PHKA2 mutation spectrum in Korean patients with glycogen storage disease type IX: prevalence of deletion mutations

Background Molecular diagnosis of glycogen storage diseases (GSDs) is important to enable accurate diagnoses and make appropriate therapeutic plans. The aim of this study was to evaluate the PHKA2 mutation spectrum in Korean patients with GSD type IX. Methods Thirteen Korean patients were tested for PHKA2 mutations using direct sequencing and a multiplex polymerase chain reaction method. A comprehensive review of the literature on previously reported PHKA2 mutations in other ethnic populations was conducted for comparison. Results Among 13 patients tested, six unrelated male patients with GSD IX aged 2 to 6 years at the first diagnostic work-up for hepatomegaly with elevated aspartate transaminase (AST) and alanine transaminase (ALT) were found to have PHKA2 mutations. These patients had different PHKA2 mutations: five were known mutations (c.537 + 5G > A, c.884G > A [p.Arg295His], c.3210_3212delGAG [p.Arg1072del], exon 8 deletion, and exons 27–33 deletion) and one was a novel mutation (exons 18–33 deletion). Notably, the most common type of mutation was gross deletion, in contrast to other ethnic populations in which the most common mutation type was sequence variant. Conclusions This study expands our knowledge of the PHKA2 mutation spectrum of GSD IX. Considering the PHKA2 mutation spectrum in Korean patients with GSD IX, molecular diagnostic methods for deletions should be conducted in conjunction with direct sequence analysis to enable accurate molecular diagnosis of this disease in the Korean population.

improve with age [4]. However, several GSD IX patients possessing PKHA2 mutations have been reported to progress to liver cirrhosis, or even present with cirrhosis at the time of diagnosis [4,5].
GSD is a clinically and genetically heterogeneous group of diseases. Different types of GSD can sometimes be clinically indistinguishable; for example, both GSD VI and GSD IX often manifest with hepatomegaly with short stature [3]. Molecular diagnosis of GSD provides the advantage of avoiding invasive liver biopsy [6] and is an important starting point for appropriate therapeutic and monitoring plans [7]. Furthermore, molecular diagnosis allows determination of the inheritance pattern (autosomal recessive vs. X-linked) and DNA testing for other family members when the mutation in the proband is known.
The PHKA2 gene is located at chromosomal locus Xp22.2-p22.1 and consists of 33 exons spanning more than 65 kb and encoding a protein of 1,235 amino acids [8]. Sequence analysis of this X-linked gene can detect small intragenic deletions/insertions and missense, nonsense, and splice site mutations in males and females, and lack of amplification by PCR prior to sequence analysis can suggest a putative exonic or whole-gene deletion on the X chromosome in affected males [9]. Understanding the mutation spectrum of causative genes in different ethnic groups could provide background knowledge to develop a tailored diagnostic approach for different patient populations [7].
To date, data on PHKA2 mutations in Korean patients with GSD IX are available for only two cases. Thus, the aim of this study was to evaluate the mutation spectrum in Korean patients with GSD IX and compare it with previously reported mutation spectra in other ethnic populations.

Study population
From May 2010 to April 2015, blood samples from 13 unrelated Korean children were collected for PHKA2 mutation analysis at Samsung Medical Center. The study population included two female patients with hepatomegaly and elevated aspartate transaminase (AST) and/ or alanine transaminase (ALT) who were undergoing PHKA2 sequencing because no pathogenic mutations in G6PC (for GSD Ia), which is known to a common causative gene of glycogen storage disease in the Korean population, were identified during diagnostic work up for their hepatomegaly [10,11]. Blood samples from healthy individuals, who visited the health promotion center at Samsung Medical Center for regular health checkups without any clinical symptoms or signs of illnesses, and who volunteered for blood sampling, were also collected as negative controls for identified mutations. This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving human subjects were approved by the Institutional Review Board of Samsung Medical Center. Written informed consent was obtained from all subjects and/or their parents.

