- Email: [email protected]
UDP-galactose pyrophosphorylase in mice with galactose-1-phosphate uridyltransferase deficiency
Molecular Genetics and Metabolism 85 (2005) 21–27 www.elsevier.com/locate/ymgme
UDP-galactose pyrophosphorylase in mice with galactose-1-phosphate uridyltransferase deficiency Nancy Leslie a
a,*
, Claire Yager b, Robert Reynolds b, Stanton Segal
b
Division of Human Genetics, Cincinnati ChildrenÕs Hospital Research Foundation, Cincinnati, OH 45229, USA b Metabolic Research Laboratory, The ChildrenÕs Hospital of Philadelphia, Department of Pediatrics, University of Pennsylvania School of Medicine, Philadelphia, PA 19104, USA Received 11 November 2004; received in revised form 12 January 2005; accepted 12 January 2005 Available online 23 February 2005
Abstract UDP-glucose pyrophosphorylase (E.C. 2.7.7.9), encoded by ugp, provides UDP-glucose which is critical to the synthesis of glycogen, and also catalyzes the reaction between UTP and galactose-1-phosphate, yielding UDP-galactose. This activity of UDP-gal pyrophosphorylase (UDP-galPP) suggests a role in an alternate pathway for galactose metabolism in patients with deficiency of galactose-1-phosphate uridyltransferase (GALT). We examined the effects of GALT deficiency and dietary galactose on UDP-glucose pyrophosphorylase (UDP-gluPP) and UDP-galactose pyrophosphorylase activity and ugp expression in liver of mice with homozygous deletion of the critical regions of galt. Activity with glucose-1-phosphate as substrate was significantly higher than that with galactose-1-phosphate. In liver from mice with GALT deficiency (G/G), UDP-galPP activity appeared to be lower than that measured in liver from control (N/N) animals. This difference disappeared when the N/N tissue homogenate was dialyzed to remove residual UDP-glucose, confirming that careful elimination of residual GALT activity is necessary, since GALT has 1000-fold greater activity toward galactose-1-phosphate than that of UDP-galPP in liver homogenates. Prior exposure to conventional mouse chow, high galactose chow, and high glucose chow did not alter UDP-glu PP or UDP-galPP activity. Steady state UGP mRNA levels were determined in tissues from normal and G/G animals. UGP expression was highest in liver, and did not differ by genotype or exposure to high galactose chow. UDP-galPP activity may account for unexplained ability to oxidize galactose in animals with no GALT activity, but is insufficient to alter accumulation of galactose metabolites. 2005 Elsevier Inc. All rights reserved. Keywords: Galactose-1-phosphate uridyltransferase; UDP-glucose; UDP-galactose; Galactose
Introduction UDP-glucose pyrophosphorylase (E.C. 2.7.7.9), encoded by ugp, occupies a pivotal position in the synthesis of glycogen by forming UDP-glucose from glucose-1phosphate and UTP (Fig. 1) This enzyme also catalyzes the reaction between UTP and galactose-1-phosphate, yielding UDP-galactose [1,2]. The existence of this catalytic activity was convincingly demonstrated by
*
Corresponding author. Fax: +1 513 636 2261. E-mail address: [email protected] (N. Leslie).
1096-7192/$ - see front matter 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.ymgme.2005.01.004
Isselbacher in 1957 [3]. Although unaware of this proteinÕs dual role in both galactose and glucose metabolism, his studies in rat liver clearly showed activity toward galactose-1-phosphate that was separate from galactose-1-phosphate uridyltransferase (EC 2.7.7.10) (GALT), establishing evidence for a possible alternative pathway for galactose metabolism in patients with GALT-deficient galactosemia [3]. UDP-galactose pyrophosphorylase was further characterized in normal human liver by Abraham and Howell [4]. Their studies did not directly measure UDP-galPP activity in an individual with galactosemia, but did show that the contribution of UDP-galPP to total nmoles [14C]UDP-
22
N. Leslie et al. / Molecular Genetics and Metabolism 85 (2005) 21–27
UDP galPP activity in liver could account for the observed ability of GALT-deficient mice to oxidize galactose to expired CO2.
Materials and methods Animals
Fig. 1. UDP-galactose and UDP-glucose metabolism through GALT and UGP catalyzed pathways. The dark gray curved arrows indicate the active pathway of GALT. The straight arrows in the center indicate the major pathway of UDP-gluPP as well as the minor UDP-galPP activity. The diagonal arrows show the relationships between the substrates and products of both pathways.
galactose formed was much less than 1% of that attributed to GALT. This low contribution of UDP-galPP to UDP-gal synthesis was confirmed in normal human liver by Shin et al [5]. The functional significance of UDP-glucose pyrophosphorylase (UTP:a-D-glucose-1-phosphate uridyltransferase) has been demonstrated in bacteria as well as yeast and cultured CHO cells with a point mutation in ugp. Bacteria with absent UDP-gluPP have little UDP-glu content [6]. Yeast with deficient UDP-gluPP activity have low glycogen stores and impaired growth [7]. CHO cells with deficient UDP-gluPP activity have decreased steady state UDP-glucose levels, and functionally show differential sensitivity to the phospholipase C of Clostridium perfringens [8]. The importance of UDP-galPP in yeast deficient in GALT activity is clear when these strains are grown on galactose as a sole carbon source. Selection for revertants capable of growing in galactose medium revealed strains with up-regulated expression of UGP 1 [9]. Similarly, cultured human GALT-deficient fibroblasts transfected with a plasmid expressing ugp showed improved growth in galactose containing medium and higher steady state levels of both UDP-glucose and UDP-galactose than sham transfected cells [10]. However, there are no known mammalian models for deficiency of UGP, and only limited investigation of UGP activity in humans with GALT deficiency. Indeed, the only data on human liver from patients with GALT deficiency relate to its UDP-gluPP activity [11] and there are no data on the role of UDP-galPP to circumvent the block in galactose metabolism in human galactosemia. In this report, we have examined the effects of GALT deficiency and dietary galactose on UDP-gluPP and UDP-galPP activity and ugp expression in liver of mice with homozygous deletion of the critical regions of galt. These experiments directly assess the role of UDP-galPP in a mammalian model of GALT deficiency, and seek to answer whether
The 30–50 g mice used in these studies correspond to those employed in our previous reports [12–14]. The GALT-deficient mice (G/G), bred to a Balb/C background, were from well-established colonies in Philadelphia and Cincinnati. Their genotype was determined by polymerase chain reaction of DNA extracted from tail snips at 21 days of age. The normal (N/N) Balb/c mice were derived from the Charles River Laboratories colony (Wilmington, MA). Mice were weaned at 21 days of age to either chow 5015 or a 40% galactose or 40% glucose diet in which the carbohydrate was replaced by the hexose. All diets were purchased from PMI Feed (Richmond, IN). These diets were analyzed previously: chow 5015 contains small amounts of bioavailable galactose and the 40% glucose diet is galactose free [14]. Enzyme analysis Animals were killed by cervical dislocation, and the livers were removed and homogenized or flash-frozen with metal tongs cooled in liquid nitrogen. Frozen specimens were placed in plastic bags and stored at 80 C until analyzed. A piece of liver, fresh or frozen, was weighed and homogenized in four volumes of 0.1 M KCl with a motor-driven homogenizer with a teflon tip in a glass Potter-Elvehjem tissue grinding tube. The homogenate was then centrifuged for 10 min at 750g at 0 C. The supernatant from the low speed spin was then centrifuged further at 20,000g at 4 C for 30 min. The supernatant below the lipid layer was removed and used for the enzyme assays or subjected to dialysis. The dialysis procedure was carried out in a Slide-a-lyzer (Pierce Biotechnology, Rockford, IL) cassette with a ratio of 1–900 ml dialyzing solution of Tris–HCl, pH 8, 5 mmol/L, MgCl2 10 mmol/L, and EDTA 0.1 mmol/L with three exchanges of 900 ml each over a 24-h period. Protein concentrations were determined using the BioRad assay (Bio-Rad Laboratories, CA). Each assay contained 20 lmol of 1 M Tris buffer (pH 8.1), 5 lmol of MgCl2, 2 lmol dithiothreitol (ClelandÕs reagent), 6.46 nmol of unlabelled glucose-1-phosphate or galactose-1-phosphate, 1.7 lmol UTP, and 0.05 lCi 14 C-a-D-galactose-1-phosphate or 14C-a-D-glucose-1phosphate. Approximately 1 mg. of liver homogenate protein was added for UDP-galPP determination. Since the UDP-gluPP was many-fold more active with glu-1-P as substrate, the homogenate was diluted so that 9 lg
