Main

OAT (ornithine-oxo-acid aminotransferase, EC 2.6.1.13) is a nuclear-encoded mitochondrial matrix enzyme involved in the metabolic disposal of ornithine. OAT catalyzes reversible transamination of ornithine to glutamate-γ-semialdehyde, which is further converted to P5C. OAT is expressed in nearly all tissues. Depending on the tissue and prevailing conditions, OAT participates in arginine catabolism, proline biosynthesis, or de novo ornithine and arginine biosynthesis(1,2).

In man(1), cat(3), and transgenic mouse(2), OAT deficiency leads to gyrate atrophy of the choroid and retina, a progressive blinding chorioretinal degeneration. Patients develop early cataracts and progressive atrophy of type 2 skeletal muscle cells with tubular aggregates. However, muscle symptoms occur rarely. The concentration of ornithine is increased 10-15-fold above normal in all body fluids. In most patients with GA, OAT activity in skin fibroblast extracts is decreased to <5% of normal values, and the values are often under detection limit of the conventional assays(1). Pyridoxal phosphate is a cofactor of OAT(4,5), and in few patients pharmacologic doses of pyridoxine alleviate hyperornithinemia and partially restore fibroblast OAT activity(6).

More than 60 different mutations of the OAT locus, including missense and nonsense mutations, microdeletions and insertions, as well as nucleotide substitutions, have been identified(1,2,79). Most mutant alleles produce normal amounts of normal-sized mRNA, but premature termination of translation leads to reduced mRNA levels. A single amino acid change may alter conformation of the mature protein, affect the transport into and processing in the mitochondria, decrease stability of OAT, or interact with the function of the essential cofactor pyridoxine(1). Intrafamilial clinical and biochemical variation in GA is smaller than interfamiliar, indicating that the genetic heterogeneity probably plays a role in the phenotypic variability(10).

Residual OAT activity may differ in different mutations and explain familial differences in the progression of the clinical symptoms. We thus developed a rapid and sensitive HPLC modification of the OAT assay for lymphocytes isolated from peripheral blood, and investigated OAT activity in 43 Finnish patients with GA, representing different mutations. To assess the suitability of the assay for carrier identification, we compared enzyme activities in obligatory GA heterozygotes with those in healthy controls.

METHODS

Subjects. Forty-three patients with GA (20 female patients), from 31 families, with known OAT gene mutations participated in the study. They all had hyperornithinemia and characteristic clinical findings of the disease. Eleven patients had received experimental therapy for GA before the study; two children had been on low arginine diet (15-20 mg of arginine·kg-1d-1) for 2-18 mo, four adults on creatine supplementation for 8-15 y (1500-2000 mg/d), and five patients had received creatine precursors guanidinoacetate (330-880 mg/d) and methionine(440-1120 mg/d) for 0.6-8 y(11,12). Nine obligatory heterozygotes (son, daughter, 4 fathers, and 3 mothers of the GA patients) and 15 healthy volunteers (6 female subjects) were also investigated. The ages (mean ± SD; range) of the GA patients, carrier heterozygotes, and healthy control subjects were 33 ± 16(4-74), 32 ± 17 (3-50), and 38 ± 12 (25-65) y, respectively.

Plasma amino acid analysis. Plasma amino acid concentrations after an overnight fast were measured with an automated amino acid analyzer(LKB α Plus, Munich, Germany)(13).

Preparation of lymphocytes. Peripheral blood mononuclear cells were isolated from heparinized venous blood using Ficoll-Paque(Pharmacia, Uppsala, Sweden). The cells were washed twice in PBS, pH 7.2, suspended in 0.1 M phosphate buffer, pH 8.0, and disrupted by sonication(five 10-s bursts) on an ice bath.

