Holocarboxylase synthetase deficiency: A biotin-responsive organic acidemia

May 1980 The Journal of P E D I A T R I C S

845

Holocarboxylase synthetase deficiency: A biotin-responsive organic acidemia The clinical and biochemical features of an infant affected by holocarboxylase synthetase deficiency are presented. The patient was the sibling of the deceased child in whose cultured skin fibroblasts the precise enzymatic disorder was first determined. This fact permitted administration of specific therapy in the form of oralbiotin, resulting in immediate improvement from impending respiratory failure and shock. The clinical response to biotin was accompanied by recovery of the biochemical meehanisms known to be biotin-dependent, as manifeste d by disappearance of intermediates in urine and blood. The variability of biotin responsiveness and the diversity of clinical presentation in the patients originally thought to have a deficiency of beta methylerotonyiCoA carboxylase, a biotin-dependent enzyme, raises the question of a separate, specific apoearboxylase defect.

K. S. Roth,* W. Yang, J. W. Foreman, R. Rothman, and S. Segal, Philadelphia, Pa.

A N U MBER of patients have been reported to date with a putative fl-CH3 crotonyl CoA carboxylase deficiency.1"6 The clinical findings in these patients have been protean and no uniform pattern of clinical disease has emerged. Enzymatic confirmation of the diagnosis in these patients has yielded variable results, leading to confusion as to the biochemical basis for the illness. Since our initial report of a patient with beta methylcrotonic acidemia, 6 we have been involved in a collaborative effort to document the molecular defect. The results of these investigations revealed that our patient had a defect in the enzyme that conjugates biotin into apoprotein (holocarboxylase synthetase) leading to impairment of multiple biotin-dependent pathways.' Since our original patient also had severe cyanotic heart disease and died, an adequate in vivo trial of biotinFrom the Departmeflts of Pediatrics and Medicine, University of Pennsylvania School of Medicine and the Children's Hospital of Philadelphia. Supported in part by Grant HDO7107 from the National Institutes of Health, and 1-638from the National Foundation-March of Dimes. *Recipient of a Research Career Development Award K04 HDOO257from the National Institutes of Health. Reprint address: Dr. K. S. Roth, Children's Hospital of" Philadelphia, Division of Biochemistry Development and Molecular Disease, 34th St. & Civic Center Blvd., Philadelphia, PA 19104.

0022-3476/80/050845+05500.50/0 9 1980 The C. V. Mosby Co.

responsiveness could not be undertaken. Nonetheless, the in vitro findings' clearly indicated that the disease should respond to administration of biotin, as in other vitamindependent disorders? We have had the opportunity, owing to the birth of another affected infant in the same family, to document the clinical characteristics of the disease and to observe the improvement attributable to large doses of biotin. Abbreviation used DIC: disseminated intravascular coagulation

CASE REPORT The patient was a male 2,220 gm product of an uncomplicated pregnancy and elective cesarean delivery. The mother, a healthy 25-year-old gravida 4, para 5 abortus 0 woman, was given supplemental biotin in the last trimester of pregnancy because of the history of two previously affected infants in this family, The results of Organic acid analyses by gas chromatography of cord blood and urine were normal during the first seven days of Life. Consequently, the infant was not given biotin following birth. The baby was discharged with the mother on the seventh hospital day on a combination of breast and proprietary formula feedings. CJrowih and development proceeded normally until 3 months of age. Over the week prior to admission, the parents noted increasing irritability accompanied by "fussy feeding" with the development of a "skin rash" and the odor of "cat's urine"

VoL 96, No.5 pp. 845-849

846

Roth et al.

The Journal of Pediatrics May 1980

Table I

Urine Plasma (mg/ dl) Free acids Acetic Propionic Hydroxyacetic Isobutyric Isovaleric Derivatized Lactic Alpha hydroxybutyric Beta hydroxybutyric Acetoacetic Beta hydroxyisovaleric Beta hydroxypropionic Beta ketovaleric Hippuric Beta CH~-glutaconic Beta OH-beta-CH~-glutaric Beta CHa crotonylglycine Alpha CH~-betahydroxybutyric p-OH-phenyllactic p-OH-phenylpyruvic p-OH-phenylacetic Glutaric Adipic