N. Leslie et al. / Molecular Genetics and Metabolism 85 (2005) 21–27
was added to give a linear rate for a 5-min assay. The total volume of each assay is reached by adding H2O to a level of 430 ll. The tubes were preincubated for 3 min at 37 C. The reaction was initiated with the addition of the radioactive substrate. Reactions were run for the appropriate time period at 37 C. The enzyme reactions were stopped by immersing the tubes into a 100 C water bath for 2 min. The tubes were cooled in ice for 5 min and then centrifuged refrigerated for 10 min at 1000g. Background tubes containing enzyme homogenates were first pre-incubated at 100 C for 3 min to inactivate any catalytic activity, then handled as described above. After centrifugation, a 20 ll sample from each assay tube was spotted in duplicate on DEAE paper strips measuring 1.5 cm in width and 28 cm in length. The sample was spotted at the origin, 7 cm from one end. After drying, the papers were developed by descending chromatography in closed tanks using 0.05 M LiCl as a solvent. The papers were removed from the tank just before the solvent front reached the bottom of the papers and then were hung to dry. The papers were then cut into appropriate sections and counted in scintillation fluid. The first four centimeters contained the origin and any unmigrated substances. Radioactivity in the next section contained the UDP-gal or UDP-glu. The length of this section varied slightly and was determined each time by cutting a strip into 22 1 cm sections and counting them to determine the optimal length of this portion. The next section contained the galactose1-phosphate or glucose-1-phosphate. The remaining 3 cm contained the counts for galactose or glucose. Enzyme activity was calculated as the nmol of product formed per mg protein. The calculations are performed by determining the ratio of radioactivity in the product/total radioactivity on the chromatographic strip. The ratio of radioactivity in the same regions of the background incubation was then subtracted from that of the experimental assays. The resulting value is multiplied by the initial concentration of the substrate and then divided by the protein content in the assay tube, as depicted in the equation below. nmol=min=mg ¼ f½ðcpm product=total cpm per stripÞ ðblank cpm product=total cpm per stripÞ ðnmol=L substrateÞg=mg protein
23
Ambion, Austin, TX) and quantitated by UV absorption (Beckman DU 640). Total RNA was separated on a 1% agarose/formaldehyde gel and transferred to a Hybond-N+ membrane (Amersham–Pharmacia, Piscataway, NJ). A 359 bp mUGP cDNA fragment was synthesized by rtPCR from mouse placental RNA, using primers selected from sequences common to the human UGP (NM 001001521), bovine UGP (NM 174212), and mouse EST sequences available in GenBank (forward: GCCAGAGGCTTGCCTGATAACATATC, reverse: CCCTGAATAAGACACATCCTTTGC) and the identity of the resulting product was confirmed by automated sequencing. A template appropriate for T7 transcription was synthesized using an antisense primer (TAATACGACTCACTATAGGAGGGTACCTGCT TTGATTG) incorporating a T7 tag coupled with antisense sequences 50 bp 5 0 of the original reverse amplification primer. Hybridization antisense riboprobes for mUGP or mouse TriActin (Ambion) were transcribed with T7 polymerase and [32P]UTP (800 Ci/mmol) using the Maxiscript kit (Ambion).
Reagents Uniformly labeled a-D-[14C]galactose-1-phosphate diammonium salt was prepared from D-galactose 14 C(U). It was purchased from American Radiolabeled Chemicals (St. Louis, MO) with an initial specific activity of 55 mCi/mmol. a-D-[U14C]Glucose-1-phosphate, potassium salt were purchased from Amersham Life Science (Arlington Heights, Illinois), 284 mCi/mmol. Unlabeled galactose-1-phosphate or glucose-1-phosphate, Tris, magnesium chloride, and dithiothreitol were purchased from Sigma (St. Louis, MO). Lithium chloride was obtained from Baker Chemical Company (Phillipsburg, NJ). DEAE–cellulose ion exchange paper was a product of Whatman (Clifton, NJ). Oligonucleotides were purchased from Gibco/BRL (Gaithersburg, MD).
Statistics The significance of differences was determined from the mean and standard error of the mean by StudentÕs t test.
Northern analysis of mouse UGP
Results
Adult g/g mice were exposed to 1 week of chow feeding or 1 week of 40% galactose feeding, then sacrificed by cervical dislocation after induction of CO2 narcosis. Tissues were quickly removed and frozen in liquid nitrogen. Total RNA was extracted by guanidinium extraction of pulverized tissue powder (Totally RNA,
UDP-gluPP activity The protein product of ugp catalyzes formation of both UDP-glucose and UDP-galactose. To determine the effect of GALT genotype on UDP glucose formation, UDP-glucose produced from Glu-1-P was
24
N. Leslie et al. / Molecular Genetics and Metabolism 85 (2005) 21–27
Fig. 2. Effect of time on glucose-1-P pyrophosphorylase activity in normal mouse liver (–) and in GALT-deficient mouse liver (j–j). Each data point is the average of two determinations.
measured in N/N and G/G liver (Fig. 2). There was no difference in the UPDgluPP activity between the two genotypes. The reaction was rapid and equilibrium was reached after 10 min when 7.5 lg of homogenate protein was employed. Because of the high activity, the original supernatant was diluted until the reaction was linear for the 5-min incubation as shown in the figure. UDP-galPP activity The activity of UDP-galPP for N/N and G/G liver is shown in Fig. 3. When fresh undialyzed supernatants were compared, there was a marked difference in apparent UDP-galPP activity between the two genotypes. In Fig. 3A the activity curve for the N/N mice is linear for 10 min and reaches equilibrium at 30 min incubation of 3.73 nmol/mg protein. In contrast, the activity curve for undialyzed G/G liver (Fig. 3B) was slow and linear over the same time period with an activity of 0.1 nmol/mg protein at 40 min. At 5 min, the apparent activity in the N/N homogenate was over 40 times that of the G/G. This discrepancy has two possible explanations. Accumulated gal-1-P in the liver of the G/G mice could inhibit UDP-galPP activity. An alternative hypothesis can be derived from the careful work of Isselbacher [3] who demonstrated the importance of dialysis for in order to measure UDP-galPP activity in
normal rat liver. In the rat liver, as in N/N mouse liver, GALT activity is present, and residual UDP-glu in the preparation will react with labeled Gal-1-P to produce labeled UDP-gal, the product measured in UDP-galPP reaction (see Fig. 1). As shown in Fig. 3A, dialysis of the N/N liver homogenate, which would decrease the UDP-glu present, markedly decreases the apparent UDP-galPP levels. The activity at 10 min is reduced from 1.97 to 0.05 nmol/mg protein and the curve (Fig. 3A) resembles that of undialyzed G/G supernate (Fig. 3B). However, dialysis does not alter the observed UDP-galPP activity in G/G liver as shown in Fig. 3B. These experiments demonstrate that dialysis decreases apparent UDP-galPP activity in GALT intact animals and does not increase apparent UDP-galPP activity in GALT-deficient animals. We therefore conclude that the difference in UDP-galPP activity between the two genotypes is not due to inhibition of catalytic activity in the G/G liver by accumulated inhibitors, such as gal-1-P, but rather the overestimation of UDP-gallPP activity in N/N liver due to contaminating GALT activity. Indeed, the activity of UDP-galPP in the G/G mouse homogenate is the true activity of the enzyme with Gal-1-P as substrate. The ratio of UDP-gluPP to UDP-galPP and ratio of UDP-galPP to GALT A comparison of the rate of UDP-glu formation in N/N and G/G liver is shown in Fig. 2. UDP-glu formation after a 5 min reaction for N/N and G/G liver, as well as for dialyzed N/N and G/G homogenates, is shown in Table 1. The UDP-gluPP activity is about 4000 times that of UDP-galPP. The activity of GALT in N/N mice is 10 nmol/min/ mg protein [12]. The UDP-galPP in mice is 0.0037 nmol/min/mg protein using the 5-min value of dialyzed G/G mouse value of Table 1. The GALT activity is 2700 times that of UDP-galPP confirming that in normal liver formation of UDP-gal via GALT and the Leloir pathway is the predominant route for metabolism of galactose.
