Novel Heterozygous Mutations in TALDO1 Gene Causing Transaldolase Deficiency and Early Infantile Liver Failure

Transaldolase (TALDO) deficiency (OMIM #606003), a recently recognized new inborn error of the pentose phosphate pathway (PPP), has been reported to date in only 10 patients from 6 families, presenting primarily with liver disease and variable clinical course (1–6). The PPP is a multifunctional pathway that plays a pivotal role in supplying ribose-5-phosphate for nucleic acid biosynthesis and for generation of reducing equivalents, which regulate the cellular redox potential via maintenance of the reduced glutathione (GSH) pool and neutralization of reactive oxygen intermediates (ROIs). TALDO is one of the key enzymes responsible for this neutralization. TALDO deficiency was first described in 2001 (1) and has a broad phenotypic heterogeneity, ranging from fetal hydrops (3) to slowly progressive liver cirrhosis (1). The leading symptoms in the neonatal period in all 10 patients reported were bleeding tendencies with thrombocytopenia, abnormal liver function tests, hepatosplenomegaly, hemolytic anemia, and dysmorphic features (antimongoloid slant, low-set ears, and cutis laxa). Interestingly, mental development and motor development were normal in the majority who were assessed (3 patients died before 6 months of age). The biochemical profile of all of the affected patients is similar: elevated urine erythritol, arabitol, ribitol, sedoheptitol, perseitol, sedoheptulose, mannoheptulose, and sedoheptulose-7P (1,2,7). Measurement of TALDO activity in fibroblasts, lymphoblasts or liver tissue, and sequence analysis of the TALDO1 gene confirm the diagnosis. We report an additional patient with TALDO deficiency with the hope of increasing awareness of this rare disorder that should be considered in the differential diagnosis of pediatric liver diseases. We also review the literature on TALDO deficiency. CASE REPORT The patient was the first child of nonconsanguineous Chinese parents delivered by emergency cesarean section at 36 weeks gestation for fetal distress with a birth weight of 2.85 kg. The pregnancy was otherwise uncomplicated. He presented at birth with hydrops and mild pallor, and was initially suspected to have fetomaternal transfusion. His hemoglobin was 12.7 g% and albumin, 20 g/L (normal 30–54). Albumin was infused twice and discharged well 2 days later from the primary care hospital. He presented again at 2.5 months of age with a febrile illness and was noted to have splenomegaly. Baseline investigations including liver function, blood indices, and kidney functions, which were normal. He was discharged after 3 days, only to be readmitted 3 weeks later with seizures. He was mildly jaundiced and had generalized edema. Serum bilirubin was raised, albumin was low, and transaminases were elevated (data unavailable). Albumin was infused twice and he was treated for presumed sepsis. He was discharged 4 days later, but was readmitted shortly thereafter and subsequently referred to our tertiary center for further management. Clinically he was pale, jaundiced, and had discrete dysmorphic features (Fig. 1), cutis laxa with visible vascular network, generalized edema, and a left inguinal hernia. The abdomen was grossly distended with dilated visible veins. He had a grossly enlarged spleen palpable approximately 8 cm below the left costal margin and a liver that was felt just below the right costal margin. He required ventilatory support for respiratory distress and cardiac failure secondary to a moderate-sized atrial septal defect. He had Coombs negative hemolytic anemia (hemoglobin was 5.6 g/dL), thrombocytopenia 62,000/μL, and hepatic dysfunction with coagulopathy; activated partial thromboplastin time 127.2 seconds (control 30.9–45.9 seconds), prothrombin time 62.5 seconds (control 11.9–14 seconds), fibrinogen 172 mg/dL (normal 212–457), D dimmer, 2.90 μg/mL (normal 0.0–0.8), serum albumin 22 g/L (normal 30–54), and mildly elevated transaminases; and alanine aminotransferase 110 U/L (normal 0–54), aspartate aminotransferase 110 U/L (normal 0–82). Maximum total bilirubin was 795 μmol/L (normal 0–17) with conjugated bilirubin at 339 μmol/L and raised alkaline phosphatase at 1194 U/L (normal 