diseases, neurotransmitter
graphic technique most commonly used in the studies of autism metabolites, nutritional and metabolic status, oxidative stress
markers. As opposed to GC, liquid chromatography enables an and environmental impact.
analysis of polar and thermally labile compounds (Agilent Finding interesting metabolites requires a search for statis-
Technologies I2, 2007). Using a photodiode array detector or mass tically significant variations in abundance within a set of
spectrometer detector coupled with HPLC it is possible to obtain experimental and control samples. The next step in the
structural information, which is essential in the identification of study of metabolic profiles is identification, which means
unknown compounds (Kwong, 2002). For each HPLC system the determination of the chemical structure of these metabolites
achievement of high resolution at high flow rates is characteristic. after profiling and finally interpretation, which involves find-
Important advantages of HPLC are better resolution, reproducibil- ing connections between the metabolites discovered and
ity and excellent within- and between-run precision (Gooding and the biological processes or conditions. Statistically significant
Regnier, 2002). variations can only be seen by analyzing large numbers of
Appropriate sample preparation is essential for the successful samples, which is an additional challenge in the case of
chromatographic analysis of specimens which include biological autism research. The principal chromatographic technologies
samples such as blood and urine. Body fluids in particular need (GC and HPLC) that have been developed to study metabolic
to be processed prior to analysis in order to remove any interfer- profiles are usually based on spectroscopy and mass spec-
ing compounds that could cause either contamination or block- trometry (MS).
age of the column (especially blood), giving inaccurate results.
Sample preparation can involve pH adjustment, ultrafiltration,
enzymatic reactions, extraction methods like liquid–liquid ex-
traction (solvent extraction), solid-phase extraction, solid-phase
microextraction, and a very important step in the GC-MS analy-
sis, derivatization (Hempel, 2004). According to the reports,
blood is the most common body fluid in the chromatographic
studies (Fig. 2), despite its complex matrix.
This review presents a necessary theoretical introduction and
examples of applications of chromatographic studies of disorder
markers in autism.

The metabolic profile, a link between the

biological system and diagnosis
With well-prepared samples of body fluids, chromatographic
techniques offer a possibility of measuring metabolic end points Figure 3. The scheme of work in the study of metabolic profiles.

Figure 2. Diagram of report frequency (2005–2012) using chromatography in analysis of biological fluids. Bibliographic research related to the use of

chromatography in blood, plasma, urine and cerebrospinal fluid analysis (PubMed). A total of 64,697 reports were found.

Biomed. Chromatogr. 2013; 27: 1273–1279 Copyright © 2013 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/bmc
E. Żurawicz et al.