Fig. 3. (A) Effect of time on galactose-1-P pyrophosphorylase activity in normal mouse liver (–) and in dialyzed normal mouse liver (j–j). Each data point is the average of 2–5 determinations. (B) Effect of time on galactose-1-P pyrophosphorylase activity in GALT-deficient mouse liver (}–}) and in dialyzed GALT-deficient mouse liver (h–h).
N. Leslie et al. / Molecular Genetics and Metabolism 85 (2005) 21–27 Table 1 Comparison of UDP-gluPP to UDP-galPP Homogenate
UDP-gluPP (nmol/5 min)
UDP-galPP (nmol/5 min)
N/N N/N G/G G/G
126 113 113 132
0.88 0.75 0.023 0.037
a b
undialyzed dialyzed undialyzed dialyzed
(2) (2) (2) (2)
Ra
(4) ± 0.09b (3) ± 0.005b (4) ± 0.006b (3) ± 0.004b
4913 3568
UDP-gluPP/UDP-galPP. Mean (number of determinations) ± SEM.
25
for a small but statistically significant increase in 30-min chow values. This finding was not reproduced with increasing galactose content of the diet. A comparison of the 30-min measurements of dialyzed N/N preparations with that of the G/G animals, dialyzed and undialyzed, shows a consistent increase in apparent UDPgalPP activity in the N/N livers. It is possible that a small presence of GALT activity remains despite extensive dialysis to rid the homogenate of UDP-glu.
Effect of diet on UDP-galPP
Effect of GALT genotype on UGP expression
Previous studies of the ability of G/G mice to oxidize galactose to CO2 showed that the percent of the dose oxidized in 4 h increased from a chow-fed groupÕs 5.5– 12.9%, when the animals were fed either a 40% galactose or glucose diet for 4 weeks [13]. Normal animals which oxidized 39% of the administered dose in 4 h showed no such increase. To determine if the feeding of galactose or glucose diets enhanced galactose metabolism by augmenting UDP-galPP, a study was performed to measure the liver enzyme activity after mice had been weaned to mouse chow, 40% galactose, or 40% glucose diets for 4 weeks. The results are shown in Table 2 for both N/N and G/G liver homogenates before and after dialysis. There was no difference in the UDP-galPP activities measured in livers from animals fed chow, galactose, or glucose, if comparisons were made between homogenates from the same genotype or dialysis condition. The data suggest that there is no induction or enhancement of UDP-galPP activity in animals fed a high galactoseor a high glucose-containing diet. As expected, there was a marked difference in apparent UDP-galPP activity between undialyzed and dialyzed preparations of N/N mouse homogenates across all diets. These data support those shown in graphic form in Fig. 3 which shows the time vs UDP-galPP activity. In G/G homogenates, undialyzed or dialyzed preparations did not differ except
Steady state levels of UGP message were unchanged when mice of G/G genotype were exposed to high galactose chow (Fig. 4). Although this manipulation increases gal-l-levels considerably in these tissues, no increase in steady state transcript levels was found. Of these tissues (brain, liver, kidney, and ovary), the highest expression of UGP was detected in liver. Similar results were found for N/N tissues (data not shown).
Fig. 4. Northern analysis of tissues from adult G/G mice consuming conventional mouse chow or 40% galactose chow. The upper panel is hybridized to a mouse UGP riboprobe. The lower panel is hybridized to mouse tri-actin. Tissues include brain (B), liver (L), kidney (K), and ovary (O).
Table 2 Effect of diet on UDP-galactose pyrophosphorylase activity in liver of normal and GALT-deficient mouse Incubation time/min
DIET (nmoles/mg protein) Chow Undialyzed
Galactose Dialyzed
Glucose
Undialyzed
Dialyzed
Undialyzed
Dialyzed
N/N mouse 5 30
0.88 (4)a ± 0.09b 3.36 (7) ± 0.24
0.075 (3) ± 0.005 0.16 (4)c ± 0.009
0.87 (9) ± 0.08 3.26 (12) ± 0.14
0.053 (9) ± 0.006 0.14 (8) ± 0.013
0.68 (7) ±0.04 3.39 (9) ± 0.35
0.07 (4) ± 0.04 0.14 (8)c ± 0.01
G/G mouse 5 30
0.023 (4) ± 0.006 0.073 (5) ± 0.004
0.043 (3) ± 0.016 0.118 (4)d ± 0.009
0.033 (3) ± 0.004 0.07 (3) ± 0.004
0.037 (3) ± 0.008 0.103 (3) ± 0.023
0.025 (2) 0.11 (3) ± 0.004
0.043 (3) ± 0.02 0.11 (3) ± 0
Number of specimens are indicated in the parentheses. a Mean. b SEM. c Difference from the 30 min G/G value p < 0.02. d Difference from the 30 min G/G undialyzed value, p < 0.01.
26
N. Leslie et al. / Molecular Genetics and Metabolism 85 (2005) 21–27
Discussion The present study is the first to report UDP-galPP activity in liver of an animal with complete absence of GALT activity. We found the same low activity in GALTless liver as in normal tissue. The exposure of the G/G animals to a high galactose diet did not alter catalytic activity or steady state levels of ugp message in tissues from GALT-deficient mice. Our results emphasize the importance of complete elimination of the GALT activity in normal tissue if an accurate estimate of the UDP-galPP contribution is to be made. Based on our findings in dialyzed wild type liver homogenates, the estimated UDP-galPP activity responsible for galactose conversion to UDP-gal is 0.1% that of GALT. In our previous studies, we observed that GALT-deficient adult mice could oxidize 4–12% of an administered dose of 7 mg galactose per kg over 4 h to CO2 [14]. Does UDP-galPP activity account for this level of oxidation? In a 50 g mouse, this is approximately 0.02–0.05 lmol galactose oxidized per hour. With UDP-galPP activity measured at 0.08 lmol/mg protein in 40 min in a mouse liver weighing 1.2–1.5 g and with 200 mg protein per gm wet weight, the UDP-galPP activity could account for metabolism of 0.03 lmol galactose per hour. Therefore, within the limits of measurement, UDP-galPP may account for a large portion of the limited galactose oxidation in mice without GALT activity [13,14]. The contribution of liver UDP-galPP activity to the ability of galactosemic patients to oxidize isotopic galactose to CO2 is unknown [15,16]. Humans with the classical form of GALT deficiency can oxidize up to 50% of administered [13C]galactose to CO2 in 24 h [16]. Segal and Cuatrecasas [17,18] on the basis of differences in the oxidation of 1- and 2-14C-labeled galactose suggested that half the oxidation could be due to direct oxidation via the galactonate pathway with the remainder being due to residual GALT or another pathway such as UDP-galPP. Convincing data that a pathway other than GALT is responsible comes from the study of a patient with a homozygous deletion of exons 1–10 of GALT which revealed considerable oxidation (17)% of a [13C]galactose dose in a 24-h period [19]. Our data in mice with a deletion of three exons of galt, together with direct evidence of UGP expression in humans, suggests that UDP-galPP activity may explain at least some of the measured [13C]galactose oxidation. Although UDP-galPP action may be sufficient to explain much of the slow phase of galactose oxidation in GALT-deficient mice,its expression in liver clearly does not prevent accumulation of galactose metabolites in these mice [12–14]. There is sufficient data in human tissues to demonstrate existence of similar activity in liver, muscle [5] but this alternate pathway does not prevent
accumulation of galactose metabolites in tissues of patients with GALT deficiency, nor does it alter endogenous production or long term morbidity in these patients [20,21]. The forward reaction UTP + Gal-1-P fi UDPgal + PP is the prominent direction of the reversible reaction [3]. However, Gitzelmann [22] showed evidence for the pyrophosphorylitic cleavage of UDP-gal reacting with PP to form UTP and Gal-1-P. Based on this, he proposed that the reaction was responsible for the continuous endogenous production of Gal-1-P, thereby resulting in self-intoxication of the patient and an explanation for the long-term, diet-independent complications of the disorder. On the other hand, Oliver [23] presented evidence that Gal-1-P inhibited UDP-gluPP in liver and brain and postulated that a resulting deficiency of UDP-glu would have consequences in the formation of glucuronides, glycogen and complex carbohydrates. In contrast, Pourci et al. [24] reported that galactosemic fibroblasts grow normally when incubated with galactose if inosine is present even when Gal1-P levels are very high. They concluded that a lack of energy prevented cell growth in the presence of galactose alone and that the ribose moiety of inosine provided the fuel for growth. The combined presence of UGP and epimerase explains why cultured galactosemic fibroblasts grown in glucose media have normal UDP-gal and UDP-glu concentrations [25] and that there is no deficit of nucleotide sugars in red blood cells [26] or leucocytes [27] of galactosemic patients. The inhibition of multiple enzymes by high cellular Gal-1-P levels is well established [23] but the findings of in vitro models with cells exposed to galactose as a sole substrate may not parallel those of cells in vivo which never experience exposure to galactose alone. Suckling GALT-deficient mice and adult mice fed 40% galactose accumulate 2–3 mM hepatic Gal-1-P and do not have a galactosemic phenotype [13]. In chow-fed G/G mice with hepatic Gal-1-P of 1.75 mM, there is no abnormality in the liver levels of UDP-glu or UDP-gal [14]. The lack of a galactose toxicity syndrome in GALT-deficient mice in the face of a markedly elevated Gal-1-P demonstrates that in mice clinical morbidity does not have a simple relationship to expression of UGP or altered activity of its protein product. If a small amount of UDP-galPP activity allows oxidation of galactose to CO2, would overexpression of UGP by some means rescue the human phenotype in GALT deficiency? Such possibilities are suggested by the in vitro rescue studies of Lai et al. [9,10]. However, in yeast systems, overexpression of UGP causes a growth restricted phenotype and UGP activity is regulated by PAS kinase-mediated phosphorylation [28]. Both UDP-glucose and UDP-galactose bind to the P2Y receptor and may have roles as extracellular signal-
N. Leslie et al. / Molecular Genetics and Metabolism 85 (2005) 21–27
ing intermediates [29]. In complex mammalian systems, the potential for overexpression of UGP to alter pathways other than those involved in galactose metabolism deserves careful consideration.