0–522). Gamma-glutamyl transferase was normal at 19 U/L (normal 11–50). Maximum ammonia was 140 μmol/L.FIGURE 1: Discrete dysmorphic features including hirsutism (forehead), hypertelorism, and prominent philtrum.He received ventilation separately for a possible cytomegalovirus (CMV) pneumonitis and respiratory syncytial virus (RSV) bronchopneumonia. He had recurrent infections including rotavirus acute gastroenteritis and several episodes of septicemias with group B streptococci, acinetobacter, and flavobacterium, respectively. Unfortunately, we had not investigated for a possible association with immunodeficiency. Galactosemia was initially considered a differential diagnosis because the galactose-1-phosphate uridyl transferase (GALT) was mildly reduced at 2.6 U/g hemoglobin (normal >3), but the total galactose was normal at 91 μmol/L (normal <540). These were posttransfusion results; hence, it was speculated that the total galactose may not have been raised secondary to the transfusion. A diagnostic workup for early infantile liver failure was performed and included normal transferrin isoelectric focusing, serum ferritin, serum copper, ceruloplasmin, urine porphyrin, and serum lactate. No evidence was found for tyrosinemia, defects in organic and amino acid metabolism, or fatty acid oxidation defects. Infective screens were nonreactive for hepatitis B, C, HIV, immunoglobulin(Ig)G, and IgM for rubella and toxoplasma. CMV IgM was positive. Ultrasound of the abdomen demonstrated liver cirrhosis and portal hypertension with splenomegaly and presence of splenorenal collaterals. Doppler examination excluded any portal vein thrombosis. The kidneys were normal. Brain ultrasonography did not show any obvious bleeding. Opthalmological examination was normal. The clinical course was characterized by an intractable liver failure and recurrent septicemias. He finally succumbed to flavobacterium sepsis and multiorgan failure at age 4.5 months. Postmortem liver biopsy demonstrated features of fulminant liver failure with marked distortion of its architecture, liver fibrosis, and the presence of pseudoglandular formations with loss of interlobular bile ducts. Steatosis, iron accumulation, or inclusions were absent (Fig. 2A and B). Urine collected antemortem at 4.5 months was stored and subsequently analyzed for polyols and sugars.FIGURE 2: (A) Marked distortion of liver architecture with vague nodule formation consisting of degenerating and regenerating hepatocytes, the latter displaying pseudoglandular formation. Note presence of hepatocellular and canalicular cholestasis (H&E; original magnification ×200). (B) Extensive fibrosis noted within the parenchyme MT; original magnification ×100.MATERIALS AND METHODS Metabolite Analyses Urine sugars and polyols were measured according to published methods (7,8) using liquid chromatography-tandem mass spectrometry and gas chromatography with flame ionization detection. Mutation Analysis Mutation analysis of the TALDO1 gene (GenBank Accession No. L1943.2) was performed by direct sequence analysis of genomic DNA from blood (1). RESULTS Metabolite Analyses In urine obtained at 4.5 months of age, an abnormal sugar and polyol profile was detected with highly increased excretions of erythritol, arabitol, ribitol, erythrose, xylulose, and ribose. More important, elevated excretion of the 7-carbon sugars suggested TALDO deficiency. Mutation Analysis Analysis of the TALDO1 (NM_006755.1) gene revealed 2 novel heterozygous missense mutations. One mutation consists of a deletion of 3 nucleotides in exon 7 (c.895_897delAAC) that resulted in the deletion of asparagine at position 299 (p.Asn299del). This mutation was inherited from the father; the other mutation consists of a G > A transition (c.931G>A) in exon 7, which resulted in the substitution of glycine by arginine at position 311 (p.Gly311Arg) and was inherited from the mother. Asparagine at position 299 and glycine at position 311 are highly conserved in evolution. Both mutations were not detected in 210 control alleles. These data suggest that the mutations are pathogenic. DISCUSSION The PPP is a series of interconversions of sugar phosphates that operates exclusively in the cytoplasm. The pathway has 2 distinct phases: the