Chromatographic analysis in autism healthy ones were found, which indicates a reduced risk of
free radical disease. In younger autistic subjects, similarly high
Neurotransmitter metabolites plasma vitamin concentrations to those in older cases were
observed. The results indicate that consumption of fruit and
An analysis of catecholamines can be a problem because of their
vegetables in autism was optimal in this study group. The
presence at low concentrations, their susceptibility to oxidation
results also showed that average plasma concentrations of
and the fact that they are unstable at basic pH. The use of HPLC
vitamins E and A in autistics were below threshold values,
allows the determination of neurotransmitters and the levels of
which indicates lower consumption of food rich in vitamins A
their metabolites in platelets, blood and CSF in autistic subjects
and E (Krajcovicova-Kudlackova et al., 2009). Sensitive liquid
(Israngkun et al., 1986; Moretti et al., 2005; Coutinho et al.,
chromatography with tandem mass spectrometry detection
2007). In the case of 13 autistic subjects under study the levels
(LC-MS/MS) made it possible to detect significant changes in
of plasma dopamine, norepinephrine, and epinephrine concen-
leukocyte sulfur amino acids metabolism in autistic children.
trations were significantly higher compared to controls, also
Leukocytes from autistic children contained significantly lower
whole blood serotonin, tested on the same subjects, was found
concentrations of S-adenosylmethionine (SAM), and elevated
to be significantly higher in autistic subjects (Israngkun et al.,
levels of intracellular homocysteine content. Additionally, the
1986). With HPLC analysis high urinary serotonin and low
levels of intracellular total cysteine and glutathione were
whole blood serotonin in autism and age effects on the levels
reduced. The results pointed to altered metabolism in immune
of serotonin were found (Hérault et al., 1996). Further
cells and involvement of inflammation in autism (Suh et al.,
studies showed that urinary excretion of serotonin metabolite
2008). The next study provides an application of LC-MS for
5-hydroxyindoleacetic acid (5-HIAA) in autistic patients under
the analysis of plasma amino acids (AA) levels in autistic
study was similar to that reported for healthy individuals. An
children. Seven of the 20 measured AA were significantly
analysis of 5-HIAA in the urine of 20 autistics was made using
decreased in autistic children: glutamine, threonine, aspara-
high-performance liquid chromatography with fluorescence
gine, citrulline, serine, tyrosine and leucine. Glutamate was
detection (HPLC-FLD; Mulder et al., 2005). The results of GC-MS
the only AA for which the levels were higher in the autistic
analysis of urine demonstrate significant differences in the level
group compared with the control group. An age-dependent
of homovanillic acid and vanillylmandelic acid between 20
decrease in plasma glutamate as well as aspartic acid was
autistic and 36 healthy children. These results explain some
observed only in healthy children. Effects of gender, ethnicity,
disruptions of functioning of the dopaminergic system, such as
time of blood draw, medications, cognitive function on AA
repeated behaviors, mood disorders, aggression attacks and
levels within each group or of the diagnosis (i.e. autism vs per-
disorders of social relationships, observed in the case of autistic
vasive developmental disorder – not otherwise specified), as
children (Kałużna-Czaplińska et al., 2010a). Two further studies
well as any associations between clinical scores, namely Autism
applied chromatography to test the relationship between child-
Diagnostic Interview, Revised and Autism Diagnostic Observation
hood and brain opioid activity. In the first study urinary opioid
Schedule, and any AA levels in the autistic group were not
peptides were examined by HPLC coupled with matrix-assisted
observed in this particular study group (Tirouvanziam et al.,
laser desorption ionisation time-of-flight mass spectrometry
2012). Another chromatographic HPLC-FLD analysis of plasma
(Cass et al., 2008), while in the second study they were tested
amino acids was performed in the group of autistic and Asperger
by liquid chromatography coupled with ultraviolet and mass
Syndrome patients, their siblings and parents. In this group six of
spectrometric detection (Hunter et al., 2003). Opioid peptides
the 17 measured amino acids were raised. These were glutamic
were not detected in autistic subjects, which is why these
acid, phenylalanine, asparagine, tyrosine, alanine and lysine
studies are in opposition to the opioid peptide excess theory
(Aldred et al., 2003). In another study of plasma amino acid
for the cause of autism. Another study shows the use of HPLC
analysis was performed by HPLC coupled with tandem mass
with fluorescence detection in the analysis of CSF levels of
spectrometry (MS/MS) technique. Twenty primary amino acids
indoleacetic acid, which is the metabolite of tryptamine, a
(proteinogenic) and 21 secondary amino acids with amino acid
central neuromodulator. There was no significant difference
metabolites were analyzed. Considering the primary amino
in the levels of CSF indoleacetic acid between the autistic
acids, the autistic children had significantly lower levels of tryp-
and healthy groups (Anderson et al., 1988).
tophan and higher levels of glutamate, slightly higher serine,
and slightly lower tyrosine and phenylalanine in plasma. The
analysis of secondary amino acids showed that the autistic
Nutritional and metabolic status
group had a higher level of b-amino isobutyrate and a lower
Feeding habits of children with autism are very specific. These level of taurine in plasma (Adams et al., 2011). GC-MS technique
children may have problems in three areas: food selectivity made it possible to determine the levels of homocysteine in
based on type, color and texture; food refusal; and disruptive the urine of autistic and healthy children. The results showed
mealtime behaviors. A selective, deficient diet can significantly increased levels of homocysteine in the urine of autistic children.
exacerbate the symptoms of the gastro-intestinal tract and The higher level of homocysteine in autistic children may indicate
nervous system. Chromatographic techniques allow the quanti- deficiencies of folic acid and vitamins B6 and B12 in nutrition of
tative determination of both nutrients and their metabolites, these children (Kałużna-Czaplińska et al., 2011a). GC-MS analysis
which in turn makes it possible to set an individual uptake of was used to determine the level of urinary tryptophan in autistic
nutrients and to find metabolic abnormality markers. Selected children. It showed that they had significantly lower urinary
antioxidants were analyzed with HPLC for both autistic and levels of tryptophan compared with healthy children. Defi-
healthy children. The study included vitamins C, E, A, carotenoids ciency in tryptophan can lead to the worsening of autistic
b-carotene and lycopene. In older autistic children significantly symptoms such as mild depression and increased irritability

