Development and reproducibility of a novel high-performance liquid-chromatography monolithic column...

Research Article

Received: 27 November 2008, Revised: 10 February 2009, Accepted: 12 February 2009 Published online in Wiley Interscience: 29 April 2009

(www.interscience.wiley.com) DOI 10.1002/bmc.1231

Copyright © 2009 John Wiley & Sons, Ltd. Biomed. Chromatogr. 2009; 23: 1108–1115

11

08

John Wiley & Sons, Ltd.

Development and reproducibility of a novel high-performance liquid-chromatography monolithic column method for the detection and quantification of trans-indolyl-3-acryloylglycine in human urine

Kevin Carr,a Paul Whiteleya and Paul Shattocka,b

ABSTRACT: Elevated levels of trans-indolyl-3-acryloylglycine (IAcrGly) have been reported in the urine of people with variousconditions including pervasive developmental disorders (PDDs) such as autism and Asperger syndrome. Reversed-phasehigh-performance liquid chromatography with ultra-violet detection using traditional particle silica-based columns subse-quent to solid-phase extraction (SPE) has been the preferred assay method; requiring long analytical run times, high flowrates and high solvent usage. Recent developments in monolithic HPLC column technology facilitated the development of anovel analytical method, for the detection and quantification of urinary IAcrGly. The revised method eliminates the require-ment for SPE pre-treatment, reduces sample run-time and decreases solvent volumes. Five urine samples from people diag-nosed with PDD were run in quadruplicate to test the intra- and inter-day reliability of the new method based on retentiontime, peak area and peak height for IAcrGly. Detection was by UV with IAcrGly confirmation by MS/MS-MS. Relative standarddeviations showed significant improvement with the new method for all parameters. The new method represents a majoradvancement in the detection and quantification of IAcrGly by reducing time and cost of analysis whilst improving detectionlimits and reproducibility. Copyright © 2009 John Wiley & Sons, Ltd.

Keywords: high-performance liquid-chromatography; mass spectrometry; monolithic column; urine; indole; tryptophan; autism

Introduction

Trans-indolyl-3-acryloylglycine (IAcrGly) (Fig. 1) is believed to be

a metabolite of the indole amino acid tryptophan and is found

in the urine of most people at differing levels. Concentrations

can be affected not only by existing medical conditions but also

by exposure to sunlight and dietary variables (Marklova, 1999;

Whiteley et al., 1999). Abnormal levels of IAcrGly have been

reported in a number of conditions including the photoder-

matitic condition, polymorphous light eruption (Marklova et al.,1975) and Hartnup disease (Jepson, 1965). Based on differing

symptom phenotypes, various levels of IAcrGly have been

detected in the urine of people with pervasive developmental

disorder (PDD) (Shattock and Whiteley, 2002; Alcorn et al., 2004),

although the precise relationship to the syndrome is unknown.

PDD includes a number of heterogeneous conditions including

autism (Kanner, 1943) and Asperger syndrome (Asperger, 1944).

It is characterized by a triad of cognitive and behavioural impair-

ments including problems with language and reciprocal social

interaction and the presence of a restricted repertoire of activi-

ties (World Health Organization, 1992).

Analytical methods used for the detection and identification of

IAcrGly in biological fluids have principally relied on the use of

reversed-phase high-performance liquid-chromatography (RP-

HPLC) with ultra-violet (UV) detection (Mills et al., 1998; Ander-

son et al., 2002). A gradient RP-HPLC method has also been used

for the detection of small-chain polypeptides and related

material in urine from people with PDD (Solaas et al., 2002).

Whilst urine analysis by RP-HPLC is used by a number of laboratories

Figure 1. Structure of indolyl-3-acryloylglycine (IAcrGly).

* Correspondence to: K. Carr, Autism Research Unit, Department of Pharmacy,

Health & Well-being, School of Applied Sciences, University of Sunderland,

Sunderland SR1 3SD, UK. E-mail: [email protected]

a Autism Research Unit, Department of Pharmacy, Health and Well-being, Fac-

ulty of Applied Sciences, University of Sunderland, Sunderland SR1 3SD, UK

b Education and Services for People with Autism (ESPA), 2A Hylton Park Road,

Sunderland SR5 3HD, UK

Abbreviations used: IAcrGly, trans-indolyl-3-acryloylglycine; PDDs, perva-

sive developmental disorders; TFA, trifluoroacetic acid.