0.4 5.1 Trace Trace

mgldl

Postbiotin (mg/dO

0.7 0.6 0.3

86.2 6.7 40.4 53.9 10.2 0.8 6.8

emanating ]'rom soiled diapers. The baby was referred to Chil dren's Hospital because of lethargy and tachypnea unaccompanied by. fever, vomiting, or diarrhea. Physical examination on admission revealed a moribund, ashen male infant with marked tachypnea (80/minute) and chest wall retractions. His rectal temperature was 34~ and blood pressure was 45/20. Otherwise, his physical examination was unremarkable, except for a generalized erythematous, exfoliative dermatitis, particularly marked over chest, face, and scalp, and poor skin turgor. Laboratory data on admission were as follows: arterial blood pH 6.99, Pr 14 torr, base excess-38 mmoles, serum Na + 142 mEq, K + 4:4 mEq, C 1 118 mEq, total CO22 mmoles, BUN 20 mg/dl, glucose 170 rng/dl, hemoglobin 10.8 gm, WBC 10,900 cells/ram 3, platelets 220,000; CSF protein 94 mg/dl, glucose 95, and CSF cell count unremarkable. Marked ketonuria was noted on urinalysis. Serum lactate concentration was 8.42 mM, pyruvate 0.368 mM, acetoacetate 3.6 mM, and fl-OH-butyrate 2.44 mM. The infant was given fluids containing sodium bicarbonate intravenously and 2.5 mg of biotin every six hours via a nasogastric tube. A diagnosis of bacterial sepsis was initially considered, but was ruled out when cultures of blood, CSF, and urine were found to be sterile. After the dehydration was corrected, the hemoglobin concentration was noted to have fallen to 9.0 gm/dl; because of the clinical picture of shock, a packed erythrocyte transfusion was given. The platelet count fell to 51,000/mm ~. After 24 hours of alkali and biotin therapy, his respirations had decreased from 80/minute to 40/minute, and he was alert and vigorously taking

Prebiotin (mg/dl)

238.9 29.1 172.1 Trace 118.1 6.5

0.20 Trace 0.24 Trace 2.29 Trace

Trace 6.4 0.8 10.5 6.8 3.1 0.7 4.8 3.8 5.7

Trace Trace Trace Trace Trace Trace Trace Trace Trace Trace

oral feedings. Ptatelet count returned to normal without platelet transfusions. There was a similarly dramatic improvement in his acid-base values. After 48 hours of hospitalization he was given a dilute formula deficient in branch chain amino acids. Feeding was cautiously advanced to a standard proprietary formula, which he tolerated without recurrence of acidosis. He was then discharged, receiving biotin 2.5 mg four times a day and a proprietary formula. Over the ensuing two weeks, his body weight increased 800 gm. Six months after discharge, he was thriving and developing normally.

METHODS Organic acids a n d metabolites were identified with gas c h r o m a t o g r a p h y - m a s s spectrometry ( G C - M S ; LKB9000). Q u a n t i t a t i o n of free s h o r t : c h a i n fatty acids from a n ethyl ether extract o f acidified urine was p e r f o r m e d o n a Hewlett Packard 402A gas c h r o m a t o g r a p h e q u i p p e d with U l t r a b a n d o n 100/120 C h r o m a s o r b W A W c o l u m n (Applied Science Laboratory, State Co!lege, Pa.)? Hydroxy, keto, dicarboxylic, a n d glycine-conjugated acidic m e t a b olites were extracted a n d silylated 9 with N,O-bis (trimethylsilyl) trifluoroacetamide (Regis Chemicals Co., M o r t o n Grove, Ill.) a n d trimethylchlorosilane (Supelco, Bellefonte, Pa.) a n d q u a n t i t a t i o n was p e r f o r m e d o n Hewiett Packard 5711 gas c h r o m a t o g r a p h e q u i p p e d with 3% OV-1 o n 100/120 Gas C h r o m Q c o l u m n ( A p p l i e d Science L a b o r a t o r y ) ? Crotonic a n d h e n d e c a n e d i o i c acids were

Volume 96 Number 5

Holocarboxylase synthetase deficiency

847

I00

80

8O 7O

7O

6O

6O 5O

501

I

13

4oi

40 t5 ~0

II

3O !

