Aromatic L-Amino Acid Decarboxylase Deficiency Is a Cause of Long-Fasting Hypoglycemia

CLIN.

CHEM.

38/12,

Aromatic Keith

2405-2410

L-Amino

Hyland’

and

(1992)

Acid Decarboxylase

Peter

Deficiency:

urine,

3-methoxytyrosine,

plasma,

creased

and

and 5-hydroxytryptophan

cerebrospinal

cerebrospinal

fluid together

fluid concentrations

that allow these metabolites to be measured in a single chromatographic run and provide positive identification of 3MT in plasma and urine frorn the patients by gas chromatography/mass spectrometry (GC-MS). The patients, monozygote twins, presented with features of biogenic amine deficiency. Decreased CSF concentrations of homovanillic acid (HVA) and 5-hydroxyindoleacetic acid (5HIAA) together with low wholeblood serotonin and plasma catecholamines confirmed a

in

with

de-

of homovanil-

lic acid and 5-hydroxyindoleacetic acid. In addition, concentrations of vanillactic acid are increased in the urine.

Specific

HPLC

and gas chromatography-mass

spectrom-

severe deficiency ofdoparnine and serotonin in both the central and peripheral nervous systems (1, 2). The measurement ofHVA and 5HIAA in CSF by HPLC with electrochemical detection is standard for investigating neurological diseases of unknown origin in our laboratory (6). We have now adapted the assay to detect 3MT

etry methods are described that permit the identification and measurement of these metabolites in the above body fluids. Simplified assays for human plasma L-dopa decar-

boxylase and liver L-dopa and 5-hydroxytryptophan decarboxylase, used to demonstrate the enzyme deficiency, are also reported.

in the same chromatographic sure the biogenic amine

Keyphrases: chromatography, reversed-phase gas chromatography/mass spectrometr,’ heritable disorders neurotransmitters dopamine serotonin L-dOpa

-

Recently (1, 2), we described the biochemical and clinical features of the first reported cases of deficiency in aromatic L-amino acid decarboxylase (AADC; dopa decarboxylase, hydroxytryptophan decarboxylase, EC 4.1.1.28).2 enzyme converts L-3,4-dihydroxyphenyl(L-dopa)

(5HTP) deficiency

to serotonin of

neurological developed

to dopamine

these

problems several new

and

5-hydroxytryptophan

(3); lack of the enzyme leads to neurotransmitters and to severe (2). During our investigations we

methods

specifically

for diagnos-

and monitoring this disease. Diagnostic markers for AADC deficiency creases of 5HTP, L-dopa, and 3-methoxytyrosine

in urine, cerebrospinal L-Dopa and 5HTP are and the 3MT accumulates

fluid (CSF), the substrates

and

in-

(3MT)

plasma (1, 2). for the enzyme,

after the catechol 0-methylmethylation of L-dopa (4). 3MT has a very long half-life in plasma and CSF, and high concentrations of3MT accumulate when L-dopa concentrations are increased (5). Here we describe methods transferase-catalyzed

of Child Health, 1EH, UK.

Department

WC1N 1

Current

address

(and

Institute

address

Disease Center, Baylor University Ave., Dallas, TX 75246.

of

Child

Health,

London

for correspondence): Metabolic Medical Center, 3500 Gaston

2 Nonstandard abbreviations: AADC, aromatic L-amuno acid docarboxylase; L-dopa, L-3,4-dihydroxyphenylalanine; 5HTP, 5-hydroxytryptophan; 3MT, 3-methoxytyrosine; CSF, cerebrospinal fluid; GCMB, gas chromatography/mass spectrometry-, HVA, homovanillic acid; 5}UAA, 5-hydroxyundoleacetic acid; VIA, vanillacticacid VPA, vanilpyruvic acid ODS, octadecylsilyl; rrrBarFA, N-methyl-N-(tertbutyldimethylsilyl)-trifluoroacetamide; and TBDMS, tert-butyldi-

methylsilyl. Received

April

27, 1992; accepted

August

3, 1992.

is used to meaThe new protocol ofAADC deficiency from other to decreased turnover of dopa-

the differentiation conditions that may lead mine or serotonin. We also method for the assay ofAADC or liver. The new mine or serotonin

method

that

developed a simplified activity in either plasma

allows

the

detection

of dopa-

(the products ofthe reaction) in either plasma or liver samples with the only sample cleanup step being the acidic removal of protein. All methods required for the initial detection and positive diagnosis of AADC deficiency are described here. MaterIals

and Methods

Matenals The following

ing

are

run metabolites.

