Portal de Periódicos da CAPES

Introduction to the Maastricht workshop: lessons from the past and new directions in galactosemia

2010; Springer Science+Business Media; Volume: 34; Issue: 2 Linguagem: Inglês

10.1007/s10545-010-9232-1

1573-2665

Gerard T. Berry, Louis J. Elsas,

Neonatal Health and Biochemistry

Hereditary galactosemia is an autosomal recessive genetic disorder of carbohydrate metabolism (OMIM 230400; Fridovich-Keil and Walter 2008; Elsas 2010; Berry and Walter 2011). The mutated gene encodes a protein, galactose-1-phosphate uridyltransferase (GALT, EC 2.7.7.12), that catalyzes the conversion of galactose-1-phosphate and UDP-glucose to UDP-galactose and glucose-1-phosphate (Fig. 1). In the newborn period, a life-threatening disease with multiorgan involvement emerges, particularly in those infants ingesting lactose in breast milk and proprietary baby formulas containing the ingredients of cow's milk (Berry and Walter 2011). The first observation of this nutritional toxicity state involving the neonate and breast milk was by von Reuss in 1908 (von Reuss 1908). In the next decade, Göppert documented the presence of excess galactose in the urine of a similarly affected infant (Göppert 1917). The first well characterized infant with hypergalactosemia and galactosuria who responded to a lactose-restricted diet was described by Mason and Turner in 1935 (Mason and Turner 1935). GALT-catalyzed conversion of D-galactose-1-phosphate and UDP-glucose to UDP-galactose and D-glucose-1-phosphate This paper, describing an African-American infant with a variant form of galactosemia, served to reveal marked hypergalactosemia as an integral component of GALT deficiency. However, establishment of the abnormal biochemistry awaited the discovery of Schwarz in 1956 (Schwarz et al. 1956) that the substrate, galactose-1-phosphate, was elevated in erythrocytes from galactosemic patients exposed to galactose, and the demonstration later that year in the Kalkar laboratory that GALT enzyme activity was absent (Isselbacher et al. 1956). This rare Mendelian disorder with a world-wide frequency of 1/40,000 to 1/60,000 newborn infants entered the modern era of molecular biology when Reichardt and Berg cloned a GALT cDNA from a human liver library (Reichardt and Berg 1988), and its correct cDNA sequence enabled cloning and delineation of the structure of the GALT gene (Flach et al. 1990; Leslie et al. 1992). Soon thereafter, many causative gene mutations were identified producing a heterogeneous array of impaired pGALT function (Elsas and Lai 1998; Tyfield 2000; Calderon et al. 2007). It was clear in the decades following the landmark Mason and Turner paper that restriction of galactose intake in the affected newborn infant would usually permit survival and allow the following early infantile complications to remit, resolve or disappear: poor growth, poor feeding, emesis, jaundice, liver enlargement and dysfunction that includes hyperbilirubinemia, transaminasemia, hypofibrinoginemia with bleeding diathesis, cataracts, encephalopathy, including lethargy, irritability and hypotonia, hyperchloremic metabolic acidosis, albuminuria, generalized aminoaciduria and anemia. Yet, between 1970 and 1990, there was a growing awareness that there were “clouds over galactosemia” (Anonymous 1982; Holton and Leonard 1994), that patients were not faring as well as physicians had expected (Komrower and Lee 1970; Lee 1972; Fishler et al. 1972, 1980; Komrower 1982; Waisbren et al. 1983; Gitzelmann and Steinmann 1984), especially given the fact that lactose restriction largely eliminates death from E. coli sepsis (Levy et al. 1977). And, this was all the more poignant, as infants treated very early in life, especially after the introduction of newborn screening for galactosemia in the U.S. in the early 1960s, appeared to be indistinguishable from normal, healthy, thriving infants, or at least until approximately 18–36 months of age when language acquisition and speech become conspicuous landmarks of nervous system development. This naïveté came to an end once and for all when Kaufman et al. published their paper in 1981 on the prevalence of primary ovarian insufficiency (POI) in female patients on lactose-restricted diets (Kaufman et al. 1981), followed by the 1990 retrospective analysis by Waggoner et al., summarizing the outcome of 350 patients from around the world (Waggoner et al. 1990). The latter paper is the most comprehensive of its type. A review of patients’ records from many centers indicated that they had suffered developmental delay involving language acquisition and speech defects as well as POI in females even though they had been on a lactose-restricted diet. Furthermore, these late-onset apparent diet-independent complications occurred even if diet therapy began on day 1 of life (Waggoner et al. 1990; Hughes et al. 2009). In addition to the above well known diet-independent complications (Nelson et al. 1991; Schweitzer et al. 1993; Guerrero et al. 2000; Robertson et al. 2000; Webb et al. 2003; Gubbels et al. 2008; Berry 2008; Potter et al. 2008), we now know that additional complications include poor growth and/or short stature, reduced bone mineral density, deficits in personality construct, mood disorder, tremor, cerebellar ataxia and extrapyramidal movement disorders (Kaufman et al. 1993; Rubio-Gozalbo et al. 2002; Leslie 2003; Bosch 2006; Panis et al. 2007). In some instances, discrete neurological, imaging and neuropathological findings have been recorded (Crome 1962; Huttenlocher et al. 1970; Haberland et al. 1971; Jan and Wilson 1973; Nelson et al. 1992; Belman et al. 1986; Bohles et al. 1986; Koch et al. 1992; Kaufman et al. 1995; Ridel et al. 2005; Hughes et al. 2009). What we don't know is how frequently each of these entities occurs in patients