Molecular Genetics and Metabolism 111 (2014) 467–476
Contents lists available at ScienceDirect
Molecular Genetics and Metabolism journal homepage: www.elsevier.com/locate/ymgme
Mouse model of glycogen storage disease type III Kai-Ming Liu a,b, Jer-Yuarn Wu a, Yuan-Tsong Chen a,c,⁎ a b c
Institute of Biomedical Sciences, Academia Sinica, 128 Academia Road, Section 2, Nankang, Taipei 115, Taiwan Institute of Clinical Medicine, National Yang-Ming University, 155, Sec.2, Linong Street, Taipei 112, Taiwan Department of Pediatrics, Duke University Medical Center, Box 3528, Durham, NC 27710, USA
a r t i c l e
i n f o
Article history: Received 4 January 2014 Received in revised form 3 February 2014 Accepted 3 February 2014 Available online 18 February 2014 Keyword: GSD III Glycogen storage disease type III GDE Glycogen debranching enzyme Mouse model
a b s t r a c t Glycogen storage disease type IIIa (GSD IIIa) is caused by a deficiency of the glycogen debranching enzyme (GDE), which is encoded by the Agl gene. GDE deficiency leads to the pathogenic accumulation of phosphorylase limit dextrin (PLD), an abnormal glycogen, in the liver, heart, and skeletal muscle. To further investigate the pathological mechanisms behind this disease and develop novel therapies to treat this disease, we generated a GDE-deficient mouse model by removing exons after exon 5 in the Agl gene. GDE reduction was confirmed by western blot and enzymatic activity assay. Histology revealed massive glycogen accumulation in the liver, muscle, and heart of the homozygous affected mice. Interestingly, we did not find any differences in the general appearance, growth rate, and life span between the wild-type, heterozygous, and homozygous affected mice with ad libitum feeding, except reduced motor activity after 50 weeks of age, and muscle weakness in both the forelimb and hind legs of homozygous affected mice by using the grip strength test at 62 weeks of age. However, repeated fasting resulted in decreased survival of the knockout mice. Hepatomegaly and progressive liver fibrosis were also found in the homozygous affected mice. Blood chemistry revealed that alanine transaminase (ALT), aspartate transaminase (AST) and alkaline phosphatase (ALP) activities were significantly higher in the homozygous affected mice than in both wild-type and heterozygous mice and the activity of these enzymes further increased with fasting. Creatine phosphokinase (CPK) activity was normal in young and adult homozygous affected mice. However, the activity was significantly elevated after fasting. Hypoglycemia appeared only at a young age (3 weeks) and hyperlipidemia was not observed in our model. In conclusion, with the exception of normal lipidemia, these mice recapitulate human GSD IIIa; moreover, we found that repeated fasting was detrimental to these mice. This mouse model will be useful for future investigation regarding the pathophysiology and treatment strategy of human GSD III. © 2014 Elsevier Inc. All rights resered.
1. Introduction Glycogen storage disease type III (GSD III, OMIM) is a rare autosomal recessive metabolic disorder with an estimated incidence of 1:100,000 live births. GSD III, also known as Cori's or Forbes disease for the biochemical and clinical descriptions of the disease [1,2], is caused by a deficiency of the glycogen debranching enzyme (GDE), which is encoded by the Agl gene. GDE has two independent catalytic activities, oligo1,4-1,4-glucantransferase (EC 2.4.1.25) and amylo-1,6-glucosidase (EC 3.2.1.33), both of which are responsible for removing glycogen branches during cytoplasmic glycogenolysis. When GDE loses either one or both of its enzyme activities, glycogen with short outer chains, called phosphorylase limit dextrin (PLD), accumulates in tissues. Thus, this disease is also known as “limit dextrinosis”. Most GSD III patients (~85%) exhibit
⁎ Corresponding author at: Institute of Biomedical Sciences, Academia Sinica, 128 Academia Road, Section 2, Nankang, Taipei 11529, Taiwan. Fax: +886 27899085. E-mail address: [email protected] (Y.-T. Chen).
http://dx.doi.org/10.1016/j.ymgme.2014.02.005 1096-7192/© 2014 Elsevier Inc. All rights resered.
liver and muscle symptoms and are classified as type IIIa. The few patients (~ 15%) who present only liver symptoms are classified as type IIIb. In rare cases, loss of GDE glucosidase or transferase activity is found and is classified as type IIIc or type IIId, respectively. Hepatomegaly, hypoglycemia, hyperlipidemia, and short stature are common disease features of GSD III in affected infants and children. Liver symptoms usually improve with age, especially after puberty; however, some GSD III patients will continue to develop progressive liver fibrosis, cirrhosis, which ultimately results in liver failure [3–5]. In addition, some GSD III patients have reported hepatic adenoma and carcinoma [6–10]. In GSD IIIa, muscle weakness is usually minimal in young patients, but can develop with age after the third or fourth decade of life [11–14]. The onset of heart symptoms is variable in GSD IIIa; asymptomatic or symptomatic cardiomyopathy can be observed [15], and some patients exhibit severe heart involvement including heart failure and arrhythmias that can lead to death [16,17]. In agreement with the organs affected, enhanced alanine transaminase (ALT), aspartate transaminase (AST), alkaline phosphatase (ALP), and/or creatine phosphokinase (CPK) activities are frequently present in the serum
468
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
Fig. 1. Generation of GDE knockout mice. (A) Targeting strategy to disrupt the Agl gene. The mouse engrailed 2-gene splice acceptor (En2 SA) and poly A tail (pA) cassette is inserted between exons 5 and 6. After splicing, the mRNA does not contain exons downstream of exon 5. Arrowheads (F, forward; R1, reverse1; and R2, reverse2) indicate the primers used for genotyping. (B) Genotyping of the GDE knockout mice. Two PCR fragments, 522 bp and 393 bp in length, are amplified from genomic DNA of wild-type (+/+) and homozygous (−/−) mice, respectively. Both fragments are observed with the heterozygous (+/−) genomic DNA. M, 100-bp DNA ladder.
of GSD III patients. Additional symptoms of GSD III include osteoporosis and polycystic ovaries [18,19] which have not been shown to reduce fertility [20,21]. To treat and maintain normoglycemia, GSD III patients need to be given carbohydrate supplement through frequent meals, cornstarch intake, and nocturnal gastric drip-feeding [15]. A high protein diet is also beneficial for GSD III patients and, for some patients, has been shown to improve delayed growth, myopathy, and cardiomyopathy [22–24]. However, dietary management does not prevent many long-term complications of GSD III and some patients had to undergo liver transplantation (LT) due to liver cirrhosis and/or dysfunction [5,7,25,26]. In addition, LT does not prevent the progressive myopathy and cardiomyopathy characteristic of GSD III [25]. Currently, there is no specific or
adequate therapy for GSD III. In this study, we reported the first GDE deficiency mouse model that recapitulates human-like GSD IIIa features and pathological progression. 2. Materials and methods 2.1. Generation of GSD III mice Embryonic stem cells (ES cell) with a knockout of the Agl gene were purchased from the European Conditional Mouse Mutagenesis Program (EUCOMM, Germany). ES injections and the generation of chimeric mice were performed by Transgenic Mouse Models Core (TMMC, Taiwan). After receiving the chimera mice in our animal facility,
Fig. 2. Immunoblots and enzymatic activity of GDE. GDE immunoblotting (A) and an enzymatic activity assay (B) were performed in the liver, heart, and muscle of the wild-type (+/+), heterozygous (+/−), and homozygous (−/−) mice at 4 weeks of age. GDE protein expression was determined using an anti-Agl antibody and actin was used as an internal loading control. GDE enzyme activity was measured in 5 mice for each group.
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
heterozygous and homozygous affected mice were generated by mating the chimera and wild-type mice, and intercrossing the heterozygous mice. All animal experiments complied with the animal protocol approved by the Institutional Animal Care and Use Committees (IACUC, Academia Sinica, Taiwan). 2.2. Tissue homogenization and protein quantification Mouse tissues and an appropriate amount of distilled water (20 μl to 100 μl ddH2O for 1 mg of tissue) were added to a Green-Beads® tube and homogenized with the MagNa lyser (Roche). The homogenized samples were transferred to new Eppendorf tubes and sonicated on ice using a Sonifier 150 (Branson) for 1 s on and 2 s off for 6 cycles. The samples were then centrifuged at 16,000 ×g for 10 min at 4 °C to separate the supernatant fraction that contained the cellular protein from the cell debris. The protein sample concentration was determined using the BCA Protein Assay Kit (Pierce Biotechnology). 2.3. Western blot analysis A western blot protocol employing standard methods was utilized in this study. A rabbit polyclonal anti-AGL (Abcam) primary antibody was
469
used at a 1/2000 dilution to detect AGL and a mouse anti-Actin antibody (Millipore) primary antibody was used at a 1/10,000 dilution to detect Actin. An HRP-conjugated goat anti-mouse secondary antibody (Millipore) was applied at a 1/5000 dilution to detect AGL and actin primary antibodies.
