Gerald Lancaster

Quantitative analysis of branched-chain α-keto acids as their trimethylsilylated oximes

Abstract The oximes of the branched-chain α-keto acids have extraction efficiencies and gas Chromatographic properties after trimethylsilylation superior to those of the original acids. Comparison of the chromatograms of extracted serum and urine before and after treatment with hydroxyiamine permits unequivocal identification and correction for co-eluting peaks even at low concentrations (about 10 μM). The mass spectra of these derivatives have abundant even-numbered fragment ions making possible their detection in the nanogram range. The method is useful for the investigation of branched-chain α-keto acid metabolism. Address requests for reprints to: Dr. O.A. Mamer, Royal Victoria Hospital, Montreal 112, Quebec, Canada.

A γ-aminobutyrate pathway in mammalian kidney cortex☆

Abstract Rat kidney cortex converts l -glutamate to γ-aminobutyrate by a decarboxylation reaction which differs from the corresponding reaction in brain. Renal l -glutamate decarboxylase has two apparent K m values for glutamate in homogenates (0.4 and 2.5 mM). γ-Aminobutyrate is converted by a transaminase whose capacity appears to exceed the decarboxylase. γ-Aminobutyrate is converted ultimately to succinate and CO 2 . γ-Aminobutyrate stimulates respiration of kidney cortex slices in vitro and the compound crosses cell membranes in kidney by a respiration-linked, mediated process. Chronic acidosis lowers renal γ-aminobutyrate in the rat; brain γ-aminobutyrate is unaffected by acidosis. Glutamic acid decarboxylase and γ-aminobutyrate transaminase activities are unchanged in acidosis. α-Methylglutamate, an inhibitor of renal glutaminase, lowers the γ-aminobutyrate and glutamate content of rat kidney in normal and acidotic states. Aminooxyacetic acid in vivo , an inhibitor of γ-aminobutyrate transaminase, causes a striking increase in renal γ-aminobutyrate during chronic acidosis. At concentrations of glutamate in vitro , which are similar to the tissue glutamate content in vivo , the γ-aminobutyrate pathway accounts for approximately one-fourth of glutamate disposal in rat kidney cortex slices.

Branched-chain alpha-keto acids isolated as oxime derivatives: Relationship to the corresponding hydroxy acids and amino acids in Maple Syrup Urine Disease

Abstract Branched-chain α-keto acids can be isolated from serum and urine with good recoveries when treated first with hydroxylamine. Silylation of the keto acid oximes then yields clearly identified peaks after gas chromatography. Serum branched-chain keto acids are increased ten fold above normal in patients with Maple Syrup Urine Disease (MSUD) due to branched-chain α-keto acid dehydrogenase deficiency, when the level of branched-chain amino acids are between the normal range (0.05-0.2 m M ) and 1 m M . Keto acids rise another 10–20-fold (i.e., up to 200 times normal), when serum amino acid levels are permitted to rise only twofold above 0.8–1.0 m M . The acute clinical syndrome of MSUD coincides with the latter severe accumulation of keto acids. Removal of branched-chain keto acids by transamination in vivo to the corresponding amino acids seems to be most effective when serum amino acid concentrations are below 1 m M . Keto acids can also be excreted into urine, a mechanism in which they have preference over the corresponding amino and hydroxy acids. Accumulation of branched-chain α-hydroxy acids in MSUD is significant only for α-hydroxy- is ovaleric acid, as reported previously by others. Thiamine-responsive MSUD is characterized by greater accumulation of α-keto-B-methyl-valerate relative to the two other branched-chain keto acids when compared with keto acid ratios in classical MSUD. It has been shown by Elsas et al. 21 that thiamine preferentially stimulates the oxidation of isoleucine and its keto acid in cultured skin fibroblasts. Branched-chain keto acid concentrations are sufficient in uncontrolled MSUD to permit inhibition of pyruvate dehydrogenase according to current evidence in vitro. However, there is no abnormal retention of pyruvate or lactate in vivo in the acute clinical syndrome. Dietary control which permits serum branched-chain amino acids to rise above the normal range and yet remain below 0.8–1.0 m M may be as effective as more stringent dietary measures.