with classic and variant galactosemia at different stages of fetal and postnatal maturation including adulthood. Classic and variant galactosemia were clinical terms used to define the outcome of older patients with impaired GALT function. With the advent of newborn screening, pGALT enzyme and GALT gene mutational analysis, quantitation of erythrocyte galactose-1-phosphate and urinary galactitol in the newborn, particularly in response to galactose-restricted diets and assessment of total body oxidation of galactose to CO2 in breath. These terms must be extended with further prospective research so that we can predict and define clinical outcome in the newborn, infant, child, and adult. Classic galactosemia is predicted in the newborn to occur in cases of absent or barely detectable GALT enzyme activity, persistent elevations of galactose-1-phosphate (erythrocyte range above 20 mg% in the newborn with peak >120 mg% and persisting with diet between 1 and 5 mg%) and galactitol (urine excretion range above 1000 μmol/mmol of creatinine in the newborn with peak greater than 50,000 μmol/mmol of creatinine and persisting with diet between 100 and 400 μmol/mmol of creatinine), and evidence of or the propensity for a neonatal life-threatening multiorgan disorder particularly with liver failure and the risk of lethal E. coli sepsis. Additional defining features include markedly diminished oxidation of [1-13C]galactose to 13CO2 in a 2–5 h breath test (Berry et al. 1995a, 1997, 2000; Barbouth et al. 2007), elevated serum FSH and decreased anti-Müllerian hormone (AMH) levels in female infants and children (Sanders et al. 2009), and brain MRI lesions (white matter hyperintensities, cortical atrophy and cerebellar atrophy). Most of these patients will possess severe GALT gene mutations such as Q188R, K285N, Δ5.2 kb deletion and L195P either in homozygous or compound heterozygous states. Clinical variations occur producing heterogeneous outcomes, such as the newborn who may manifest complications such as poor growth, hepatomegaly, jaundice, anemia, developmental delay and cognitive impairment if exposed to dietary lactose, but if screened and put on galactose restriction will have a normal outcome. This is seen in the S135L/S135L genotype (Lai et al. 1996; Elsas and Lai 1998; Manga et al. 1999). The S135L allele is common in Africa and is present in the majority of African-Americans in the U.S. (Lai et al. 1996). Their total body galactose oxidation is normal in later childhood (Berry et al. 1997). These clinically significant variants differ from the largely benign Duarte D2 compound heterozygous condition that seems to be only a biochemical variant, not a clinical variant (Ficicioglu et al. 2008). The most common genotype example in the Western world may be the Q188R/4 bp deletion-cis-N314D. The AGCT 4 bp deletion reduces GALT gene expression (Elsas et al. 2001). Before one can successfully use the new technologies available to uncover the mechanisms of disease in hereditary galactosemia, it is imperative that we first define the phenotypes in patients with diverse forms of galactosemia and at different stages of development. There have been many outstanding papers published delineating the prevalence of one or more long-term complications such as DQ, IQ and brain MRI findings in a cohort of patients, but few are prospective in nature and almost none are comprehensive. Obviously, more well-defined, careful, prospective clinical research needs to be performed to sharpen our focus on exactly what are the classic and clinical variant galactosemic phenotypes and how they change with time (Schadewaldt et al. 2010; Doyle et al. 2010). Ideally, these prospective studies should be performed with the same instruments of investigation and properly trained professionals with like-minded effective approaches, should incorporate NBS variables and should be international and multicenter in nature. The current therapy for classic galactosemia is less than ideal. Unlike treated patients with phenylketonuria, many older individuals suffer a poor quality of life (Bosch et al. 2004, 2009). Some even appear to be quite handicapped. Is a strict diet after late infancy or puberty partly to blame for the CNS disease? By contrast does the chronically elevated galactose-1-phosphate continue to produce toxicity (Guerrero et al. 2000; Robertson et al. 2000; Webb et al. 2003; Lai et al. 2009)? There is a great need to answer this question. It is very likely that there will not be simple and singular answers. It is also likely that we will not be able to effect better therapies and improve our patients’ lives unless we better delineate disease mechanisms. An understanding of the critical global pathophysiology as well as local perturbations in cell biology within target tissues such as the CNS and ovary are of tremendous importance. From a broader perspective, several over-all mechanisms of disease can be considered. To date, most of the data suggest that cognitive impairment and speech defects, as well as POI, have their origin in prenatal life; neonatal lactose exposure may magnify these toxicities. Why do we need to consider a priori each disturbance as being promulgated at the level of the genome, the protein, and the metabolome? This is not simply an exercise in being thorough. The answer has to do with the enigmatic GALT knockout mouse (Leslie et al. 1996; Ning et al. 2000, 2001). Although this murine “model” is without GALT mRNA and protein levels yet displays evidence of severe hypergalactosemia (serum galactose may be > 20 mmol/L) and increased concentrations of galactose-1-phosphate, galactitol and galactonate in tissues, the mice are essentially without relevant clinical disease (no cataracts, CNS pathology, POI, etc.). They usually manifest an osmotic diuresis as