2.4. Quantification of glycogen and assessment of GDE enzymatic activity Glycogen concentration in the organs of mice was quantified using the Glycogen Assay Kit (BioVision). The supernatant of the protein samples were diluted by the appropriate fold amounts (5–100×) to fit within the measurement range of glycogen provided by the Glycogen Assay Kit. All steps of the assay followed the commercial protocol. For the GDE enzyme activity assay, β-limit dextrin (Megazyme) was used as a substrate to quantify the combined enzymatic activities of glucantransferase and α-1,6-glucosidase of GDE [27]. Each protein sample at a concentration of 1 to 2 mg/ml was divided into two equal parts. β-Limit dextrin (5%) was dissolved in distilled water and an equal volume of distilled water was utilized as a control and analyzed. We employed a different sample for each respective time-point studied, which included 0, 30, 60, 90, and 200 min of incubation at 30 °C. At each respective time-point, the reaction was stopped by heating for 5 min at 100 °C. After chilling on ice,
Fig. 3. General appearance and growth. The appearance of the mouse (A, top) and femur (A, bottom) of wild-type (+/+), heterozygous (+/−), and homozygous (−/−) mice at 4 weeks of age. The body weight of wild type, heterozygous, and homozygous male and female mice from 3 to 30 weeks of age (B). Ratio of liver over the body weight of wild-type and homozygous mice at 4 weeks of age (C). The survival rate of wild type, heterozygous, and homozygous affected mice with 24-h fasting every 4 weeks from 4 weeks of age until 24 weeks and then at 40 weeks of age (D). Means ± s.d. are shown.
470
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
the samples were centrifuged at 13,000 rpm for 10 min to remove any precipitate due to heating. The amount of glucose in the supernatant of each sample at each time point was quantified using the D-GLUCOSE– HK kit (Megazyme). The linear regression slope and trend line were generated from the amount of glucose in each of the five samples and represented GDE enzyme activity. 2.5. Genotyping and PCR Genomic DNA from mice tails was extracted using the Gentra Puregene kit (Qiagen). Three primers, 5′-CTT CTG TCC ATA CTG CAT CAG ATA GG-3′ (forward), 5′-AAC AAA CCC TTG GGT GAA GAT GGG3′ (reverse 1) and 5′-CAA CGG GTT CTT CTG TTA GTC C-3′ (reverse 2), were added at the same time to the genomic DNA in a PCR reaction. The primers were designed from EUCOMM.
aminotransferase (ALT), creatinine kinase (CPK), and alkaline phosphatase (ALP) were measured using a Fuji Dri-Chem Clinical Chemistry Analyzer FDC 3500. 2.8. Assessment of muscle strength Muscle strength was measured using a grip strength meter (MK-380, Muromachi). The forelimb and hind legs of male and female mice were assessed. Mice were trained 3 times prior to the formal grip strength test. Each measurement in the mouse was performed in triplicate. 2.9. Statistical analyses Student's t-test was used for all analysis comparing each experimental group in this study. A P-value of b0.05 was considered statistically significant.
2.6. Histology Organs and tissues were fixed in 4% paraformaldehyde solution (pH 7.4) and embedded in paraffin wax. The sections were stained with hematoxylin–eosin (H&E), Periodic acid-Schiff (PAS, for tissue glycogen analysis), Masson's and Sirius red (both for liver fibrosis analysis) stain. All procedures were performed following standard protocols. 2.7. Measurement of serum biochemistry Serum was separated from mice tail blood by centrifugation at 3,000 ×g for 10 min. The serum concentrations of glucose (GLU), triglycerides (TG), total-cholesterol (TCHO), and high-density lipoprotein (HDL), and the activities of aspartate aminotransferase (AST), alanine
3. Results 3.1. Generation of a GDE deficiency mouse model The knockout mouse cassette contained the mouse engrailed 2 gene splice acceptor (En2 SA) and poly(A) tail inserted between exon5 and exon6 of the Agl gene. After splicing, the exons downstream of Agl exon5 are not present in the mRNA thus resulting in a C-terminal deletion of GDE that contains 1310 amino acids (Fig. 1A). The deleted region of GDE contains the putative transferase and glucosidase catalytic domains [28]. The genotype of each mouse was confirmed by PCR prior to each experiment. Two PCR products, 522 bp and 393 bp in length, were amplified from the WT allele using primer F and R1 and from
Fig. 4. Liver histology of GSD III knockout mice. Liver sections of a wild-type (A), a heterozygous (B), and a homozygous affected (C) mouse at 4 weeks of age (H&E staining). Sirius red (D–F), Masson's (G–I) staining (both stain for tissue collagen) of the liver of affected mice at different ages of 4 weeks (D, G), 18 weeks (E, H), and 28 weeks (F, I). CV, central vein. Scale bar = 100 μm (A–C) and 200 μm (D–I).
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
the mutant allele using primer F and R2, respectively (Fig. 1B). Western blot revealed the GDE protein in the liver, muscle, and heart of the wildtype and heterozygous mice, but not the knockout mice (Fig. 2A). As expected, GDE enzyme activity was significantly reduced in the mutant organs (liver, 0.33 ± 0.09; heart, 0.35 ± 0.07; and muscle, 0.29 ± 0.13 nmole glucose/min/mg protein, n = 5 for each) as compared to wild-type organs (liver, 2.90 ± 0.18; heart, 1.59 ± 0.10; and muscle, 2.49 ± 0.61 nmole glucose/min/mg protein, n = 5 for each) or
471
heterozygous mice (liver, 2.62 ± 0.45; heart, 1.50 ± 0.29; and muscle, 2.09 ± 0.12 nmole glucose/min/mg protein, n = 5 for each) (Fig. 2B). 3.2. General appearance and growth The homozygous affected mice exhibited a normal appearance and femur length (Fig. 3A, at 4 weeks of age), and life span. However, with repeated fasting, 50% of mice died before 40 weeks of age (Fig. 3D).
Fig. 5. Tissue glycogen in GSD III knockout mice. Liver (A–C), diaphragm (D–F), gastrocnemius (G–I), quadriceps (J–L), and heart (M–O) sections of wild-type (A, D, G, J, M), heterozygous (B, E, H, K, N), and homozygous affected (C, F, I, L, O) mice at 4 weeks of age with 24-h fasting. All sections were stained with PAS staining, with dark purple color representing glycogen. CV, central vein. RA, Right atrium. LA, Left atrium. RV, Right ventricle. LV, Left ventricle. A, Aorta. Scale bar = 100 μm (A–L), 1 mm (M–O).
472
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
The body weight in both males and females did not significantly differ among the wild-type, heterozygous, and homozygous affected mice from 4 to 30 weeks of age (Fig. 3B). Hepatomegaly was observed in 4week-old homozygous affected mice as compared to age-matched wild-type mice (Fig. 3C).
3.3. Histology and levels of glycogen in tissues
Fig. 6. Quantification of tissue glycogen in GSD III knockout mice. Glycogen contents of the liver, diaphragm, gastrocnemius, quadriceps, and heart from wild-type, heterozygous, and homozygous affected mice at 4 weeks of age with 24-h fasting were quantified (n = 5 for each group). The weight percentages of glycogen contents were all significantly higher in the 5 types of affected tissues compare to the same type tissue of wild-type and heterozygous (all of P values b 0.0005). Means ± s.d. are shown.
Enlarged hepatocytes with pale cytoplasm were observed in the liver sections of homozygous affected mice at 4 weeks of age (Fig. 4C). Progressive liver fibrosis was observed after 4 weeks of age with mild fibrosis, periportal fibrosis presenting at 18 weeks of age (Figs. 4E,H), and an expanded area of fibrosis by 28 weeks of age (Figs. 4F,I). Hepatic adenoma and carcinoma were not observed in the homozygous affected mice up to 40 weeks of age. PAS staining revealed severe glycogen accumulation in the affected liver, heart, diaphragm, quadriceps, and gastrocnemius sections. In contrast, glycogen accumulation was rarely observed in tissue from heterozygous and wild-type mice after fasting (Fig. 5). We further quantified the amount of glycogen content (n = 5 in each group). The percentage of glycogen contents by weight in affected liver (13.72 ± 1.19%), heart (1.25 ± 0.53%), gastrocnemius (1.1 ± 0.22%), quadriceps (1.48 ± 0.47%), and diaphragm (2.48 ± 0.21%) were all significantly higher than tissues from wild-type and heterozygous mice, which were all less than 0.05% (Fig. 6).
Fig. 7. Ad-lib feeding serum biochemistries in GSD III knockout mice. Ad-lib feeding of serum AST, ALT, ALP, and CPK were measured from wild-type (+/+), and homozygous affected (−/−) mice at 4, 16, and 30 weeks of age. Number in the bracket indicates the number of mice examined. Means ± s.d. are shown.
