Identification of indolyl-3-acryloylglycine in the urine of people with autism

Article abstract of DOI:10.1211/0022357021778349

HPLC analysis of the urine of autistic subjects indicated the presence of an unidentified component in greatly increased concentrations. We have reported the isolation of this component by HPLC and its identification. Mass spectrometry, NMR and UV spectroscopy identified the peak as corresponding to indolyl-3-acryloylglycine (IAG, 3), and this has been confirmed by an independent synthesis.

Full text of DOI:10.1211/0022357021778349

Journal of
JPP 2002, 54: 295–298
# 2002 The Authors
Received July 17, 2001
Accepted September 11, 2001
ISSN 0022-3573
IdentiŽ cation of indolyl-3-acryloylglycine in the urine
of people with autism
Rosaleen J. Anderson, David J. Bendell, Ian Garnett, Paul W.
Groundwater, W. John Lough, Malcolm J. Mills, Dawn Savery
and Paul E. G. Shattock
Abstract
HPLC analysis of the urine of autistic subjects indicated the presence of an unidentiŽed
component in greatly increased concentrations. We have reported the isolation of this
component by HPLC and its identiŽcation. Mass spectrometry, NMR and UV spectroscopy
identiŽed the peak as corresponding to indolyl-3-acryloylglycine (IAG, 3), and this has been
conŽ rmed by an independent synthesis.
Introduction
Autism is a lifelong pervasive developmental disorder that manifests as problems
with communication, social skills and imagination. It aŒects 80000 people in the
UK (data from the National Autistic Society, London 1997) and, at present, can
only be diagnosed by reference to behavioural and psychological characteristics.
For many years there has been interest in the analysis of body ¯ uids from people
with autism in the search for unusual correlates which could be used as markers for
the disorder, and which would provide evidence for a physiological dysfunction
(Panksepp 1979; Gillberg 1988; Adrien et al 1989; Knivsberg et al 1990; Shattock
et al 1990). As a result of an ongoing programme in this laboratory, screening the
urine from people with autism by high-performance liquid chromatography
(HPLC), it was noted that there was a particular peak present in all the
chromatograms from the sample population (Mills et al 1998). This peak was often
present in greatly increased concentrations, based on peak height and area by
HPLC, compared with those found in controls. We have reported the isolation and
structural elucidation of this compound.
Institute of Pharmacy, Chemistry
and Biomedical Sciences,
University of Sunderland,
Sunderland SR1 3SD, UK
Rosaleen J. Anderson, David J.
Bendell, Ian Garnett, Paul W.
Groundwater, W. John Lough
Autism Research Unit, School of
Sciences, University of
Sunderland, Sunderland
SR1 3SD, UK
Malcolm J. Mills, Dawn Savery,
Paul E. G. Shattock
Correspondence: R. J. Anderson,
Institute of Pharmacy, Chemistry
and Biomedical Sciences,
Materials and Methods
Materials
University of Sunderland,
Sunderland SR1 3SD, UK. E-mail:
roz.anderson!sunderland.ac.uk
Chemicals and solvents were obtained from Aldrich-Sigma-Fluka unless indicated
otherwise. UV spectroscopy was carried out on a Shimadzu SP6-AD UV-vis
detector, which was set to 326 nm for HPLC detection. Melting points were
Acknowledgements: We thank
Terry Noctor and Darren Thomas
at Corning Hazelton for some of
the mass spectral data.
1
determined on a Gallenkamp apparatus and are uncorrected. H and 13C NMR
spectra were recorded on a BRUKER AVANCE 300 NMR spectrometer at 300
295