allows

AdditIonal

alanine

Methodology

T. Clayton

Aromatic L-amino acid decarboxylase (EC. 4.1 .1 .28) deficiency is a newly described inborn error of metabolism that affects serotonin and dopamine biosynthesis. The major biochemical markers for this disease are increases

of L-dopa,

Diagnostic

UK:

Poole,

compounds

5-HIAA,

pargyline,

HVA,

a-methyltyrosine,

nyllactic

acid

were obtained from 3MT, L- and D-dopa, 4-hydroxy-3-methoxyphe-

Sigma, 5HTP,

(vanillactic

acid, VLA), 4-hydroxy-3(vanilpyruvic acid, VPA), and pyridoxyl 5’-phosphate. Octyl sodium sulfate was obtained from Kodak (Liverpool, UK) and HPLC-grade methanol, dl-n-butylarnine, and acetonitrile were from acid

methoxyphenylpyruvic

BDH

(Poole,

tion

reactions

UK).

was Bond

The acetonitrile used for derivitizadried over a 5-gm molecular sieve

SCX Elut cation-exchange columns were from Analytichern (Harbor City, CA). N-Methyl-N-(tertbutyldimethylsilyl)-trifluoroacetarnide (Mms’FFA) and ThSil/BSA [N,0-bis(trimethylsilyl)acetamide in pyridine] were from Pierce Europe B.V. (Oud Beijerland, The Netherlands). All other chemicals were of Analar (Sigma).

grade

or higher.

Biological

Samples

The first 0.5 ml of lumbar CSF collected was frozen on solid CO2 at bedside. Samples were stored at -70 #{176}C until with

analyzed. diverse

Control neurological

CLINICAL

CSF

samples diseases

CHEMISTRY,

were

who

from

had

patients

no clinical

Vol. 38, No. 12, 1992

2405

or biochemical abnormalities.

indications The 24-h

into 6 molfL collected into

heparin-containing

separated

and

biopsies and

HC1

from

and

samples at -20

stored

tubes,

stored at -70 #{176}C within the infants were obtained

without

frozen

of dopamine urine

delay

at -70

or serotonin were collected #{176}C. Blood was and plasma was 15 mm. Liver percutaneously

#{176}C.

Procedures Measurement

For HPLC

ofneurotransmitter

we used

metabolites

phosphate buffer, of methanol, 5 mmol of octyl sodium A.rnol of EDTA; an Isochrom isocratic

Physics,

San

sulfate, pump

and 48 (Spectra(Cotati, CA) 7125 (i.d.) Apex 5-gm column (Jones

Jose, CA), a Rheodyne a 25 cm x 4.5 mm

injector; and octadecylsilyl

in CSF.

a mobile phase of0.05 molIL sodium pH 2.2, containing, per liter, 220 ml

(ODS)

reversed-phase

Chromatography, Llanbradach, 1-cm-long Spherisorb 5-sm ODS

by a (HPLC Technology, Macclesfield, UK). HVA, 5HIAA, and 3MT were detected by using the second electrode of an ESA 5010 dual-cell electrochemical detector (Bedford, MA) with the analytical V. The gain was

was

calibrated

500 nmol

containing

protected column

set at +0.05

V and

+0.35

set at 1000 and the output (to a Chromjet integrator) at 10 mV. The 1.3 mL/min and the column temperature at 35 #{176}C by using a column oven. The

Spectra-Physics flow rate was was maintained system

electrodes

UK) guard

by injecting

of 5HIAA,

10 L

HVA,

and

of a solution 3MT in 0.1

mol/L HCL. Measurement assay of L-dopa

oIAADC. Incubation procedures for the decarboxylase and 5HTP decarboxylase in plasma and liver were adapted from previously described methods (7, 8). PkLsma L-dopa decarboxylase. Pyridoxal 5’-phosphate (50 tL of a 0.7 mmolfL solution) and 100 pL of plasma

were

added

p1 of sodium phosphate buffer (167 mmol/L, pH 7.0), containing, per liter, 0.167 mrnol of EDTA and 39 mmol of dithioerythritol. The contents were mixed and incubated at 37 #{176}C for 2 h. The assay was started by adding 50 p1 of2O mrnol/L L-dopa in 0.01 mmol/L stopped Blanks

to 300

HC1. After 90 mm at 37 #{176}C, the reaction by adding 500 pL of 0.8 molfL perchloric

were included in which was omitted or in which

plasma

was acid.