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
3.4. Serum biochemistry When mice were fed ad-lib, serum AST, ALT, and ALP were significantly higher (1.4–6.2 fold) in the homozygous (AST 123–283 U/l, ALT 69–286 U/l, and ALP 449–1222 U/l) than in the wild-type mice (AST 51–53 U/l, ALT 34–46 U/l, and ALP 259–846 U/l) at all ages measured (4, 16, and 30 weeks); however, there was no significant difference in CPK values (wild-type, 109–228 U/l; homozygous, 118–227 U/l) (Fig. 7). Notably, liver transaminases (AST and ALT) further increased with age in the homozygous affected mice. When blood chemistries were obtained after fasting in 7-week-old homozygous affected mice, activities of AST, ALT, and ALP were further significantly elevated (8, 5.7, and 1.3 folds) in the affected mice (Fig. 8A; AST, 1181 ± 377 U/l; ALT, 679 ± 29 U/l; ALP, 1207 ± 146 U/l) as compared to mice fed ad libitum at 6 weeks of age (AST, 147 ± 21 U/l; ALT, 120 ± 14 U/l; ALP, 895 ± 134 U/l). Interestingly, CPK was also elevated (18.2 folds) with fasting (CPK, 2290 ± 1098 U/l) as compared to ad lib feeding (CPK, 126 ± 17 U/l). These enzymatic activities were rapidly attenuated when re-fed at 8 weeks of age (AST, 164 ± 22 U/l; ALT, 132 ± 23 U/l; ALP, 801 ± 29 U/l; CPK, 271 ± 119 U/l). Examination of blood glucose levels revealed that the weaning homozygous affected mice (3-week-old) exhibited hypoglycemia (Fig. 8B) with blood glucose significantly reduced in the homozygous affected mice (61 ± 17 mg/dl) as compared to the wild-type (102 ± 18 mg/dl) after 12 h of fasting. However, there was no significant difference in blood glucose between the same batch of homozygous affected (115 ± 32 mg/dl) and wild-type (111 ± 24 mg/dl) mice at 8 weeks of age after 12 h of fasting. Even with 24 h fasting, blood glucose was reduced
473
but was not down to the hypoglycemic range (Fig. 8A). Both triglycerides and total cholesterol were normal in homozygous affected mice at 3 and 8 weeks of age after 12 h of fasting (Fig. 8B). Next we investigated the long-term effects of repeated fasting on blood biochemistry in these mice. Mice were subjected to 24 h of fasting every 4 weeks from 4 weeks of age until 24 weeks of age, and then at 40 weeks of age, they were fasted again (Fig. 9). We performed longitudinal studies using the same batch of wild-type (n = 7), heterozygous (n = 7), and homozygous affected (n = 8) mice. Four homozygous affected mice died at 17, 28, 34, and 36 weeks of age, respectively. Thus, the numbers of homozygous affected mice reduces with aging. We observed that, at different time points of fasting, AST (483–1914 U/l), ALT (191–785 U/l), and ALP (645–1547 U/l) activities were significantly higher (1.4–13.8 folds) in the homozygous affected mice than the agematched wild-type (AST, 84–212 U/l; ALT, 43–64 U/l; ALP, 297–1103 U/l) or heterozygous (AST, 64–157 U/l; ALT, 31–57 U/l; ALP, 354–850 U/l) mice. Notably, both AST and ALT continued to increase until 40 weeks of age. A similar pattern was observed for ALP and CPK except higher activities were already observed in younger mice. 3.5. Muscle weakness Signs of muscle weakness was first observed in 1-year-old homozygous and manifested as reduced motor activity. Muscle strength of wildtype (3 females and 3 males), heterozygous (5 females and 4 males), and homozygous affected (3 females and 5 males) mice were measured at 62 weeks of age using the grip test (Fig. 10). The muscle strength of the forelimb and hind legs were significantly lower in homozygous
Fig. 8. Fasting serum biochemistries in GSD III knockout mice. A. Fasting effect on serum biochemistry in GSD III knockout homozygous affected mice. Serum GLU, AST, ALT, ALP, and CPK levels of 4 littermate affected mice were measured at 6 and 8 weeks of age with ad-lib feeding and 7 weeks of age with 24 h of fasting. B. Fasting blood glucose and lipid levels in GSD III knockout mice. Serum GLU, TG, and TCHO levels were measured from the same batch of wild-type (+/+) and homozygous affected (−/−) mice after 12 h of fasting at 3 and 8 weeks of age. Means ± s.d. are shown.
474
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
Fig. 9. Serum biochemistry in GSD III knockout mice subjected to repeated fasting. Twenty-four hour fasting serum AST, ALT, ALP, and CPK levels were measured from the wild-type (+/+, n = 7), heterozygous (+/−, n = 7), and homozygous affected (−/−, n = 8) mice at 4, 8, 12, 16, 20, 24, and 40 weeks of age. The initial number of affected mice in each group is 8. Due to early death in some affected mice, the number was reduced to 7 by 20 weeks and to 4 by 40 weeks of age. P-value of each group comparison is shown below each figure. Means ± s.d. are shown.
affected male and female mice than in gender-matched wild-type and heterozygous mice. 4. Discussion GSD III is an autosomal recessive metabolic disorder of glycogenolysis that primarily affects the liver, heart, and skeletal muscle. Until now, only a canine model of GSD III has been reported [29]. Affected canines carried a frame-shift mutation in the Agl gene that results in the deletion of a C-terminal 126-amino acid-long sequence in the GDE protein. Phenotypes of affected canines include glycogen accumulation in the liver, muscle, and adipocytes; elevated fasting serum ALT, AST, and CPK; progressive liver fibrosis, cirrhosis, and disruption of the contractile apparatus; and fraying of myofibrils. However, the growth curves, fasting serum glucose, triglyceride, and cholesterol levels of affected canines are normal. A mouse model is an advantageous tool to study disease due to their shorter generation time, small size, and low maintenance cost. Our mouse model exhibits pathologic accumulation of glycogen in the liver, heart, and muscles with hepatomegaly and progressive liver fibrosis and muscle weakness. However, homozygous affected mice with ad-lib feeding did not exhibit symptomatic cardiomyopathy including left ventricular hypertrophy up to one and half years of age (data not shown). Serum ALT, AST, ALP, and CPK were significantly greater in the homozygous affected mice than in the wildtype and heterozygous mice. Our mouse model recapitulated features that mimic human GSD IIIa with the exception of normal lipidemia and a milder form of hypoglycemia. The hypoglycemia in our mouse model only occurred in very young mice (3 weeks of age) after 12 h of fasting. Notably, this was not observed in older mice (from 4 weeks
up to 40 weeks of age) after 12 h or 24 h of fasting. In human GSD, hypoglycemia causes an increase in the β-oxidation of fats; this is a compensatory mechanism and resulted in increased triglycerides and total cholesterol production [30]. Interestingly, the GSD canine model does not exhibit hypoglycemia, and similarly, we did not observe hypoglycemia in our mouse model. This discrepancy between the animal models and the human disorder may reflect differences in endogenous glucose production or gluconeogenesis between species. Cardiomyopathy is variable in human GSD III. We performed echocardiography analysis in our mice when they were fed ad-lib. Homozygous affected mice (n = 6) exhibited the normal parameters including ejection fraction and left ventricular mass as compared to heterozygous (n = 8) and wild-type mice (n = 5) up to one and half years of age (data not shown). The cardiomyopathy might be more likely to develop in the mice after repeating fasting. However, we did not have enough number of mice to do echo after repeated fasting because of the decreased survival. In our mouse model, liver transaminases (AST and ALT) were elevated (2.3 and 1.8 folds) when mice (4 weeks old) were fed ad lib, these enzyme activities further increased (6.4–9 folds and 4.3–12.2 folds) with fasting, aging, and repeated fasting (the transaminases of repeated fasting reached to thousands at 40 weeks of age) indicating progressive liver damage. During fasting, blood glucose levels decreased, but continued to remain within the normal range (Fig. 8A). The repeated fasting also resulted in early death in these mice. Thus, in our study, fasting detrimentally affects our model. These findings re-emphasize the importance of close monitoring of the dietary regimen in human GSD III patients. Specifically, patients should avoid fasting and aggressive intravenous glucose therapy when oral intake is compromised. It also
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
475
Fig. 10. Muscle strength of aged GSD III knockout mice. Forelimb and hind leg muscle strength for male and female wild-type (+/+), heterozygous (+/−), and homozygous (−/−) mice at 62 weeks of age were measured by grip strength test. Means ± s.d. are shown.
suggested that maintenance of “normoglycemia” may not be sufficient to avoid liver damage that is presumably due to the accumulation of phosphorylase limit dextrin, an abnormal glycogen with short outer braches. In conclusion, we have generated a knock-out mice model with a GDE deficiency that recapitulates the key characteristics of human GSD type III. This mouse model will be useful for future investigation of the pathophysiology and treatment strategy of human GSD III. Acknowledgments This study was supported by the Academia Sinica Genomic Medicine Multicenter Study (40-05-GMM), the National Research Program for Genomic Medicine, National Science Council, Taiwan (National Center for Genome Medicine, NSC101-2319-B-001-001). We would like to thank the EUCOMM Consortium for the production of the ES cell used to generate the Agl gene knockout mice, the technical services provided by the “Transgenic Mouse Model Core Facility of the National Core Facility Program for Biotechnology, National Science Council”, the “Gene Knockout Mouse Core Laboratory of National Taiwan University Center of Genomic Medicine”, the Sequencing Core Facility at the Scientific Instrument Center at Academia Sinica for DNA sequencing, and the Taiwan Mouse Clinic funded by the National Research Program for Biopharmaceuticals, National Science Council, Taiwan, for the technical support. References [1] B. Illingworth, G.T. Cori, Structure of glycogens and amylopectins. III. Normal and abnormal human glycogen, J. Biol. Chem. 199 (1952) 653–660. [2] G.B. Forbes, Glycogen storage disease; report of a case with abnormal glycogen structure in liver and skeletal muscle, J. Pediatr. 42 (1953) 645–653.