oxidative, nonreversible branch that results in NADPH and pentose phosphate generation; and the nonoxidative branch, which creates a reversible link between the PPP and glycolysis (9) (Fig. 3). TALDO is the second enzyme of the nonoxidative phase and a deficiency of this enzyme results in elevated concentrations of polyols and 7-carbon chain carbohydrates in body fluids.FIGURE 3: Schematic illustration of the pentose phosphate pathway (PPP) and conversions to sugars and polyols. The dashed arrows indicate unverified, presumed reactions in humans. NADP = nicotinamide adenine dinucleotide phosphate; NADPH = reduced form of NADP; RPE = ribulose-5-phosphate epimerase; RPI = ribose-5-phosphate isomerase.TALDO deficiency has been diagnosed in only 6 families (10 patients: 3 girls and 7 boys) to date. The patients were born to consanguineous Turkish, United Arab Emirates', Pakistani, and Polish couples (1–6). Verhoeven et al (1) described the first patient with TALDO deficiency who presented in the neonatal period with low birth weight, aortic coarctation, mild bleeding problems, and enlarged clitoris. Within several months, this girl developed hepatosplenomegaly. At the age of 2 years, a liver biopsy showed micronodular liver cirrhosis with no specific characteristics. She was diagnosed at the age of 10 years by the analysis of sugars and polyols in urine, enzymatic assay in erythrocytes, and mutation studies of the TALDO gene. She had thrombocytopenia, mildly increased coagulation times, and intermittently elevated ammonia. Her neurological and intellectual development has been normal. The second patient presented with a more severe clinical course. She was the first of dizygotic female twins born to consanguineous Turkish parents. She presented with neonatal liver failure and Coombs negative hemolytic anemia and displayed generalized edema, moderate hypotonia, and discrete dysmorphic signs (down-slanting palpebral fissures, low-set ears, and increased intermamillary distance). Karyotype analysis demonstrated XX/XO mosaicism. The clinical course was characterized by intractable liver failure, progressive myocardial hypertrophy, and finally respiratory failure and severe lactic acidosis at the terminal phase. She died from bradycardic heart failure at age 18 months (2). TALDO deficiency has subsequently been reported in a further 8 patients (Table 1). The severities of the clinical manifestations vary greatly between patients; however, liver disease was present in all. The predominant features included hepatosplenomegaly, hemolytic anemia, dysmorphic features (eg, antimongoloid slant, low-set ears, and cutis laxa), neonatal edema, congenital heart defects, renal problems (tubulopathy, renal failure, nephrocalcinosis), and/or intermittent hypoglycemia. Interestingly, mental development and motor development were normal in the majority where assessment was possible (4 patients died before the age of 6 months). Liver pathology includes hepatocellular damage and degeneration of hepatocytes or hyperplastic hepatocytes with fibrosis (9). TALDO has been recognized as a regulator of apoptotic signal processing (10), which may contribute to the pathogenesis of the liver disease. The study by Banki et al (10) provides evidence that TALDO plays an essential role in regulating sensitivity to cell death programs dependent on the formation of ROIs, via regulation of the PPP and reduced GSH production.TABLE 1: Clinical features of the 10 TALDO-deficient patients diagnosed to date in comparison to our patientIn a few individual cases, enlarged clitoris, microphallus, deafness, and rickets were noted. Of interest, 2 cases (4), including our patient, were diagnosed as having CMV pneumonitis, a viral pneumonia that occurs less frequently in immunocompetent hosts. In addition, our patient had various bacterial septicemias. A possible association with cellular and/or humoral immunodeficiency would, however, require further elucidation. The resulting GSH depletion, which has also been implicated in a wide range of viral infections (11), could offer a reasonable explanation. Many in vitro studies have demonstrated the role of GSH in attacking viruses having different replicative mechanisms and, inhibiting viral replication at various stages (12). The biochemical profile in all of the patients diagnosed to date is similar. TALDO deficiency exhibits elevated urinary erythritol, arabitol, ribitol, sedoheptitol, perseitol, sedoheptulose, mannoheptulose, and sedoheptulose-7P excretion (1,2,7). Elevations of polyols and 7-carbon chain carbohydrates are most prominent in the neonatal period, improving slightly with age (9). The criterion standard of diagnosis is the measurement of TALDO activity in vitro and sequence analysis of the TALDO1 gene. The human TALDO1 gene (OMIM #606003) is located on chromosome 11p15.5-p15.4, and a pseudogene is located on chromosome 1p34.1-p33. Banki et al (13) cloned the TALDO1 gene in 1994 and determined its organization in 5 exons. To date, 3 homozygous mutations have been detected (Table 1). Allelic homozygosity and inheritance pattern confirm autosomal-recessive inheritance. All of the previously reported cases had homozygous mutations and were born to consanguineous parents. We identified 2 heterozygous novel missense mutations in our patient. The paternal mutation consists of a deletion of 3 nucleotides (c.895_897delAAC) in exon 7, resulting in the deletion of asparagine at position 299 (p.Asn299del). The maternal mutation consists of a G>A transition (c.931G>A) in exon 7, with the substitution of glycine by arginine at position 311 (p.Gly311Arg). Asparagine at position 299 and glycine at position 311 are highly conserved in evolution, and are proposed to participate in the dimmer interface (14), indicating its importance for enzyme activity and its pathogenicity. The polyol biochemical pathway is highly active during the first trimester of pregnancy and appears essential for rapidly proliferating tissues such as embryonic tissues, as suggested by the increased incidence of major congenital abnormalities found in rats when glucose-6-phosphate dehydrogenase, the enzyme catalyzing the first step of the pathway, is experimentally decreased (15). The presentation of hydrops fetalis and a polymalformative syndrome in a 28-week-old fetus with TALDO deficiency (3) further indicates its relevance during fetal development (6). With the enhanced awareness of these inherited metabolic defects in the PPP, and with improved reporting of newly diagnosed patients, more elaborate knowledge garnered by unraveling the specific pathophysiology and biochemical and clinical phenotypes for these disorders can aid in deciphering the remaining questions pertaining to potential therapeutic options that remain undefined. Treatment should be aimed to restore the altered pathophysiology or to replace the affected organs (9). Because only the liver seems to be affected in TALDO deficiency, liver transplantation may be beneficial (2,9). Hypothetical approaches to treatment strategies include inducing the production of reduced GSH with N-acetylcysteine, decreasing oxidative stress with antioxidants (eg, vitamins C or E), and reducing polyol or sugar-P accumulation by using specific inhibitors of the PPP (9). In conclusion, a wide range of inborn errors of metabolism lead to biochemical disturbances within the liver, either directly or indirectly. Although individually rare, collectively these disorders make up a significant proportion of liver disease, particularly in infants and children. Recognizing such disorders is important for guiding prognostication, genetic counseling, and therapy. TALDO deficiency is the new addition to the expanding list of metabolic causes of pediatric liver disease. At present, routine diagnostic workup of patients with liver impairment or failure does not entail the assessment of urinary sugars and polyols. Being only a recently described disorder, TALDO deficiency is likely to be underdiagnosed. In addition, the limited availability of specific analytic methods required for estimation of the appropriate biomarkers further hampers diagnostic capabilities. We suggest that TALDO deficiency be included in the differential diagnosis of neonates, infants, and children affected with unexplained liver disease, hepatosplenomegaly, hemolytic anemia or fetal hydrops, especially when associated with excessive weight gain in the pregnant mother. Acknowledgment The authors thank Ana Pop for performing the DNA sequence analysis.