Diseases Associated with Primary Abnormalities in Carbohydrate Metabolism

Published on 13/04/2015 by admin

Filed under Neurology

Last modified 13/04/2015

Print this page

rate 1 star rate 2 star rate 3 star rate 4 star rate 5 star
Your rating: none, Average: 0 (0 votes)

This article have been viewed 5600 times

Chapter 34 Diseases Associated with Primary Abnormalities in Carbohydrate Metabolism

Abnormalities of Galactose Metabolism


Galactosemia describes a family of autosomal-recessive disorders characterized by increased blood levels of galactose. Galactose cannot be used directly for glycolysis, and must be converted to glucose-1-phosphate. Five enzymes are involved in this interconversion in most species: galactose mutarotase, galactokinase, galactose-1-phosphate uridyltransferase, uridine diphosphogalactose-4-epimerase, and phosphoglucomutase [Sellick et al., 2008]. Mutations in the GALK, GALT, and GALE genes that encode the second, third, and fourth enzymes, respectively, cause deficiency or absence of these enzymes, with consequent galactosemia [Fridovich-Keil, 2006]. These enzymes comprise the Leloir pathway. Human galactose mutarotase deficiency has not been described.

Galactose-1-Phosphate Uridyltransferase Deficiency

Galactose-1-phosphate uridyltransferase (GALT) deficiency is by far the most common cause of galactosemia. The incidence of galactosemia in Western Europe varies between 1:23,000 and 1:44,000 [Bosch, 2006]. Neonatal screening programs have found population incidence rates as high as 1 in 19,700 in Estonia [Ounap et al., 2010].


The precise link between the metabolic abnormality and the neuropathologic condition remains unknown. Galactose-1-phosphate uridyltransferase is present in the brain in low concentrations. Studies of rat brain reveal no significant site-specific differences in enzyme activity [Rogers et al., 1992]. Hypoglycemia may contribute significantly to the pathologic findings in many cases. The toxic effect of galactitol accumulation is not fully understood but is clearly relevant to adverse outcomes in the brain and lens. Animal data suggest that osmotic and oxidative stress in the lens activates the unfolded protein response [Mulhern et al., 2006]. Great variety and widespread distribution of the accompanying lesions have been documented.

Only two autopsy reports have been published; these were reviewed by Ridel et al. [2005]. The two patients were severely impaired and died at 8 and 25 years, respectively. In both cases, there was diffuse white matter gliosis, with focal areas of infarction. There was marked depletion of cerebellar Purkinje cells, with sparing of the granular layer and neuronal loss in the dentate nuclei and inferior olives. Cortical neuronal degeneration depleted the entire cerebral cortex, including Ammon’s horn, albeit to a variable extent and in different patterns. Spongiform changes are pronounced in some areas. Other findings included mild pigmentary changes in the surviving neurons, perineuronal satellitosis, and mild, diffuse microglial activation. Sclerotic and atrophic white matter were evident, as were dense accumulations of iron-containing and non-iron-containing pigment material in the reticular zone of the substantia nigra and the globus pallidus. Histochemical staining of the pigment indicated the presence of lipoprotein and polysaccharides. Pigment accumulation was accompanied by neuronal degeneration in the substantia nigra and dysmyelination in the globus pallidus. Bright pink eosinophilic hyaline-like bodies resembling axonal spheroids were present in the thalamus.


The primary abnormality in galactosemia is the deficiency of activity of galactose-1-phosphate uridyltransferase (Figure 34-1) that leads to accumulation of galactose-1-phosphate in red blood cells, liver, and brain [Lai et al., 2009].

Galactose is metabolized through the following four possible pathways [Fridovich-Keil, 2006]:

In classic galactosemia, the fourth pathway is obstructed; the other pathways function normally.

Shih and colleagues assigned the GALT locus to 9p13 by gene dosage [Shih et al., 1984]. A number of mutations produce abnormal enzymes with little or no galactose-1-phosphate uridyltransferase activity [Reichardt, 1991; Reichardt et al., 1992]. It is therefore not surprising that various genetic forms of galactosemia result from the presence of inefficient isoenzymes of galactose-1-phosphate uridyltransferase (Table 34-1). Most patients are compound heterozygotes, not true molecular homozygotes [Elsas et al., 1995].

The isoenzymes are separated and identified by electrophoresis. Cross-reactivity patterns for the enzyme variants have been studied using rabbit antibodies to purify human placental galactose-1-phosphate uridyltransferase [Andersen et al., 1984]. Transferase activity is absent in homozygous classic galactosemia (Q188R – the most common mutation in the United States, arising in Western Europe about 20,000 years ago [Flanagan et al., 2010]) and the African American variant (S135L – accounting for 8.4 percent of U.S. cases [Lai and Elsas, 2001]), virtually absent in the Rennes variant [Schapira and Kaplan, 1969], and abnormally low in the Chicago, Indiana, and Duarte (N314D) variants [Beutler et al., 1965; Lai et al., 1998; Levy et al., 1978]. Activity is normal in the Los Angeles variant (L218L) [Applegarth et al., 1976; Ng et al., 1973]. A screening method utilizing a standard newborn screening blood spot has been described that detects the four most common GALT alleles (Q188R, S135L, K285N, and L195P) and the N314D Duarte variant with a turnaround time of less than 2 hours [Dobrowolski et al., 2003]. By January 2010, 239 sequence variants had been described in the GALT mutation database [Calderon et al., 2007], most of which were missense mutations. The precise molecular mechanisms of galactosemia are not yet known, but a computational biology approach has yielded insights into the structure–function relationships of many mutations [Facchiano and Marabotti, 2010].

The precise mechanisms of central nervous system dysfunction and developmental aberrations remain unexplained. Hypoglycemia associated with galactosemia may be responsible for some of the neurologic (including higher cortical function) abnormalities observed in these patients, but the accumulation of unusual metabolites also may contribute to the pathogenesis of the disorder. Hypoglycemia may be the result of inhibition of phosphoglucomutase and accompanying failure of hepatic glycogenesis. The high concentration of galactose may increase both insulin production and release, and the increased galactose also may compete with glucose for transport sites and carrier mechanisms across the blood–brain barrier.