14

zO 16

20 I

20

\

'01

I0 O

0 x 10240 ~

~3~o

q

b

.32o

I

Fig. 1. Gas chromatograph of TMS/BSTFA derivatives of urinary organic acids and metabolites. A, Urine obtained on admission (prebiotin therapy) and B, urine obtained 72 hours after initiation of biotin therapy. The urine was processed as outlined in Methods section. Gas chromatographic method utilized is identified in Methods section with temperature program of 60~ isothermal for 4 minutes followed by 4 ~ increase to final temperature of 180~ The following peaks were identified by comparison of their mass spectra with literature values. Peak 1 = Lactic, Peak 2 = 2hydroxybutyric, Peak 3 = 3-hydroxypropionic, Peak 4 = 3-hydroxybutyric, Peak 5 = 3-hydroxy-2-methylbutyric (acetoacetic-minor component), Peak 6 = 3-hydroxyisovaleric (+ methylamalonic), Peak 7 = 3-hydroxyvaleric, Peak 8 = Unknown, MW 246, Peak 9 = gultaric, Peaks 10 and 11 = 3-methylglutaconic, Peak 12 = adipic, Peak 13 = 3-methyl-(and 2-methyl) crotonyl gtycine, Peak 14 = 3-hydroxy-3-methylglutaric, Peak = 15 4-hydroxyphenylacetic, Peak 16 = 3-hydroxydecanoic, Peak 17 -~ suberic, Peak, 18 = homovanillic, Peak 19 = vanillylmandelic, Peak 20 = 4-hydroxyphenyllactic, Peak 21 = hendecanedioic (internal standard = 10 mg/dl), Peak 22 = 4-hydroxyphenylpyruic. Chromatogram A was run with initial attenuation set at 10240x with later change to 320x while chromatognim B was run with attenuation set at 320x at all times. The identity of Peak 13 as 3-CH~ crotonylglycine was confirmed by alkaline hydrolysis of the urine and repeat analysis of the free short-chain acids.

used as internal standards. Serum and C S F were deproteinized first and otherwise treated similarly to urine specimens. A m i n o acids were identified and quantitated by ion-exchange chromatography on a Beckman 119 automated amino acid analyzer. RESULTS A plasma sample, obtained immediately upon admission and prior to initiation o f biotin therapy, was examined by the procedure outlined in the Methods section. The short-chain fatty acids, including propionic, hydroxypropionic, and beta hydroxyisovaleric acids, which were found in this plasma sample are shown in Table I. The predominant short-chain acid was lactate, consistent with the clinical diagnosis of lactic acidosis. The ketone bodies were also prominent, which was reflected clinically by large amounts of urinary ketones. Unlike his sibling, this child had no free beta methylcrotonic acid identifiable in the plasma. ~ Examination of a simultaneously collected random urine revealed the presence of a great variety o f highly significant organic acids (Table I and Fig. 1). Chief among these was beta methylcrotonylglycine, the glycine conju-

gate of free beta methylcrotonic acid. Of equal import, however, was the finding of beta methylglutaconic and beta hydroxy-beta methylglutaric acids, metabolites whose presence indicates an incomplete enzymatic block at the beta methylcrotonylCoA carboxylase step (Fig. 2), as well as a partial or complete block at the acetylCoA carboxylase step. There was a marked elevation of branch chain amino acid levels in the CSF (Table 1I), most likely related to the elevation of plasma levels. Unfortunately, the admission plasma sample for amino acids was lost, but after 12 hours of biotin therapy, the plasma levels of branch chain a m i n o acids were normal. DISCUSSION The initial description of a defect in COx-fixation in leucine metabolism by Eldjarn et aP was characterized by the presence of beta-methylcrotonylglycine and beta hydroxyisovaleric acid in the urine. The disease was given the name beta methylcrotonylglycinuria and attributed presumptively to a defect in the beta methylcrotonylCoA carboxylase enzyme. At the time of our report of the first patient in the United States, 6 there had been two addition-

848

Roth et al.

The Journal o f Pediatrics May 1980 GLUCOSE

LEUCINE

ISOLEUCINE

VALINE

cc-KETOISOCAP ROATE ISOVALERYL-CoA

j

~-MET.YLCR0"ONY'-CoA-...~ B-METHYLGLUTACONYL'CoA

LACTATE"- PYRUVATE

ACETYfLICoA S

PROPIONYLCoA

,

TCA CYCLE

Fig. 2. Metabolic pathways affected by ffolocarboxylasesynthetase defects. Numbers refer to specific enzymes partially or completely blocked in this disorder. 1 = fi-methylcrotonic CoA carboxylase; 2 = pyruvate carboxylase; 3 = acetyl CoA carboxylase; 4 = propionyl CoA carboxylase.