Contract/grant sponsor: The Robert Luff Foundation.

Detection and quantification of trans-indolyl-3-acryloylglycine in human urine

Biomed. Chromatogr. 2009; 23: 1108–1115 Copyright © 2009 John Wiley & Sons, Ltd. www.interscience.wiley.com/journal/bmc

11

09

as a research tool for investigations into PDD (Whiteley and

Shattock, 2003; Bull et al., 2003), there remain several issues

regarding reproducibility, length of analysis time, high flow rates

and the volume of solvent used in the existing method.

The present method (Anderson et al., 2002) relies on the use

of solid-phase extraction (SPE) of urine samples prior to HPLC

analysis. SPE removes particulates from the sample which would

block traditional silica particle packed columns. Monolithic col-

umns, by comparison, are composed of a straight rod of highly

porous silica with a bimodal pore structure. The material has a

silica-gel skeleton containing mesopores with diameters of

approximately 13 nm and macropores with diameters of approx-

imately 2 μm. The macropores allow easy solvent flow with low

back pressure whilst the mesopores provide a large surface area

for compound retention (Novakova et al., 2004).

We report on the development of a novel gradient RP-HPLC

method using a monolithic column with direct injection for the

analysis of IAcrGly in urine samples. Intra- and inter-day reliabil-

ity for retention time, peak area and peak height is reported

based on UV detection accompanied by mass spectrometry

(MS) confirmation.

Experimental

Chemicals and Reagents

Deionized water was supplied in-house by an Elga Purelab

Option (Elga, Buckinghamshire, UK) producing water at a con-

ductivity >18 MΩ. Methanol, acetonitrile, isopropyl alcohol

(used for line cleaning), trifluoroacetic acid (TFA) and formic acid

used in this study were all HPLC grade (Sigma, UK). Thymol, used

as an anti-bacterial for urine samples, was also supplied by

Sigma.

Collection of Urine Samples

A fasting morning mid-stream urine sample was collected in a

30 mL universal tube (NELS, UK) containing a small amount of

thymol. Raw urine samples and SPE fractions were stored over-

night at −20°C.

Solid-phase Extraction Sample Preparation for Existing Method

Sample fractions were extracted by vacuum elution using an IST

VacMaster using Varian 5 mL 200 mg C18 Bond Elut LRC SPE car-

tridges (Phenomenex, UK). SPE cartridges were preconditioned

with 2 mL of methanol followed by 2 mL deionized water and

4 mL 0.1% (v/v) aqueous TFA. A 5 mL sample was vacuum

extracted and washed with 2 mL of 0.1% (v/v) aqueous TFA fol-

lowed by 1 mL of a mixture of 10% acetonitrile containing 0.1%

TFA: 90% of 0.1% (v/v) aq. TFA. Final sample collection was made

in aliquots of 1 mL of 40% acetonitrile containing 0.1% TFA: 60%

of 0.1% (v/v) aq. TFA.

HPLC Conditions

All HPLC experiments were performed using an Agilent 1100 LC

with a vacuum degasser (G1379A), a quaternary pump (G1311A),

an autosampler (G1313A), a thermostated column compartment

(G1316A) and a diode array detector (DAD) (G1315B). Analysis

was performed using Agilent Chemstation v. 10 (Agilent Tech-

nologies UK Ltd, Berkshire, UK). Mass spectrometry was carried

out on a PE Sciex API 2000 triple quadrupole with ESI TurboIon-

Spray. MS analysis was performed using Analyst v. 1.4.1 (Applied

Biosystems, Warrington, UK). Both the C18 Jupiter column and

the Onyx column had very low levels of use prior to the start of

the comparison test.

Existing reversed-phase gradient high performance liquid

chromatography. The existing method comprised RP-HPLC

carried out on an integrated Agilent 1100 HPLC as described

previously, using a C18 Jupiter (Phenomenex, UK) column

(25 cm × 4.6 mm i.d., 300 Å pore diameter, 5 μm particle diame-

ter). The column temperature was internally regulated at 27°C.