Galactose-1-phosphate accumulates in erythrocytes, liver, lenses, and kidneys, and may be influential in disruption of both tissue structure and function, but galactitol accumulation likely is a more important factor in cataract development and perhaps other tissue disruptions and cerebral dysfunction.

Galactokinase or transferase activity appears to be lower than expected in many women who give birth to nongalactosemic children with cataracts [Harley et al., 1974].

Clinical Characteristics

Infants usually are normal at birth, except for a slight decrease in birth weight [Hsia and Walker, 1961]. Symptoms become apparent when milk feedings begin. Jaundice usually develops between 4 and 10 days of age and persists for a longer period than does physiologic jaundice [Donnell et al., 1967]. Progressive hepatic involvement in the first several weeks causes edema, hepatomegaly, and hypoprothrombinemia. Renal dysfunction is accompanied by generalized aminoaciduria, proteinuria, and acidosis. Escherichia coli sepsis is a common complication [Levy et al., 1977]. Mild hypoglycemia also is common [Donnell et al., 1967]. Cataracts appear between 4 and 8 weeks, reflecting the accumulation of galactose-1-phosphate or galactitol. Studies in animals show that apoptosis of lens epithelial cells and cataract formation are directly related to galactitol accumulation and are prevented by aldose reductase inhibitors [Murata et al., 2001]. Apoptosis may be triggered by altered p53 expression secondary to the accumulating metabolites [Takamura et al., 2003].

Central nervous system impairment is manifested by lethargy and hypotonia, often associated with cerebral edema [Huttenlocher et al., 1970; Welch and Milligan, 1987], and may be documented by computed tomography (CT) scanning [Belman et al., 1986]. Cerebral edema in an encephalopathic neonate with galactose-1-phosphate uridyltransferase deficiency has been correlated with increased brain galactitol on magnetic resonance (MR) spectroscopy [Berry et al., 2001]. A subgroup of patients may develop marked ataxia and tremor, which does not correlate with cognitive abilities or dietary restriction [Ridel et al., 2005]. Seizures have been reported in two siblings with galactosemia [Ridel et al., 2005], and chorea in a single case [Shah et al., 2009]. After the appearance of neurologic symptoms, the patient may experience anorexia, vomiting, or diarrhea.

Galactosemia may prove fatal at any time. Cirrhosis progresses inexorably in patients who do not receive treatment. If the disease progresses more slowly and therapy is not optimal, mild gastrointestinal symptoms and failure to thrive persist.

Gonadal function in women with galactosemia is abnormal and usually manifests as primary ovarian failure. Both hypergonadotropic hypogonadism and abnormal response to gonadotropin-releasing hormone may be present. The mechanisms of ovarian failure are not yet fully understood, but likely include direct toxicity of galactose and its metabolites, incomplete galactosylation of glycoconjugates, oxidative stress and activation of apoptosis [Forges et al., 2006]. Some women with galactosemia do become pregnant; neither they nor their offspring seem to be affected by elevated levels of galactose, at least in the short term [Gubbels et al., 2008].

If treatment is not instituted, moderate to severe intellectual and motor retardation ensue in a majority of cases. Even in patients who are treated adequately, cognitive disability is common, but usually does not progress over time [Schadewaldt et al., 2010]. Language impairment is prominent [Hansen et al., 1996], and takes the form of verbal dyspraxia in about 50 percent of patients with galactose-1-phosphate uridyltransferase deficiency [Webb et al., 2003]. The risk of dyspraxia is associated with elevated mean galactose-6-phosphate and urinary galactitol concentrations, and with impaired total body galactose metabolism, as assessed by a carbon dioxide breath test [Hansen et al., 1996; Webb et al., 2003]. The characteristics of retardation resulting from untreated galactosemia are nonspecific. Pregnant women at risk of giving birth to galactosemic infants require strict restriction of galactose intake and close metabolic monitoring to prevent congenital cataracts in offspring. One study of 33 subjects aged 4–16 years found that children with galactosemia and speech disorders had a 4–6 times greater risk for language impairment than children with early speech disorders of unknown origin. Notwithstanding a negative effect of early dietary lactose exposure, the data suggested an antenatal origin of language disorder in most cases [Potter et al., 2008]. The incidence of speech and language disorders also appears to be increased in children with Duarte galactosemia, in whom galactose was restricted in the first 12 months [Powell et al., 2009].

Prenatal diagnosis is possible by means of galactose-1-phosphate uridyltransferase assay using cultured amniotic fluid cells or chorionic villus biopsy specimens, and by galactitol estimation in amniotic fluid supernatant [Holton et al., 1989]. When both mutations in an index case have been identified, direct molecular analysis is the preferred method of prenatal diagnosis [Elsas, 2001].

Occasional patients with genotypes and residual enzyme activity usually associated with severe phenotypes present with mild manifestations despite lack of dietary restriction. In one such case, the patient’s markedly reduced galactitol production, presumably reflecting limited aldose reductase activity, was identified as a major factor in preventing neurologic injury despite her classic Q188R missense mutations [Lee et al., 2003; Segal, 2004].

Clinical Laboratory Tests

Biochemical tests for galactosemia screen for elevated levels of small molecules and directly assay the enzymes in the Leloir pathway. Detailed protocols have been published outlining the methods, including their rationale and interpretation [Cuthbert et al., 2008].

A fluorescent spot test for erythrocyte transferase activity (the Beutler test) is quite sensitive [Beutler and Baluda, 1966], but can yield false-negative results if the subject has received a blood transfusion up to 120 days earlier, because of residual galactose-1-phosphate uridyltransferase activity in the transfused blood. Enzyme activity is relatively low in heterozygous persons; this phenomenon aids in identifying carriers.