OAT assay. Our OAT assay was modified from the HPLC method developed by O'Donnel et al.(14). In principle, the end product of the OAT reaction, P5C, was allowed to react with o-aminobenzaldehyde, and the resulting dihydroquinozolinium compound was separated by HPLC and detected spectrophotometrically. The standard enzyme reaction mixture contained 300 µL of cell lysate supernatant, 30 µL of L-ornithine (0.5 M; Fluka, Buchs, Germany), 20 µL of α-ketoglutarate (0.1 M; Fluka), and 25 µL of pyridoxal phosphate (0.064 mg/mL; Fluka). After 4 h of incubation at 37°C, protein was removed by trichloroacetic acid precipitation. P5C, the end product of the OAT reaction, was then allowed to react with o-aminobenzaldehyde to obtain a dihydroquinozolinium derivative. A 50-µL aliquot of sample or P5C standard (Sigma Chemical Co., St. Louis, MO) was incubated with 10 µL of o-aminobenzaldehyde (7.5 mg/mL; Sigma Chemical Co.) for 60 min at room temperature. The acquired dihydroquinozolinium compound was separated by HPLC, and its absorbance was measured at 254 nm.

The HPLC system consisted of an LKB solvent conditioner (LKB, Bromma, Sweden), LKB 2150 HPLC pump, LKB 2151 LC-controller, Rheodyne 7125 injector with 20 µL of sample loop, and LKB Variable Wavelength Monitor with LKB Spherisorb ODS 2.5 mm, 4 × 250-mm analytical reversed phase columm. Chromatography was performed with a mobile phase of methanol:water:phosphate buffer (0.01 M), pH 3.0, at 1:1:2, with a flow rate of 1.5 mL/min.

The protein concentration of the cell lysate supernatant was determined with a photometric assay (Bio-Rad, Hercules, CA), and OAT activity was expressed as micromoles of P5C formed·min-1 mg protein-1.

OAT mutations. The OAT mutations were determined (D.V.) at the Howard Hughes Medical Institute, Baltimore, as described previously(9).

Statistical analysis. OAT activities of the GA homozygotes, carrier heterozygotes, and healthy controls were compared using a two-sample t test with unequal variances. Confidence intervals at 95% were also calculated. The values were regarded significant at p < 0.05.

Ethics. The study was approved by the Joint Ethics Committee of the University of Turku and the Turku University Central Hospital. The patients and control subjects and/or their guardians gave written informed consent for the study.

RESULTS

Optimization of the OAT assay. The OAT enzyme reaction was essentially complete in 4 h (Fig. 1A), and the derivatization reaction in 2 h, respectively (Fig. 1B). To shorten the time of analysis, we routinely injected the samples into the HPLC system after 60-min derivatization. The retention time of the dihydroquinozolinium compound formed was approximately 2.5 min. No other peaks except the solvent and reagent peak were detected in the chromatograms of the P5C standard. The linearity of the method was excellent at the tested range of 2 to 200 mM P5C standard (r = 0.999). The interassay variation of the method was 3% (n = 6), and the detection limit for P5C with a 30-min derivatization time was 40 pmol. If the derivatization time was increased, the detection limit decreased even further (data not shown).

Figure 1
figure 1

Characteristics of the OAT assay.(A) Effect of reaction time on the formation of P5C. Supernatant of peripheral blood mononuclear cell sonicate of a healthy subject was used as the enzyme source. The reaction time of o-aminobenzaldehyde-P5C derivatization was 60 min. AU, arbitrary units. (B) Effect of reaction time of o-aminobenzaldehyde-P5C derivatization on the peak height of the dihydroquinozolinium reaction product. The supernatant of mononuclear cell sonicate of a healthy subject was used as the enzyme source. The reaction proceeded for 4 h.

Genotypes of the GA patients and carrier heterozygotes. Thirty-five GA patients were homozygous for the OAT allele L402P, and eight patients were compound heterozygotes (R180T/L402P, N89K/L402P, and L402P/x; x = previously unknown allele). The known alleles all represent missense mutations: L402P has T → G change in bp 1205 in exon 11, R180T G → C change in bp 539 in exon 6, N89K C → A change in bp 4, respectively(1). The obligatory heterozygotes investigated all carried the most common Finnish mutation L402P.