296
Rosaleen J. Anderson et al
}
and 75.5 MHz, respectively, or on a JEOL GSX NMR 243 (M®1) in negative ion mode; UV (0.1% v v
spectrometer at 270 MHz for proton spectra. Tetra- aqueous formic acid or 0.1 m NaOH) 226, 277 and
methylsilane was used as an internal standard and 326 nm; dH (CD3CN) 3.91 (2H, d, J 5.6, CH2), 6.49 (1H,
coupling constants were reported in Hz. The molecular d, J 15.8, ?CHa), 6.69 (1H, br s, NHCO), 7.14 (2H, m,
}
weight was determined using a triple quadruple LC
H-5 and H-6), 7.40 (1H, d, J 5.9, H-7), 7.52 (1H, d, J 3,
}
MS MS PERKIN-ELMER mass spectrometer (Sciex H-2), 7.69 (1H, d, J 15.8, ?CHb), 7.85 (1H, d, J 6.5,
Instruments, UK) in LC-APCI mode. Accurate mass H-4), 11.80 (1H, br s, H-1).
determinations were carried out on a BRUKER APEX
II FTMS (electrospray ionization). Preparative reverse-
phase TLC was carried out on Whatman pre-coated
glass plates, KC18F 60A SiO2, with C18 bonded phase
at a loading of 50-mg sample per plate. Solid-phase
extraction was conducted using C2, C8, and C18 phenyl
and cyanopropyl phases on 1 mL¬100 mg Bond Elut
SPE cartridges from Phenomenex. Several diŒerent
HPLC columns from Phenomenex and mobile phase
systems were used to isolate, purify and characterize the
compound of interest as indolyl-3-acryloylglycine
(IAG). HPLC system 1: 150¬4.6 mm i.d. Nucleosil
aminopropyl (5 lm), mobile phase: acetonitrile:5 mm
Molecular weight determination
The major component extracted from the urine samples
wasfreeze-dried, thenreconstituted in acetonitrile:0.1%
}
}
(v v) aqueous formic acid (18:82, v v) and injected
into the mass spectrometer (Mills et al 1998).
Synthesis of indolyl-3-acryloylglycine methyl ester 2
Indolyl-3-acrylic acid 1 (0.5 g, 2.67 mmol), 2-chloro-
1-methylpyridinium iodide (0.819 g, 3.2 mmol), tri-n-
butylamine (1.98 g, 2.55 mmol) and glycine methyl ester
hydrochloride (0.339 g, 2.7 mmol) were dissolved in dry
dichloromethane (12 cm3) and the reaction mixture was
stirred for 24 h. The organic phase was washed with 1 m
aq. HCl (20 cm3), followed by water (20 cm3), dried
(MgSO4), and the solvent removed under reduced press-
ure. Puri®cation by ¯ ash chromatography on silica,
1
-
ammonium acetate (95:5, v v), ¯ ow rate: 1.0 mL min .
}
HPLC system 2: 150¬4.6 mm i.d. Nucleosil C18 (5 lm)
with a 10¬4.6 mm i.d. Nucleosil (10 lm) guard column,
}
mobile phase: methanol:0.1% (v v) aqueous formic
acid(40:60,v v), ¯ ow rate: 1.5 mL min . HPLC system
1
-
}
3 for LC-MS: 150 mm¬2.1 mm i.d. (5 lm) Nucleosil
}
ODS2 column, mobile phase: 0.1% v v aqueous formic
acid:acetonitrile (82:18, v v), ¯ ow rate: 400 lL min .
1
-
}
}
eluting with ethyl acetate petrol 40±60 (50:50±100:0),
followed by recrystallization from chloroform, gave E-
indolyl-3-acryloylglycine methyl ester 2 as white needle-
like crystals (0.49 g, 71% ); mp 126±127°C [lit. (InhoŒen
et al 1963) mp 161±163°C; isomer not speci®ed].
Isolation of indolyl-3-acryloylglycine from
biological samples for spectroscopy
+
(Found : MH 259.1077. Calc. for C14H15N2O3 : MH,
Urine wasobtained from selected volunteers withautism
who were able to give informed consent. It was passed
through three separate solid-phase extraction Bond Elut
cartridges and lyophilized before reconstitution into
5% aqueous acetonitrile (0.75 mL) and injection onto
HPLC system 1. The compound that gave rise to the
peak of interest was then collected and lyophilized.
+
259.1076. Found: MNa , 281.0897. Calc. for C14H14
N2O3Na: MNa, 281.0896.) mmax 3345 (NH), 1735 (ester
1
-
C?O), 1660 (amide C?O), and 1604 (C?C) cm
;
dH(300 MHz, CDCl3) 3.81 (3H, s, CO2CH3), 4.22 (2H, d,
J 5.2, CH2NH), 6.12 (1H, br t, NHCO), 6.50 (1H, d,
J15.6 ?CHa), 7.26 (2H, m, H-5 and H-6), 7.43 (1H, dd,
J 6.9 and 1.9, H-7), 7.46 (1H, d, J 2.7, H-2), 7.89 (1H, d,
J 15.6,?CHb), 7.90 (1H, dd, J 5.6 and 2.3, H-4), 8.78
(1H, br s, H-1); dC (75.5 MHz, CDCl3) 41.6 (CH3), 41.7
(CH2), 113.0 (quat.), 113.1 (CH), 116.4 (CH), 120.7
(CH), 121.2 (CH), 123.0 (CH), 125.9 (quat.), 131.0 (CH),
134.5 (CH), 138.3 (quat.), 167.6 (C?O), 171.5 (C?O).
}
Once dry, the sample was reconstituted in 5% v v
acetonitrile in 0.1% aqueous formic acid (0.5 mL) and
microcentrifuged for 10 min. The resulting supernatant
was then injected onto HPLC system 2 and the peak of
interest collected. Due to the instability of the compound
in a strongly acidic environment, the sample was buf-
fered with ammonium acetate before a second lyophili-
zation. Any traces of ammonium acetate were removed
via a ®nal solid-phase extraction procedure (using a C18
Synthesis of indolyl-3-acryloylglycine (3)
Bond Elut cartridge) and the product was eluted with Indolyl-3-acryloylglycine methyl ester
acetonitrile to give indolyl-3-acryloylglycine (1.5 mg); 0.36 mmol) was dissolved in phosphate buŒer (20 mm,
2 (100 mg,
m z (LC-APCI MS) 245 (M-1) in positive ion mode, pH 8, 18 cm3) and acetone (2 cm3) and pig liver esterase
}