Galactose tolerance testing is potentially dangerous and should be undertaken only for well-planned investigational purposes [Donnell et al., 1967]. Bedside urine testing is positive for reducing substances (Benedict’s test) but negative for glucose by the glucose oxidase method. Chromatography definitively identifies galactose as the abnormal metabolite.

Antibiotics may interfere with neonatal screening [Clemens et al., 1986]. Urinary galactose may be present in neonates with hepatic dysfunction in disorders other than galactosemia. Repeated attempts at detecting the reducing substance must be made because galactosuria may be inconstant as a result of fluctuating galactose or lactose ingestion. Generalized aminoaciduria, proteinuria, and abnormalities on liver function tests are common.

Screening tests that use E. coli bacteriophage assay of galactose and galactose-1-phosphate in dried blood samples are available and are useful for large-volume applications [Jinks et al., 1987; Schulpis et al., 1997].

Magnetic resonance imaging (MRI) in 67 transferase-deficient galactosemic patients revealed that 22 had mild cerebral atrophy, 8 had cerebellar atrophy, and 11 had multiple small hyperintense lesions in the cerebral white matter on T2-weighted images. The patients with classic galactosemia (those without measurable transferase activity) older than 1 year of age did not manifest the expected maturational decrease in peripheral white matter signal intensity on intermediate- and T2-weighted images. Interference with normal galactocerebroside formation may explain these findings [Nelson et al., 1992]. A study of MRI and MR spectroscopy in 14 sibling pairs found that delayed or absent myelination of the deep white matter was the most common finding; some individuals showed cerebellar atrophy, others ventricular dilatation [Hughes et al., 2009]. There was no correlation between dietary control or clinical status and the imaging findings. An infant was studied with MRI and MR spectroscopy, and was found to have increased signal in the cerebral white matter associated with increased diffusion. MR spectroscopy showed a peak at 3.7 ppm, consistent with galactitol. NAA/Cr, Cho/Cr, mI/Cr ratios were also decreased [Cakmakci et al., 2009].


Because galactose is a nonessential nutrient, exclusion from the diet is relatively easy and without complication. Milk, the primary galactose-containing fluid, can be avoided by the use of vegetable product substitutes [Hansen, 1969]. Unfortunately, certain fruits and vegetables contain relatively high concentrations of galactose, including bell peppers, dates, tomatoes, papaya, and watermelon [Gross and Acosta, 1991]. Appropriate diets are available. The widespread use of cow’s milk in the newborn diet makes early diagnosis essential. Intellectual and personality impairment is most successfully prevented with early treatment [Fishler et al., 1972]. Studies of the long-term outcome of therapy have been relatively disappointing, particularly in regard to central nervous system and ovarian dysfunction [Widhalm et al., 1997]. A report based on a German cross-sectional study was more encouraging, suggesting that infants given appropriate treatment by 5 days of age achieved better outcomes than did those in whom treatment was begun later [Schweitzer-Krantz, 2003].

Most acute sequelae of the disease, including cirrhosis, are ameliorated with therapy, even if briefly delayed [Donnell et al., 1967]. Cataracts also may recede or disappear. Cataract formation in severely galactosemic rats has been prevented by inhibitors of aldose reductase. Signs or symptoms of sepsis should be investigated with blood, urine, and cerebrospinal fluid cultures to detect E. coli.

Because dietary therapy does not uniformly alleviate many of the sequelae of the disease, new strategies for therapy are necessary. The use of folic acid has been advocated as a supplement to galactose restriction to enhance transferase activity [Segal and Rogers, 1990]. Uridine administration demonstrated some promise in enhancing galactose transformation [Holton, 1990], but a trial of oral uridine did not reveal any evidence of benefit to neurocognitive functioning in treated versus untreated cases [Manis et al., 1997]. Neurologic complications may occur as late as the fourth decade of life [Friedman et al., 1989].

Unfortunately, galactose-1-phosphate appears to accumulate in the galactosemic fetus in spite of maternal milk restriction [Irons et al., 1985].

Uridine Diphosphogalactose Epimerase Deficiency

Uridine diphosphogalactose epimerase (GALE) deficiency has conventionally been separated into peripheral and generalized forms, implying a dichotomy between levels of enzyme activity in blood and other tissues. Ten children in a study were diagnosed with peripheral GALE deficiency as neonates were found to have a range of GALE activity in lymphoblasts (i.e., nonperipheral tissue), that was correlated with metabolic abnormalities in some patients, implying that this is a spectrum disorder, and not a binary condition as suggested in the older literature [Openo et al., 2006]. Children previously recognized with generalized (severe) deficiency of GALE (see Figure 34-1) [Bowling et al., 1986; Garibaldi et al., 1986; Sardharwalla et al., 1988; Walter et al., 1999] had manifestations resembling those in classic galactosemia. Most survivors were dysmorphic and deaf. The GALE locus is at 1p36–p35; the human GALE gene is about 4 kilobases (kb) in length and contains 11 exons [Maceratesi et al., 1998]. The coding sequence of the GALE gene and screening for mutations in epimerase-deficient persons have been reported by the same investigators. The patients are either homozygotes or compound heterozygotes for mutations. Two forms of enzyme deficiency were originally described, one type benign (with expression restricted to the lens) and the other severe [Quimby et al., 1997], but an intermediate form is now recognized [Openo et al., 2006]. Three mutations (S81R, T150M, and P293L) have been reported in children with this intermediate form of GALE deficiency [Chhay et al., 2008]. Patients with GALE deficiency require exogenous galactose for the synthesis of glycolipids and glycoproteins.