[3] I.W. Fellows, J.S. Lowe, A.L. Ogilvie, A. Stevens, P.J. Toghill, M. Atkinson, Type III glycogenosis presenting as liver disease in adults with atypical histological features, J. Clin. Pathol. 36 (1983) 431–434. [4] A.J. Markowitz, Y.T. Chen, J. Muenzer, E.A. Delbuono, M.R. Lucey, A man with type III glycogenosis associated with cirrhosis and portal hypertension, Gastroenterology 105 (1993) 1882–1885. [5] E.B. Haagsma, G.P. Smit, K.E. Niezen-Koning, A.S. Gouw, L. Meerman, M.J. Slooff, Type IIIb glycogen storage disease associated with end-stage cirrhosis and hepatocellular carcinoma. The Liver Transplant Group, Hepatology 25 (1997) 537–540. [6] P. Labrune, P. Trioche, I. Duvaltier, P. Chevalier, M. Odievre, Hepatocellular adenomas in glycogen storage disease type I and III: a series of 43 patients and review of the literature, J. Pediatr. Gastroenterol. Nutr. 24 (1997) 276–279. [7] D. Matern, T.E. Starzl, W. Arnaout, J. Barnard, J.S. Bynon, A. Dhawan, J. Emond, E.B. Haagsma, G. Hug, A. Lachaux, G.P. Smit, Y.T. Chen, Liver transplantation for glycogen storage disease types I, III, and IV, Eur. J. Pediatr. 158 (Suppl. 2) (1999) S43–S48. [8] M. Siciliano, E. De Candia, S. Ballarin, F.M. Vecchio, S. Servidei, R. Annese, R. Landolfi, L. Rossi, Hepatocellular carcinoma complicating liver cirrhosis in type IIIa glycogen storage disease, J. Clin. Gastroenterol. 31 (2000) 80–82. [9] A. Cosme, I. Montalvo, J. Sanchez, E. Ojeda, J. Torrado, E. Zapata, L. Bujanda, A. Gutierrez, I. Arenas, Type III glycogen storage disease associated with hepatocellular carcinoma, Gastroenterol. Hepatol. 28 (2005) 622–625. [10] E. Demo, D. Frush, M. Gottfried, J. Koepke, A. Boney, D. Bali, Y.T. Chen, P.S. Kishnani, Glycogen storage disease type III-hepatocellular carcinoma a long-term complication? J. Hepatol. 46 (2007) 492–498. [11] R.A. Coleman, H.S. Winter, B. Wolf, Y.T. Chen, Glycogen debranching enzyme deficiency: long-term study of serum enzyme activities and clinical features, J. Inherit. Metab. Dis. 15 (1992) 869–881. [12] S. Lucchiari, D. Santoro, S. Pagliarani, G.P. Comi, Clinical, biochemical and genetic features of glycogen debranching enzyme deficiency, Acta Myol. 26 (2007) 72–74. [13] P.S. Kishnani, Y. Suzuki, Y.-T. Chen, Chapter 71: Glycogen Storage Diseases, McGraw Hill Professional, New York, 2009. [14] S. DiMauro, G.B. Hartwig, A. Hays, A.B. Eastwood, R. Franco, M. Olarte, M. Chang, A.D. Roses, M. Fetell, R.S. Schoenfeldt, L.Z. Stern, Debrancher deficiency: neuromuscular disorder in 5 adults, Ann. Neurol. 5 (1979) 422–436. [15] P.S. Kishnani, S.L. Austin, P. Arn, D.S. Bali, A. Boney, L.E. Case, W.K. Chung, D.M. Desai, A. El-Gharbawy, R. Haller, G.P. Smit, A.D. Smith, L.D. Hobson-Webb, S.B. Wechsler, D.A. Weinstein, M.S. Watson, ACMG, glycogen storage disease type III diagnosis and management guidelines, Genet. Med. 12 (2010) 446–463. [16] S.W. Moses, K.L. Wanderman, A. Myroz, M. Frydman, Cardiac involvement in glycogen storage disease type III, Eur. J. Pediatr. 148 (1989) 764–766.
476
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
[17] P.J. Lee, J.E. Deanfield, M. Burch, K. Baig, W.J. McKenna, J.V. Leonard, Comparison of the functional significance of left ventricular hypertrophy in hypertrophic cardiomyopathy and glycogenosis type III, Am. J. Cardiol. 79 (1997) 834–838. [18] H.R. Mundy, J.E. Williams, P.J. Lee, M.S. Fewtrell, Reduction in bone mineral density in glycogenosis type III may be due to a mixed muscle and bone deficit, J. Inherit. Metab. Dis. 31 (2008) 418–423. [19] P.J. Lee, A. Patel, P.C. Hindmarsh, A.P. Mowat, J.V. Leonard, The prevalence of polycystic ovaries in the hepatic glycogen storage diseases: its association with hyperinsulinism, Clin. Endocrinol. 42 (1995) 601–606. [20] S. Bhatti, E. Parry, Successful pregnancy in a woman with glycogen storage disease type III, Aust. N. Z. J. Obstet. Gynaecol. 46 (2006) 168–169. [21] A. Mendoza, N.C. Fisher, J. Duckett, J. McKiernan, M.A. Preece, A. Green, P.J. McKiernan, G. Constantine, E. Elias, Successful pregnancy in a patient with type III glycogen storage disease managed with cornstarch supplements, Br. J. Obstet. Gynaecol. 105 (1998) 677–680. [22] A.E. Slonim, R.A. Coleman, W.S. Moses, Myopathy and growth failure in debrancher enzyme deficiency: improvement with high-protein nocturnal enteral therapy, J. Pediatr. 105 (1984) 906–911. [23] V. Valayannopoulos, F. Bajolle, J.B. Arnoux, S. Dubois, N. Sannier, C. Baussan, F. Petit, P. Labrune, D. Rabier, C. Ottolenghi, A. Vassault, C. Broissand, D. Bonnet, P. de Lonlay, Successful treatment of severe cardiomyopathy in glycogen storage disease type III With D, L-3-hydroxybutyrate, ketogenic and high-protein diet, Pediatr. Res. 70 (2011) 638–641.
[24] A.I. Dagli, R.T. Zori, H. McCune, T. Ivsic, M.K. Maisenbacher, D.A. Weinstein, Reversal of glycogen storage disease type IIIa-related cardiomyopathy with modification of diet, J. Inherit. Metab. Dis. 32 (Suppl. 1) (2009) S103–S106. [25] M.K. Davis, D.A. Weinstein, Liver transplantation in children with glycogen storage disease: controversies and evaluation of the risk/benefit of this procedure, Pediatr. Transplant. 12 (2008) 137–145. [26] P.P. Liu, V.H. de Villa, Y.S. Chen, C.C. Wang, S.H. Wang, Y.C. Chiang, B. Jawan, H.K. Cheung, Y.F. Cheng, T.L. Huang, H.L. Eng, F.R. Chuang, C.L. Chen, Outcome of living donor liver transplantation for glycogen storage disease, Transplant. Proc. 35 (2003) 366–368. [27] D.H. Brown, B.I. Brown, [88] Enzymes of glycogen debranching: amylo-1,6-glucosidase (I) and oligo-1,4 → 1,4-glucantransferase (II), in: V.G. Elizabeth, F. Neufeld (Eds.), Methods in Enzymology, Academic Press, 1966, pp. 515–524. [28] J.J. Shen, Y.T. Chen, Molecular characterization of glycogen storage disease type III, Curr. Mol. Med. 2 (2002) 167–175. [29] H. Yi, B.L. Thurberg, S. Curtis, S. Austin, J. Fyfe, D.D. Koeberl, P.S. Kishnani, B. Sun, Characterization of a canine model of glycogen storage disease type IIIa, Dis. Model. Mech. 5 (2012) 804–811. [30] A.V. Bernier, C.P. Sentner, C.E. Correia, D.W. Theriaque, J.J. Shuster, G.P. Smit, D.A. Weinstein, Hyperlipidemia in glycogen storage disease type III: effect of age and metabolic control, J. Inherit. Metab. Dis. 31 (2008) 729–732.