their sole evidence of clinical disease! Considering man vs. mouse, there are several etiologic considerations. An interesting explanation was recently offered when the signaling message for Aplysia ras homolog I (ARHI) was found increased in galactose-stressed human GALT-deficient (Q188R/Q188R and ∆5.2 kb/∆5.2 kb genotypes) fibroblasts (Lai et al. 2008, 2009). The mouse has lost the gene for ARHI during evolution, but if ARHI is overexpressed in the mouse using transgenic techniques, the mouse will have growth restriction, ovarian failure, and and a twirler phenotype. Another possible pathological reason for the mouse-man differences may be related to a lesion in the GALT genomic space of man that would have unexpected effects producing loss of miRNA and neighboring effects on the interleukin-11 alpha-subunit receptor gene expression (Magrangeas et al. 1998). Thirdly, GALT may be a “moonlighting” protein in man and, unlike in the mouse, may play another role in human metabolism. Only the absence of a murine ARHI gene currently explains why the knockout mouse fails to exhibit both acute neonatal and chronic long-term complications, and the metabolomic stress induced by excess accumulation of galactose-1-phosphate. It would be very appropriate to determine whether the administration of a safe and specific galactokinase inhibitor to the patient with classic disease could alter the phenotype (Wierenga et al. 2008; Lai et al. 2009; Tang et al. 2010). Appropriate murine models with GALT deficiency and expression of the ARHI gene would be necessary for initial evaluation of galactokinase inhibitors. The more one contemplates the conundrum that is galactosemia, the more its disparate array of organ-specific and highly variable features is understood to describe a complex genetic disease whose pathophysiological mechanisms offer insight into more common disorders such as ovarian insufficiency, cerebellar development, bone mineralization, growth and neurological development. In other words, although galactosemia is a Mendelian single gene disease it becomes an important human disorder of “metabolomics” where modifier genes and epigenetic effects will explain the highly variable clinical complications and offer therapeutic interventions for more common disorders. In fact the search has begun in yeast and humans for answers to the effects of disturbing a single protein required for life through evolution (Ideker et al. 2001; Coman et al. 2010). There is still much work to be done to help change the lives of patients with this orphan disease. The new directions must include initiatives to correct the genetic defect, cellular transplantation and pharmacologic treatment to temper the toxicity of galactose-1-phosphate accumulation. With eyes wide open, we fully realize the tremendous hurdles that we face, not the least of which is the harsh reality that no treatment may be effective if we cannot somehow correct the GALT gene defect or replace the GALT enzyme in developing neurons and/or glial cells in the affected fetus. This may be true for ovarian cells as well as the bulk of data suggest that granulosa cells and oocytes are dysfunctional very early in life. But this is not a reason to abandon aims to prevent ongoing damage to target tissues in postnatal life especially in the newborn period. Earlier diagnosis and implementation of lactose restriction in the neonatal period should be a priority. It is very unfortunate that a suitable animal model does not exist. One may argue that the galactose-1-phosphate and/or galactitol levels in the brain, lens, and liver of the GALT knockout mouse do not rise to the levels seen in toxic newborn infants with massive hypergalactosemia. But we may still use the murine GALT deficiency model to establish proof of priniciples regarding the biochemical efficacy and safety of galactokinase and aldose reductase inhibitors and the requirement for stress-related responses that are organ specific. A rodent model may also be used in the future to study the effects of gene transfer in diverse tissues or in toto, as well as determine the effects of stem cell transplantation pre- and postnatally with and without ex vivo gene transfer. Still lacking is a definitive delineation of the phenotype of patients, especially adults, with classic and variant forms of galactosemia. We must know the genotype-phenotype relationships before rational decisions can be made about the targets and goals of our interventions. A complete understanding of the classic phenotype over time will also help us understand the magnitude of pre- versus postnatal toxicity and whether the complications of galactosemia worsen with time, because it is fundamentally a progressive disease, in part a neurodegenerative disorder. Most of the data suggest that galactosemia is not a neurodegenerative disease such as a sphingolipidosis but this still remains to be established with certainity. Some physicians are concerned that we have transformed galactosemia into a progressive disease through the very use of a chronic strict diet therapy that limits galactose intake and thus creates further deficiency of UDP-galactose and UDP-glucose in some target tissues (Lai et al. 2003). These concerns need to be addressed as soon as possible. This workshop at the Maastricht University was an important first step in our quest to improve the care of these patients, to rectify errors in our management, to eliminate biases or prejudices in our thinking about this enigmatic disease, to open our minds to novel concepts of the mechanisms of disease and pathophysiology and, most importantly, to initiate a broad cooperative international study of galactosemia. By working together in a collegial fashion, we can maximize our individual strengths and solve these seemingly unsolvable problems.