Acknowledgments The authors gratefully acknowledge the assistance of Janice Malseed, Terri Wallace, and Shuzhen Bai. This work was supported by R01-DK HD 60768-01A1 to S. Segal.
References [1] J.K. Knop, R.G. Hansen, Uridine diphosphate glucose pyrophosphorylase. IV. Crystallization and properties of the enzyme from human liver, J. Biol. Chem. 245 (1970) 2499–2504. [2] R.L. Turnquist, T.A. Gillett, R.G. Hansen, Uridine diphosphate glucose pyrophosphorylase. Crystallization and properties of the enzyme from rabbit liver and species comparisons, J. Biol. Chem. 249 (1974) 7695–7700. [3] K.J. Isselbacher, A mammalian uridinediphosphate galactose pyrophosphorylase, J. Biol. Chem. 232 (1958) 429–444. [4] H.D. Abraham, R.R. Howell, Human hepatic uridine diphosphate galactose pyrophosphorylase. Its characterization and activity during development, J. Biol. Chem. 244 (1969) 545–550. [5] Y.S. Shin, H.P. Niedermeier, W. Endres, J. Schaub, S. Weidinger, Agarose gel isoelectrofocusing of UDP-galactose pyrophosphorylase and galactose-1-phosphate uridyltransferase. Developmental aspect of UDP-galactose pyrophosphorylase, Clin. Chim. Acta 166 (1987) 27–35. [6] T.A. Sundararajan, A.M. Rapin, H.M. Kalckar, Biochemical observations on E. coli mutants defective in uridine diphosphoglucose, Proc. Natl. Acad. Sci. USA 48 (1962) 2187–2193. [7] J.M. Daran, W. Bell, J. Francois, Physiological and morphological effects of genetic alterations leading to a reduced synthesis of UDP-glucose in Saccharomyces cerevisiae, FEMS Microbiol. Lett. 153 (1997) 89–96. [8] M. Flores-Diaz, A. Alape-Giron, R.W. Titball, et al., UDPglucose deficiency causes hypersensitivity to the cytotoxic effect of Clostridium perfringens phospholipase C, J. Biol. Chem. 273 (1998) 24433–24438. [9] K. Lai, L.J. Elsas, Overexpression of human UDP-glucose pyrophosphorylase rescues galactose-1-phosphate uridyltransferase-deficient yeast, Biochem. Biophys. Res. Commun. 271 (2000) 392–400. [10] K. Lai, S.D. Langley, F.W. Khwaja, E.W. Schmitt, L.J. Elsas, GALT deficiency causes UDP-hexose deficit in human galactosemic cells, Glycobiology 13 (2003) 285–294. [11] E.P. Anderson, K.J. Isselbacher, H.M. Kalckar, Defect in uptake of galactose-1-phosphate into liver nucleotides in congenital galactosemia, Science 125 (1957) 113–114.
27
[12] N.D. Leslie, K.L. Yager, P.D. McNamara, S. Segal, A mouse model of galactose-1-phosphate uridyl transferase deficiency, Biochem. Mol. Med. 59 (1996) 7–12. [13] C. Ning, R. Reynolds, J. Chen, et al., Galactose metabolism in mice with galactose-1-phosphate uridyltransferase deficiency: Sucklings and 7-week-old animals fed a high-galactose diet, Mol. Genet. Metab. 72 (2001) 306–315. [14] C. Ning, R. Reynolds, J. Chen, et al., Galactose metabolism by the mouse with galactose-1-phosphate uridyltransferase deficiency, Pediatr. Res. 48 (2000) 211–217. [15] G.T. Berry, R.H. Singh, A.T. Mazur, et al., Galactose breath testing distinguishes variant and severe galactose-1-phosphate uridyltransferase genotypes, Pediatr. Res. 48 (2000) 323–328. [16] G.T. Berry, R.A. Reynolds, C.T. Yager, S. Segal, Extended [13C]galactose oxidation studies in patients with galactosemia, Mol. Genet. Metab. 82 (2004) 130–136. [17] S. Segal, P. Cuatrecasas, Oxidation of 14C galactose by patients with congenital galactosemia: Evidence for a direct oxidation pathway, Am. J. Med. 44 (1968) 340. [18] P. Cuatrecasas, S. Segal, Galactose conversion to D-xylulose: An alternate route of galactose metabolism, Science 153 (1966) 549– 551. [19] G.T. Berry, N. Leslie, R. Reynolds, C.T. Yager, S. Segal, Evidence for alternate galactose oxidation in a patient with deletion of the galactose-1-phosphate uridyltransferase gene, Mol. Genet. Metab. 72 (2001) 316–321. [20] G.T. Berry, I. Nissim, Z. Lin, A.T. Mazur, J.B. Gibson, S. Segal, Endogenous synthesis of galactose in normal men and patients with hereditary galactosaemia, Lancet 346 (1995) 1073– 1074. [21] S. Segal, Komrower Lecture. Galactosaemia today: The enigma and the challenge, J. Inherit. Metab. Dis. 21 (1998) 455–471. [22] R. Gitzelmann, Galactose-1-phosphate in the pathophysiology of galactosemia, Eur. J. Pediatr. 154 (1995) S45–49. [23] I.T. Oliver, Inhibitor studies on uridine diphosphoglucose pyrophosphorylase, Biochim. Biophys. Acta 52 (1961) 75–81. [24] M.L. Pourci, M. Mangeot, T. Soni, A. Lemonnier, Culture of galactosaemic fibroblasts in the presence of galactose: Effect of inosine, J. Inherit. Metab. Dis. 13 (1990) 819–828. [25] K.S. Ornstein, E.J. McGuire, G.T. Berry, S. Roth, S. Segal, Abnormal galactosylation of complex carbohydrates in cultured fibroblasts from patients with galactose-1-phosphate uridyltransferase deficiency, Pediatr. Res. 31 (1992) 508–511. [26] J.B. Gibson, R.A. Reynolds, M.J. Palmieri, et al., Comparison of erythrocyte uridine sugar nucleotide levels in normals, classic galactosemics, and patients with other metabolic disorders, Metabolism 44 (1995) 597–604. [27] J.B. Gibson, R.A. Reynolds, M.J. Palmieri, B. States, G.T. Berry, S. Segal, Uridine diphosphate hexoses in leukocytes and fibroblasts of classic galactosemics and patients with other metabolic diseases, Pediatr. Res. 36 (1994) 613–618. [28] J. Rutter, B.L. Probst, S.L. McKnight, Coordinate regulation of sugar flux and translation by PAS kinase, Cell 111 (2002) 17–28. [29] J.K. Chambers, L.E. Macdonald, H.M. Sarau, et al., A G protein-coupled receptor for UDP-glucose, J. Biol. Chem. 275 (2000) 10767–10771.
