4-Aminobutyric

Demonstration of 4-Aminobutyric Acid Aminotransferase Deficiency in Lymphocytes and Lymphoblasts
K.M.GIBSON,L.SWEETMAN,W.L.NYHAN and I.JANSEN
Departments of Pediatrics and Medicine,University of California, San Diego,La Jolla, CA92093,USA
J.JAEKEN
Department of Pediatrics, University of Leuten, Universitair Ziekenhuis Gasthuisberg,
Herestraat 49,B-3000 Leuten,Belgium
Lysates of lymphocytes, isolated from whole blood,and Epstein-Barr virus transformed cultured lymphoblasts catalysed the transamination of 4-aminobutyric acid with 2-oxoglutaric acid as co-substrate. 4-Aminobutyric acid aminotransferase activity in lymphocyte and lymphoblast sonicates derived from 12 unrelated control individuals (6each) was 39 ±19pmol min-‘ (mgprotein-‘)(mean ±1 SD). Activitie lysates of both types of cell derived from a Flemish patient were lessthan 3% of control.4-Aminobutyric acid aminotransferase activity in sonicates derived from the parents and a healthy sibling were 15-37% of the control mean for lymphocytes and 13-20% of the control mean in lymphoblasts, respectively. Km values in a control lymphoblast sonicate were 0.63 and 0.08 mmolL-1 for 4-aminobutyric and 2-oxoglutaric acids, respectively. These data indicate that the parents and heaIthy sibling are heterozygous and the patient is homozygous fora defective gene responsible for 4-aminobutyric acid aminotransferase deficiency,and that inheritance is autosomal recessive.
Three enzymes are directly involved in the mammalian metabolism of 4-aminobutyric acid (GABA) (Figure 1). Glutamic acid decarboxylase(EC 4.1.1.15)catalyses the production of GABA in brain through the irreversible decarboxylation of glutamic acid. GABA is then sequentially degraded by the action of two catabolic enzymes,4-aminobutyric acid aminotransferase(GABA transaminase,EC2.6.1.19),which converts GABA to succinic semialdehyde and stoichiometrically produces glutamic acid from 2-oxoglutaric acid,and succinic semialdehyde dehydrogenase(SSADH,EC1.2.1.24), which catalyses the oxidation of succinic semialdehyde to succinic acid,thus allowing entry into the citric acid cycle.The concentrations of glutamic acid and GABA in brain are high (10 and 0.8 mmol L-1, respectively),and for this reason these enzymes have been considered to play a significant role in overall oxidation in the brain (Metzler,1977).

Deficiency of glutamic acid decarboxylase has been reported in a single patient (Yoshida et al., 1971) and six patients have been described (Gibson et al.,1983,1984a) in whom 4-hydroxybutyric aciduria was caused by a deficiency of succinic semialdehyde dehydrogenase. Recently a patient has been described(Jaeken et al., 1984) in whom a deficiency of GABA transaminase was demonstrated in biopsied liver. Cardinal clinical manifestations included severe psychomotor retar-dation,hypotonia and hyperreflexia. Amino acid analysis of the cerebrospinal fluid showed highl elevated concentrations of GABA,homocarnosine and β-alanine.
We have shown that GABA transaminase activity is expressed in lymphocytes isolated from whole blood and in transformed human lymphoblasts (Gibson et al., 1983, 1984a). This activity was demonstrated in a coupled assay system designed to measure succinic