either the L-dopa or the the L-dopa was replaced

by D-dopa. After 10 mm the samples were centrifuged in an Eppendorfmicrofuge, and 25 L ofthe supernate was iijected onto the HPLC. Liver L-dopa decarboxylase and 5HTP decarboxylase. Liver samples were thawed and homogenized (1 rng/100 /LL wet weight) in 167 mmol/L sodium phosphate buffer containing EDTA (0.167 mmol/L). The assay conditions

for L-dopa

decarboxylase

were

the

same

as for plasma

that the buffer volume was increased to 350 L and 50 i.iL of homogenate was used. The preincubation time was 10 mm and the incubation time after the

except

addition and the

ofL-dopa supernate

done for plasma. similarly except 2406

CLINICAL

was 20 mm. The reaction was stopped was prepared before HPLC as was Liver 5HTP decarboxylase was assayed that the incubation buffer was 167

CHEMISTRY,

Vol. 38, No. 12, 1992

rnol/L was

sodium

phosphate

by adding

started

Chromatography.

buffer, pH 8.0, and the reaction 50 L of 5 mmol/L L-5HTP. For HPLC we used a mobile phase

of0.1 molIL sodium phosphate buffer, pH 3.45, containing, per liter, 250 mL of methanol, 5 mmol of octyl sodium sulfate, and 48 junol of EDTA and an SP8770 isocratic pump (Spectra-Physics); the injector, reversedphase column, and guard column were as described above. Dopamine and serotonin were detected by using an ESA 5011 high-sensitivity dual-cell electrochemical detector operated in the redox mode, with the analytical electrodes set at +0.15 V and -0.3 V. The gain was set between 10 x 100 and 95 x 100, depending on the sample, and the output (to a Spectra-Physics SP 4270 integrator) mLlmin

and

was set to the column

30 #{176}C for detection

using

25 p1

mine or serotonin Measurement ofL-dopa, and CSF. Chromatography

bile phase ing, per

flow rate was was maintained

1.3 at

of dopamine (45 #{176}C for serotonin) by oven. The system was calibrated by of a 1 mol/L solution of either dopain 0.4 molIL perchloric acid.

a column

iijecting

10 mV. The temperature

5HTP, was

and 3MT in plasma performed with a rno-

of0.05 molIL acetate buffer, liter, 0.048 rnol of EDTA

pH 4.85, containand 2 unol of

dl-n-butylarnine; a Spectra-Physics isocratic SP8770 isocratic pump; a Rheodyne 7125 injector; and a 25 cm x 4.5 mm (i.d.) Techsphere 5-pin ODS reversed-phase column protected by a 1-cm-long Spherisorb 5-rn ODS guard column. Detection was by an LS3 spectrofluorom-

eter tion

(Perkin-Elmer and emission

Ltd., Beaconsfield, UK) with excitawavelengths set at 278 and 325 nm, Peaks were quantified by using a Spectra-

respectively.

Physics

was

SP4270

1.3

computing

mL/min,

and

integrator.

the

column

The

temperature

flow

rate was The

maintained at 35 #{176}C by using a column oven. system was calibrated by injecting 100 L of a solution containing, per liter, 0.45 p.rnol (100 pg) of 5HTP, 1.09 nnol (250 g) of3MT, and 1.27 pmol (250 p.g) of L-dopa. CSF was injected directly onto the column. Plasma was prepared for chromatography by adding an equal volume of 0.8 mol/L perchloric acid. After 10 mimi the protein was removed by centrifugation, and 100 p.L of the supernate was injected onto the column without

delay. and 3MT in urine. We of a-rnethyltyrosine to 1 ml of urine as an internal standard. The pH was adjusted to 1.5 with 2 rnol/L HC1, and the urine was applied to a 3-mL SCX Bond Elut column that had been Measurement

added

ofL-dopa,

5HTP,

20 p1 of a 1 gIL solution

preconditioned

of 10 milL

with

acetic

3 ml

acid.

ml ofmethanol followed The retained compounds saturated barium chloride

followed by 3 ml was washed with 1 by 3 ml of 10 milL acetic acid. were then eluted with 10 ml of