higher plasma values of vitamin C and b-carotene than in the (Kałużna-Czaplińska et al., 2010b). Two discordant reports show

wileyonlinelibrary.com/journal/bmc Copyright © 2013 John Wiley & Sons, Ltd. Biomed. Chromatogr. 2013; 27: 1273–1279
Chromatographic methods in the study of autism

the use of chromatographic techniques for analysis of indolyl- prepubertal boys. The results showed that the laboratory was able
3-acryloylglycine (IAG) in the urine of autistic subjects. In the to identify correctly the urine of boys with autism in over half of
first study IAG levels were detected with HPLC coupled with the cases. It was also found that the results were markedly
UV detection. The results of analysis showed very significant improved when the analysis was restricted to those of normal abil-
increases in the levels of urinary IAG in the autistic group. ity or mild learning disability (Alcorn et al., 2004). Owing to the
According to the authors, urinary IAG may constitute an objec- chromatographic analysis of plasma sterols, it was possible to de-
tive diagnostic indicator for autism (Bull et al., 2003). On the termine the prevalence of autism in children with Smith–Lemli–
other hand, the results of other studies carried out with the Opitz Syndrome (SLOS). Plasma cholesterol, 7-dehydrocholesterol
use of HPLC-MS/MS suggest that there is no association and 8-dehydrocholesterol were measured by capillary-column
between the presence of IAG in urine and autism, so it can gas chromatography. Most children with SLOS (characterized by
be used neither for diagnosis nor for therapeutic intervention abnormal levels of sterols in plasma) had some variant of autism,
in the diet (Wright et al., 2005). GC-MS was used for the analysis which may mean that SLOS appears to have the most consistent
of urinary dicarboxylic (succinic, adipic and suberic) acids as relationship with autism of any single gene disorder (Sikora
markers of metabolic disorders in children with autism before et al., 2006). Liquid chromatography–electrospray ionization-
and after vitamin B2, vitamin B6 and magnesium supplementa- mass spectrometry (LC-ESI-MS) on time-of-flight instruments
tion. The levels of all acids were significantly higher in the made it possible to discover proteins altered by the disease. A
group of children before supplementation. The reduction of total of 6348 serum peptide components were analyzed. As a
urinary dicarboxylic acids was associated with the improve- result, five peptide components showed the biggest differences
ment of some autistic symptoms such as the ability to concen- between the autistic and typical groups, four of which corre-
trate and make eye contact (Kałużna-Czaplińska et al., 2011b). spond to three proteins involved in the complement pathway
Plasma fatty acid profiles of autistic patients and age-matching (Corbett et al., 2007). GC-MS was also used for the analysis
control subjects were obtained with a gas chromatograph with a of the D-arabinitol, L-arabinitol and D-arabinitol/L-arabinitol
flame ionization detector. Fatty acids in the plasma of Saudi (DA/LA) ratio in the urine of autistic children as a biomarker of
autistic patients were increased, especially in most of the satu- candidiasis. The study included quantification before and after
rated fatty acids except for propionic acid, and decreased in most the probiotic therapy. The results show that the level of DA
of the polyunsaturated fatty acids. The results were related to was significantly higher in the urine of autistic children before
oxidative stress, mitochondrial dysfunction and a high concen- and after probiotic supplementation. The probiotic supplemen-
tration of lead (Pb) previously reported in Saudi autistic patients tation caused a significant decrease in DA and DA/LA and an
(El-Ansary et al., 2011). The aim of the next study was to investi- improvement in the ability for concentration and carrying out
gate brain energy metabolism in autistic children through mea- orders (Kałużna-Czaplińska and Błaszczyk, 2012).