297
IdentiŽcation of indolyl-3-acryloylglycine in autistic subjects’ urine
(2000 U) was added. The reaction mixture was stirred and 326 nm, were suggestive of a heteroaromatic system
for ®ve days and ®ltered before freeze-drying. Puri®ca- with further conjugation giving rise to a shift to longer
tion by reverse phase TLC, eluting with acetonitrile: wavelength. Treatment with NaOH made no diŒerence
water (1:3), gave indolyl-3-acryloylglycine (3) as a col- to the UV spectrum. Finally, the 1H NMR of the
+
ourless semi-solid (35 mg, 37% ). (Found: MNa , compound, recorded in CD3CN as this had been the
267.0741. Calc. for C13H12N2O3Na: MNa, 267.0740.) solvent used in the chromatography, indicated the pres-
kmax (0.02 m phosphate-borate buŒer with 0.1 m sodium ence of an indole nucleus (indole C2-H, doublet at d7.5,
cholate, pH 9) 232, 280, and 330 nm; dH (300 MHz, d6- J 3 Hz) linked through a trans acrylic acid (aCH?,
DMSO) 3.55 (2H, d, J 4.7, CH2), 6.82 (1H, d, J 15.8, d6.48, J 15.8 Hz; bCH?, d7.69, J 15.8 Hz) to a glycine
?CHa), 7.15 (2H, m, H-5 and H-6), 7.44 (1H, d, J 7.7, unit (CH2, d3.91, J 5.6 Hz). The full structure was
H-7), 7.55 (1H, t, J 5.0, NHCO), 7.57 (1H, d, J 15.8, elucidated using the spectral data obtained and the
?CHb), 7.71 (1H, s, H-2), 8.02 (1H, d, J 7.3, H-4), 11.80 compound identi®ed as indolyl-3-acryloylglycine (3).
(1H, br s, H-1); dC (75.5 MHz, d6-DMSO) 45.2 (CH2),
To con®rm the identity of the compound, we synthe-
113.0 (CH), 113.2 (CH), 118.0 (quat.), 118.5 (CH), 120.9 sized indolyl-3-acryloylglycine independently. The syn-
(CH), 121.1 (CH), 122.5 (CH), 122.8 (CH), 125.9 (quat.), thesis was achieved via the Mukaiyama coupling (Bald
133.0 (quat.), 166.4 (C?O), 174.5 (C?O).
et al 1975) of indoleacrylic acid (1) with glycine methyl
ester to give the ester 2, which yielded the trans isomer
1
exclusively, as con®rmed by the H NMR spectrum.
Comparison of the melting point of the ester 2 with that
reported in the literature (InhoŒen et al 1963) suggested
that the isomer prepared was the cis isomer. Hydrolysis
Results and Discussion
1
Mass spectrometry, H NMR and UV spectroscopy of the ester group gave the carboxylic acid 3 (Figure 1).
were used to identify the compound responsible for this Thishydrolysis was accomplished using pig liveresterase
peak. The mass spectrum, using LC-MS (atmospheric (E.C. 3.1.1.1)due to the sensitivity of indolyl-3-acryloyl-
pressure chemical ionization), indicated a molecular glycine (3) to acid, precluding either acid-catalysed
mass of 244. UV spectra were recorded at high and low hydrolysis or acidi®cation after a base-catalysed re-
}
pH to investigate the presence of acidic and or basic action. Puri®cation by lyophilization, followed by pre-
groups, via bathochromic or hypsochromic shifts. At parative reverse phase TLC gave a sample that was
low pH, the absorptions in the UV spectrum, at 226, 277 spectroscopically and chromatographically identical to
Figure 1 Independent synthesis of indolyl-3-acryloylglycine (3). PLE, pig liver esterase.