Galactokinase Deficiency

Galactokinase deficiency was first detected in the Bulgarian gypsy (Romany) population. Its birth incidence varies, ranging from a high of 1 in 52,000 in Bulgaria to 1 in 2,200,000 in Switzerland [Kalaydjieva et al., 1999]. Deficiency of galactokinase activity causes a clinical condition similar to that in classic galactosemia. Patients have cataracts and accumulation of galactose. As in galactosemia, galactitol, a reduction metabolite of galactose, is found in the urine and tissues [Egan and Wells, 1966]. The existence of two GALK genes is likely. The GALK1 gene is located at 17q24 [Bergsma et al., 1996]. GALK2 may reside on chromosome 15 [Lee et al., 1992]; its metabolic role is unknown, but it does not appear to be necessary for galactose metabolism. No mutations in GALK2 have been described. The structure of GALK1 has been solved and its relationship to other members of the GHMP kinase superfamily defined, as well as its role as an essential componenet of the ‘switch’ permitting expression of the Leloir pathway genes in the presence of galactose [Holden et al., 2004]. More than 20 mutations in GALK1 have been described, most in compound heterozygotes for private mutations [Sangiuolo et al., 2004]. Only one common mutation, P28T, has been recognized in the Romany population [Kalaydjieva et al., 1999].


Galactokinase deficiency causes the accumulation of galactose, which eventually is metabolized to galactitol (see Figure 34-1). Enzyme activity is reduced rather than absent in erythrocytes [Xu et al., 1989]. Galactose-1-phosphate does not accumulate. Large amounts of circulating galactose result in urinary excretion of galactose, which causes positive results on copper sulfate screening tests for urinary reducing substances. Specific assays for galactose-1-phosphate or galactose-1-phosphate uridyltransferase are necessary to differentiate between deficiency of galactose-1-phosphate uridyltransferase and deficiency of galactokinase.

Clinical Characteristics

The clinical manifestations of 55 patients reported in the literature were reviewed in 2002 [Bosch et al., 2002]. Cataract was present in all cases, except for those detected by newborn screening. Thirty-five percent of the patients had other manifestations; only mental retardation and pseudotumor occurred in more than one patient in this series. The mental retardation was thought to be unrelated to the GALK deficiency. Forty percent of the patients were of Roma ancestry and 26 percent were the product of consanguineous unions. One child deteriorated following the onset of epilepsy at 17 years.

Cataracts, the only consistent manifestation of galactokinase deficiency, form in the first months of life. Cataract formation is likely related to the production and accumulation of the osmotically active sugar alcohol galactitol from galactose through the activity of aldose reductase (see earlier).

Abnormalities of Fructose Metabolism

Hereditary Fructose Intolerance


Fructose is rapidly absorbed from the gut, facilitated by the glucose transporters GLUT 2 and GLUT5, and is metabolized in the liver by the fructokinase pathway, through which it is linked to glycolysis, gluconeogenesis, glycogenolysis, and lipid metabolism. Fructose can also be synthesized endogenously from sorbitol, an important point in management [Bouteldja and Timson, 2010]. Hereditary fructose intolerance was first described by Chambers and Pratt [1956]. This condition results from a deficiency of hepatic fructose-1-phosphate aldolase B [Hers and Joassin, 1961]. (Isoenzyme A is found in most vertebrate tissues and isoenzyme C in brain.) The gene encoding this enzyme maps to 9q22 [Henry et al., 1985]. Normal fructokinase activity results in the accumulation of large amounts of fructose-1-phosphate in the liver and kidneys. Fibroblasts from patients with hereditary fructose intolerance consume less glucose, produce less lactate, and contain less glycogen compared with control cells [Lemonnier et al., 1987]. A radioisotopic method for fructose-1-phosphate assay is available [Shin et al., 1983]. The enzyme deficiency is inherited as an autosomal-recessive trait and has an estimated prevalence in central Europe of 1:26,100 (95 percent confidence interval 1: 12,600–79,000) [Santer et al., 2005].

Urinary fructose excretion also is present in a harmless metabolic variant resulting from fructokinase deficiency that should not be confused with hereditary fructose intolerance.

Fructose-1,6-Diphosphatase Deficiency

Another inborn error of metabolism, fructose-1,6-diphosphatase (i.e., fructose-1,6-bisphosphatase [FBPase]) deficiency, also is characterized by hypoglycemia after fructose ingestion [Baker and Winegrad, 1970; Hulsmann and Fernandes, 1971]. About half of the cases manifest in infants, with life-threatening episodes of hypoglycemia and metabolic acidosis. The causal relationship between the enzyme deficiency and hypoglycemia in this abnormality is not fully explained. Some data, however, suggest that α-glycerol phosphate, fructose-1-phosphate, and fructose-1,6-diphosphate all inhibit phosphorylase a activity [Kaufmann and Froesch, 1973]. This relationship may explain the hypoglycemic episodes in both hereditary fructose intolerance and fructose-1,6-diphosphatase deficiency. The role of the enzyme deficiency in glycolysis and gluconeogenesis requires further clarification [Adams et al., 1990].

The diagnosis can be established by measuring FBPase activity and mutational analysis in cultured monocytes, without the necessity for liver biopsy [Kikawa et al., 2002]. A retrospective study of Japanese patients with FBPase deficiency treated with intravenous glycerol, which contains fructose in Asian countries, found a relationship between the infusion of glycerol and the onset of cerebral edema in some patients [Hasegawa et al., 2003]. In one case FBPase deficiency was associated with a prolonged prothrombin time, which corrected with intravenous glucose and bicarbonate [Nitzan et al., 2004]. Another patient gave birth to normal children after three uncomplicated pregnancies, but developed subsequent hearing loss and cognitive impairments despite careful metabolic monitoring [Krishnamurthy et al., 2007].

Glycogen Storage Diseases

The biochemistry of the glycogen storage diseases (GSDs) illustrates the diverse effects of genetically determined enzymatic deficiencies along a single metabolic pathway. In spite of a few inconsistencies and a number of unexplained conditions, a logical approach to these diseases is practical. The GSDs are a family of diseases sui generis, with the exception of at least two disorders that can be included under the rubric of lysosomal storage diseases – Pompe’s disease and Danon’s disease. Indeed, the first lysosomal storage disease defined as such was Pompe’s disease [Hers, 1963]. General characteristics of this disease family are discussed in Chapter 36. Patients have also been described who accumulate glycogen in autophagic vacuoles but who do not appear to have an enzymatic deficiency [Danon et al., 1981]. This phenotype, named Danon’s disease, is known to result from deficiency of lysosomal-associated membrane protein 2 [Nishino et al., 2000]. This X-linked dominant disorder has multisystem effects, most consistently involving the heart and skeletal muscle [Sugie et al., 2002].