Contents lists available at ScienceDirect
Molecular Genetics and Metabolism journal homepage: www.elsevier.com/locate/ymgme
Mouse model of glycogen storage disease type III Kai-Ming Liu a,b, Jer-Yuarn Wu a, Yuan-Tsong Chen a,c,⁎ a b c
Institute of Biomedical Sciences, Academia Sinica, 128 Academia Road, Section 2, Nankang, Taipei 115, Taiwan Institute of Clinical Medicine, National Yang-Ming University, 155, Sec.2, Linong Street, Taipei 112, Taiwan Department of Pediatrics, Duke University Medical Center, Box 3528, Durham, NC 27710, USA
a r t i c l e
i n f o
Article history: Received 4 January 2014 Received in revised form 3 February 2014 Accepted 3 February 2014 Available online 18 February 2014 Keyword: GSD III Glycogen storage disease type III GDE Glycogen debranching enzyme Mouse model
a b s t r a c t Glycogen storage disease type IIIa (GSD IIIa) is caused by a deficiency of the glycogen debranching enzyme (GDE), which is encoded by the Agl gene. GDE deficiency leads to the pathogenic accumulation of phosphorylase limit dextrin (PLD), an abnormal glycogen, in the liver, heart, and skeletal muscle. To further investigate the pathological mechanisms behind this disease and develop novel therapies to treat this disease, we generated a GDE-deficient mouse model by removing exons after exon 5 in the Agl gene. GDE reduction was confirmed by western blot and enzymatic activity assay. Histology revealed massive glycogen accumulation in the liver, muscle, and heart of the homozygous affected mice. Interestingly, we did not find any differences in the general appearance, growth rate, and life span between the wild-type, heterozygous, and homozygous affected mice with ad libitum feeding, except reduced motor activity after 50 weeks of age, and muscle weakness in both the forelimb and hind legs of homozygous affected mice by using the grip strength test at 62 weeks of age. However, repeated fasting resulted in decreased survival of the knockout mice. Hepatomegaly and progressive liver fibrosis were also found in the homozygous affected mice. Blood chemistry revealed that alanine transaminase (ALT), aspartate transaminase (AST) and alkaline phosphatase (ALP) activities were significantly higher in the homozygous affected mice than in both wild-type and heterozygous mice and the activity of these enzymes further increased with fasting. Creatine phosphokinase (CPK) activity was normal in young and adult homozygous affected mice. However, the activity was significantly elevated after fasting. Hypoglycemia appeared only at a young age (3 weeks) and hyperlipidemia was not observed in our model. In conclusion, with the exception of normal lipidemia, these mice recapitulate human GSD IIIa; moreover, we found that repeated fasting was detrimental to these mice. This mouse model will be useful for future investigation regarding the pathophysiology and treatment strategy of human GSD III. © 2014 Elsevier Inc. All rights resered.
1. Introduction Glycogen storage disease type III (GSD III, OMIM) is a rare autosomal recessive metabolic disorder with an estimated incidence of 1:100,000 live births. GSD III, also known as Cori's or Forbes disease for the biochemical and clinical descriptions of the disease [1,2], is caused by a deficiency of the glycogen debranching enzyme (GDE), which is encoded by the Agl gene. GDE has two independent catalytic activities, oligo1,4-1,4-glucantransferase (EC 2.4.1.25) and amylo-1,6-glucosidase (EC 3.2.1.33), both of which are responsible for removing glycogen branches during cytoplasmic glycogenolysis. When GDE loses either one or both of its enzyme activities, glycogen with short outer chains, called phosphorylase limit dextrin (PLD), accumulates in tissues. Thus, this disease is also known as “limit dextrinosis”. Most GSD III patients (~85%) exhibit
⁎ Corresponding author at: Institute of Biomedical Sciences, Academia Sinica, 128 Academia Road, Section 2, Nankang, Taipei 11529, Taiwan. Fax: +886 27899085. E-mail address: [email protected] (Y.-T. Chen).
http://dx.doi.org/10.1016/j.ymgme.2014.02.005 1096-7192/© 2014 Elsevier Inc. All rights resered.
liver and muscle symptoms and are classified as type IIIa. The few patients (~ 15%) who present only liver symptoms are classified as type IIIb. In rare cases, loss of GDE glucosidase or transferase activity is found and is classified as type IIIc or type IIId, respectively. Hepatomegaly, hypoglycemia, hyperlipidemia, and short stature are common disease features of GSD III in affected infants and children. Liver symptoms usually improve with age, especially after puberty; however, some GSD III patients will continue to develop progressive liver fibrosis, cirrhosis, which ultimately results in liver failure [3–5]. In addition, some GSD III patients have reported hepatic adenoma and carcinoma [6–10]. In GSD IIIa, muscle weakness is usually minimal in young patients, but can develop with age after the third or fourth decade of life [11–14]. The onset of heart symptoms is variable in GSD IIIa; asymptomatic or symptomatic cardiomyopathy can be observed [15], and some patients exhibit severe heart involvement including heart failure and arrhythmias that can lead to death [16,17]. In agreement with the organs affected, enhanced alanine transaminase (ALT), aspartate transaminase (AST), alkaline phosphatase (ALP), and/or creatine phosphokinase (CPK) activities are frequently present in the serum
468
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
Fig. 1. Generation of GDE knockout mice. (A) Targeting strategy to disrupt the Agl gene. The mouse engrailed 2-gene splice acceptor (En2 SA) and poly A tail (pA) cassette is inserted between exons 5 and 6. After splicing, the mRNA does not contain exons downstream of exon 5. Arrowheads (F, forward; R1, reverse1; and R2, reverse2) indicate the primers used for genotyping. (B) Genotyping of the GDE knockout mice. Two PCR fragments, 522 bp and 393 bp in length, are amplified from genomic DNA of wild-type (+/+) and homozygous (−/−) mice, respectively. Both fragments are observed with the heterozygous (+/−) genomic DNA. M, 100-bp DNA ladder.
of GSD III patients. Additional symptoms of GSD III include osteoporosis and polycystic ovaries [18,19] which have not been shown to reduce fertility [20,21]. To treat and maintain normoglycemia, GSD III patients need to be given carbohydrate supplement through frequent meals, cornstarch intake, and nocturnal gastric drip-feeding [15]. A high protein diet is also beneficial for GSD III patients and, for some patients, has been shown to improve delayed growth, myopathy, and cardiomyopathy [22–24]. However, dietary management does not prevent many long-term complications of GSD III and some patients had to undergo liver transplantation (LT) due to liver cirrhosis and/or dysfunction [5,7,25,26]. In addition, LT does not prevent the progressive myopathy and cardiomyopathy characteristic of GSD III [25]. Currently, there is no specific or
adequate therapy for GSD III. In this study, we reported the first GDE deficiency mouse model that recapitulates human-like GSD IIIa features and pathological progression. 2. Materials and methods 2.1. Generation of GSD III mice Embryonic stem cells (ES cell) with a knockout of the Agl gene were purchased from the European Conditional Mouse Mutagenesis Program (EUCOMM, Germany). ES injections and the generation of chimeric mice were performed by Transgenic Mouse Models Core (TMMC, Taiwan). After receiving the chimera mice in our animal facility,
Fig. 2. Immunoblots and enzymatic activity of GDE. GDE immunoblotting (A) and an enzymatic activity assay (B) were performed in the liver, heart, and muscle of the wild-type (+/+), heterozygous (+/−), and homozygous (−/−) mice at 4 weeks of age. GDE protein expression was determined using an anti-Agl antibody and actin was used as an internal loading control. GDE enzyme activity was measured in 5 mice for each group.
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
heterozygous and homozygous affected mice were generated by mating the chimera and wild-type mice, and intercrossing the heterozygous mice. All animal experiments complied with the animal protocol approved by the Institutional Animal Care and Use Committees (IACUC, Academia Sinica, Taiwan). 2.2. Tissue homogenization and protein quantification Mouse tissues and an appropriate amount of distilled water (20 μl to 100 μl ddH2O for 1 mg of tissue) were added to a Green-Beads® tube and homogenized with the MagNa lyser (Roche). The homogenized samples were transferred to new Eppendorf tubes and sonicated on ice using a Sonifier 150 (Branson) for 1 s on and 2 s off for 6 cycles. The samples were then centrifuged at 16,000 ×g for 10 min at 4 °C to separate the supernatant fraction that contained the cellular protein from the cell debris. The protein sample concentration was determined using the BCA Protein Assay Kit (Pierce Biotechnology). 2.3. Western blot analysis A western blot protocol employing standard methods was utilized in this study. A rabbit polyclonal anti-AGL (Abcam) primary antibody was
469
used at a 1/2000 dilution to detect AGL and a mouse anti-Actin antibody (Millipore) primary antibody was used at a 1/10,000 dilution to detect Actin. An HRP-conjugated goat anti-mouse secondary antibody (Millipore) was applied at a 1/5000 dilution to detect AGL and actin primary antibodies.