Figure 1 Metabolic pathways involving GABA. The block in this family is indicated by the striped rectangle
204
Journal of Inherited Metabolic Disease.ISSN 0141-8955.Copyright©SSIEM and MTP Press Limited,Queen Square,Lancaster, UK.Printed in The Netherlands.
4-Aminobutyric Acid Aminotransferase Deficiency
semialdehyde dehydrogenase activity using U-14C-GABA as precursor. Carrying out the assay in the absence of NAD+ permits the quantification of succinic semialdehyde production, and the semialdehyde is measured as the methoxime derivative.The recognition of the child with deficient activity of hepatic GABA transaminase stimulated the development of a reliable and accurate assay for measurement of this enzyme and the investigation of the optimal conditions for the assay and for the measurement of kinetic properties of the enzyme. It was of considerable interest to determine whether this deficiency could be demonstrated in lysates of lymphocytes and lymphoblasts. If an accurate assay procedure could be employed using cultured lympho-blasts,it was hoped that carrier detection could be achieved and the mode of inheritance identified.This report summarizes the results of these studies.
MATERIALS AND METHODS
Uniformly radiolabeled GABA (specific activity 238mCimmol-‘) was purchased from New England Nuclear (Boston,MA) and all other reagents were obtained from Sigma Chemical Co. (St. Louis, MO). [U-14C]GABA was purified using a two column batchwise procedure. Radiolabeled GABA (100μCi aliquots in 0.1 molL-‘ HCI) was applied to a 0.9cm i.d.x10cm column of AG 50W-X4 resin(200-400 mesh,H+form) equilibrated with distilled water (Bio-Rad,Richmond, CA,USA).The column was washed with 4 bed volumes of water and the [U-14C]GABA eluted with 2molL-1 ammonium hydroxide. Ammonia was removed under nitrogen and the GABA solution concentrated by lyophilization. The concentrated GABA was then applied to a column of AG2-X10 (200-400 mesh,acetate form) of the same dimensions as the first column equilibrated with distilled water, and the [U-14C]GABA was eluted with distilled water.Purified [U-14CJGABA was concentrated by lyophilization. This procedure served to reduce the blank in GABA transaminase assays lacking cell sonicate from approxi-mately 3000-4000d.p.m. to 200-300d.p.m. SSADH was assayed in white cell sonicates (Gibson et al.,1984b) using U-14C-succinic semialdehyde prepared as pre-viously described (Gibson and Sweetman, 1983).
Clinical details of the patient have been reported (Jaeken et al., 1984). Whole venous blood (10ml)from patient, parents, sibling and six control individuals was added to 1.5 ml acid-citrate dextrose and 8.5 ml RPMI tissue culture medium and transported at ambient temperature from Leuven, Belgium to San Diego, California,USA.Approximate time between shipment and the assay of lymphocytes was 72h.Lymphocytes were isolated under sterile conditions from whole blood by density centrifugation on Ficoll gradients(Boyum, 1968).Isolated lymphocytes were divided equally to allow assay of SSADH and GABA transaminase in one set and transformation to permanent lymphoid lines in the remaining cells.Transformation of lymphocytes was performed using Epstein-Barr virus according to established procedures (Sly et al.,1976).Lymphoblast lines derived from parents, sibling, patient and six

205
control individuals were maintained in RPMI 1640 medium supplemented with 10%(v/v) fetal bovine serum.Except during initial transformation, antibiotics were not used. Established lines were checked for mycoplasma contamination using mycotrim TC flasks (HANA Biologics Inc.,Berkeley,CA,USA).
Isolated lymphocytes or lymphoblasts harvested by centrifugation were washed with phosphate-buffered saline and pelleted. Cells were lysed by sonic disruption at a concentration of 10×106 cellsml-1 complete buffer with two 10s bursts and an intermediate 10s pause at 28W using a Sonifer Cell Disruptor(Heat Systems-Ultrasonics Inc.,Model W-185,Plainview, NY).Cell suspensions were maintained in an ice bath during sonication. Lysates were centrifuged for 5 min at 4°C and 8800g in a Beckman Model B microfuge (Palo Alto,CA) to remove unbroken cells and debris.The resultant supernatant fluid was used for enzyme assay.
GABA transaminase was assayed in lysates of lymphocytes and lymphoblasts using a modification of the procedure of White (1979) for the platelet enzyme. The final assay components included 50mmolL-1 potassium phosphate buffer,pH 8.0,supplemented with 0.25mmolL-1’dithiothreitol, 0.05mmolL-1 Na2 EDTA and 0.1 mmol L-‘ pyridoxal-5’-phosphate(com-plete buffer);1.35mmolL-1 2-oxoglutaric acid (pot-assium salt),1mmoiL-1 14C-GABA (final speci activity 5 mCimmol-‘)and lymphocyte or lymphoblast sonicate in a volume of 100 μl. All assay solutions except purified radiolabeled GABA were prepared in complete buffer.
Incubations were carried out for 60 min at 37°C.In all experiments assays without lysate or 2-oxoglutaric acid served as blank. The activity of SSADH was determined as a control enzyme.Following incubation,assays were terminated by addition of 10 μl of 4.2 mol L-1 perchloric acid.Denatured protein was removed by centrifugation, and the resultant supernatant fluid neutralized by the addition of 7μl of 6molL-‘ potassium hydroxide. Insoluble potassium perchlorate was removed by centrifugation.The supernatant was added to a mixture of 25μl of aqueous 28% methoxylamine hydrochloride and 15μl of14%(w/v)sodium hydroxide and incubated for 30min, and 100μl aliquots of this mixture were analyzed by reverse-phase high performance liquid chromatography,as described below.
Baseline separation and quantification of radio-labeled precursors and products for assays of SSADH and GABA transaminase were carried out by reverse-phase high performance liquid chromatography. A guard column (0.41 cm i.d. x 2 cm)filled with SC 30-40 micron reverse phase C18 ODS pellicular particles was connected to a 0.46 cm i.d.x25 cm Spherisorb ODS II, 5 micron column (Custom LC Inc.,Houston,TX).The mobile-phase solvents were 0.05molL-1 potassium phosphate buffer,pH 2.1(buffer A)and methanol.For GABA transaminase assays, the column was eluted isocratically with 85% buffer A/15% methanol,and for SSADH assays isocratic elution was with 80% buffer A/20% methanol.All chromatographic elutions were for 30min. Separations were performed at room temperature and a flow rate of 0.44 ml min-1,and 1 min
206