Dominantly inherited osteogenesis imperfecta in man: an examination of collagen biosynthesis.

Extract: We have examined control subjects and patients in an effort to discover a metabolic basis for dominantly inherited osteogenesis imperfecta (OI). Studies were carried out in vitro with cultured skin fibroblasts obtained from OI patients, and in vivo on peptide-bound hydroxyproline excretion in urine. Urinary hydroxyproline excretion (milligrams/24 hr) adjusted for age is essentially normal in OI patients, although the mean excretion rate is below average. The latter finding is presumably a reflection of the smaller body mass of OI patients.The OI skin fibroblasts, matched for age of donor, site of biopsy, phase of growth, and generation number in culture, incorporate L-proline into hot trichloroacetic acid (TCA)-soluble protein (collagen) at normal rates. The rate of conversion of proline to hydroxyproline in the nascent polypeptide is also normal in OI. Incorporation of L-lysine was also normal in OI. These findings indicate that peptide synthesis of collagen is not impaired in OI.Rates of galactose incorporation into collagen and the extractability of collagen into normal saline or 0.2 M citric acid were all normal both in OI cells and in the culture medium recovered from the monolayer. These findings, in combination with the urinary data on hydroxyproline excretion in vivo reveal that cross-linking and export of collagen in OI is essentially normal.The elution profile after ion exchange chromatography of fibroblast collagen on carboxymethyl (CM)-Sephadex was also examined. The normal 2/1 ratio of peak 1 (largely α1(I) chains) to peak 2 (largely α2 chains) was found in OI fibroblast extracts, which implies that synthesis and initial aggregation of the two types of polypeptide to yield [α1(I)]2α2 collagen composition is not abnormal in OI.Despite the negative biochemical findings, a consistent defect in the morphology of OI cells was identified in the log phase and the confluent phase of monolayer cultures. The finding is characterized by irregular packing of the aggregated cells and by an irregular tessellated appearance of the individual OI fibroblast. This observation reassures us that the inherited defect is expressed in vitro.Speculation: An abnormality in the primary sequence of polypeptide chain in collagen would be compatible with all of our findings and with the genetics of OI. The mutant allele would affect only about half the products, under the control of only one of the loci determining the polypeptide sequences in collagen chains. Because the OI allele is not expressed in cartilage, a tissue without α2 collagen chains, the defect in OI would perhaps be found in the α2 polypeptide. However, since the α1(II) chain of cartilage differs in amino acid composition and in hydroxylysine-linked carbohydrate from the α1(I) chains of noncartilagenous structures, a defect in α1(I) chains at the nonhomologous residues will also require investigation.

Secreted Collagen Ratios in Normal Human and Osteogenesis Imperfecta Skin Fibroblasts

We examined the effects of several variables on the ratio of type I:type III collagen secreted by human caucasian skin fibroblasts in normal and osteogenesis imperfecta (OI) phenotypes. Isotopically labelled collagen extracted from fibroblast medium was analyzed by DEAE-cellulose chromatography and identified by appropriate methods. Type I procollagen was the major form of collagen secreted into the medium by normal cells cultured from one mid-term fetus, infants (n = 3), children (n = 3), adolescents (n = 2), and adults (n = 3). Interstrain differences in collagen production under standardized conditions were significantly greater than intrastrain variation (anova, p = 0.0051). There was no significant alteration in the type I:type III collagen ratio due to variation in: phase of cell growth, doublings (between 13th and 22nd), rate of isotope incorporation, labelling time (24-72 hrs) in the presence of ascorbic acid (50 micrograms/ml), age of donor (with the possible exception of adolescence), and site of biopsy (genital and non-genital sites). Variable conversion of type I procollagen to collagen did not perturb the type I:type III collagen ratio. Cell strains from OI patients (Sillence classification): type I (one strain); type II, III and IV (3 strains each) had greater interstrain than intrastrain variation in the collagen ratio (p = 0.0149). Interstrain differences were greater in OI cell strains relative to normal cell strains (p less than 0.01). In the aggregate, OI cells had significantly lower type I collagen production relative to type III (I/III ratio = 1.18) when compared with normal cells (I/III ratio = 2.90; t test, p less than 0.0001). These findings imply abnormal synthesis, secretion or stability of type I procollagen and greater phenotypic heterogeneity in OI skin fibroblasts relative to normal cells.