2.4. Quantification of glycogen and assessment of GDE enzymatic activity Glycogen concentration in the organs of mice was quantified using the Glycogen Assay Kit (BioVision). The supernatant of the protein samples were diluted by the appropriate fold amounts (5–100×) to fit within the measurement range of glycogen provided by the Glycogen Assay Kit. All steps of the assay followed the commercial protocol. For the GDE enzyme activity assay, β-limit dextrin (Megazyme) was used as a substrate to quantify the combined enzymatic activities of glucantransferase and α-1,6-glucosidase of GDE [27]. Each protein sample at a concentration of 1 to 2 mg/ml was divided into two equal parts. β-Limit dextrin (5%) was dissolved in distilled water and an equal volume of distilled water was utilized as a control and analyzed. We employed a different sample for each respective time-point studied, which included 0, 30, 60, 90, and 200 min of incubation at 30 °C. At each respective time-point, the reaction was stopped by heating for 5 min at 100 °C. After chilling on ice,
Fig. 3. General appearance and growth. The appearance of the mouse (A, top) and femur (A, bottom) of wild-type (+/+), heterozygous (+/−), and homozygous (−/−) mice at 4 weeks of age. The body weight of wild type, heterozygous, and homozygous male and female mice from 3 to 30 weeks of age (B). Ratio of liver over the body weight of wild-type and homozygous mice at 4 weeks of age (C). The survival rate of wild type, heterozygous, and homozygous affected mice with 24-h fasting every 4 weeks from 4 weeks of age until 24 weeks and then at 40 weeks of age (D). Means ± s.d. are shown.
470
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
the samples were centrifuged at 13,000 rpm for 10 min to remove any precipitate due to heating. The amount of glucose in the supernatant of each sample at each time point was quantified using the D-GLUCOSE– HK kit (Megazyme). The linear regression slope and trend line were generated from the amount of glucose in each of the five samples and represented GDE enzyme activity. 2.5. Genotyping and PCR Genomic DNA from mice tails was extracted using the Gentra Puregene kit (Qiagen). Three primers, 5′-CTT CTG TCC ATA CTG CAT CAG ATA GG-3′ (forward), 5′-AAC AAA CCC TTG GGT GAA GAT GGG3′ (reverse 1) and 5′-CAA CGG GTT CTT CTG TTA GTC C-3′ (reverse 2), were added at the same time to the genomic DNA in a PCR reaction. The primers were designed from EUCOMM.
aminotransferase (ALT), creatinine kinase (CPK), and alkaline phosphatase (ALP) were measured using a Fuji Dri-Chem Clinical Chemistry Analyzer FDC 3500. 2.8. Assessment of muscle strength Muscle strength was measured using a grip strength meter (MK-380, Muromachi). The forelimb and hind legs of male and female mice were assessed. Mice were trained 3 times prior to the formal grip strength test. Each measurement in the mouse was performed in triplicate. 2.9. Statistical analyses Student's t-test was used for all analysis comparing each experimental group in this study. A P-value of b0.05 was considered statistically significant.
2.6. Histology Organs and tissues were fixed in 4% paraformaldehyde solution (pH 7.4) and embedded in paraffin wax. The sections were stained with hematoxylin–eosin (H&E), Periodic acid-Schiff (PAS, for tissue glycogen analysis), Masson's and Sirius red (both for liver fibrosis analysis) stain. All procedures were performed following standard protocols. 2.7. Measurement of serum biochemistry Serum was separated from mice tail blood by centrifugation at 3,000 ×g for 10 min. The serum concentrations of glucose (GLU), triglycerides (TG), total-cholesterol (TCHO), and high-density lipoprotein (HDL), and the activities of aspartate aminotransferase (AST), alanine
3. Results 3.1. Generation of a GDE deficiency mouse model The knockout mouse cassette contained the mouse engrailed 2 gene splice acceptor (En2 SA) and poly(A) tail inserted between exon5 and exon6 of the Agl gene. After splicing, the exons downstream of Agl exon5 are not present in the mRNA thus resulting in a C-terminal deletion of GDE that contains 1310 amino acids (Fig. 1A). The deleted region of GDE contains the putative transferase and glucosidase catalytic domains [28]. The genotype of each mouse was confirmed by PCR prior to each experiment. Two PCR products, 522 bp and 393 bp in length, were amplified from the WT allele using primer F and R1 and from
Fig. 4. Liver histology of GSD III knockout mice. Liver sections of a wild-type (A), a heterozygous (B), and a homozygous affected (C) mouse at 4 weeks of age (H&E staining). Sirius red (D–F), Masson's (G–I) staining (both stain for tissue collagen) of the liver of affected mice at different ages of 4 weeks (D, G), 18 weeks (E, H), and 28 weeks (F, I). CV, central vein. Scale bar = 100 μm (A–C) and 200 μm (D–I).
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
the mutant allele using primer F and R2, respectively (Fig. 1B). Western blot revealed the GDE protein in the liver, muscle, and heart of the wildtype and heterozygous mice, but not the knockout mice (Fig. 2A). As expected, GDE enzyme activity was significantly reduced in the mutant organs (liver, 0.33 ± 0.09; heart, 0.35 ± 0.07; and muscle, 0.29 ± 0.13 nmole glucose/min/mg protein, n = 5 for each) as compared to wild-type organs (liver, 2.90 ± 0.18; heart, 1.59 ± 0.10; and muscle, 2.49 ± 0.61 nmole glucose/min/mg protein, n = 5 for each) or
471
heterozygous mice (liver, 2.62 ± 0.45; heart, 1.50 ± 0.29; and muscle, 2.09 ± 0.12 nmole glucose/min/mg protein, n = 5 for each) (Fig. 2B). 3.2. General appearance and growth The homozygous affected mice exhibited a normal appearance and femur length (Fig. 3A, at 4 weeks of age), and life span. However, with repeated fasting, 50% of mice died before 40 weeks of age (Fig. 3D).
Fig. 5. Tissue glycogen in GSD III knockout mice. Liver (A–C), diaphragm (D–F), gastrocnemius (G–I), quadriceps (J–L), and heart (M–O) sections of wild-type (A, D, G, J, M), heterozygous (B, E, H, K, N), and homozygous affected (C, F, I, L, O) mice at 4 weeks of age with 24-h fasting. All sections were stained with PAS staining, with dark purple color representing glycogen. CV, central vein. RA, Right atrium. LA, Left atrium. RV, Right ventricle. LV, Left ventricle. A, Aorta. Scale bar = 100 μm (A–L), 1 mm (M–O).
472
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
The body weight in both males and females did not significantly differ among the wild-type, heterozygous, and homozygous affected mice from 4 to 30 weeks of age (Fig. 3B). Hepatomegaly was observed in 4week-old homozygous affected mice as compared to age-matched wild-type mice (Fig. 3C).
3.3. Histology and levels of glycogen in tissues
Fig. 6. Quantification of tissue glycogen in GSD III knockout mice. Glycogen contents of the liver, diaphragm, gastrocnemius, quadriceps, and heart from wild-type, heterozygous, and homozygous affected mice at 4 weeks of age with 24-h fasting were quantified (n = 5 for each group). The weight percentages of glycogen contents were all significantly higher in the 5 types of affected tissues compare to the same type tissue of wild-type and heterozygous (all of P values b 0.0005). Means ± s.d. are shown.
Enlarged hepatocytes with pale cytoplasm were observed in the liver sections of homozygous affected mice at 4 weeks of age (Fig. 4C). Progressive liver fibrosis was observed after 4 weeks of age with mild fibrosis, periportal fibrosis presenting at 18 weeks of age (Figs. 4E,H), and an expanded area of fibrosis by 28 weeks of age (Figs. 4F,I). Hepatic adenoma and carcinoma were not observed in the homozygous affected mice up to 40 weeks of age. PAS staining revealed severe glycogen accumulation in the affected liver, heart, diaphragm, quadriceps, and gastrocnemius sections. In contrast, glycogen accumulation was rarely observed in tissue from heterozygous and wild-type mice after fasting (Fig. 5). We further quantified the amount of glycogen content (n = 5 in each group). The percentage of glycogen contents by weight in affected liver (13.72 ± 1.19%), heart (1.25 ± 0.53%), gastrocnemius (1.1 ± 0.22%), quadriceps (1.48 ± 0.47%), and diaphragm (2.48 ± 0.21%) were all significantly higher than tissues from wild-type and heterozygous mice, which were all less than 0.05% (Fig. 6).
Fig. 7. Ad-lib feeding serum biochemistries in GSD III knockout mice. Ad-lib feeding of serum AST, ALT, ALP, and CPK were measured from wild-type (+/+), and homozygous affected (−/−) mice at 4, 16, and 30 weeks of age. Number in the bracket indicates the number of mice examined. Means ± s.d. are shown.