Primary detection of IAcrGly was by UV absorbance (0.1 min

peak width, 4 nm slit) at 326 nm (reference wavelengths

360 ± 100 nm). Mobile phase A was aq. 0.1% TFA and mobile

phase B was 0.1% TFA in acetonitrile. A gradient of 5–50% (v/v)

mobile phase B over 8–40 min at a flow-rate 2 mL/min was used

for analysis followed by a continued gradient of 50–80% (v/v)

mobile phase B (40–55 min) and a post-run time (5 min) back to

original gradient conditions. Sample vials (2 mL; Agilent, UK)

containing 1 mL of total SPE urine sample were injected (10 μL)

by autosampler.

Modified method reversed-phase gradient high perfor-

mance liquid chromatography. The revised method was per-

formed using the same HPLC system with a C18 Onyx monolithic

(Phenomenex, UK) column (10 cm × 3 mm i.d.). The column tem-

perature was regulated at 27°C. Detection was as previously

described. The acid modifier in the mobile phase was changed

from TFA to formic acid to aid MS detection. Mobile phase A was

aqueous 0.1% formic acid and mobile phase B was 0.1% formic

acid in acetonitrile. A gradient method was used as shown in

Table 1. Sample vials containing 1 mL of untreated urine, diluted

1:1 with deionized water (>18 MΩ) were loaded onto the system

and 10 μL of sample injected by autosampler.

Mass Spectrometer Conditions

The output solvent from the DAD was fed directly into the MS

without the need for splitting. The source block temperature

was set at 500°C in positive ion mode with a capillary voltage

5000 eV. Nitrogen was used as a nebulizer gas (45 psi), auxiliary

gas (60 psi), collision gas (7 psi), and curtain gas (50 psi). Mass

parameters for all analysis were (Q1, 100–1000 amu), product

ion (Q3, 10–250 amu), declustering potential (DP, 40 eV),

Table 1. Table showing gradient method used for Onyx

column analysis

Time

(min)

% Water + 0.1%

formic acid

% Acetonitrile + 0.1%

formic acid

Flow

(mL/min)

0.00 99.00 1.00 0.40

1.00 99.00 1.00 0.40

20.00 70.00 30.00 0.40

25.00 20.00 80.00 0.40

25.50 10.00 90.00 1.50

27.50 99.00 1.00 1.00

30.00 99.00 1.00 0.40

31.00 99.00 1.00 0.40

K. Carr et al.

www.interscience.wiley.com/journal/bmc Copyright © 2009 John Wiley & Sons, Ltd. Biomed. Chromatogr. 2009; 23: 1108–1115

11

10

entrance potential (EP, 400 eV), cell exit potential (CXP, 10 eV),

collision energy (CE, 45 eV) and collision gas (CG, 3 psi).

Procedure

Because of the photo-instability of indolyl-3-acryloylglycine all

experiments were performed in a laboratory with no natural

light and reduced levels of artificial light. All urine preparation

was performed as quickly as possible with samples being kept in

the dark wherever possible. Sample vials for the HPLC were

made of brown glass.

Five urine samples, from people diagnosed with PDD, were

randomly selected from a sample bank. A proportion of each

sample was prepared by SPE. A separate portion was allowed to

settle and 1 mL of the top fraction was removed via pipette and

diluted 1:1 with deionized water. The samples treated by SPE

were analysed by HPLC with both the Jupiter and Onyx columns,

based on the relevant methods described. The diluted untreated

urine samples were analysed using the Onyx column only. The

five samples were each run in quadruplicate and RSD% calcu-

lated from the results (Tables 2 and 3). The same five samples

(both SPE and untreated) were run on three separate days on

the onyx column; the SPE samples were run for just one day on

the Jupiter column. Both the inter-day and intra-day RSD% were

calculated. Confirmation of the IAcrGly peak was made by MS

(positive parent ion) and MS-MS (relevant daughter ions).

To determine linearity a set of IAcrGly standards (2, 3, 4, 5, 6, 7,

8, 9 and 10 μg/mL) were prepared in deionized water and analysed

with the Onyx column and associated method. The integrated

area under the curve results was plotted. The concentrations of

IAcrGly used for the test were chosen because they represent

the levels of IAcrGly routinely found in urine by our laboratory.