298
Rosaleen J. Anderson et al
that originally obtained from the urine of the subjects
with autism. The UV and NMR spectroscopy of syn-
thetic indolyl-3-acryloylglycine were carried out in more
conventional spectroscopic solvents than had been used
to investigate the structure of the unknown compound.
References
Adrien, J. L., Barthelemy, C., Lelord, G., Muh, J. P. (1989) Use of
bioclinical markers for the assessment and treatment of children
Neuropsychobiology
with pervasive developmental disorders.
117±124
22:
}
It is known that indolylacrylic acid (IAcrA) and or
Bald, E., Saigo, K., Mukaiyama, T. (1975) Facile synthesis of carbox-
amides by using 1-methyl-2-halopyridinium iodides as coupling
Chem Lett
. 1163±1166
indolyl-3-acryloylglycine levels are elevated in a number
of disease states, including Hartnup Disease (Milne et al
1960), light sensitive dermatitis (Kimmig et al 1958)
and phenylketonuria (Strohmayer 1964), but levels of
indolyl-3-acryloylglycine had not previously been re-
ported as being elevated in autism. Previous reports had
shown that 71% of the normal population excreted
indolyl-3-acryloylglycine in varying amounts (Mandell
& Rubin 1965; Marklova et al 2000). The source of
indolyl-3-acryloylglycine in autistic subjects is not
clearly understood but has been linked to defects in
tryptophan metabolism (Marklova 1999). It is assumed
that the indolyl-3-acryloylglycine is a metabolic product
from the precursor indolyl-3-acrylic acid, which could
be produced by gut ¯ ora or by endogenous routes (Smith
et al 1968).
reagents.
.
J
Child
.
Gillberg, C. (1988) The neurobiology of infantile autism.
Psychol Psychiatry
.
29: 257±266
InhoŒen, H. H., Nordsiek, K. H., Schafer, H. (1963) Structure de-
N
termination of an indole dye from -[3-(3-indolyl)acryloyl]glycine.
Justus Liebigs Ann Chem
.
. 668: 104±121
Kimmig, J., Sticherlung, W., Tschesche, R., Urbach, H.-G. (1958)
b
Isolation of N-[ (3-indolyl)acryloyl]glycine from urine and its
Hoppe-Seyler’s Z. Physiol. Chem.
synthesis.
311: 234±238
Knivsberg, A.-M., Wiig, K., Lind, G., Nùdland, M., Reichelt, K.-L.
Brain Dysfunction
(1990)Dietaryinterventioninautisticsyndromes.
3: 315±327
Mandell, A. J., Rubin, R. T. (1965) Indolylacryloylglycine excretion
Life Sci
in man.
. 4: 1657±1664
Marklova, E. (1999) Where does indolylacrylic acid come from?
Amino Acids
17: 401±413
Marklova, E., Makovickova, H., Krakorova, I. (2000) Screening for
J Chromatogr A
870: 289±293
defects in tryptophan metabolism.
.
.
Mills, M. J., Savery, D., Shattock, P. E. G. (1998) Rapid analysis of
low levels of indolyl-3-acryloylglycine in human urine by high-
J Chromatogr B
performance liquid chromatography. .
.
712: 51±58
Conclusions
Milne, M. D., Crawford, M. A., Girao, C. B., Loughridge, L. (1960)
Q J Med
The metabolic disorder in Hartnup disease.
Panksepp, J. (1979)A neurochemicaltheory of autism.
sci
.
.
. 29: 407±421
Trends Neuro
-
The compound noted as being present in markedly
increased amounts in the urine of autistic subjects was
isolated by HPLC and was identi®ed as indolyl-3-
acryloylglycine using UV and NMR spectroscopy, and
MS spectrometry. An independent chemical synthesis
of indolyl-3-acryloylglycine con®rmed the structure as
indolyl-3-acryloylglycine as did comparison of the spec-
tral features with the urine-derived indolyl-3-acryloyl-
glycine.
. 2: 174±177
Shattock,P. E. G., Kennedy,A., Rowell, F. J., Berney, T. (1990)Role
of neuropeptides in autism and their relationships with classical
Brain Dysfunction
neurotransmitters.
3: 328±345
Smith, H. G., Smith, W. R. D., Jepson, J. B. (1968) Interconversion
of indolic acids by bacteria and rat tissue. Possible relevance to
Clin Sci
. 34: 333±342
Hartnup disease.
.
Strohmayer, G. (1964) Nachweis von indolylacrylsaure bei der
Arch Klin Exp Dermatol
Phenylketonurie.
.
.
.
. 219: 743±748