There is general agreement on the numeric designations of GSDs I to VI, but the nomenclature beyond that is confusing. For example, GSD types VIII and X were originally considered distinct conditions, but are now classified with GSD VI by many authors.

Clinical manifestations of GSD often result from glucose deficiency, with ensuing hypoglycemia occurring separately or in association with increased glycogen storage. The location of the enzymatic block in the pathway determines whether the configuration of the glycogen is normal or abnormal.

GSDs result in the accumulation in various tissues of increased concentrations of glycogen of normal or abnormal configuration (Table 34-2). These diseases result from a deficiency or absence of specific enzyme activity in the metabolic pathway of glycogen.

The glucose molecule is the prime building block in the multistep synthesis of glycogen (Figure 34-2; see also Table 34-2). Glycogen synthesis occurs in many tissues, predominantly in liver, kidney, and muscle. Glucose transported in the blood enters the cell, facilitated by a glucose transporter [Scheepers et al., 2004], is phosphorylated in a reaction catalyzed by the enzyme hexokinase, and becomes glucose-6-phosphate. In the next step, the enzyme phosphoglucomutase mediates the transformation to glucose-1-phosphate. Glucose-1-phosphate, in association with uridine triphosphate, is transformed to uridine diphosphate-glucose with the participation of uridine diphosphate glucose pyrophosphorylase. The glucose portion of this molecule is then attached by a 1,4 linkage to a terminal glucosyl unit. This reaction is facilitated by the active form of the enzyme glycogen synthase (uridine diphosphate–glucose-glycogen glucosyl transferase). Glycogenin is a protein primer that initiates glycogen synthesis by covalently attaching individual glucose residues to tyrosine 194. This process occurs by autoglycosylation to form a short priming chain of glucose residues that are a substrate for glycogen synthase [Hurley et al., 2006]. When the glucosyl chain becomes 6–12 units long, this section is transferred and affixed to another glucosyl chain by a 1,6linkage as a result of the action of the branching enzyme α-1,4-glucan: α-1,4-glucan-6-glucosyl transferase. A 1,6 linkage constitutes the branch point, a final stage in glycogen formation. Glycogen exists in the cell in association with proteins (including the enzymes described above) as organelles known as glycosomes. Glycosomes may occur free in the cytosol (lyoglycosomes) or in association with other structures (desmoglycosomes), including myofibers, mitochondria, and sarcoplasmic reticulum cisterna.


Fig. 34-2 Pathways of glycogen metabolism depicting sites of metabolic block that lead to glycogen storage disease.

See Table 34-2 for description of abnormalities denoted by Arabic numerals enclosed in boxes. *See Figure 34-3 for phosphorylase activation sequence. **Other defects of terminal glycolysis.

During the degradation process, the phosphorylase enzymes split the 1,4 linkages, which results in formation of free glucose-1-phosphate molecules. Both muscle and hepatic phosphorylase isoenzymes exist. Activation of phosphorylase takes place through a cascade of reactions ultimately involving phosphorylase b kinase (Figure 34-3). As the cleavage of 1,4 bonds moves near the 1,6 branching point, “three-glucose” residues are removed in a block by oligo-1,4 alpha 1,4-glucan transferase, and the 1,6 linkage is disrupted by the debranching enzyme (amylo-1,6 glucosidase), with the resultant release of a free glucose molecule. Approximately 8 percent of glucose in glycogen is involved at 1,6 branch points and may be released in this free form. This process continues along the branches of the glycogen molecule. Therefore, both glucose-1-phosphate and free glucose molecules result from this series of degradation reactions.


Fig. 34-3 Activation sequence of phosphorylase.

See Table 34-2 for description of abnormalities denoted by the Arabic numerals enclosed in boxes. ADP, adenosine diphosphate; ATP, adenosine triphosphate; cAMP, cyclic adenosine monophosphate.

(Modified from Goldberg NB. Vigilance against pathogens. Hosp Pract 1974;9:127.)

The glucose-1-phosphate molecules subsequently are converted to glucose-6-phosphate by the action of the enzyme phosphoglucomutase. Phosphate is released from glucose-6-phosphate in liver and kidney by glucose-6-phosphatase, and free glucose molecules result. (Glucose-6-phosphatase activity is absent in skeletal muscle.) The free glucose is transported by the circulation to other organs, where it is used. In other organs and in muscle the glucose-6-phosphate formed on entry is phosphorylated to fructose-1,6-diphosphate, which is transformed by the metabolic steps in the Embden–Meyerhof pathway that leads eventually to the formation of pyruvic acid and lactic acid.

Glucose-6-Phosphatase Deficiency (Von Gierke’s Disease, Glycogen Storage Disease Type I, Hepatorenal Glycogenosis)


Two distinct subgroups of glycogen storage disease type I have been identified: those with primary glucose-6-phosphatase deficiency (type Ia) and those phenocopies with additional features of immune impairment (neutropenia and neutrophil adherence defects), now designated as glycogen storage disease type I non-a [Moses, 2002]. Glycogen storage disease type I non-a disorders originally were thought to result from defects in a multicomponent translocase system responsible for transporting glucose-6-phosphatase into microsomes [Annabi et al., 1998]. This model postulated three transport proteins, T1, T2, and T3, to chaperone glucose, glucose-6-phosphatase, phosphate, and pyrophosphate across the endoplasmic reticulum membrane. Cloning of the glucose-6-phosphatase translocase gene (G6PT) demonstrated that the previously proposed subtypes b, c, and d all were associated with mutations in G6PT, producing different kinetic variants [Matern et al., 2002; Moses and Parvari, 2002]. The G6PC gene that codes for glucose-6-phosphatase is located at 17q21 [Brody et al., 1995]. A number of allelic variants have been described. Glucose-6-phosphatase comprises at least five different polypeptides. The G6P locus is at 11q23 [Annabi et al., 1998]; several mutations have been characterized [Matern et al., 2002].