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
3.4. Serum biochemistry When mice were fed ad-lib, serum AST, ALT, and ALP were significantly higher (1.4–6.2 fold) in the homozygous (AST 123–283 U/l, ALT 69–286 U/l, and ALP 449–1222 U/l) than in the wild-type mice (AST 51–53 U/l, ALT 34–46 U/l, and ALP 259–846 U/l) at all ages measured (4, 16, and 30 weeks); however, there was no significant difference in CPK values (wild-type, 109–228 U/l; homozygous, 118–227 U/l) (Fig. 7). Notably, liver transaminases (AST and ALT) further increased with age in the homozygous affected mice. When blood chemistries were obtained after fasting in 7-week-old homozygous affected mice, activities of AST, ALT, and ALP were further significantly elevated (8, 5.7, and 1.3 folds) in the affected mice (Fig. 8A; AST, 1181 ± 377 U/l; ALT, 679 ± 29 U/l; ALP, 1207 ± 146 U/l) as compared to mice fed ad libitum at 6 weeks of age (AST, 147 ± 21 U/l; ALT, 120 ± 14 U/l; ALP, 895 ± 134 U/l). Interestingly, CPK was also elevated (18.2 folds) with fasting (CPK, 2290 ± 1098 U/l) as compared to ad lib feeding (CPK, 126 ± 17 U/l). These enzymatic activities were rapidly attenuated when re-fed at 8 weeks of age (AST, 164 ± 22 U/l; ALT, 132 ± 23 U/l; ALP, 801 ± 29 U/l; CPK, 271 ± 119 U/l). Examination of blood glucose levels revealed that the weaning homozygous affected mice (3-week-old) exhibited hypoglycemia (Fig. 8B) with blood glucose significantly reduced in the homozygous affected mice (61 ± 17 mg/dl) as compared to the wild-type (102 ± 18 mg/dl) after 12 h of fasting. However, there was no significant difference in blood glucose between the same batch of homozygous affected (115 ± 32 mg/dl) and wild-type (111 ± 24 mg/dl) mice at 8 weeks of age after 12 h of fasting. Even with 24 h fasting, blood glucose was reduced
473
but was not down to the hypoglycemic range (Fig. 8A). Both triglycerides and total cholesterol were normal in homozygous affected mice at 3 and 8 weeks of age after 12 h of fasting (Fig. 8B). Next we investigated the long-term effects of repeated fasting on blood biochemistry in these mice. Mice were subjected to 24 h of fasting every 4 weeks from 4 weeks of age until 24 weeks of age, and then at 40 weeks of age, they were fasted again (Fig. 9). We performed longitudinal studies using the same batch of wild-type (n = 7), heterozygous (n = 7), and homozygous affected (n = 8) mice. Four homozygous affected mice died at 17, 28, 34, and 36 weeks of age, respectively. Thus, the numbers of homozygous affected mice reduces with aging. We observed that, at different time points of fasting, AST (483–1914 U/l), ALT (191–785 U/l), and ALP (645–1547 U/l) activities were significantly higher (1.4–13.8 folds) in the homozygous affected mice than the agematched wild-type (AST, 84–212 U/l; ALT, 43–64 U/l; ALP, 297–1103 U/l) or heterozygous (AST, 64–157 U/l; ALT, 31–57 U/l; ALP, 354–850 U/l) mice. Notably, both AST and ALT continued to increase until 40 weeks of age. A similar pattern was observed for ALP and CPK except higher activities were already observed in younger mice. 3.5. Muscle weakness Signs of muscle weakness was first observed in 1-year-old homozygous and manifested as reduced motor activity. Muscle strength of wildtype (3 females and 3 males), heterozygous (5 females and 4 males), and homozygous affected (3 females and 5 males) mice were measured at 62 weeks of age using the grip test (Fig. 10). The muscle strength of the forelimb and hind legs were significantly lower in homozygous
Fig. 8. Fasting serum biochemistries in GSD III knockout mice. A. Fasting effect on serum biochemistry in GSD III knockout homozygous affected mice. Serum GLU, AST, ALT, ALP, and CPK levels of 4 littermate affected mice were measured at 6 and 8 weeks of age with ad-lib feeding and 7 weeks of age with 24 h of fasting. B. Fasting blood glucose and lipid levels in GSD III knockout mice. Serum GLU, TG, and TCHO levels were measured from the same batch of wild-type (+/+) and homozygous affected (−/−) mice after 12 h of fasting at 3 and 8 weeks of age. Means ± s.d. are shown.
474
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
Fig. 9. Serum biochemistry in GSD III knockout mice subjected to repeated fasting. Twenty-four hour fasting serum AST, ALT, ALP, and CPK levels were measured from the wild-type (+/+, n = 7), heterozygous (+/−, n = 7), and homozygous affected (−/−, n = 8) mice at 4, 8, 12, 16, 20, 24, and 40 weeks of age. The initial number of affected mice in each group is 8. Due to early death in some affected mice, the number was reduced to 7 by 20 weeks and to 4 by 40 weeks of age. P-value of each group comparison is shown below each figure. Means ± s.d. are shown.
affected male and female mice than in gender-matched wild-type and heterozygous mice. 4. Discussion GSD III is an autosomal recessive metabolic disorder of glycogenolysis that primarily affects the liver, heart, and skeletal muscle. Until now, only a canine model of GSD III has been reported [29]. Affected canines carried a frame-shift mutation in the Agl gene that results in the deletion of a C-terminal 126-amino acid-long sequence in the GDE protein. Phenotypes of affected canines include glycogen accumulation in the liver, muscle, and adipocytes; elevated fasting serum ALT, AST, and CPK; progressive liver fibrosis, cirrhosis, and disruption of the contractile apparatus; and fraying of myofibrils. However, the growth curves, fasting serum glucose, triglyceride, and cholesterol levels of affected canines are normal. A mouse model is an advantageous tool to study disease due to their shorter generation time, small size, and low maintenance cost. Our mouse model exhibits pathologic accumulation of glycogen in the liver, heart, and muscles with hepatomegaly and progressive liver fibrosis and muscle weakness. However, homozygous affected mice with ad-lib feeding did not exhibit symptomatic cardiomyopathy including left ventricular hypertrophy up to one and half years of age (data not shown). Serum ALT, AST, ALP, and CPK were significantly greater in the homozygous affected mice than in the wildtype and heterozygous mice. Our mouse model recapitulated features that mimic human GSD IIIa with the exception of normal lipidemia and a milder form of hypoglycemia. The hypoglycemia in our mouse model only occurred in very young mice (3 weeks of age) after 12 h of fasting. Notably, this was not observed in older mice (from 4 weeks
up to 40 weeks of age) after 12 h or 24 h of fasting. In human GSD, hypoglycemia causes an increase in the β-oxidation of fats; this is a compensatory mechanism and resulted in increased triglycerides and total cholesterol production [30]. Interestingly, the GSD canine model does not exhibit hypoglycemia, and similarly, we did not observe hypoglycemia in our mouse model. This discrepancy between the animal models and the human disorder may reflect differences in endogenous glucose production or gluconeogenesis between species. Cardiomyopathy is variable in human GSD III. We performed echocardiography analysis in our mice when they were fed ad-lib. Homozygous affected mice (n = 6) exhibited the normal parameters including ejection fraction and left ventricular mass as compared to heterozygous (n = 8) and wild-type mice (n = 5) up to one and half years of age (data not shown). The cardiomyopathy might be more likely to develop in the mice after repeating fasting. However, we did not have enough number of mice to do echo after repeated fasting because of the decreased survival. In our mouse model, liver transaminases (AST and ALT) were elevated (2.3 and 1.8 folds) when mice (4 weeks old) were fed ad lib, these enzyme activities further increased (6.4–9 folds and 4.3–12.2 folds) with fasting, aging, and repeated fasting (the transaminases of repeated fasting reached to thousands at 40 weeks of age) indicating progressive liver damage. During fasting, blood glucose levels decreased, but continued to remain within the normal range (Fig. 8A). The repeated fasting also resulted in early death in these mice. Thus, in our study, fasting detrimentally affects our model. These findings re-emphasize the importance of close monitoring of the dietary regimen in human GSD III patients. Specifically, patients should avoid fasting and aggressive intravenous glucose therapy when oral intake is compromised. It also
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
475
Fig. 10. Muscle strength of aged GSD III knockout mice. Forelimb and hind leg muscle strength for male and female wild-type (+/+), heterozygous (+/−), and homozygous (−/−) mice at 62 weeks of age were measured by grip strength test. Means ± s.d. are shown.
suggested that maintenance of “normoglycemia” may not be sufficient to avoid liver damage that is presumably due to the accumulation of phosphorylase limit dextrin, an abnormal glycogen with short outer braches. In conclusion, we have generated a knock-out mice model with a GDE deficiency that recapitulates the key characteristics of human GSD type III. This mouse model will be useful for future investigation of the pathophysiology and treatment strategy of human GSD III. Acknowledgments This study was supported by the Academia Sinica Genomic Medicine Multicenter Study (40-05-GMM), the National Research Program for Genomic Medicine, National Science Council, Taiwan (National Center for Genome Medicine, NSC101-2319-B-001-001). We would like to thank the EUCOMM Consortium for the production of the ES cell used to generate the Agl gene knockout mice, the technical services provided by the “Transgenic Mouse Model Core Facility of the National Core Facility Program for Biotechnology, National Science Council”, the “Gene Knockout Mouse Core Laboratory of National Taiwan University Center of Genomic Medicine”, the Sequencing Core Facility at the Scientific Instrument Center at Academia Sinica for DNA sequencing, and the Taiwan Mouse Clinic funded by the National Research Program for Biopharmaceuticals, National Science Council, Taiwan, for the technical support. References [1] B. Illingworth, G.T. Cori, Structure of glycogens and amylopectins. III. Normal and abnormal human glycogen, J. Biol. Chem. 199 (1952) 653–660. [2] G.B. Forbes, Glycogen storage disease; report of a case with abnormal glycogen structure in liver and skeletal muscle, J. Pediatr. 42 (1953) 645–653.