To calculate recovery, the same IAcrGly concentrations used in

the linearity test were prepared in aliquots of a urine sample and

analysed as described previously. After subtracting the amount

of IAcrGly originally contained in the urine, the results were used

to plot a graph of IAcrGly detected against IAcrGly added.

The protocols for the collection and analysis of urine samples

were carried out in accordance with the Declaration of Helsinki

and have been ratified by the University of Sunderland. Parents

Table 2. Table showing inter-day RSD% for Onyx and Jupiter columns for SPE treated and diluted untreated urine samples

SPE samples Diluted neat urine samples

Onyx column RSD% Jupiter column Onyx column RSD%

Day 1 Day 2 Day 3 Day 1 Day 1 Day 2 Day 3

Sample 1Time 0.09 0.05 0.01 0.13 0.11 0.05 0.12

Area 1.67 3.04 2.67 26.58 4.82 0.91 3.15

Height 1.78 3.32 3.22 23.25 3.71 0.3 2.77

% Total area 1.76 2.25 2.65 22.01 2.64 8.52 2.3

% Total height 1.22 1.27 3.92 15.61 4.52 4.76 0.61

Sample 2Time 0.04 0.05 0.03 0.05 0.39 0.04 0.09

Area 0.68 1.89 6.86 25.71 4.23 0.32 3.44

Height 1.9 1.4 7.07 22.08 3.79 0.25 2.42

% Total area 1.39 1.31 3.81 23.96 3.02 0.63 10.63

% Total height 0.15 0.72 1.32 12.57 1.49 3.79 5.51

Sample 3Time 0.09 0.06 0.01 0.02 0.1 0.07 0.13

Area 2.94 3.19 2.33 18.23 3.42 0.07 1.84

Height 3.75 3.77 3.19 19.03 2.97 1.12 0.9

% Total area 4.97 5.06 2.68 23.96 2.66 0.81 5.69

% Total height 3.57 3.46 2.66 23.96 2.07 1.06 1.02

Sample 4Time 0.07 0.02 0.04 0.02 0.14 0.06 0.12

Area 1.65 1.62 5.15 28.77 4.11 0.28 3.68

Height 1.86 1.95 5.33 25.35 4.26 0.11 3.38

% Total area 1 1.11 1.1 23.96 2.81 2.52 2.11

% Total height 1.83 1.58 1.46 23.96 0.52 0.94 1.94

Sample 5Time 0.02 0.05 0.05 0.04 0.03 0.08 0.06

Area 1.16 1.22 1.5 23.18 5.11 3.17 3.54

height 0.8 2.44 1.3 21.99 3.09 1.88 3.41

Height 2.28 4.79 5.8 23.96 2.26 2.15 0.97

% Total height 3.02 4.39 0.81 23.96 3.62 3.42 1.26



11

11

of all participants provided consent for the supply of urine sam-

ples as stipulated by the Ethics Committee of the University of

Sunderland.

Results

Limit of detection (LOD) and limit of quantification (LOQ) were

assessed using IAcrGly standards produced in house, as per

Anderson et al. (2002). Standards were dissolved in deionized

water at concentrations of between 50 ng/mL and 2.5 mg/mL

and analysed as per the Onyx procedure (2.4.2) with MS detec-

tion. HPLC gave one major peak at 17.55 ± 0.1 min which corre-

sponded to a peak on the mass spectrometer of 245.10 amu at

an m/z of 17.71 ± 0.1 min (Fig. 2); this matches with an M + 1 of

IAcrGly (244.1 amu). Analysis of results from the standards of

IAcrGly gave an LOD (S/N = 3) of 75 ng/mL and an LOQ (S/

N = 10) of 250 ng/mL, with respect to peak area by diode array

detection.

The analysis was also performed by MS/MS of the main peak

at 245.1 amu. This produced two daughter ions at 170.2 and

142.1 amu and corresponded to expected results, as shown in

Fig. 3.