Gibson,Sweetman,Nyhan,Jansen and Jaeken
SUCCINIC SEMIALDEHYDE METHOXIME(pmolsx10-2)
INCUBATION TIME(min.)
Figure 2 Linearity of the assay for the activity of GABA transaminase in normal lymphoblast lysates with protein content(A)and with time(B)
fractions were collected.Radioactivity was determined by liquid scintillation counting using 5ml of fluor. In GABA-transaminase analyses,unreacted radiolabeled GABA eluted with the solvent front at 8min,and radiolabeled succinic semialdehyde methoxime eluted at 22 min.The presence of the methoxy moiety served to separate the small peak of radiolabeled succinic semialdehyde from the very large radioactive GABA peak. Similarly, for the assay of SSADH radiolabeled succinic acid eluted at 10min while unreacted radio-labeled succinic semialdehyde methoxime eluted at 19min. In all experiments,assays containing purified [14C]succinic acid or succinic semialdehyde and all

assay components with denatured protein were anal-yzed to assess product recovery.Protein concentration was determined with the Folin-phenol reagent (Lowry et al.,1951).Determination of Michaelis constants was performed as previously described (Wilkinson, 1961).
RESULTS
The production of succinic semialdehyde, measured as the methoxime derivative, was linear with protein concentration up to 0.4 mg(Figure 2A)and with time up to 90min (Figure 2B).The results of assays for GABA transaminase and SSADH are shown in Table 1.Mean
Table 1 Activities of GABA-transaminase and SSADH in cell sonicates derived from control,patient and family members
Lymphocyte* Lymphoblast*
Cell line SSADH GABA-T SSADH GABA-TT
Control(San Diego)t 436 74.0
Controls 283 15.6 675 44.6,55.0,80.6
Control 335 45.4 991 49.2,40.4,54.9
Control 326 969 200 46.0,38.1,52.5
Control 328 38.0 595 37.5,42.7,60.8
Control 29.2 704 8.6,10.2,16.9
Control 233 33.7 877 22.2,16.0,19.8
Mean control±1SD 301±43 38.6±18.2 674±273 38.7±18.8
Mother 177 6.2 659 4.2,5.8,5.5
Father 260 14.3 550 7.4,6.4,6.6
Sibling 241 8.5 445 8.0,7.0,8.7
Patient 142 0.8 329 0.0,0.0,0.8
*Lymphocyte and lymphoblast controlIs were non-identical. All values are expressed in units of pmolmin-‘ (mg protein)-‘
tAll lymphoblast lines were analysed in parallel for GABA-transaminase(GABA-T) activity on three separate occasions.Lymphoblast SSADH activity was analysed once in parallel with the second(middle) GABA-T analysis
Lymphocytes were obtained from a freshly drawn blood sample in San Diego.Enzyme activities for this control were not included in the calculation of mean control activities
All control lymphocytes(excluding the San Diego Control)were simultaneously shipped with patient and family members’blood.
Calculated using the mean activity from three individual determinations foreach control,ie.the mean of means
4-Aminobutyric Acid Aminotransferase Deficiency
activity of GABA transaminase in sonicates of six control lymphoblast lines assayed on three different occasions was 38.7 ± 18.8 pmol min-‘ (mg protein)-1. Identical activities were obtained in lysates of lympho-cytes obtained from six control individuals. In contrast, the patient displayed less than 3% of control GABA transaminase activity in lysates of both cell types (Figure 3). The parents and healthy sibling of the patient displayed activities of GABA transaminase that ranged from 16 to 37% of the control mean in lymphocyte sonicates and from 13 to 20% of the control mean in lymphoblast sonicates assayed on three separate occasions. The activity of SSADH in lymphoblast sonicates derived from the patient and family members

Figure 3 Quantification of [‘4C]succinic semialdehyde production (measured as the methoxime derivative) from [U-
14C]GABA in lysates of lymphoblasts derived from control, patient and blank by reverse-phase high performance liquid chromatography.