Lymphocyte OAT activity. Lymphocyte OAT activities of all patients with GA were less than 10% of the control values(Fig. 2). The OAT activities of 33 of the 42 patients were below the detection limit of the assay (1 pmol·min-1 mg protein-1; see "Optimization of the OAT assay"). Five male and two female patients of the 33 with the most common Finnish mutation L402P/L402P had measurable OAT activity (2-10 pmol·min-1 mg protein-1). The activity was above the detection limit also in a female patient with R180T/L402P mutation and in a female patient with N89K/L402P mutation (2-10, 13, and 17 pmol·min-1 mg protein-1, respectively). In the two patients representing mutation L402P/x, the OAT activities were below the detection limit of the assay (Fig. 3).

Figure 2
figure 2

Lymphocyte OAT activity in patients with GA, carrier heterozygotes, and healthy control subjects. Horizontal bars = mean values of the carriers and controls.

Figure 3
figure 3

Lymphocyte OAT residual activity in GA patients with different mutations. The values of the two children on an arginine-restricted diet are marked with solid symbols. The dotted line shows the detection limit of the assay when the reaction product derivatization time with OAB was standardized to 60 min (see"Optimization of the assay").

Residual OAT activity was not influenced by gender (range <1-10 and<1-17 pmol·min-1 mg-1 in male and female patients, respectively), age, or experimental therapy (data not shown). Fasting plasma ornithine concentration (mean ± SD) of the 40 GA patients who were on a free diet was 928 ± 169 µmol/L. The two patients whose arginine intake had been restricted had markedly lower plasma ornithine concentrations(183 and 516 µmol/L). Both values, however, clearly exceeded the upper reference limit (115 µmol/L). Obviously, plasma ornithine concentration showed no correlation with the type of mutation (Fig. 4A) or with the residual OAT activity (Fig. 4B).

Figure 4
figure 4

Plasma ornithine concentrations of patients with different OAT mutations (A) and as a function of residual enzyme activity (B). The values of the two children on arginine restricted diet are marked with solid symbols. The upper normal limit of plasma ornithine concentration is indicated with a dotted line.

Lymphocyte OAT activity in the obligatory heterozygotes who carried the most common Finnish GA allele L402P was (mean ± SD) 70 ± 50 pmol·min-1 mg protein-1 (range 33-193). Plasma ornithine concentrations of the heterozygotes ranged from 82 to 144µmol/L, with most of the values within the reference range. OAT activity of the control subjects was 184 ± 60 (range 85-297) pmol·min-1 mg protein-1 (Fig. 2).

The OAT activities of the patients with GA differed clearly from those of carrier heterozygotes and normal controls, with no overlapping values. Lymphocyte OAT activities of the carrier heterozygotes were below those of the control subjects (p < 0.001). Confidence intervals of 95% for the carriers and the control group were 34-106 and 156-212, respectively, indicating that the method was able to differentiate most GA carriers from normal control subjects. However, some individual values overlapped(Fig. 2).

DISCUSSION

OAT deficiency is expressed in most tissues including cultured skin fibroblasts(14,15), lymphocytes(16), and liver(17). Peripheral blood lymphocytes provide convenient material for clinical enzyme studies, as blood samples are easily obtained and readily transported. Skin and liver biopsies are more invasive procedures, and time is needed before sufficient amounts of fibroblasts have developed in culture. Presumably, the accuracy of the lymphocyte assay might be further improved by phytohemagglutinin-stimulated transformation, which increases specific activity of lymphocyte OAT approximately 15-fold(16).