suring plasma polyunsaturated fatty acids (PUFAs), serum
carnitine and plasma lactate, which is an index of mitochondrial
Oxidative stress
function, in relation to autistic severity and clinical manifestations.
The determination of the levels of polyunsaturated fatty acids in Oxidative stress is a general term used to describe an imbalance
plasma was made by gas chromatography. Autistic patients had between the production and manifestation of reactive oxygen
significantly lower plasma PUFAs (except linoleic acid) than con- species and the ability of the biological system to readily detox-
trols. Severely autistic patients had lower values of plasma PUFAs ify the reactive intermediates or to repair the resulting damage.
than patients with mild and moderate autism, but these differences When the normal redox state of tissue is affected, the system
were statistically insignificant. Hypotonic autistic children (20%) can respond through the toxic effect, that is by the production
had significantly lower serum docosahexaenoic acid than of peroxides and free radicals, which damage all components
normotonic autistics (Mostafa et al., 2005). With GC-MS it was pos- of the cell, including proteins, lipids and DNA. The task of
sible to detect a marked increase in analogs of Krebs cycle metab- chromatographic techniques is the identification and quantifica-
olites and arabinose in the urine of two brothers with autistic tion of markers of oxidative stress and antioxidants in the body
features. The metabolites which were elevated were citramalic, fluids and tissues. An example of application of chromatography
tartaric (3-OH-malic), and 3-oxoglutaric acids and compounds ten- in the study of oxidative stress in autism is LC-MS/MS analysis
tatively identified as a citric acid analog and partially identified as a of reduced and oxidized glutathione in plasma. Also, high-
phenylcarboxylic acid. The authors present three explanations for performance liquid chromatography with fluoroscence detec-
such results. The first is that these siblings have a new genetic dis- tion (HPLC-FLD) analysis of plasma taurine was applied to study
ease in which some abnormal metabolites characteristic of yeast the plasma of autistic patients. The results show significantly
metabolism are excreted coincidentally but are not causally decreased levels of the plasma taurine, reduced glutathione
related to autism. The second explanation is that these siblings and increased levels of the transsulfuration metabolite of oxi-
can be infected with one or more yeasts and/or bacteria but this/ dized glutathione. Increased levels of plasma oxidized glutathi-
these infection(s) are secondary to immune deficiency. The third one were correlated with increasing mercury-associated urinary
explanation is the infection of siblings with one or more microor- porphyrins (Geier et al., 2008). HPLC coupled with an electro-
ganisms. These microorganisms produce metabolites that are chemical detector was applied for the analysis of amino acids
causally related to the disease through inhibition of mitochondrial to investigate the dynamics of an integrated metabolic pathway,
Krebs cycle activity by citramalic, carboxycitric, 3-oxoglutaric and essential for cellular antioxidant and methylation capacity.
tartaric acids, and (or) by interference in normal carbohydrate The results show that methionine and SAM were significantly
metabolism caused by high concentrations of arabinose (Shaw decreased, and S-adenosylhomocysteine (SAH) and adenosine
et al., 1995). The aim of another study was to establish whether were significantly increased in the case of autistic children.
autism might have a distinctive chromatographic profile on uri- The SAM/SAH ratio, an indicator of methylation capacity, was

nary analysis. Metabolic profiles were performed with HPLC over decreased in children with autism. In the case of total cysteine,

Biomed. Chromatogr. 2013; 27: 1273–1279 Copyright © 2013 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/bmc
E. Żurawicz et al.