UDP-galactose pyrophosphorylase in mice with galactose-1-phosphate uridyltransferase deficiency Nancy Leslie a
a,*
, Claire Yager b, Robert Reynolds b, Stanton Segal
b
Division of Human Genetics, Cincinnati ChildrenÕs Hospital Research Foundation, Cincinnati, OH 45229, USA b Metabolic Research Laboratory, The ChildrenÕs Hospital of Philadelphia, Department of Pediatrics, University of Pennsylvania School of Medicine, Philadelphia, PA 19104, USA Received 11 November 2004; received in revised form 12 January 2005; accepted 12 January 2005 Available online 23 February 2005
Abstract UDP-glucose pyrophosphorylase (E.C. 2.7.7.9), encoded by ugp, provides UDP-glucose which is critical to the synthesis of glycogen, and also catalyzes the reaction between UTP and galactose-1-phosphate, yielding UDP-galactose. This activity of UDP-gal pyrophosphorylase (UDP-galPP) suggests a role in an alternate pathway for galactose metabolism in patients with deficiency of galactose-1-phosphate uridyltransferase (GALT). We examined the effects of GALT deficiency and dietary galactose on UDP-glucose pyrophosphorylase (UDP-gluPP) and UDP-galactose pyrophosphorylase activity and ugp expression in liver of mice with homozygous deletion of the critical regions of galt. Activity with glucose-1-phosphate as substrate was significantly higher than that with galactose-1-phosphate. In liver from mice with GALT deficiency (G/G), UDP-galPP activity appeared to be lower than that measured in liver from control (N/N) animals. This difference disappeared when the N/N tissue homogenate was dialyzed to remove residual UDP-glucose, confirming that careful elimination of residual GALT activity is necessary, since GALT has 1000-fold greater activity toward galactose-1-phosphate than that of UDP-galPP in liver homogenates. Prior exposure to conventional mouse chow, high galactose chow, and high glucose chow did not alter UDP-glu PP or UDP-galPP activity. Steady state UGP mRNA levels were determined in tissues from normal and G/G animals. UGP expression was highest in liver, and did not differ by genotype or exposure to high galactose chow. UDP-galPP activity may account for unexplained ability to oxidize galactose in animals with no GALT activity, but is insufficient to alter accumulation of galactose metabolites. 2005 Elsevier Inc. All rights reserved. Keywords: Galactose-1-phosphate uridyltransferase; UDP-glucose; UDP-galactose; Galactose
Introduction UDP-glucose pyrophosphorylase (E.C. 2.7.7.9), encoded by ugp, occupies a pivotal position in the synthesis of glycogen by forming UDP-glucose from glucose-1phosphate and UTP (Fig. 1) This enzyme also catalyzes the reaction between UTP and galactose-1-phosphate, yielding UDP-galactose [1,2]. The existence of this catalytic activity was convincingly demonstrated by
*
Corresponding author. Fax: +1 513 636 2261. E-mail address: [email protected] (N. Leslie).
1096-7192/$ - see front matter 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.ymgme.2005.01.004
Isselbacher in 1957 [3]. Although unaware of this proteinÕs dual role in both galactose and glucose metabolism, his studies in rat liver clearly showed activity toward galactose-1-phosphate that was separate from galactose-1-phosphate uridyltransferase (EC 2.7.7.10) (GALT), establishing evidence for a possible alternative pathway for galactose metabolism in patients with GALT-deficient galactosemia [3]. UDP-galactose pyrophosphorylase was further characterized in normal human liver by Abraham and Howell [4]. Their studies did not directly measure UDP-galPP activity in an individual with galactosemia, but did show that the contribution of UDP-galPP to total nmoles [14C]UDP-
22
N. Leslie et al. / Molecular Genetics and Metabolism 85 (2005) 21–27
UDP galPP activity in liver could account for the observed ability of GALT-deficient mice to oxidize galactose to expired CO2.
Materials and methods Animals
Fig. 1. UDP-galactose and UDP-glucose metabolism through GALT and UGP catalyzed pathways. The dark gray curved arrows indicate the active pathway of GALT. The straight arrows in the center indicate the major pathway of UDP-gluPP as well as the minor UDP-galPP activity. The diagonal arrows show the relationships between the substrates and products of both pathways.
galactose formed was much less than 1% of that attributed to GALT. This low contribution of UDP-galPP to UDP-gal synthesis was confirmed in normal human liver by Shin et al [5]. The functional significance of UDP-glucose pyrophosphorylase (UTP:a-D-glucose-1-phosphate uridyltransferase) has been demonstrated in bacteria as well as yeast and cultured CHO cells with a point mutation in ugp. Bacteria with absent UDP-gluPP have little UDP-glu content [6]. Yeast with deficient UDP-gluPP activity have low glycogen stores and impaired growth [7]. CHO cells with deficient UDP-gluPP activity have decreased steady state UDP-glucose levels, and functionally show differential sensitivity to the phospholipase C of Clostridium perfringens [8]. The importance of UDP-galPP in yeast deficient in GALT activity is clear when these strains are grown on galactose as a sole carbon source. Selection for revertants capable of growing in galactose medium revealed strains with up-regulated expression of UGP 1 [9]. Similarly, cultured human GALT-deficient fibroblasts transfected with a plasmid expressing ugp showed improved growth in galactose containing medium and higher steady state levels of both UDP-glucose and UDP-galactose than sham transfected cells [10]. However, there are no known mammalian models for deficiency of UGP, and only limited investigation of UGP activity in humans with GALT deficiency. Indeed, the only data on human liver from patients with GALT deficiency relate to its UDP-gluPP activity [11] and there are no data on the role of UDP-galPP to circumvent the block in galactose metabolism in human galactosemia. In this report, we have examined the effects of GALT deficiency and dietary galactose on UDP-gluPP and UDP-galPP activity and ugp expression in liver of mice with homozygous deletion of the critical regions of galt. These experiments directly assess the role of UDP-galPP in a mammalian model of GALT deficiency, and seek to answer whether
The 30–50 g mice used in these studies correspond to those employed in our previous reports [12–14]. The GALT-deficient mice (G/G), bred to a Balb/C background, were from well-established colonies in Philadelphia and Cincinnati. Their genotype was determined by polymerase chain reaction of DNA extracted from tail snips at 21 days of age. The normal (N/N) Balb/c mice were derived from the Charles River Laboratories colony (Wilmington, MA). Mice were weaned at 21 days of age to either chow 5015 or a 40% galactose or 40% glucose diet in which the carbohydrate was replaced by the hexose. All diets were purchased from PMI Feed (Richmond, IN). These diets were analyzed previously: chow 5015 contains small amounts of bioavailable galactose and the 40% glucose diet is galactose free [14]. Enzyme analysis Animals were killed by cervical dislocation, and the livers were removed and homogenized or flash-frozen with metal tongs cooled in liquid nitrogen. Frozen specimens were placed in plastic bags and stored at 80 C until analyzed. A piece of liver, fresh or frozen, was weighed and homogenized in four volumes of 0.1 M KCl with a motor-driven homogenizer with a teflon tip in a glass Potter-Elvehjem tissue grinding tube. The homogenate was then centrifuged for 10 min at 750g at 0 C. The supernatant from the low speed spin was then centrifuged further at 20,000g at 4 C for 30 min. The supernatant below the lipid layer was removed and used for the enzyme assays or subjected to dialysis. The dialysis procedure was carried out in a Slide-a-lyzer (Pierce Biotechnology, Rockford, IL) cassette with a ratio of 1–900 ml dialyzing solution of Tris–HCl, pH 8, 5 mmol/L, MgCl2 10 mmol/L, and EDTA 0.1 mmol/L with three exchanges of 900 ml each over a 24-h period. Protein concentrations were determined using the BioRad assay (Bio-Rad Laboratories, CA). Each assay contained 20 lmol of 1 M Tris buffer (pH 8.1), 5 lmol of MgCl2, 2 lmol dithiothreitol (ClelandÕs reagent), 6.46 nmol of unlabelled glucose-1-phosphate or galactose-1-phosphate, 1.7 lmol UTP, and 0.05 lCi 14 C-a-D-galactose-1-phosphate or 14C-a-D-glucose-1phosphate. Approximately 1 mg. of liver homogenate protein was added for UDP-galPP determination. Since the UDP-gluPP was many-fold more active with glu-1-P as substrate, the homogenate was diluted so that 9 lg