The first OAT assay methods were based on colorimetric measurement of the end product of the OAT reaction, P5C, which was first allowed to react with o-aminobenzaldehyde to produce a yellow dihydroquinozolinium compound(18,19). Another more sensitive ninhydrin-based colorimetric method has also been reported(20). To improve sensitivity, radiochemical OAT assays were developed with [14C]ornithine as a substrate. Separation of the end product, however, requires complicated and time-consuming steps, and sufficient specific activity of the [14C]ornithine may be a problem(21,22). We used a modification of an assay developed by O'Donnel et al.(14), and separated P5C by isocratic reverse phase HPLC and then used the o-aminobenzaldehyde reaction for detection. The HPLC method is technically simple, and its sensitivity is comparable with that of the radioactive assays.

All types of OAT mutations characterized so far in Finland were included in this study. Two compound heterozygotes carried a so far uncharacterized allele in combination with the predominant Finnish mutant allele L402P. All three known mutations reported here are situated in the well conserved region of the OAT gene(1) and give rise to normal amounts of normal-sized mRNA. L402P and N89K apparently destabilize the structure of the enzyme, leading to decreased amounts of immunoreactive protein, whereas R180T has been reported to completely inactivate the function of the enzyme without reducing the amount of OAT antigen(1). As all patients had at least one L402P allele, it may have diluted the effects of the other alleles, explaining the residual enzyme activity in one patient with the R180T/L402P genotype. No distinct association between the clinical course of the disease and the mutation type was evident. Lymphocyte OAT activities that were below the detection limit of the assay were measured in all mutation types, whereas values up to 9% of the mean value of the healthy controls were found in patients with L402P/L402P, R180T/L402P, and N89K/L402P mutations. The two patients with L402P/x mutation showed no detectable OAT activity. The enzyme activities varied equally widely in all three defined GA genotypes, indicating that residual enzyme activity, at least in peripheral blood lymphocytes, is influenced also by other factors in addition to the mutation. OAT activity is under complex tissue-specific hormonal and nutritional regulation(2325), and part of the interindividual variation in residual OAT activities in patients with GA may in fact be caused by normal physiologic control mechanisms. Interestingly, none of the patients who were receiving experimental therapy showed any measurable OAT activity, suggesting that dietary arginine restriction had no effect on OAT activity, even though plasma ornithine concentrations decreased markedly.

Most tissues constitutively express some OAT activity. Liver, kidney, and retina are particularly OAT-rich tissues. Liver and kidney OAT probably take care of most of the overall ornithine degradation in the body, and thus may strongly modulate plasma ornithine concentrations. In GA, retinal OAT activity may ultimately determine the rate of clinical progression of the characteristic ocular manifestations. Minor differences in residual OAT activities in these key tissues may markedly influence the progression of the disease, whereas they may be less evident in other tissues with inherently lower OAT activities, e.g. fibroblasts or lymphocytes. This might explain the apparent lack of correlation between plasma ornithine concentration, OAT activity, GA genotype, and phenotype.

Mean lymphocyte OAT activity of the GA heterozygotes was approximately 50% of the control values, as expected in a disease with autosomal recessive inheritance. However, individual values of the heterozygotes and the healthy controls overlapped. Plasma ornithine concentrations also fail to differentiate GA carriers from healthy controls. Thus, OAT alleles have to be determined for definite carrier identification. However, as several OAT mutations remain uncharacterized, OAT activity has a role in identification of the heterozygotes and possibly also in prenatal diagnosis of GA in individual cases. Our OAT assay is sensitive enough for analysis of villus biopsy cell lysates, where the benefits of rapid and sensitive assay are essential. Furthermore, early identification of neonates with GA is important, as temporary early hypo-ornithinemia with arginine deficiency and hyperammonemia may develop in man as well as in the transgenic mice, and the symptoms may warrant specific treatment(2).

In conclusion, our HPLC assay of OAT activity in lymphocytes is technically easy, rapid, and sensitive. It recognizes GA homozygotes reliably, but OAT activities of carrier heterozygotes and healthy controls overlap, restricting the value of the assay in carrier detection. The various Finnish OAT mutations investigated (with the exception of L402P/x) showed similar ranges of residual OAT activities, suggesting that other factors in addition to the genotype regulate OAT activity in human tissues.