To ensure assay precision, within-day repeatability (n = 4) and

between-day repeatability (n = 3) were assessed for the 10 sam-

ples (five SPE samples and five untreated samples) using the

Onyx monolithic column. The results are summarized in Tables 2

and 3. Intra-day RSDs ranged from 0.01 to 0.39% for retention

time and from 0.07 to 6.86% for peak area. Inter-day repeatabil-

ity of the SPE samples gave an RSD for retention time of 0.06–

0.15% and an RSD of 6.51–7.95% for peak area. The inter-day

repeatability of the untreated samples gave an RSD for retention

time of 0.09–0.15% and an RSD of 5.46–7.92% for peak area. RSD

was also calculated for peak height and percentage IAcrGly as a

ratio of total chromatogram peak area and height (Tables 2 and

3). Confirmation of IAcrGly was performed by MS/MS with the

same results as per standards.

The five SPE samples were also analysed with the Jupiter

5 μm C18 packed column. This produced a within-day RSD for

retention time of between 0.02 and 0.13% and an RSD of

18.23–28.77% for peak area. The results are summarized in

Table 2. Analysis with the Jupiter column was not performed

between days.

The proposed analytical method was validated with regard to

linearity using a set of nine IAcrGly standards (2–10 μg/mL). The

Table 3. Table showing intra-day RSD% for Onyx column for SPE treated and diluted

untreated urine samples

SPE samples RSD% Diluted neat urine samples RSD%

Sample 1Time 0.15 0.13

Area 6.51 7.92

Height 6.29 6.52

% Total area 7.13 9.15

% Total height 7.29 9.26


Area 7.54 6.48

Height 6.94 5.34




Area 7.95 5.46

height 6.49 4.69




Area 6.74 5.69

Height 6.59 6.25




Area 6.54 6.94

Height 6.98 5.49



K. Carr et al.


11

12

calibration curve was constructed and fitted to a linear equa-

tion. Each point of the calibration curve is the average of three

peak–area measurements. The method showed satisfactory lin-

earity with a determination coefficient of 0.9979 (Fig. 5).

The linear equation constructed from the standards was used

to determine the amount of IAcrGly in relation to the HPLC inte-

grated area under the curve. To calculate the percentage recov-

ery, the same set of nine IAcrGly standards (2–10 μg/mL) were

prepared in aliquots of a urine sample and refrigerated (4°C) for

48 h. These were then analysed by the same method as the stan-

dards. The amount of IAcrGly originally contained in the urine

sample was subtracted from all the results. Using the linear

equation produced from the standards, the amount of IAcrGly

detected was determined for each of the IAcrGly concentrations.

These results were plotted against the concentration of IAcrGly

added.

The graph produced a slope of 0.9950, demonstrating good

correlation between the amount of IAcrGly added and the

amount detected, with an error of less than 1%. Linearity was

good with a determination coefficient of 0.9983 (Fig. 6).

Discussion

Results from our experiments showed that the Onyx column and

revised method performed significantly better than the existing

method (Anderson et al., 2002). RSD for peak area and height

showed on average an improvement by a factor of 10 with the

new method compared with the existing one. Other advantages

include the elimination of the SPE clean-up procedure, reduc–

tion in HPLC analysis time and reduced solvent usage. The new

method can also be applied directly to mass spectrometry fol-

lowing changes to the acid modifier from TFA to formic acid.

Total analysis time has been reduced from 60 to 31 min with the

IAcrGly peak appearing at 17.5 min for the new method com-

pared with 21 min for the existing method (Fig. 4).

Results showed excellent linearity between the concentrations

of 2 and 10 μg/mL for the new Onyx method and a recovery rate

in excess of 99%. Although this method was devised purely for

the analysis of IAcrGly, removal of the SPE clean-up allows for

the analysis of other compounds which may have been lost

by the SPE method and possibly retained by the Jupiter column.

Figure 2. Plot of MS Total Ion Count (TIC) from IAcrGly standard.



11

13

Whilst the new HPLC method shows excellent levels of detec-

tion, quantification of IAcrGly in human urine relating to actual

levels of IAcrGly present in the body is still problematic. Initial

studies of creatinine, the normal reference to determine urine

concentration, not being a reliable standard for quantification

of urinary compounds from samples from people with PDD

(Whiteley et al., 2006), require further confirmation. Studies by our

laboratory are on-going to confirm if creatinine is a viable refer-

ence or if other compounds may produce more accurate results.

It is intended that this new method of HPLC detection and

quantification of IAcrGly will be used in studies to determine lev-

els of IAcrGly in people displaying differing characteristics of

PDD alongside the potential effects that diet, sunlight and

accompanying disease states may have on its production.