All forms share common clinical manifestations that are attributable to abnormal metabolism of glucose-6-phosphate. In type Ia, glucose-6-phosphatase deficiency results in storage of glycogen of normal configuration in the liver and kidneys. The glycogen concentration usually exceeds 4 percent by weight. The enzyme activity frequently is absent or extremely low [Cori and Cori, 1952]. Glucose-6-phosphatase is important in regulating the entry of free glucose into the circulation from the liver. Because of this pivotal role, deficiency of the enzyme produces hypoglycemia.

Ethanol causes decreased blood lactate and pyruvate content, presumably by diverting carbon to triglyceride formation [Sadeghi-Nejad et al., 1975]. A study of insulin secretion in five adult patients with glucose-6-phosphatase deficiency found that their capacity to increase blood insulin was significantly less than normal. As patients with this condition mature, they become normoglycemic and characteristically have abnormal glucose tolerance curves. These studies suggest that the increasing clinical stability noted with age and the associated tendency toward normoglycemia reflect a decrease in insulin responsiveness that may develop as an adaptive process [Lockwood et al., 1969]. Diabetes mellitus has been reported in GSD type 1 [Spiegel et al., 2005]. The intricacies of the glucose-6-phosphatase system and its role in glucose metabolism have been reviewed [Foster and Nordlie, 2002].

Clinical Characteristics

Hypoglycemia causes much of the morbidity during the first year of life. Seizures are frequent and almost invariably are the presenting complaint of affected children. Hypoglycemia may result in severe, chronic neurologic impairment, including hemiplegia [Fine et al., 1969]. Hepatomegaly and the failure to thrive syndrome are commonly present. An association with moyamoya disease has been described [Goutières et al., 1997]. A study of 19 patients with glycogen storage disease type I (median age 11 years) in one center found prevalence rates for epilepsy, deafness, and neuroradiologic abnormalities of 10.5 percent, 15 percent, and 57 percent, respectively, far in excess of the rates in the general population, or in children with other causes of neonatal hypoglycemia. MRI abnormalities included dilatation of occipital horns and/or hyperintensity of subcortical white matter in the occipital lobes in all patients [Melis et al., 2004]. Subcutaneous fat often is increased, especially over the buttocks, breasts, and cheeks. Xanthomas of the skin occur over the extensor surfaces of the limbs and at times over the buttocks [Hou et al., 1996; Hou and Wang, 2003]. Affected children frequently have a protuberant abdomen because of massive enlargement of the liver. Hepatomegaly may be present at birth. The liver edge is hard and not tender. Careful palpation may reveal enlarged kidneys. Hepatic adenomas develop in between one-half and three-quarters of adults with glycogen storage disease I; about 10 percent undergo malignant transformation. Some data suggest that lower frequencies are associated with better dietary control [Lee, 2002]. Hepatocellular carcinoma has been reported as complicating hepatic adenomas, and may reflect poor metabolic control [Franco et al., 2005]. Patients carrying mutations that cause relatively mild expression of the disease in childhood, often without hypoglycemia (such as 727 G>T), are associated with adult presentation of hepatocellular carcinoma [Matern et al., 2002].

Type I non-a patients typically have recurrent stomatitis, frequent infections, and chronic inflammatory bowel disease secondary to neutropenia and neutrophil dysfunction [de Parscau et al., 1988]. The neutropenia seen in GSD1b has been attributed to endoplasmic reticulum and oxidative stress secondary to the G6PT deficiency [Chou et al., 2010]. Seventy-five percent of 36 GSD type I non-a patients had chronic gastrointestinal complaints, and 28 percent had proven inflammatory bowel disease. A further 22 percent had a highly suggestive history [Dieckgraefe et al., 2002].

Clinical Laboratory Tests

The diagnosis can be made by assaying the enzyme activity in liver and peripheral white blood cells [Maire et al., 1991]. Direct assay of hepatic glucose-6-phosphatase activity in liver remains the definitive diagnostic procedure but can be replaced by mutational analysis in many patients. Just five mutant alleles account for almost 70 percent of cases of glucose-6-phosphatase deficiency, so that mutation screening is a reasonable initial diagnostic approach, avoiding the risks and discomfort of liver biopsy [Matern et al., 2002]. Molecular analysis of the G6P and G6PT genes permits rapid confirmation of the diagnosis in most cases [Janecke et al., 2001].

Severe hypoglycemia frequently occurs because of the failure of glucose formation from glucose-6-phosphate. Postprandial blood glucose concentration may be exceedingly high. Severe acidosis, which may vary in degree but usually is associated with lacticacidemia and pyruvicacidemia, and hyperuricemia are frequent. The presence of ketoacidosis has been unduly stressed.

Hyperuricemia has often been documented, but is poorly understood. Adenosine triphosphate depletion has been postulated as a causative factor [Greene et al., 1978]. Ketoacidosis and hyperthermia during anesthesia in a child with GSD type I has been described [Edelstein and Hirshman, 1980]. Blood cholesterol, fatty acids, and triglycerides are elevated; overt lipemia may be present.

Decreased bone density documented by radiography may be related to chronic metabolic acidosis. Serum inorganic phosphate may be decreased in the presence of normal alkaline phosphatase activity.

Provocation with epinephrine or glucagon demonstrates failure of the patient’s blood glucose concentration to increase within 10–20 minutes, although intracellular glucose-6-phosphate is formed quickly after the induced activation of liver phosphorylase.