207
was within the range of the six control values assayed simultaneously. SSADH activity in lysates of lympho-cytes derived from the father and healthy sibling were within the range of control values while the mother and patient displayed SSADH activities that were 59 and 47% of the mean control value. Assays of GABA transaminase were carried out in which equal volumes of control and patient lymphoblast sonicates were mixed and assayed,and the activity was 61% ofa parallel assay containing onlycontrollymphoblast sonicate.This ruled out the possibility of an endogenous inhibitor as the source of deficient GABA transaminase activity in the patient.
Table 2 displays some properties of GABA transa-minase in extracts of human brain, platelets and lymphoblasts.Mlichaelis constants for GABA and 2-oxoglutaric acid in lymphoblast sonicate were 0.63 and 0.08 mmolL-1,respectively. Both values are similar to those obtained in brain and platelet extracts.The latter was virtually identical to the platelet value while the former value was approximately 2-fold higher. Pyridoxal-5’-phosphate had been observed to stimulate GABA transaminase activity in extracts of brain and platelets, and this was also the case in sonicates of human lymphoblasts.The activity of GABA transa-minase in extracts of platelets and lymphoblasts was quite comparable at 37 and 39nmolmin-1 (g protein)-1.
DISCUSSION
These data document a severe deficiency of GABA transaminase in lysates of lymphocytes and lympho-blasts derived from the patient.This finding is consistent with the deficiency of GABA transaminase previously reported in liver of this patient. The development of an assay suitable for use with cells obtained from peripheral blood simplifies the assessment of other patients to be discovered with this disease. It has made possible genetic studies of the family.
In sonicates of lymphocytes and lymphoblasts derived from control individuals, there was an appreciable variation of values for GABA transaminase activity, ranging from approximately 15 to 70pmol min-‘ (mg protein)-‘ in lymphocytes and from 9 to 81 pmolmin-‘ (mgprotein)-‘ in lymphoblasts. This may simply reflect biochemical heterogeneity. The activities of GABA transaminase in the parents and the sibling were clearly outside of one standard deviation from the control.There was one control lymphocyte and one lymphoblast value at thelow end of the control range which were close to the values obtained in the father and sibling,respectively. Of course the incidence of heterozygosity in the general population is not known. On the basis of currently available data it may not be possible. conclusively to establish that an individual with a low normal value is not a heterozygote. However, the majority of normal cells assayed differ clearly from those of heterozygotes. In this family it is highly likely that both parents and the sibling are carriers of the defective gene. These observations indicate that GABA transaminase deficiency is inherited
208

Gibson,Sweetman,Nyhan,Jansen and Jaeken
Table 2 Comparison of human brain, platelet and lymphoblast GABA-T
Property Brain* Platelet* Lymphoblast
Km(GABA)t 0.31 mmol L-1 0.31 mmol L-1 0.63 mmol L-1
Km(2-oxoglutaric acid) 0.16 mmol L-1 0.07 mmol L-1 0.08 mmol L-1
Activity in crude tissue 167 nmol min-1 37 nmol min-1 39 nmol min-1
(g wet tissue)- (g protein)-1 (g protein)-1
*These properties of human brain and piatelet GABA-transaminase have been reported by White(1979) t For kinetic analyses in lymphoblasts,2-oxoglutaric acid concentration was varied from 0.04 to 1.35 mmol L-1 and the concentration of GABA from 0.03 to 1.0mmolL-
as an autosomal recessive trait. The occurrence of the disease in male and female siblings in this family (Jaeken et al., 1984) is consistent with this mode of inheritance.
SSADH activity was within the range of control in sonicates of lymphoblasts derived from the patient and family members.However,in lymphocytes activities were low in the patient and her mother. This is a resulIt of the fact that the total lymphocyte protein included in assays of SSADH in these individuals was somewhat beyond the limit of linearity of product formation with respect to lysate protein. This was not the case for the lymphoblast assays of SSADH.
The properties of GABA transaminase in normal lymphoblasts were strikingly similar to those reported for brain and platelets (White,1979).It seems likely that the enzyme in each of these tissues is the same. Documentation of deficiency of GABA transaminase in the patient using lymphocytes and lymphoblasts as well as liver is consistent with this conclusion.Furthermore, it suggests strongly that in the patient there is also a profound deficiency of the activity of GABA trans-aminase in tissues of the central nervous system.This is consistent with the high concentration of GABA found in the cerebrospinal fluid (Jaeken et al.,1984).It would appear intuitively to be consistent with the clinical manifestations observed in the patient. The similarity of human GABA transaminase in brain,platelets and lymphoblasts is consistent with findings in the mouse (Wu et al., 1978) in which no difference in GABA transaminase could be detected in brain,spinal cord and kidney using immunodiffusion and microcomplement fixation. The method of Wu and colleagues (1978) for assay of GABA transaminase utilizes [4C]2-oxoglutaric acid and quantifies [14C]glutamic acid production by ion-exchange chromatography. This procedure was employed to document the deficiency of GABA transaminase in biopsied liver of the patient (Jaeken et al.,1984).The assay described in this report utilizes [14C]GABA as precursor and quantifies [14C]succinic semialdehyde as the methoxime de-rivative by reverse-phase high perormance liquid chromatography. The preceding observations suggest that lymphocytes and cultured lymphoblasts provide an excellent model in which to study the metabolism of GABA.
MS receited 8.1.85
Accepted for publication 30.4.85
The authors are indebted to Dr Claude Bachmann for suggesting the study, Mary Anne Carter-Birkman for