Figure 4. Diagram of report frequency (2005–2012) using chromatography and biomarkers in the study of autism. Bibliographic research related to
the use of chromatography and biomarkers in autism investigations (PubMed). A total of 530 reports were found.

its levels were lower in children with autism. Cystine (was children with autism, which in turn enables an individual diet
considerably elevated in autistic children, and the plasma free and supplementation. The next step in the chromatographic
cysteine/cystine redox ratio was significantly lower. One can studies of the nutritional status in autism is the monitoring of
conclude that the deficit in antioxidant and methylation capacity changes in metabolism during diet and supplementation. A very
is specific to autism and may promote cellular damage and important tool in the diagnosis of autism may also be determin-
altered epigenetic gene expression (Melnyk et al., 2012). ing the influence of environmental factors on children’s develop-
ment. These factors may be compounds that are foreign to a
living organism (e.g. drugs, carcinogens and various compounds
Conclusions and significant predictions for introduced into the environment by artificial means).
the future The most common limitations of chromatographic studies
connected with autism are relatively small and uneven sizes
As autism is not a pure genetic disease, the chromatographic of patients and control groups, lack of matching by gender
analysis of metabolites makes it possible to find the source of and imprecise matching for ethnicity and age. Such limitations
the symptoms observed in this disease. However, the descrip- make the generalization of the results difficult and uncertain.
tion of the human metabolome (a set of all metabolites) will One of the priorities in the investigation of autism should be
be an even more difficult challenge than describing the genome a search for early biomarkers of this disease, and here
and proteome. The reason is the large number of metabolites, chromatographic techniques can play a key role. Rapid identi-
among which there may be one, or a few, that can answer fication of high-risk children would give the opportunity to
the question about the cause of autism. Chromatographic monitor metabolic and neurological changes that predict the
techniques, especially GC-MS, GC-MS/MS, LC-MS, LC-MS/MS beginning of behavioral symptoms of autism, thus making it
and liquid chromatography–nuclear magnetic resonance–mass possible to counter the early symptoms.
spectrometry, are very important tools in clinical chemistry. In
contrast to the conventional biochemical analysis, chromato- References
graphic techniques provide high reliability in trace analysis.
They contribute to the determination of concentrations of a Adams JB, Audhya T, McDonough-Means S, Rubin RA, Quig D, Geis E,
number of organic substances in biological samples, such as Gehn E, Loresto M, Mitchell J, Atwood S, Barnhouse S and Lee W.
Nutritional and metabolic status of children with autism vs.
body fluids, tissues and breath. Coupled techniques provide neurotypical children, and the association with autism severity. Nutri-
specificity and sensitivity of the analysis. Analytical procedures tion and Metabolism 2011; 8: 34; doi: 10.1186/1743-7075-8-34
are rapid, easy and, in most cases, do not require complicated Agilent Technologies I2. Considerations for Selecting GC/MS or LC/MS for
sample pretreatment. There is higher reliability for chromato- Metabolomics. Agilent Technologies Inc., 2007.
Alcorn A, Berney T, Bretherton K, Mills M, Savery D and Shattock P.
graphic determination of trace concentrations in biological Urinary compounds in autism. Journal of Intellectual Disability Research
samples than in classic biochemical investigations. 2004; 48: 274–278.
According to the trend shown in Fig. 4, it can be expected that Aldred S, Moore KM, Fitzgerald M and Waring RH. Plasma amino acid
the importance of chromatography and biomarker studies of levels in children with autism and their families. Journal of Autism
and Developmental Disorders 2003; 33: 93–97.
autism will continue to grow. So far, in the studies of autism,
American Psychiatric Association. Diagnostic and statistical manual
chromatographic techniques have been applied in many areas. of mental disorders. Fourth Edition, Text Revision. American Psy-
An example is the identification and quantitative determination chiatric Association: Washington, DC, 2000; doi: 10.1176/appi.
of markers of neurological and metabolic disorders. With the books.9780890423349
chromatographic techniques, it is possible to determine the Anderson GM, Ross DL, Klykylo W, Feibel FC and Cohen DJ. Cerebrospinal
fluid indoleacetic acid in autistic subjects. Journal of Autism and
progression of the disease and to find relationships between Developmental Disorders 1988; 18: 259–262.
metabolite levels and the age of autistic children. It is becoming Bruner F. Gas Chromatographic Environmental Analysis: Principles, Tech-