N. Leslie et al. / Molecular Genetics and Metabolism 85 (2005) 21–27
was added to give a linear rate for a 5-min assay. The total volume of each assay is reached by adding H2O to a level of 430 ll. The tubes were preincubated for 3 min at 37 C. The reaction was initiated with the addition of the radioactive substrate. Reactions were run for the appropriate time period at 37 C. The enzyme reactions were stopped by immersing the tubes into a 100 C water bath for 2 min. The tubes were cooled in ice for 5 min and then centrifuged refrigerated for 10 min at 1000g. Background tubes containing enzyme homogenates were first pre-incubated at 100 C for 3 min to inactivate any catalytic activity, then handled as described above. After centrifugation, a 20 ll sample from each assay tube was spotted in duplicate on DEAE paper strips measuring 1.5 cm in width and 28 cm in length. The sample was spotted at the origin, 7 cm from one end. After drying, the papers were developed by descending chromatography in closed tanks using 0.05 M LiCl as a solvent. The papers were removed from the tank just before the solvent front reached the bottom of the papers and then were hung to dry. The papers were then cut into appropriate sections and counted in scintillation fluid. The first four centimeters contained the origin and any unmigrated substances. Radioactivity in the next section contained the UDP-gal or UDP-glu. The length of this section varied slightly and was determined each time by cutting a strip into 22 1 cm sections and counting them to determine the optimal length of this portion. The next section contained the galactose1-phosphate or glucose-1-phosphate. The remaining 3 cm contained the counts for galactose or glucose. Enzyme activity was calculated as the nmol of product formed per mg protein. The calculations are performed by determining the ratio of radioactivity in the product/total radioactivity on the chromatographic strip. The ratio of radioactivity in the same regions of the background incubation was then subtracted from that of the experimental assays. The resulting value is multiplied by the initial concentration of the substrate and then divided by the protein content in the assay tube, as depicted in the equation below. nmol=min=mg ¼ f½ðcpm product=total cpm per stripÞ ðblank cpm product=total cpm per stripÞ ðnmol=L substrateÞg=mg protein
23
Ambion, Austin, TX) and quantitated by UV absorption (Beckman DU 640). Total RNA was separated on a 1% agarose/formaldehyde gel and transferred to a Hybond-N+ membrane (Amersham–Pharmacia, Piscataway, NJ). A 359 bp mUGP cDNA fragment was synthesized by rtPCR from mouse placental RNA, using primers selected from sequences common to the human UGP (NM 001001521), bovine UGP (NM 174212), and mouse EST sequences available in GenBank (forward: GCCAGAGGCTTGCCTGATAACATATC, reverse: CCCTGAATAAGACACATCCTTTGC) and the identity of the resulting product was confirmed by automated sequencing. A template appropriate for T7 transcription was synthesized using an antisense primer (TAATACGACTCACTATAGGAGGGTACCTGCT TTGATTG) incorporating a T7 tag coupled with antisense sequences 50 bp 5 0 of the original reverse amplification primer. Hybridization antisense riboprobes for mUGP or mouse TriActin (Ambion) were transcribed with T7 polymerase and [32P]UTP (800 Ci/mmol) using the Maxiscript kit (Ambion).
Reagents Uniformly labeled a-D-[14C]galactose-1-phosphate diammonium salt was prepared from D-galactose 14 C(U). It was purchased from American Radiolabeled Chemicals (St. Louis, MO) with an initial specific activity of 55 mCi/mmol. a-D-[U14C]Glucose-1-phosphate, potassium salt were purchased from Amersham Life Science (Arlington Heights, Illinois), 284 mCi/mmol. Unlabeled galactose-1-phosphate or glucose-1-phosphate, Tris, magnesium chloride, and dithiothreitol were purchased from Sigma (St. Louis, MO). Lithium chloride was obtained from Baker Chemical Company (Phillipsburg, NJ). DEAE–cellulose ion exchange paper was a product of Whatman (Clifton, NJ). Oligonucleotides were purchased from Gibco/BRL (Gaithersburg, MD).
Statistics The significance of differences was determined from the mean and standard error of the mean by StudentÕs t test.
Northern analysis of mouse UGP
Results
Adult g/g mice were exposed to 1 week of chow feeding or 1 week of 40% galactose feeding, then sacrificed by cervical dislocation after induction of CO2 narcosis. Tissues were quickly removed and frozen in liquid nitrogen. Total RNA was extracted by guanidinium extraction of pulverized tissue powder (Totally RNA,
UDP-gluPP activity The protein product of ugp catalyzes formation of both UDP-glucose and UDP-galactose. To determine the effect of GALT genotype on UDP glucose formation, UDP-glucose produced from Glu-1-P was
24
N. Leslie et al. / Molecular Genetics and Metabolism 85 (2005) 21–27
Fig. 2. Effect of time on glucose-1-P pyrophosphorylase activity in normal mouse liver (–) and in GALT-deficient mouse liver (j–j). Each data point is the average of two determinations.
measured in N/N and G/G liver (Fig. 2). There was no difference in the UPDgluPP activity between the two genotypes. The reaction was rapid and equilibrium was reached after 10 min when 7.5 lg of homogenate protein was employed. Because of the high activity, the original supernatant was diluted until the reaction was linear for the 5-min incubation as shown in the figure. UDP-galPP activity The activity of UDP-galPP for N/N and G/G liver is shown in Fig. 3. When fresh undialyzed supernatants were compared, there was a marked difference in apparent UDP-galPP activity between the two genotypes. In Fig. 3A the activity curve for the N/N mice is linear for 10 min and reaches equilibrium at 30 min incubation of 3.73 nmol/mg protein. In contrast, the activity curve for undialyzed G/G liver (Fig. 3B) was slow and linear over the same time period with an activity of 0.1 nmol/mg protein at 40 min. At 5 min, the apparent activity in the N/N homogenate was over 40 times that of the G/G. This discrepancy has two possible explanations. Accumulated gal-1-P in the liver of the G/G mice could inhibit UDP-galPP activity. An alternative hypothesis can be derived from the careful work of Isselbacher [3] who demonstrated the importance of dialysis for in order to measure UDP-galPP activity in
normal rat liver. In the rat liver, as in N/N mouse liver, GALT activity is present, and residual UDP-glu in the preparation will react with labeled Gal-1-P to produce labeled UDP-gal, the product measured in UDP-galPP reaction (see Fig. 1). As shown in Fig. 3A, dialysis of the N/N liver homogenate, which would decrease the UDP-glu present, markedly decreases the apparent UDP-galPP levels. The activity at 10 min is reduced from 1.97 to 0.05 nmol/mg protein and the curve (Fig. 3A) resembles that of undialyzed G/G supernate (Fig. 3B). However, dialysis does not alter the observed UDP-galPP activity in G/G liver as shown in Fig. 3B. These experiments demonstrate that dialysis decreases apparent UDP-galPP activity in GALT intact animals and does not increase apparent UDP-galPP activity in GALT-deficient animals. We therefore conclude that the difference in UDP-galPP activity between the two genotypes is not due to inhibition of catalytic activity in the G/G liver by accumulated inhibitors, such as gal-1-P, but rather the overestimation of UDP-gallPP activity in N/N liver due to contaminating GALT activity. Indeed, the activity of UDP-galPP in the G/G mouse homogenate is the true activity of the enzyme with Gal-1-P as substrate. The ratio of UDP-gluPP to UDP-galPP and ratio of UDP-galPP to GALT A comparison of the rate of UDP-glu formation in N/N and G/G liver is shown in Fig. 2. UDP-glu formation after a 5 min reaction for N/N and G/G liver, as well as for dialyzed N/N and G/G homogenates, is shown in Table 1. The UDP-gluPP activity is about 4000 times that of UDP-galPP. The activity of GALT in N/N mice is 10 nmol/min/ mg protein [12]. The UDP-galPP in mice is 0.0037 nmol/min/mg protein using the 5-min value of dialyzed G/G mouse value of Table 1. The GALT activity is 2700 times that of UDP-galPP confirming that in normal liver formation of UDP-gal via GALT and the Leloir pathway is the predominant route for metabolism of galactose.