The goal of therapy is to provide sufficient free glucose to maintain a normal blood glucose concentration. Continuous nocturnal intragastric infusion of glucose has been relatively successful [Greene et al., 1976]. Subsequently, the use of cornstarch suspensions given during the day obviated the need for nocturnal infusion in some children [Chen et al., 1984; Wolfsdorf et al., 1990]. The dietary carbohydrate must be monitored because excess glucose leads to glycogen storage in the liver and kidneys. Frequent small feedings of carbohydrates are provided. Severity of the disease reaches a plateau after the fourth or fifth year of life. Vigorous treatment is, therefore, worthwhile until the plateau is reached. A long-term study of 15 children with GSD type I, beginning in infancy, found that careful metabolic control, aiming for high to normal plasma glucose levels and normal urine lactate, was associated with normal growth and lowering, but not normalization, of plasma lipids. Hepatic adenomas or renal impairment developed in none of the patients who reached adolescence [Daublin et al., 2002; Weinstein et al., 2002].

Dietary substitution of medium-chain triglycerides for long-chain triglycerides was attempted in glucose-6-phosphatase deficiency, to alter the hyperlipemic state by means of the unique absorptive and metabolic properties of medium-chain triglycerides. The results suggested that substitution of medium-chain triglycerides for long-chain triglycerides in the diet, along with normal carbohydrate consumption, leads to significant decrease in serum lipid levels, disappearance of eruptive xanthomas, and decrease in liver mass [Cuttino et al., 1970].

The hyperglycemic agent diazoxide has been beneficial [Rennert and Mukhopadhyay, 1968]. The drug’s action is not well understood, but normal blood glucose concentration has been maintained with this drug. Phenytoin also has been used [Jubiz and Rallison, 1974].

Surgical treatment for glucose-6-phosphatase deficiency involves creation of a portacaval shunt, which increases the peripheral blood glucose by allowing portal blood to bypass the liver after absorption of glucose from the gut; excellent metabolic control can be achieved over the long term, and the operation does not preclude subsequent liver transplantation [Corbeel et al., 2000]. The postoperative course has been complicated by severe hypoglycemia, hypocalcemia, acidosis, and respiratory impairment, the last primarily because of hepatomegaly. Preoperative intravenous hyperalimentation appears to eliminate these metabolic problems, reduce the size of the liver, and provide a smoother and shorter postoperative course [Folkman et al., 1972]. Liver transplantation was reported to produce beneficial results [Malatack et al., 1983]. One report suggested that this procedure, usually indicated for management of multiple hepatic adenomas, does not itself benefit metabolic control, and indeed, careful systemic management is essential to prevent graft complications [Labrune, 2002]. A more recent report of living donor liver transplantation in four children with GSD1b described markedly improved quality of life [Kasahara et al., 2009]. Another case report described a 47-year-old woman with GSD1a, whose fasting tolerance was significantly improved after infusion of hepatocytes [Muraca et al., 2002].

Chronic inflammatory bowel disease similar to Crohn’s disease has been associated with GSD type I non-a [Dieckgraefe et al., 2002]. Initial studies suggested benefit from therapy with colony-stimulating factors [Roe et al., 1992]. A retrospective study of 57 patients with GSD type I non-a found evidence of less frequent infections and diminished severity of inflammatory bowel disease in those who received granulocyte colony-stimulating factor [Visser et al., 2002]. Splenomegaly was associated with granulocyte colony-stimulating factor therapy in this group. Renal disease also may ensue in older patients [Chen, 1991].

Brain abscess has been reported in a patient with type I non-a disease [Park et al., 1991]. Renal transplantation has been used for terminal renal failure, and occasionally, combined hepatic and renal grafting has been used. In both circumstances, meticulous systemic metabolic management is essential to successful outcome [Labrune, 2002].

Acid α-Glucosidase (GAA, Acid Maltase) Deficiency, Infantile Type (Pompe’s Disease, Idiopathic Generalized Glycogenosis, Glycogen Storage Disease Type II)


Hers [1963] first reported the deficiency of activity of the lysosomal enzyme acid maltase (α-1,4-glucosidase), located at 17q25.2–q25.3 [Kuo et al., 1996; Martiniuk et al., 1985]. Glycogen structure has consistently been normal, and its accumulation is restricted primarily to lysosomes, although lysosomal breakdown and cytoplasmic accumulation with disruption of muscle fibers occur in severe cases. Although direct injury of muscle fibers by glycogen leaking from lysosomes was thought to be the major cause of contractile dysfunction, experimental and pathological evidence suggests that the accumulation of autophagosomes is the major culprit, and that these pathologic orgenelles impair the effectiveness of enzyme replacement therapy by acting as a sink for infused enzyme [Shea and Raben, 2009; Raben et al., 2009]. Hypoglycemia is not a feature of this condition, but increased protein turnover with increased leucine flux and oxidation and increased resting energy expenditure has been found in late-onset cases [Bodamer et al., 2000].

Attempts have been made to differentiate biochemically among the infantile, late infantile, and adult-onset forms of acid maltase deficiency. Activity of α-1,4-glucosidase (acid maltase) at various pH values in infants, children, and adults with acid maltase deficiency has been studied. Only traces of neutral maltase are found in the heart, and significantly decreased neutral maltase activity was measured in the skeletal muscle and liver of an affected infant. In the late infantile form, neutral acid maltase activity is decreased only in the liver; in the adult form, neutral maltase is not deficient in any tissue. An absolute decrease of leukocyte acid maltase was found in four adults and a relative decrease in 1 of 5 adults with acid maltase deficiency. Decrease in the pH ratio of acid to neural maltase activity in leukocytes may be of diagnostic importance in adult acid maltase deficiency [Angelini and Engel, 1972].

A number of allelic variations have been described and may explain the differences in age at onset. The theoretical abnormalities that could result in a deficiency of α-glucosidase include synthesis of catalytically inactive protein, absence of messenger RNA (mRNA) for the enzyme, decreased synthesis of the precursor, lack of phosphorylation of the precursor, impaired conversion of the precursor to the mature enzyme, and synthesis of unstable precursor [Tager et al., 1987; Zhong et al., 1991]. In general, the location and nature of mutations predict the phenotype, but exceptional cases are described in which relatively mild phenotypes occur despite low levels of α-glucosidase expression in cultured cells and in the patient’s tissues [Hermans et al., 2004; Kroos et al., 2004]. Thus far, unidentified genetic modifiers and environmental factors are presumed to account for such variability.