Ontogeny of L-Glutamic Acid Decarboxylase and γ-Aminobutyric Acid Concentration in Human Kidney

Extract: Mature human renal cortex contains γ-aminobutyric acid (GABA) at concentrations on the order of 0.2 μmol/g wet wt and about half of that concentration in fetal kidney.The pyridoxal-5′-phosphate (PLP)-dependent enzyme, L-glutamic acid decarboxylase (GAD), catalyzes the conversion of L-glutamate to GABA. The PLP-saturated GAD activity in post-term (1 day–9 years) renal cortex homogenate is 0.94 ± 0.38 μmol CO2 formed/g wet wt/hr (Table 1). The corresponding GAD activity in fetal renal cortex, in the midtrimester and early third trimester, is 0.28 ± 0.19 μmol CO2 formed/g wet wt/hr (P < 0.01 vs. post-term activity). GAD activity increases during gestation to reach levels at birth equivalent to mature kidney. When corrections are made for the difference in tissue water between fetal and mature kidney, they do not ablate the post-term increase in GAD activity.Renal GAD was undersaturated with PLP in 12 of 21 specimens examined; 8 of these specimens were of fetal origin (Table 1 and Fig. 2). Undersaturation was revealed when addition of PLP (0.24 mM) to the homogenate stimulated activity by 25 300% in the 12 specimens. GAD activity in rat kidney binds PLP avidly and is always saturated with PLP under the normal laboratory conditions of vitamin B6 nutrition. Undersaturation of renal GAD in man may imply inadequate endogenous availability of vitamin B6 coenzyme.Speculation: The GABA pathway in mammalian kidney has been implicated in the maintenance of normal acid-base balance. Renal GAD is increased in the postnatal subject relative to fetal GAD activity in support of this hypothesis. Undersaturation of fetal GAD with its PLP-coenzyme in two-thirds of the fetal specimens may have implications concerning intrauterine nutrition and renal development.

Inherited Lactic Acidosis: Correction of the Defect in Cultured Fibroblasts

ABSTRACT. We report a case of familial lactic acidosis, lethal in the newborn period. Studies in intact fibroblasts identified a defect in the oxidative pathway of pyruvate metabolism. Although assay of pyruvate dehydrogenase on cell sonicates was not appreciably reduced, flux through the enzyme and other mitochondrial multienzyme dehydrogenases was severely impaired in intact cells. Deficient lactate conversion to carbon dioxide could be repaired by the addition to the incubation medium of electron acceptors such as methylene blue (25 μg/ml) or dichlorophenolindophenol (25 μg/ml).