Table II. Amino acid concentrations in cerebrospinal fluid Amino acid

Patient

Normal values ~ (Range)

Aspartate Threonine Serine Asparagine Glutamine Proline Glutamate Citrulline Glycine Alanine Valine 89Cystine Cystathionine Methionine Isoleucine Leucine Tyrosine Phenylalanine Ornithine Lysine Histidine Arginine

Trace 48 ND 6 230 Trace Trace 5 7 122 47 ND ND 5 20 48 18 10 Trace 84 19 13

1.6-7.6 9.9-51.3 13.1-70 161-533 0-4.2 0-117 1.6-19.5 12.6-36.6 3.2-26 1-4.3 2.4-14.8 5.6-17.7 1.2-11 2.4-10 3-8.2 13.4-42 2.4-31.4 5.8-29.3

Values are expressed as/*moles/liter; N D = not detected,

al reports from England of such affected children#, 3 Because of the wide disparity of clinical and biochemical findings in all four patients, we raised the possibilities of either genetically distinct diseases or acquired disorders of organic acid metabolism# We also questioned the use of beta hydroxyisovaleric aciduria as a diagnostic criterion for the presumptive enzyme defect, since beta hydroxyiso-

valerate is regularly elevated in the presence of ketoacido-

sis. l0 Subsequent work by Sweetman et a l y showing a propionylCoA carboxylase deficiency in the second patient reported by Gompertz et al, = suggested the possibility of a holocarboxylase synthetase deficiency, which would result in a secondary state of multiple biotindependent enzyme disease. Another report indicated that hypoxia in premature n e w b o r n infants can result in organicaciduria, including large amounts of p-hydroxyphenolic acids, such as we found in both of our patients.' In order to delineate the underlying mechanism in our earlier patient, enzymatic studies were performed in cultured skin fibroblasts. Results of these studies clearly indicated a holocarboxylase synthetase deficiency# The presence of massive amounts of lactate and pyruvate, as well as acetoacetate and beta hydroxybutyrate, can now be explained on the basis of a generalized decrease in CO2 fixation at steps 2 and 3 (Fig. 2). As a further consequence of a partial or complete block at step 2, it is likely that net synthesis of glucose from pyruvate is decreased, forcing pyruvate to enter the TCA cycle by conversion to acetylCoA. As a consequence of a block at step 3, entry of acetylCoA into lipid synthesis is restricted. The net effect of these metabolic derangements is to force catabolism of carbohydrate and of lipid through acetylCoA. The acetylCoA thus might be expected to slow conversion of pyruvate to acetylCoA due to feedback inhibition of the pyruvate dehydrogenase complexJ 3' 1~ Since pyruvate and acetylCoA together constitute the primary links between metabolism of carbohydrate, lipid and protein, impairment in disposition of one or both will be reflected in serious metabolic disturbances. A possible

Volume 96 Number 5 sequela of the decreased lipid synthesis is the c o m m o n finding of a typical dermatitis in these patients) "~ 16 The appearance o f beta methylglutaconic and beta hydroxy-beta methylglutaric acids neither of which were detectable in the earlier sib 6 and both of which are interposed between two CO2 fixation s t e p s - s e e m s to indicate an incomplete block at steps 1 and 3 (Fig. 2). The absence of methylmalonate is conspicuous among all the other organic intermediates present in this disease, and is attributable to the absence of normal propionylCoA carboxylase activity, which is necessary to the formation of methylmalonate. The severe metabolic acidosis may cause progressive anemia and thrombocytopenia prior to appearance of disseminated intravascular coagulation. The anemia and thrombocyt0penia in this patient are similar to the findings in our earlier patient with holocarboxylase synthetase deficiencyp in whom D I C was thought to be the cause of anemia and thrombocytopenia. However, the present patient had no clinical evidence of DIC, which suggests a different etiology for the hematologic disturbances. Patients with isovaleric acidemia frequently present with pancytopenia 17 owing to bone marrow underproduction. TM Presumably, isovaleric acid is toxic to bone marrow stem cells and therefore an excessive amount in blood would induce pancytopenia. A similar mechanism for one or several accumulated metabolites can explain the presence of anemia and thrombocytopenia in patients with holocarboxylase synthetase deficiency. The platelet count increased toward normal after the administration of biotin, paralleling the correction of metabolic acidosis. The variable clinical response to biotin in two of the previously reported patients I'~ tends to confirm our suspicion that what has been hitherto called beta methylcrotonylCoA carboxylase deficiency probably consists of at least two distinct disorders. The holocarboxylase synthetase deficiency has been confirmed in the present patient's sibling 7 and is responsive to administration of biotin. There may be a specific apocarboxylase deficiency which is unresponsive. In patients such as ours, who respond to administration of biotin, no dietary therapy would appear to be required. However, patients who fail to respond to biotin should be given a trial of a diet deficient in the branch chain amino acids, since they may represent the apocarboxylase deficiency. The rapid disappearance of urinary organic acid excretion within three days after the initiation of biotin therapy further illustrates the specific underlying biochemical defect of this disorder (Fig. 1). REFERENCES