Unexplained storage of increased neutral lipid is coupled with low carnitine concentration and reduced β-hydroxyacyl-CoA dehydrogenase in muscle [Verity, 1991].

Clinical Characteristics

Development usually is normal for several weeks to several months; then the affected infant presents with feeding difficulties, weakness, or respiratory impairment. The median age at presentation was 1.6 and 1.9 months, respectively, in 20 Dutch patients and 133 patients described in the literature [Van den Hout et al., 2003] (Figure 34-4). Little spontaneous movement occurs, and the cry is short-lived and weak. Swallowing is grossly limited, and secretions pool in the posterior oropharynx. Respiratory difficulty reflects weakness of the accessory muscles of respiration [Tanaka et al., 1979]. Massive cardiomegaly develops, and a soft systolic murmur is often heard along the left sternal border [Pompe, 1932]. Obstruction to ventricular outflow and impairment of inflow may develop [Seifert et al., 1992], and serial echocardiography reveals progressive left ventricular posterior wall diastolic thickening [Van den Hout et al., 2003]. Hepatomegaly is almost universally present. The liver has a sharp edge and a firm consistency on palpation. Subcutaneous fat over all areas of the body is sparse, and the muscles are small and firm. The tongue often is enlarged and may protrude. Intermittent cyanosis reflects respiratory and cardiac embarrassment. Deep tendon reflexes are lost by the age of 6 months. Affected infants undergo progressive debilitation, and most die, at median ages of 6 and 7.7 months in the literature cases and Dutch series, respectively. Fewer than 10 percent survive beyond 1 year; only two patients have been described who survived 18 months [Van den Hout et al., 2003], and almost all die by 2 years. A subgroup of children has been described who present later in infancy, with lesser degrees of weakness and cardiac impairment, who have survived for periods as long as 13 years with ventilatory and nutritional support [Slonim et al., 2000].

Clinical Laboratory Tests

The quantity of glycogen storage and activity of the enzyme involved in glycogen metabolism can be monitored by using skin removed with a vacuum skin-blistering technique or fibroblasts grown from skin biopsy material. Among the enzymes in skin are phosphorylase, acid maltase, and the debranching enzyme amylo-1,6-glucosidase. No glucose-6-phosphatase activity is present in skin.

Both skin fibroblasts and amniotic fluid cells can be used for assay of acid maltase (α-1,4-glucosidase) activity [Butterworth and Broadhead, 1977; Leathwood and Ryman, 1971]. Prenatal diagnosis by biochemical study of uncultured amniotic fluid cells and chorionic villus biopsy material using maltose as a substrate have been reported [Hug et al., 1984; Park et al., 1992]. Mutational analysis has superseded these techniques in some cases, and complemented enzyme analysis and ultrasound examinations in others, as in a report of a fetus with this deficiency diagnosed in the second trimester, in which glycogen accumulation was detectable in muscle, as well as a visibly enlarged tongue on prenatal ultrasound examination [Chen et al., 2004].

A simple differential immunoprecipitation assay of urinary acid and neutral α-glucosidases has been developed [Tsuji et al., 1987].

The chest x-ray reveals massive cardiomegaly [Ruttenberg et al., 1964]. The electrocardiogram contains depressed ST segments, inverted T waves, and a shortened P-R interval. These changes may be confused with those of myocarditis. Electromyography (EMG) shows myopathic changes; polyphasic potentials and a reduced interference pattern with low voltage are the usual findings. Unusual high-frequency discharges, best described as myotonic-like, are very common [Gutman et al., 1967; Hogan et al., 1969]. Muscle and liver biopsy specimens contain large amounts of structurally normal glycogen when studied by both light and electron microscopy. Changes in peripheral nerve also have been reported [Araoz et al., 1974].


GAA deficiency is inherited as an autosomal-recessive trait. The gene for human acid α-glucosidase is contained on chromosome 17 (segment q21–q23) [Martiniuk et al., 1986]. The structural gene for human acid α-glucosidase is undergoing intensive study; it is approximately 28 kb in length and contains 20 exons [Martiniuk et al., 1991]. Various mutations may result in the phenotype; missense mutations and failure of an allele to manifest mRNA expression have been reported [Zhong et al., 1991]. Prenatal diagnosis has been available since the 1970s [Hug et al., 1974]. The use of chorionic villus assay allows first-trimester diagnosis [Chowers et al., 1986]. Both adult-onset and infantile glycogenosis type II have been detected in one family. Two types of mutant alleles were identified; one leads to complete deficiency of the enzyme, and the other results in reduced net production of active α-glucosidase, resulting in partial enzyme deficiency [Hoefsloot et al., 1990].


Before the advent of enzyme replacement therapy (ERT), no practical treatment was available. Therapies previously studied included epinephrine administration, which reduced liver, but not muscle, glycogen content to normal [Hug, 1974]. Dietary supplementation with l-alanine, designed to reduce the elevated protein turnover characteristic of acid maltase deficiency, has apparently slowed progression of weakness and even reversed cardiomyopathy in some patients with late infantile and juvenile forms [Bodamer et al., 1997, 2000, 2002].

Attempts to replace the deficient enzyme date back to 1964, but effective ERT was not possible until suitable sources of receptor-targeted human recombinant acid glucosidase became available in the late 1990s. This substance was derived from both rabbit milk and Chinese hamster ovary (CHO) cells. The first clinical trial began in 1998 [Reuser et al., 2002; Winkel et al., 2004], and in 2006, ERT received Food and Drug Administration (FDA) approval for treatment of acid maltase deficiency [Koeberl et al., 2007]. Infants who received treatment early in the course of their illness demonstrated improved strength and cardiac function, with survival now extending over several years. It has become apparent that ERT is most effective at reversing cardiomyopathy and extending the life span of infants, but that skeletal muscle disease is relatively resistant to this modality [Schoser et al., 2008]. The follow-up interval has been too short to determine if anterior horn cell and glial storage of glycogen will lead to chronic weakness and impairment of cerebral function in long-term survivors.