737 DITHIOTHREITOL (DTT) PROTECTS CYSTINOTIC FIBROBLASTS IN CYSTINE-FREE MEDIUM

Nephropathic cystinosis is associated with extreme intracellular (lysosomal) storage of cystine (CySS). Reduced dithiothreitol (DTTSH) and cysteamine both decrease cellular CySS in cystinosis patients in vivo and may ameliorate the natural course of the disease. The mutant cellular function remains undefined. We studied survival of normal (n=4) and cystinotic (n=3) fibroblast lines in selective medium (CySS-free) with or without DTTSS or DTTSH in the medium. Control and cystinotic cells were matched for site of biopsy, age of donor, and passage number in culture. Cells plated at low density were washed and refed at 24 h with either control or selective media. Cell density was measured after further incubation (24 h). CySS-free medium was selective and caused detachment (>90%) of cystinotic and control cells; DTTSS (1-8 mM) and DTTSH (0.5-2 mM) both protected cystinotic cells but not control cells in the selective medium. Cystinotic cells preincubated with cysteamine (1 mM × 4 h) were not protected by DTT in CySS-free medium. These findings imply that cystinotic cells have the capability (presumably cytoplasmic in origin) to reduce DTTSS and use it to liberate CySS from their expanded intracellular pools to support growth. If CySS accumulation, or other events related to abnormal -SH metabolism are important facets of the cystinosis phenotype, there is a rationale for treatment with DTT, and like agents, in vivo.

718 EFFECT OF METHYLENE BLUE ON FIBROBLASTS IN LACTIC ACIDOSIS

A girl presented on the 9th day of life with familial lactic acidosis, euglycemia and hepatomegaly; death occurred at day 15. Ratios for plasma lactate: pyruvate (130:1) and B-hydroxy-butyrate: acetoacetate (38:1) were elevated suggesting an imbalance of tissue NAD:NADH. Plasma pyruvate (0.25mM, Norm = <.07mM) and alanine (1.7mM, Norm = < .4mM) were both elevated indicating a defect in pyruvate utilization. Organic acids other than those above were not elevated in urine. Skin fibroblasts produced excess lactic acid in culture. Evolution of 14-CO2 (nmoles/mg prot · hr ± S.D.) from 1mM L-114C lactate (control = 19.8 ± 9.4), L-114C pyruvate (control = 41.1 ± 4.7) or L-U14C alanine (control = 12.7 ± 4.1) was markedly decreased compared to control cells (< 10%, < 1% and <10% respectively). Thiamine (0.7mM) or lipoic acid (0.01mM) for 1 or 24 hours of preincubation had no effect. Dichloroacetate (50μM) produced a modest increase of lactate utilization (150% increase) in the patients cells, but the effect of methylene blue on alanine utilization (0.27mM) was dramatic (1200% increase), bringing it within the normal range (7.8-20.5nmoles CO2/mg prot · hr).Our findings suggest that methylene blue can repair lactic acidosis in vitro and may be of use in patients whose cellular regeneration of NAD is impaired.

AGE-DEPENDENT ACTIVITY OF L-GLUTAMIC ACID DECARBOXYLASE (GAD) IN HUMAN KIDNEY

The conversion of L-glutamate (L-glu) to y-amino-butyrate (GABA) in mammalian brain & kidney is catalyzed by pyridoxal-phosphate (PLP)-dependent GAD. GABA is an important neuro-inhibitor in CNS, but another role must exist for GABA metabolism in kidney cortex. We assayed GAD by an isotopic method (Lancaster et al, BBA 297:229, 1973) at normal tissue L-glu cone. (4-8mM) ±PLP (0.24mM) in frozen postmortem human kidney obtained between 15wk gestation and lOyr.GAD is stable in tissues or homogenates stored at -20°C; PLP-saturated renal GAD activity is low (<0.5nmoles CO2 formed/rag wet wt·h) before 40wks gestation, then rises at term and plateaus in the infant & child at 5-10x fetal level. Kidney GABA cone, parallels GAD activity. Endogenous GAD activity is often inappropriately low for age, but can be stimulated by PLP in vitro, indicating, undersaturation by co-enzyme in vivo; once bound in vivo, PLP is not easily removed from GAD in vitro. Augmentation of renal GAD activity at term parallels the need for human kidney to extract glutamine from plasma to support renal ammoniagenesis. The GABA pathway permits disposal of the residual glutamate carbon chain as a non-titratable acid.