Fig. 3. (A) Effect of time on galactose-1-P pyrophosphorylase activity in normal mouse liver (–) and in dialyzed normal mouse liver (j–j). Each data point is the average of 2–5 determinations. (B) Effect of time on galactose-1-P pyrophosphorylase activity in GALT-deficient mouse liver (}–}) and in dialyzed GALT-deficient mouse liver (h–h).
N. Leslie et al. / Molecular Genetics and Metabolism 85 (2005) 21–27 Table 1 Comparison of UDP-gluPP to UDP-galPP Homogenate
UDP-gluPP (nmol/5 min)
UDP-galPP (nmol/5 min)
N/N N/N G/G G/G
126 113 113 132
0.88 0.75 0.023 0.037
a b
undialyzed dialyzed undialyzed dialyzed
(2) (2) (2) (2)
Ra
(4) ± 0.09b (3) ± 0.005b (4) ± 0.006b (3) ± 0.004b
4913 3568
UDP-gluPP/UDP-galPP. Mean (number of determinations) ± SEM.
25
for a small but statistically significant increase in 30-min chow values. This finding was not reproduced with increasing galactose content of the diet. A comparison of the 30-min measurements of dialyzed N/N preparations with that of the G/G animals, dialyzed and undialyzed, shows a consistent increase in apparent UDPgalPP activity in the N/N livers. It is possible that a small presence of GALT activity remains despite extensive dialysis to rid the homogenate of UDP-glu.
Effect of diet on UDP-galPP
Effect of GALT genotype on UGP expression
Previous studies of the ability of G/G mice to oxidize galactose to CO2 showed that the percent of the dose oxidized in 4 h increased from a chow-fed groupÕs 5.5– 12.9%, when the animals were fed either a 40% galactose or glucose diet for 4 weeks [13]. Normal animals which oxidized 39% of the administered dose in 4 h showed no such increase. To determine if the feeding of galactose or glucose diets enhanced galactose metabolism by augmenting UDP-galPP, a study was performed to measure the liver enzyme activity after mice had been weaned to mouse chow, 40% galactose, or 40% glucose diets for 4 weeks. The results are shown in Table 2 for both N/N and G/G liver homogenates before and after dialysis. There was no difference in the UDP-galPP activities measured in livers from animals fed chow, galactose, or glucose, if comparisons were made between homogenates from the same genotype or dialysis condition. The data suggest that there is no induction or enhancement of UDP-galPP activity in animals fed a high galactoseor a high glucose-containing diet. As expected, there was a marked difference in apparent UDP-galPP activity between undialyzed and dialyzed preparations of N/N mouse homogenates across all diets. These data support those shown in graphic form in Fig. 3 which shows the time vs UDP-galPP activity. In G/G homogenates, undialyzed or dialyzed preparations did not differ except
Steady state levels of UGP message were unchanged when mice of G/G genotype were exposed to high galactose chow (Fig. 4). Although this manipulation increases gal-l-levels considerably in these tissues, no increase in steady state transcript levels was found. Of these tissues (brain, liver, kidney, and ovary), the highest expression of UGP was detected in liver. Similar results were found for N/N tissues (data not shown).
Fig. 4. Northern analysis of tissues from adult G/G mice consuming conventional mouse chow or 40% galactose chow. The upper panel is hybridized to a mouse UGP riboprobe. The lower panel is hybridized to mouse tri-actin. Tissues include brain (B), liver (L), kidney (K), and ovary (O).
Table 2 Effect of diet on UDP-galactose pyrophosphorylase activity in liver of normal and GALT-deficient mouse Incubation time/min
DIET (nmoles/mg protein) Chow Undialyzed
Galactose Dialyzed
Glucose
Undialyzed
Dialyzed
Undialyzed
Dialyzed
N/N mouse 5 30
0.88 (4)a ± 0.09b 3.36 (7) ± 0.24
0.075 (3) ± 0.005 0.16 (4)c ± 0.009
0.87 (9) ± 0.08 3.26 (12) ± 0.14
0.053 (9) ± 0.006 0.14 (8) ± 0.013
0.68 (7) ±0.04 3.39 (9) ± 0.35
0.07 (4) ± 0.04 0.14 (8)c ± 0.01
G/G mouse 5 30
0.023 (4) ± 0.006 0.073 (5) ± 0.004
0.043 (3) ± 0.016 0.118 (4)d ± 0.009
0.033 (3) ± 0.004 0.07 (3) ± 0.004
0.037 (3) ± 0.008 0.103 (3) ± 0.023
0.025 (2) 0.11 (3) ± 0.004
0.043 (3) ± 0.02 0.11 (3) ± 0
Number of specimens are indicated in the parentheses. a Mean. b SEM. c Difference from the 30 min G/G value p < 0.02. d Difference from the 30 min G/G undialyzed value, p < 0.01.
26
N. Leslie et al. / Molecular Genetics and Metabolism 85 (2005) 21–27
Discussion The present study is the first to report UDP-galPP activity in liver of an animal with complete absence of GALT activity. We found the same low activity in GALTless liver as in normal tissue. The exposure of the G/G animals to a high galactose diet did not alter catalytic activity or steady state levels of ugp message in tissues from GALT-deficient mice. Our results emphasize the importance of complete elimination of the GALT activity in normal tissue if an accurate estimate of the UDP-galPP contribution is to be made. Based on our findings in dialyzed wild type liver homogenates, the estimated UDP-galPP activity responsible for galactose conversion to UDP-gal is 0.1% that of GALT. In our previous studies, we observed that GALT-deficient adult mice could oxidize 4–12% of an administered dose of 7 mg galactose per kg over 4 h to CO2 [14]. Does UDP-galPP activity account for this level of oxidation? In a 50 g mouse, this is approximately 0.02–0.05 lmol galactose oxidized per hour. With UDP-galPP activity measured at 0.08 lmol/mg protein in 40 min in a mouse liver weighing 1.2–1.5 g and with 200 mg protein per gm wet weight, the UDP-galPP activity could account for metabolism of 0.03 lmol galactose per hour. Therefore, within the limits of measurement, UDP-galPP may account for a large portion of the limited galactose oxidation in mice without GALT activity [13,14]. The contribution of liver UDP-galPP activity to the ability of galactosemic patients to oxidize isotopic galactose to CO2 is unknown [15,16]. Humans with the classical form of GALT deficiency can oxidize up to 50% of administered [13C]galactose to CO2 in 24 h [16]. Segal and Cuatrecasas [17,18] on the basis of differences in the oxidation of 1- and 2-14C-labeled galactose suggested that half the oxidation could be due to direct oxidation via the galactonate pathway with the remainder being due to residual GALT or another pathway such as UDP-galPP. Convincing data that a pathway other than GALT is responsible comes from the study of a patient with a homozygous deletion of exons 1–10 of GALT which revealed considerable oxidation (17)% of a [13C]galactose dose in a 24-h period [19]. Our data in mice with a deletion of three exons of galt, together with direct evidence of UGP expression in humans, suggests that UDP-galPP activity may explain at least some of the measured [13C]galactose oxidation. Although UDP-galPP action may be sufficient to explain much of the slow phase of galactose oxidation in GALT-deficient mice,its expression in liver clearly does not prevent accumulation of galactose metabolites in these mice [12–14]. There is sufficient data in human tissues to demonstrate existence of similar activity in liver, muscle [5] but this alternate pathway does not prevent
accumulation of galactose metabolites in tissues of patients with GALT deficiency, nor does it alter endogenous production or long term morbidity in these patients [20,21]. The forward reaction UTP + Gal-1-P fi UDPgal + PP is the prominent direction of the reversible reaction [3]. However, Gitzelmann [22] showed evidence for the pyrophosphorylitic cleavage of UDP-gal reacting with PP to form UTP and Gal-1-P. Based on this, he proposed that the reaction was responsible for the continuous endogenous production of Gal-1-P, thereby resulting in self-intoxication of the patient and an explanation for the long-term, diet-independent complications of the disorder. On the other hand, Oliver [23] presented evidence that Gal-1-P inhibited UDP-gluPP in liver and brain and postulated that a resulting deficiency of UDP-glu would have consequences in the formation of glucuronides, glycogen and complex carbohydrates. In contrast, Pourci et al. [24] reported that galactosemic fibroblasts grow normally when incubated with galactose if inosine is present even when Gal1-P levels are very high. They concluded that a lack of energy prevented cell growth in the presence of galactose alone and that the ribose moiety of inosine provided the fuel for growth. The combined presence of UGP and epimerase explains why cultured galactosemic fibroblasts grown in glucose media have normal UDP-gal and UDP-glu concentrations [25] and that there is no deficit of nucleotide sugars in red blood cells [26] or leucocytes [27] of galactosemic patients. The inhibition of multiple enzymes by high cellular Gal-1-P levels is well established [23] but the findings of in vitro models with cells exposed to galactose as a sole substrate may not parallel those of cells in vivo which never experience exposure to galactose alone. Suckling GALT-deficient mice and adult mice fed 40% galactose accumulate 2–3 mM hepatic Gal-1-P and do not have a galactosemic phenotype [13]. In chow-fed G/G mice with hepatic Gal-1-P of 1.75 mM, there is no abnormality in the liver levels of UDP-glu or UDP-gal [14]. The lack of a galactose toxicity syndrome in GALT-deficient mice in the face of a markedly elevated Gal-1-P demonstrates that in mice clinical morbidity does not have a simple relationship to expression of UGP or altered activity of its protein product. If a small amount of UDP-galPP activity allows oxidation of galactose to CO2, would overexpression of UGP by some means rescue the human phenotype in GALT deficiency? Such possibilities are suggested by the in vitro rescue studies of Lai et al. [9,10]. However, in yeast systems, overexpression of UGP causes a growth restricted phenotype and UGP activity is regulated by PAS kinase-mediated phosphorylation [28]. Both UDP-glucose and UDP-galactose bind to the P2Y receptor and may have roles as extracellular signal-
N. Leslie et al. / Molecular Genetics and Metabolism 85 (2005) 21–27
ing intermediates [29]. In complex mammalian systems, the potential for overexpression of UGP to alter pathways other than those involved in galactose metabolism deserves careful consideration.