preparation of the manuscript and Dr J.Edwin Seegmiller for use of tissue culture facilities. Special thanks are extended to Ms Ocean Pellett for supplying many of the control lymphoblast lines. This work was supported in part by US Public Health Services Grant No.HD-04608 from the National Institute of Child Health and Human Development, Grant No.GM-17702 from the National Institute of General Medical Sciences and AM-13622 from the National Institutes of Arthritis, Diabetes, Digestive and Kidney Diseases, National Institutes of Health,Bethesda,Maryland.
References
Boyum,A. Separation of leukocytes from blood and bone marrow.Scand.J.Clin.Lab.Inwest.21,Suppl.97(1968) 77-89
Gibson,K.M.,Jansen,I,Sweetman,L.,Nyhan,W.L.,Rating,
D.,Jakobs,C.and Divry,P.4-Hydroxybutyric aciduria:A new inbornerror of metabolism. III. Enzymology and inheritance.J.Inher.Metab.Dis.7,Suppl.1(1984a)95-96 Gibson, K. M. and Sweetman, L. Enzymatic preparation of
radiolabeled succinic semialdehyde. Analyt.Biochem.135 (1983)436-437
Gibson,K.M.,Sweetman,L.,Nyhan,W.L.,Jakobs,C., Rating,D.,Siemes,H. and Hanefeld,F. Succinic semialdehyde dehydrogenase deficiency:An inborn error of gamma-aminobutyric acid metabolism.Clin.Chim.Acta 133 (1983)33-42
Gibson,K.M.,Sweetman,L.,Nyhan,W.L.and Rating,D. Succinic semialdehyde dehydrogenase deficiency. J. Neurogenet.1(1984b)213-218
Jaeken,J.,Casaer,P.,de Cock,P.,Corbeel,L.,Eeckels,R., Eggermont,E.,Schechter,P.J.and Brucher,J.M.Gamma-aminobutyric acid transaminase deficiency: A newly recognized inborn error of neurotransmitter metabolism. Neuropediatrics 15(1984)165-169
Lowry,O.H.,Rosebrough,N.J.,Farr,A.L.and Randall,R.J. Protein measurement with the Folin phenol reagent.J.Biol. Chem.193(1951)265-275
Metzler,D.E.Biochemistry-The Chemical Reactions of LivingCells,Academic Press,New York,1977,p.533
Sly,W.S.,Sekhon,G.S.,Kennett,R.,Bodmer,W.F.and Bodmer,J.Permanent lymphoid lines from genetically marked lymphocytes:Success with lymphocytes recovered from frozen storage. Tiss. Antigens 7 (1976) 165-172
White,H. L. 4-Aminobutyrate: 2-oxoglutarate amino-transferase in blood platelets. Science 205(1979)696-698 Wilkinson,G.N.Statistical estimations in enzyme kinetics. Biochem.J.80(1961)324-332
Wu,J.-Y.,Moss,L.G.and Chude,O.Distribution and tissue specificity of 4-aminobutyrate-2-oxoglutarate amino-transferase.Neurochem.Res.3(1978)207-219
Yoshida, T., Tada, K. and Arakawa, T. Vitamin B6-dependency of glutamic acid decarboxylase in the kidney from a patient with vitamin B6 dependent convulsion. Tohoku J.Exp.Med.104(1971)195-198