increasingly important to determine the nutritional status in niques, Instrumentation, 1st edn. Wiley-VCH: Verlagsgesellschaft, 1993.

wileyonlinelibrary.com/journal/bmc Copyright © 2013 John Wiley & Sons, Ltd. Biomed. Chromatogr. 2013; 27: 1273–1279
Chromatographic methods in the study of autism

Bull G, Shattock P, Whiteley P, Anderson R, Groundwater PW, Lough JW Kałużna-Czaplińska J, Socha E and Rynkowski J. B vitamin supplementa-
and Lees G. Indolyl-3-acryloylglycine (IAG) is a putative diagnostic tion reduces excretion of urinary dicarboxylic acids in autistic
urinary marker for autism spectrum disorders. Medical Science Moni- children. Nutrition Research 2011b; 31: 497–502; doi: 10.1016/j.
tor 2003; 9: 422–425. nutres.2011.06.002
Cass H, Gringras P, March J, McKendrick I, O’Hare AE, Owen L and Pollin C. Krajcovicova-Kudlackova M, Valachovicova M, Mislanova C, Hudecova Z,
Absence of urinary opioid peptides in children with autism. Sustrova M and Ostatnikova D. Plasma concentrations of selected
Archives of Disease in Childhood 2008; 93: 745–750; doi: 10.1136/ antioxidants in autistic children and adolescents. Bratislavské Lekárske
adc.2006.114389 Listy 2009; 110: 247–250.
Celani G. Comorbidity between autistic syndrome and biological pathol- Kwong TC. Toxicology. In McClatchey KD (ed.). Clinical Laboratory Medicine.
ogies: which implications for the understanding of the etiology? Lippincott Williams and Wilkins: New York, 2002.
Journal of Developmental and Physical Disabilities 2003; 15: 141–154; Lindon JC, Nicholson JK and Holmes E. The Handbook of Metabonomics
doi: 10.1023/A:1022875300575 and Metabolomics. Elsevier Science: Oxford, 2007.
Centers for Disease Control and Prevention. Prevalence of autism spec- Melnyk S, Fuchs GJ, Schulz E, Lopez M, Kahler SG, Fussell JJ, Bellando J,
trum disorders – Autism and Developmental Disabilities Monitoring Pavliv O, Rose S, Seidel L, Gaylor DW and James SJ. Metabolic imbal-
Network, 14 sites, United States, 2008. Morbidity and Mortality Weekly ance associated with methylation dysregulation and oxidative dam-
Report 2012; 61: 1–19. age in children with autism. Journal of Autism and Developmental
Clarke ChJ and Haselden JN. Metabolic Profiling as a tool for understand- Disorders 2012; 42: 367–377; doi: 10.1007/s10803-011-1260-7
ing mechanisms of toxicity. Toxicologic Pathology 2008; 36: 140–147; Moretti P, Sahoo T, Hyland K, Bottiglieri T, Peters S, del Gaudio D, Roa B,
doi: 10.1177/0192623307310947 Curry S, Zhu H, Finnell RH, Neul JL, Ramaekers VT, Blau N, Bacino CA,
Corbett BA, Kantor AB, Schulman H, Walker WL, Lit L, Ashwood P, Miller G and Scaglia F. Cerebral folate deficiency with developmen-
Rocke DM and Sharp FR. A proteomic study of serum from chil- tal delay, autism, and response to folinic acid. Neurology 2005; 64:
dren with autism showing differential expression of apolipopro- 1088–1090.
teins and complement proteins. Molecular Psychiatry 2007; 12: Mostafa GA, El-Gamal HA, El-Wakkad ASE, El-Shorbagy OE and Hamza