Acknowledgments The authors gratefully acknowledge the assistance of Janice Malseed, Terri Wallace, and Shuzhen Bai. This work was supported by R01-DK HD 60768-01A1 to S. Segal.
References [1] J.K. Knop, R.G. Hansen, Uridine diphosphate glucose pyrophosphorylase. IV. Crystallization and properties of the enzyme from human liver, J. Biol. Chem. 245 (1970) 2499–2504. [2] R.L. Turnquist, T.A. Gillett, R.G. Hansen, Uridine diphosphate glucose pyrophosphorylase. Crystallization and properties of the enzyme from rabbit liver and species comparisons, J. Biol. Chem. 249 (1974) 7695–7700. [3] K.J. Isselbacher, A mammalian uridinediphosphate galactose pyrophosphorylase, J. Biol. Chem. 232 (1958) 429–444. [4] H.D. Abraham, R.R. Howell, Human hepatic uridine diphosphate galactose pyrophosphorylase. Its characterization and activity during development, J. Biol. Chem. 244 (1969) 545–550. [5] Y.S. Shin, H.P. Niedermeier, W. Endres, J. Schaub, S. Weidinger, Agarose gel isoelectrofocusing of UDP-galactose pyrophosphorylase and galactose-1-phosphate uridyltransferase. Developmental aspect of UDP-galactose pyrophosphorylase, Clin. Chim. Acta 166 (1987) 27–35. [6] T.A. Sundararajan, A.M. Rapin, H.M. Kalckar, Biochemical observations on E. coli mutants defective in uridine diphosphoglucose, Proc. Natl. Acad. Sci. USA 48 (1962) 2187–2193. [7] J.M. Daran, W. Bell, J. Francois, Physiological and morphological effects of genetic alterations leading to a reduced synthesis of UDP-glucose in Saccharomyces cerevisiae, FEMS Microbiol. Lett. 153 (1997) 89–96. [8] M. Flores-Diaz, A. Alape-Giron, R.W. Titball, et al., UDPglucose deficiency causes hypersensitivity to the cytotoxic effect of Clostridium perfringens phospholipase C, J. Biol. Chem. 273 (1998) 24433–24438. [9] K. Lai, L.J. Elsas, Overexpression of human UDP-glucose pyrophosphorylase rescues galactose-1-phosphate uridyltransferase-deficient yeast, Biochem. Biophys. Res. Commun. 271 (2000) 392–400. [10] K. Lai, S.D. Langley, F.W. Khwaja, E.W. Schmitt, L.J. Elsas, GALT deficiency causes UDP-hexose deficit in human galactosemic cells, Glycobiology 13 (2003) 285–294. [11] E.P. Anderson, K.J. Isselbacher, H.M. Kalckar, Defect in uptake of galactose-1-phosphate into liver nucleotides in congenital galactosemia, Science 125 (1957) 113–114.
27
[12] N.D. Leslie, K.L. Yager, P.D. McNamara, S. Segal, A mouse model of galactose-1-phosphate uridyl transferase deficiency, Biochem. Mol. Med. 59 (1996) 7–12. [13] C. Ning, R. Reynolds, J. Chen, et al., Galactose metabolism in mice with galactose-1-phosphate uridyltransferase deficiency: Sucklings and 7-week-old animals fed a high-galactose diet, Mol. Genet. Metab. 72 (2001) 306–315. [14] C. Ning, R. Reynolds, J. Chen, et al., Galactose metabolism by the mouse with galactose-1-phosphate uridyltransferase deficiency, Pediatr. Res. 48 (2000) 211–217. [15] G.T. Berry, R.H. Singh, A.T. Mazur, et al., Galactose breath testing distinguishes variant and severe galactose-1-phosphate uridyltransferase genotypes, Pediatr. Res. 48 (2000) 323–328. [16] G.T. Berry, R.A. Reynolds, C.T. Yager, S. Segal, Extended [13C]galactose oxidation studies in patients with galactosemia, Mol. Genet. Metab. 82 (2004) 130–136. [17] S. Segal, P. Cuatrecasas, Oxidation of 14C galactose by patients with congenital galactosemia: Evidence for a direct oxidation pathway, Am. J. Med. 44 (1968) 340. [18] P. Cuatrecasas, S. Segal, Galactose conversion to D-xylulose: An alternate route of galactose metabolism, Science 153 (1966) 549– 551. [19] G.T. Berry, N. Leslie, R. Reynolds, C.T. Yager, S. Segal, Evidence for alternate galactose oxidation in a patient with deletion of the galactose-1-phosphate uridyltransferase gene, Mol. Genet. Metab. 72 (2001) 316–321. [20] G.T. Berry, I. Nissim, Z. Lin, A.T. Mazur, J.B. Gibson, S. Segal, Endogenous synthesis of galactose in normal men and patients with hereditary galactosaemia, Lancet 346 (1995) 1073– 1074. [21] S. Segal, Komrower Lecture. Galactosaemia today: The enigma and the challenge, J. Inherit. Metab. Dis. 21 (1998) 455–471. [22] R. Gitzelmann, Galactose-1-phosphate in the pathophysiology of galactosemia, Eur. J. Pediatr. 154 (1995) S45–49. [23] I.T. Oliver, Inhibitor studies on uridine diphosphoglucose pyrophosphorylase, Biochim. Biophys. Acta 52 (1961) 75–81. [24] M.L. Pourci, M. Mangeot, T. Soni, A. Lemonnier, Culture of galactosaemic fibroblasts in the presence of galactose: Effect of inosine, J. Inherit. Metab. Dis. 13 (1990) 819–828. [25] K.S. Ornstein, E.J. McGuire, G.T. Berry, S. Roth, S. Segal, Abnormal galactosylation of complex carbohydrates in cultured fibroblasts from patients with galactose-1-phosphate uridyltransferase deficiency, Pediatr. Res. 31 (1992) 508–511. [26] J.B. Gibson, R.A. Reynolds, M.J. Palmieri, et al., Comparison of erythrocyte uridine sugar nucleotide levels in normals, classic galactosemics, and patients with other metabolic disorders, Metabolism 44 (1995) 597–604. [27] J.B. Gibson, R.A. Reynolds, M.J. Palmieri, B. States, G.T. Berry, S. Segal, Uridine diphosphate hexoses in leukocytes and fibroblasts of classic galactosemics and patients with other metabolic diseases, Pediatr. Res. 36 (1994) 613–618. [28] J. Rutter, B.L. Probst, S.L. McKnight, Coordinate regulation of sugar flux and translation by PAS kinase, Cell 111 (2002) 17–28. [29] J.K. Chambers, L.E. Macdonald, H.M. Sarau, et al., A G protein-coupled receptor for UDP-glucose, J. Biol. Chem. 275 (2000) 10767–10771.