The

ofrnethanol

column

solution into tubes contain10 mg of ascorbic acid. The eluates were acidified with 50 /41 of concentrated HC1 and could be stored for at least 1 week at -20 #{176}C until analysis without loss of L-dopa, 5HTP, 3MT, or a-rnethyltyrosine. Chromatographic conditions were the same as for the analysis of ing

CSF and plasma samples; 100 L of barium chloride eluate was injected. Identification of5HTP and 3MT in urine and plasma by GC-MS. Amino acids were extracted from the acidifled urine by using an SCX column as described above. The column was then washed with 5 ml of water and the amino acids were eluted with 2 ml of 3 mollL aqueous ammonia and dried under nitrogen. The residue was then redissolved in 100 p.L of dichloromethane, dried under nitrogen, and derivatized by dissolving them in 50 pL ofacetonitrile, adding 50 L of MTBSTFA, and heating to 60 #{176}C for 20 mm. Amino acids were extracted from plasma by using a modification of the cation-exchange cleanup procedure described by Adams (9) and analyzed by GC-MS of their tert-butyldirnethylsilyl (TBDMs) derivatives (10). Plasma samples (300 p.L)

were

acidified

with

150 L

of an equivolume

mixture

phenolic

acids

(and

other

organic

acids)

were

ex-

tracted by shaking twice with 25 ml of ethyl acetate. After drying, the organic acids were derivatized by using TriSil BSA in pyridine (50 ML); 1 p.L was injected into the gas chromatograph (with the injector conditions as above). The column temperature was 80 #{176}C for 2 ruin, then was increased by 4 #{176}C/minto 300 #{176}C. VIA was also

quantified cal

by reversed-phase

detection

HPLC

with

electrochemi-

(11).

Results Measurement adapted our

of CSF HVA, standard method

5HIAA, and for measuring

We biogenic

3MT.

IA 1600j

*

A

400

.....

0

E

,s

*

300

200 100

#{149} , 2

0 0

,#{149}#{149}

I

i

4

of

water and acetic acid, supplemented with 38.5 nmol of a-methyltyrosine as internal standard, and passed through an SCX column that had been primed by washing with 5 ml of methanol and 5 ml of 10 milL acetic acid. After the column was washed with 30 ml of water, the amino acids were eluted with ammonia and dried and derivatized as described for the urine samples. GC-MS was performed with an HP 5890 gas chromatograph (with an HP 7673 autosampler) coupled to an HP 5970 mass selective detector and ChemStation data system (all from Hewlett-Packard Co., Palo Alto, CA). The column was a 30 m x 0.25 mm (i.d.) fusedsilica capillary column coated with a 025-jmi-thick chemically bonded DB1 stationary phase (J&W Scientific, Folsom, CA); the carrier gas was helium (2 mid mm). Samples of 1 L (still in the derivatizing reagent) were injected by using the fast-injection mode of the autosampler into a splitisplitless injector heated to 270 #{176}C. After 2 mm with the column at 120 #{176}C, the split valve was opened and the column oven was heated by 20 #{176}C/minto 190 #{176}C, then by 15 #{176}C/minto 190 #{176}C, and finally by 20 #{176}C/minto 270 #{176}C. The mass selective detector was used in the scan mode to produce complete mass spectra of chromatographic peaks. For quantification, we used reconstructed ion chromatograrns to compare the area of the m/z 302 peak from 3-methoxytyrosine with that of m/z 316 from a-methyltyrosine. Measurement of VLA in urine. We diluted 2 ml of urine with 8 ml of water, saturated this with sodium chloride, and adjusted the sample to pH 2.0 with HC1.

The

1700

6

AGE Fig. 1 . Effect

8

10

(Years)

of age on CSF 3MT concentration

S. control sub;

A,

infants

*,

c

with

deficiency

I

2

A

AC

3

C

50

B

LA

0L

.

Fig. 2. Electrochemlcal

detection of 5HIAA, HVA, and 3MT in CSF 1, standard, 500 nmol/L each for 5HIM (A), HVA (B), and 31ff (C); 2, normal CSF; 3, C5F from patient with MDC deficiency. Injection volume 10 L

metabolites so that we could separate 5HIAA, HVA, and 3MT from other electroactive species in CSF. To establish a reference range, we determined 3MT concentrations in CSF samples taken from 21 children with various neurological diseases (Figure 1). The concentration of 3MT decreases rapidly in the first year of life and then remains relatively constant (