Figure 3. MS/MS plot of IAcrGly parent ion (245.1 amu).

Figure 4. Plots showing urine analysis via Jupiter column and Onyx column at 326 nm (IAcrGly highlighted).

K. Carr et al.


11

14

Figure 5. Graph plotting IAcrGly standards analysed by HPLC using Onyx column.

Figure 6. Graph plotting amount of IAcrGly detected against amount of IAcrGly added.



11

15

The new method represents a major advancement in the

detection and quantification of IAcrGly by reducing the time

and cost of analysis whilst improving detection limits and

reproducibility.

Acknowledgements

The authors gratefully acknowledge the financial support pro-

vided by The Robert Luff Foundation.

References

Alcorn A, Berney T, Bretherton K, Mills M, Savery D and Shattock P.Urinary compounds in autism. Journal of Intellect Disability Research2004; 48: 274–278.

Anderson RJ, Bendell DJ, Garnett I, Groundwater PW, Lough WJ, Mills MJ,Savery D and Shattock PE. Identification of indolyl-3-acryloylglycine inthe urine of people with autism. Journal of Pharmacy andPharmacology 2002; 54: 295–298.

Asperger H. Die autistischen Psychopathen. Kindesalter Archiv fürPsychiatrie und Nervenkrankheiten 1944; 117: 76–136.

Bull G, Shattock P, Whiteley P, Anderson R, Groundwater PW, Lough WJand Lees G. Indolyl-3-acryloylglycine (IAG) is a putative diagnosticurinary marker for autism spectrum disorders. Medical ScienceMonitoring 2003; 9: CR2–CR5.

Jepson JB. In The Metabolic Basis of Inherited Disease, 2nd edn, StanburyJB, Wyngaarden JB and Frederickson DS (eds). McGraw-Hill: New York,1965; 1283–1299.

Kanner L. Autistic disturbances of affective contact. Nervous Children1943; 2: 217–250.

Marklova E. Where does indolylacrylic acid come from? Amino Acids1999; 17: 401–413.

Marklova E, Malina L and Hais IM. Urinary excretion of indolyl-3-acryloylglycine in some skin affections. Clinica Chimica Acta 1975; 64:273–280.

Mills MJ, Savery D and Shattock PEG. Rapid analysis of low levels ofindolyl-3-acryloylglycine in human urine by high-performance liquidchromatography. Journal of Chromatography B 1998; 712: 51–58.

Novakova L, Matysova L, Solichova M, Koupparis MA and Soliich P.Comparison of performance of C18 monolithic rod columns andconventional C18 particle-packed columns in liquid chromatographicdetermination of Estrogel and Ketoprofen gel. Journal ofChromatography B 2004; 813: 191–197.

Shattock P and Whiteley P. Biochemical aspects in autism spectrumdisorders: updating the opioid-excess theory and presenting newopportunities for biomedical intervention. Expert Opinion onTherapeutic Targets 2002; 6: 175–183.

Solaas KM, Skjeldal O, Gardner MLG, Kase BF and Reichelt KL. Urinarypeptides in Rett syndrome. Autism 2002; 6: 315–329.

Whiteley P, Rodgers J, Savery D and Shattock P. A gluten-free diet as anintervention for autism and associated spectrum disorders:preliminary findings. Autism 1999; 3: 45–65.

Whiteley P and Shattock P. What makes trans-indolyl-3-acrloylglycine(IAcrGly) identified by high-performance liquid-chromatography(HPLC) relevant to pervasive developmental disorders? Journal ofNutrition and Environmental Medicine 2003; 13: 1–7.

Whiteley P, Waring R, Williams L, Klovrza L, Nolan F, Smith S, Farrow M,Dodou K, Lough WJ and Shattock P. Spot urinary creatinine excretionin pervasive developmental disorders. Pediatrics International 2006;48: 292–297.

World Health Organization. Tenth Revision of the InternationalClassification of Diseases and Related Health Problems. ClinicalDescriptions and Diagnostic Guidelines. WHO: Geneva, 1992.

Development and reproducibility of a novel high-performance liquid-chromatography monolithic column...

Documents

Transcript of Development and reproducibility of a novel high-performance liquid-chromatography monolithic column...