[3] I.W. Fellows, J.S. Lowe, A.L. Ogilvie, A. Stevens, P.J. Toghill, M. Atkinson, Type III glycogenosis presenting as liver disease in adults with atypical histological features, J. Clin. Pathol. 36 (1983) 431–434. [4] A.J. Markowitz, Y.T. Chen, J. Muenzer, E.A. Delbuono, M.R. Lucey, A man with type III glycogenosis associated with cirrhosis and portal hypertension, Gastroenterology 105 (1993) 1882–1885. [5] E.B. Haagsma, G.P. Smit, K.E. Niezen-Koning, A.S. Gouw, L. Meerman, M.J. Slooff, Type IIIb glycogen storage disease associated with end-stage cirrhosis and hepatocellular carcinoma. The Liver Transplant Group, Hepatology 25 (1997) 537–540. [6] P. Labrune, P. Trioche, I. Duvaltier, P. Chevalier, M. Odievre, Hepatocellular adenomas in glycogen storage disease type I and III: a series of 43 patients and review of the literature, J. Pediatr. Gastroenterol. Nutr. 24 (1997) 276–279. [7] D. Matern, T.E. Starzl, W. Arnaout, J. Barnard, J.S. Bynon, A. Dhawan, J. Emond, E.B. Haagsma, G. Hug, A. Lachaux, G.P. Smit, Y.T. Chen, Liver transplantation for glycogen storage disease types I, III, and IV, Eur. J. Pediatr. 158 (Suppl. 2) (1999) S43–S48. [8] M. Siciliano, E. De Candia, S. Ballarin, F.M. Vecchio, S. Servidei, R. Annese, R. Landolfi, L. Rossi, Hepatocellular carcinoma complicating liver cirrhosis in type IIIa glycogen storage disease, J. Clin. Gastroenterol. 31 (2000) 80–82. [9] A. Cosme, I. Montalvo, J. Sanchez, E. Ojeda, J. Torrado, E. Zapata, L. Bujanda, A. Gutierrez, I. Arenas, Type III glycogen storage disease associated with hepatocellular carcinoma, Gastroenterol. Hepatol. 28 (2005) 622–625. [10] E. Demo, D. Frush, M. Gottfried, J. Koepke, A. Boney, D. Bali, Y.T. Chen, P.S. Kishnani, Glycogen storage disease type III-hepatocellular carcinoma a long-term complication? J. Hepatol. 46 (2007) 492–498. [11] R.A. Coleman, H.S. Winter, B. Wolf, Y.T. Chen, Glycogen debranching enzyme deficiency: long-term study of serum enzyme activities and clinical features, J. Inherit. Metab. Dis. 15 (1992) 869–881. [12] S. Lucchiari, D. Santoro, S. Pagliarani, G.P. Comi, Clinical, biochemical and genetic features of glycogen debranching enzyme deficiency, Acta Myol. 26 (2007) 72–74. [13] P.S. Kishnani, Y. Suzuki, Y.-T. Chen, Chapter 71: Glycogen Storage Diseases, McGraw Hill Professional, New York, 2009. [14] S. DiMauro, G.B. Hartwig, A. Hays, A.B. Eastwood, R. Franco, M. Olarte, M. Chang, A.D. Roses, M. Fetell, R.S. Schoenfeldt, L.Z. Stern, Debrancher deficiency: neuromuscular disorder in 5 adults, Ann. Neurol. 5 (1979) 422–436. [15] P.S. Kishnani, S.L. Austin, P. Arn, D.S. Bali, A. Boney, L.E. Case, W.K. Chung, D.M. Desai, A. El-Gharbawy, R. Haller, G.P. Smit, A.D. Smith, L.D. Hobson-Webb, S.B. Wechsler, D.A. Weinstein, M.S. Watson, ACMG, glycogen storage disease type III diagnosis and management guidelines, Genet. Med. 12 (2010) 446–463. [16] S.W. Moses, K.L. Wanderman, A. Myroz, M. Frydman, Cardiac involvement in glycogen storage disease type III, Eur. J. Pediatr. 148 (1989) 764–766.
476
K.-M. Liu et al. / Molecular Genetics and Metabolism 111 (2014) 467–476
[17] P.J. Lee, J.E. Deanfield, M. Burch, K. Baig, W.J. McKenna, J.V. Leonard, Comparison of the functional significance of left ventricular hypertrophy in hypertrophic cardiomyopathy and glycogenosis type III, Am. J. Cardiol. 79 (1997) 834–838. [18] H.R. Mundy, J.E. Williams, P.J. Lee, M.S. Fewtrell, Reduction in bone mineral density in glycogenosis type III may be due to a mixed muscle and bone deficit, J. Inherit. Metab. Dis. 31 (2008) 418–423. [19] P.J. Lee, A. Patel, P.C. Hindmarsh, A.P. Mowat, J.V. Leonard, The prevalence of polycystic ovaries in the hepatic glycogen storage diseases: its association with hyperinsulinism, Clin. Endocrinol. 42 (1995) 601–606. [20] S. Bhatti, E. Parry, Successful pregnancy in a woman with glycogen storage disease type III, Aust. N. Z. J. Obstet. Gynaecol. 46 (2006) 168–169. [21] A. Mendoza, N.C. Fisher, J. Duckett, J. McKiernan, M.A. Preece, A. Green, P.J. McKiernan, G. Constantine, E. Elias, Successful pregnancy in a patient with type III glycogen storage disease managed with cornstarch supplements, Br. J. Obstet. Gynaecol. 105 (1998) 677–680. [22] A.E. Slonim, R.A. Coleman, W.S. Moses, Myopathy and growth failure in debrancher enzyme deficiency: improvement with high-protein nocturnal enteral therapy, J. Pediatr. 105 (1984) 906–911. [23] V. Valayannopoulos, F. Bajolle, J.B. Arnoux, S. Dubois, N. Sannier, C. Baussan, F. Petit, P. Labrune, D. Rabier, C. Ottolenghi, A. Vassault, C. Broissand, D. Bonnet, P. de Lonlay, Successful treatment of severe cardiomyopathy in glycogen storage disease type III With D, L-3-hydroxybutyrate, ketogenic and high-protein diet, Pediatr. Res. 70 (2011) 638–641.
[24] A.I. Dagli, R.T. Zori, H. McCune, T. Ivsic, M.K. Maisenbacher, D.A. Weinstein, Reversal of glycogen storage disease type IIIa-related cardiomyopathy with modification of diet, J. Inherit. Metab. Dis. 32 (Suppl. 1) (2009) S103–S106. [25] M.K. Davis, D.A. Weinstein, Liver transplantation in children with glycogen storage disease: controversies and evaluation of the risk/benefit of this procedure, Pediatr. Transplant. 12 (2008) 137–145. [26] P.P. Liu, V.H. de Villa, Y.S. Chen, C.C. Wang, S.H. Wang, Y.C. Chiang, B. Jawan, H.K. Cheung, Y.F. Cheng, T.L. Huang, H.L. Eng, F.R. Chuang, C.L. Chen, Outcome of living donor liver transplantation for glycogen storage disease, Transplant. Proc. 35 (2003) 366–368. [27] D.H. Brown, B.I. Brown, [88] Enzymes of glycogen debranching: amylo-1,6-glucosidase (I) and oligo-1,4 → 1,4-glucantransferase (II), in: V.G. Elizabeth, F. Neufeld (Eds.), Methods in Enzymology, Academic Press, 1966, pp. 515–524. [28] J.J. Shen, Y.T. Chen, Molecular characterization of glycogen storage disease type III, Curr. Mol. Med. 2 (2002) 167–175. [29] H. Yi, B.L. Thurberg, S. Curtis, S. Austin, J. Fyfe, D.D. Koeberl, P.S. Kishnani, B. Sun, Characterization of a canine model of glycogen storage disease type IIIa, Dis. Model. Mech. 5 (2012) 804–811. [30] A.V. Bernier, C.P. Sentner, C.E. Correia, D.W. Theriaque, J.J. Shuster, G.P. Smit, D.A. Weinstein, Hyperlipidemia in glycogen storage disease type III: effect of age and metabolic control, J. Inherit. Metab. Dis. 31 (2008) 729–732.