292–306. MM. Polyunsaturated fatty acids, carnitine and lactate as biological
Coutinho AM, Sousa I, Martins M, Correia C, Morgadinho T, Bento C, markers of brain energy in autistic children. The International Journal
Marques C, Ataíde A, Miguel TS, Moore JH, Oliveira G and Vicente of Child Neuropsychiatry 2005; 2: 179–188.
AM. Evidence for epistasis between SLC6A4 and ITGB3 in autism eti- Mulder EJ, Oosterloo-Duinkerken A, Anderson GM, De Vries EG, Minderaa
ology and in the determination of platelet serotonin levels. Human RB and Kema IP. Automated on-line solid-phase extraction coupled
Genetics 2007; 121: 243–256. with hplc for measurement of 5-hydroxyindole-3-acetic acid in urine.
Duffus JH. Glossary for chemists of terms used in toxicology (IUPAC Clinical Chemistry 2005; 51: 1698–1703.
Recommendations 1993). Pure and Applied Chemistry 1993; 65: Rao AN, Kavitha J, Koch M and Suresh Kumar V. Inborn errors of metab-
2003–2122. olism: review and data from a tertiary care center. Indian Journal of
El-Ansary AK, Ben Bacha AG and Al-Ayahdi LY. Plasma fatty acids as diag- Clinical Biochemistry 2009; 24: 215–222; doi: 10.1007/s12291-009-
nostic markers in autistic patients from Saudi Arabia. Lipids in Health 0041-y
and Disease 2011; 10: 62; doi: 10.1186/1476-511X-10-62 Shaw W, Kassen E and Chaves E. Increased urinary excretion of analogs
Geier DA, Kern JK, Garver CR, Adams JB, Audhya T and Geier MR. A pro- of Krebs cycle metabolites and arabinose in two brothers with autistic
spective study of transsulfuration biomarkers in autistic disorders. features. Clinical Chemistry 1995; 41: 1094–1104.
Neurochemical Research 2008; 34: 386–393; doi: 10.1007/s11064- Shaw W, Baptist J and Geenens D. Immunodeficiency, gastrointestinal
008-9782-x candidiasis, wheat and dairy sensitivity, abnormal urine arabinose,
Good DM, Thongboonkerd V, Novak J, Bascands JL, Schanstra JP, Coon JJ, and autism: a case study. North American Journal of Medical Sciences
Dominiczak A and Mischak H. Body fluid proteomics for biomarker 2010; 3: 1–8.
discovery: lessons from the past hold the key to success in the future. Sikora DM, Pettit-Kekel K, Penfield J, Merkens LS and Steiner RD. The near
Journal of Proteome Research 2007; 6: 4549–4555. universal presence of autism spectrum disorders in children with
Gooding KM and Regnier FE. HPLC of Biological Macromolecules Revised Smith–Lemli–Opitz Syndrome. American Journal of Medical Genetics.
and Expanded, 2nd edn. CRC Press: New York, 2002. Part A 2006; 140A: 1511–1518.
Hempel G. Handbook of Analytical Separations, Vol. 5: Drug Monitoring Suh JH, Walsh WJ, McGinnis WR, Lewis A and Ames BN. Altered sulfur
and Clinical Chemistry. Elsevier: Oxford, 2004. amino acid metabolism in immune cells of children diagnosed with
Hérault J, Petit E, Martineau J, Cherpi C, Perrot A, Barthélémy C, Lelord G autism. American Journal of Biochemistry and Biotechnology 2008; 4:
and Müh JP. Serotonin and autism: biochemical and molecular biol- 105–113.
ogy features. Psychiatry Research 1996; 65: 33–43. Tirouvanziam R, Obukhanych TV, Laval J, Aronov PA, Libove R, Banerjee
Hertz-Picciotto I, Croen LA, Hansen R, Jones CR, van de Water J and AG, Parker KJ, O’Hara R, Herzenberg LA, Herzenberg LA and Hardan
Pessah IN. The CHARGE study: an epidemiologic investigation of AY. Distinct plasma profile of polar neutral amino acids, leucine,
genetic and environmental factors contributing to autism. Environ- and glutamate in children with autism spectrum disorders. Journal
mental Health Perspectives 2006; 114: 1119–1125. of Autism and Developmental Disorders 2012; 42: 827–836; doi:
Hunter LC, O’Hare A, Herron WJ, Fisher LA and Jones GE. Opioid peptides 10.1007/s10803-011-1314-x
and dipeptidyl peptidase in autism. Developmental Medicine and Varcoe JS. Clinical biochemistry. Techniques and Instrumentation. A Practical
Child Neurology 2003; 45: 121–128. Course. Chromatography. World Scientific: Singapore, 2001.
Israngkun PP, Newman HA, Patel ST, Duruibe VA and Abou-Issa H. Vékey K, Telekes A and Vertes A. Medical Applications of Mass Spectrometry.
Potential biochemical markers for infantile autism. Neurochemical Elsevier Science: New York, 2007.
Pathology 1986; 5: 51–70. Williams GP, Sears LL and Allard A. Sleep problems in children with
Kałużna-Czaplińska J and Błaszczyk S. The level of arabinitol in autistic autism. Journal of Sleep Research 2004; 13: 265–268; doi: 10.1111/
children after probiotic therapy. Nutrition 2012; 28: 124–126; doi: j.1365-2869.2004.00405.x
10.1016/j.nut.2011.08.002 Wright B, Brzozowski AM, Calvert E, Farnworth H, Goodall DM, Holbrook I,
Kałużna-Czaplińska J, Socha E and Rynkowski J. Determination of Imrie G, Jordan J, Kelly A, Miles J, Smith R and Town J. Is the presence
homovanillic acid and vanillylmandelic acid in urine of autistic chil- of urinary indolyl-3-acryloylglycine associated with autism spectrum
dren by gas chromatography/mass spectrometry. Medical Science disorder? Developmental Medicine and Child Neurology 2005; 47:
Monitor 2010a; 16: 445–450. 190–192.
Kałużna-Czaplińska J, Michalska M and Rynkowski J. Determination of Yates K and LeCouteur A. Diagnosing autism. Paediatrics and Child Health
tryptophan in urine of autistic and healthy children by gas chroma- 2009; 19: 55–59.
tography/mass spectrometry. Medical Science Monitor 2010b; 16: Zafeiriou DI, Ververi A and Vargiami E. Childhood autism and associated
488–492. comorbidities. Brain & Development 2007; 29: 257–272.
Kałużna-Czaplińska J, Michalska M and Rynkowski J. Homocysteine level Zecavati N and Spence SJ. Neurometabolic disorders and dysfunction in
in urine of autistic and healthy children. Acta Biochimica Polonica autism spectrum disorders. Current Neurology and Neuroscience
2011a; 58: 31–34. Reports 2009; 9: 129–136.

Biomed. Chromatogr. 2013; 27: 1273–1279 Copyright © 2013 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/bmc