1. Eldjarn L, Jellum E, Stokke O, Pande H, and Walter PE: /3-hydroxyisovaleric aciduria and fl-methylcrotonylglycin-

Holocarboxylase synthetase deficiency

2. 3.

4.

5.

6.

7.

8. 9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

849

uria: a new inborn error of metabolism, Lancet 2:521, 1970. Gompertz D, Draffan GH, Watts JL, and Hull D: Biotinresponsive/~-methylcrotonylglycinuria, Lancet 2:22, 1971. Gompertz D, Bartlett K, Blair D, and Stern CMM: Child with a defect in leucine metabolism associated with fl-hydroxyis0valeric aciduria and fl-methylcrotonylglyeinuria, Arch Dis Child 48:975, 1973. Keeton BR, and Moosa A: Organic aciduria: Treatable cause of floppy infant syndrome, Arch Dis Child 51:636, 1976. Finnie MDA, Cottrall K, Seakins JWT, and Snedden, W: Massive excretion of 2-oxoglutaric acid and 3-hydroxyisovaleric acid in a patient with a deficiency of 3-methylcrotonyl-CoA carboxylase, Clin Chim Acta 73:513, 1976. Roth KS, Cohn R, Yandrasitz J, Preti G, Dodd P, and Segal S: Beta-methylcrotonic aciduria associated with lactic acidosis, J PEDIATR88:229, 1976. Saunders M, Sweetman L, Robinson B, Roth K, Cohn R, and Gravel RA: Genetic complementation and multiple carboxylase deficiencies in biotin responsive organicademias, J Clin Invest 64:1695, 1979. Hitlman RE: Megavitamin responsive aminoacidopathies, Pediatr Clin North Am 23:557, 1976. Cohn RM, Updegrove S, Yandrasitz JR, Rothman R, and Tomer K: Evaluation of continuous solvent extraction of organic acids from biological fluids, Clin Biochem 11:126, 1978. Landaas S: Increased urinary excretion of 3-hydroxyisovaleric acid in patients with ketoacidosis, Clin Chim Acta 54:39, 1974. Sweetma L, Bates SP, Hull D, and Nyhan WL: PropionylCoA carboxylase deficiency in a patient with biotin-responsive 3-methylcrotonylglycinuria, Pediatr Res 11:1144, 1977. Bakkeren JAJM, Sengers RCA, Trijbels JMF, and Engels PHATh: Organic aciduria in hypoxic premature newborns simulating an inborn error of metabolism, Eur J Pediatr 127:41, 1977. Garland PB, and Randle PJ: Control ofpyruvate dehydrogenase in the perfused rat heart by the intracetlular concentration of acetyl-coenzyme A, Biochem J 91:6c, 1964. Batenburg J J, and Olson MS: The inactivation of pyruvate dehydrogefiase by fatty acid in isolated rat liver mitochondria, Biochem Biophys Res Commun 66:533, 1975. Cowan M J, Packman S, Wara DW, Ammann AJ, Yoshino M, Sweetman L, and Nyhan W: Multiple biotin-dependent carboxylase deficiencies associated with defects in T-cell and B-cell immunity, Lancet 2:115, 1979. Charles BM, Hosking G, Green A, Pollitt R, Bartlett K, and Taitz LS: Biotin-responsive alopecia and developmental regression, Lancet 2:118, 1979. Levy HL, Erickson AM, Lott IT, and Kiertz DJ: Isovaleric acidemia: results of family study and dietary treatment, J P~DIAIR 52:83, 1973. Kelleher JE Jr, Yudkoff M, Hutchinson R, August CS, and Cohn RM: The pancytopenia of isovaleric acidemia, Pediatrics (in press). Scriver CR, and Rosenberg LE: Amino acid metabolism and its disorders, Philadelphia, 1973, WB Sannders Company, p 54.