PhK enzyme activity
PhK enzyme activity in erythrocytes was measured according to a previously described method [12,13]. The results of enzyme activity were expressed as μmol/minute (min)/g hemoglobin (gHb). The reference range of phosphorylase b kinase enzyme activity in the laboratory was 100.0 to 250.0 μmol/hr/g Hb [12], which was validated using erythrocytes from healthy subjects. Each test was performed simultaneously in duplicate and with erythrocytes obtained from two different healthy subjects. Test results were accepted when the results of healthy control samples were within the reference range and the coefficient of variation of duplicated results was less than 10 %.

Quantitation of liver glycogen
Liver glycogen was measured according to a previously described method [14]. Fresh liver samples, 45 μL in volume, were homogenized, heated at 95°C for 5 min, and mixed with 45 μL of 0.1 M sodium acetate buffer (pH 4.8), and 9 μL of amyloglucosidase. Each homogenate was sonicated and incubated at 37°C for 30 min and then at 95°C for 5 min. Glycogen levels per gram of wet liver were calculated as glucose released by amyloglucosidase; i.e., glucose level after incubation with amyloglucosidase minus that without enzyme. The results were expressed as % glycogen/g wet liver weight. The reference range of liver glycogen in the laboratory was 1.0 to 6.0 % per gram of wet liver weight.

PHKA2 mutation analysis
Human genomic DNA was prepared from frozen white blood cells using a Wizard genomic DNA purification kit (Promega, Madison, WI, USA) according to the manufacturer's recommendations. All 33 exons and the flanking regions of the PHKA2 gene were amplified by polymerase chain reaction (PCR) using primers designed by the authors (Table 1)  The pathogenicity of missense variants was evaluated by in silico analyses using Sorting Intolerant from Tolerant (SIFT) (http://sift.jcvi.org/), and Polymorphism Phenotyping v.2 (PolyPhen-2) (http://genetics.bwh.harvard.edu/ pph2/) prediction programs. Human Splice Finder software (http://www.umd.be/HSF3/) was used to predict splicing signals [15].

Results
During the study period, six patients were identified as having PHKA2 mutations. All six patients were male. Clinical and biological information at the time of diagnosis are detailed in Tables 2 and 3. Median age at first diagnostic work-up was 3 years (range 2-4 years). All patients presented with hepatomegaly and elevated AST and/or ALT as the first clinical features. Short stature, defined as height below the 3rd percentile for age by Korean Children and Adolescents Growth Standard [16], was observed in 33.3 % (2/6 patients, case 3 and 4) of GSD IX patients. The height of case 2, who had a missense mutation of c.884G > A (p.Arg295His), was within the 25th to 50th percentile. Among these six patients, only two had been tested for PhK enzyme activity in their erythrocytes, and both showed decreased PhK activity. Liver biopsies had been performed for five patients. Among them, the results of four (patients 3, 4, 5, and 6) were compatible with glycogen storage disease: Periodic acid-Schiff (PAS) staining was positive and PAS diastase stain was negative. One patient who was referred from an outside hospital was reported to be PAS + in liver biopsy, but limited information was available because the liver biopsy had been performed at the outside hospital. All patients initiated uncooked cornstarch therapy after diagnosis of GSD. None of the patients experienced symptomatic fasting hypoglycemia at the time of diagnosis or during the follow-up period. Glucose monitoring, in which blood glucose was measured upon waking in the morning in a fasting state, identified one episode of asymptomatic fasting hypoglycemia (<70 mg/dL) [17] in each of case 2 and case 6.
Each of the six patients had a different PHKA2 mutation. Among the six mutations, five were known mutations-c.537 + 5G > A [6], c.884G > A (p.Arg295His) [8], c.3210_3212delGAG (p.Arg1072del) [18,19], exon 8 deletion [13], and exons 27-33 deletion (Fig. 1) [6]-and one was a novel mutation (exons 18-33 deletion). None of these mutations was observed in control population databases. The novel deletion mutation identified in case 5 was the largest deletion mutation (16 exons) ever reported among GSD IX patients, except for patients reported in the literature with full PHKA2 deletion who had no paternal X-chromosome (LOVD). However, there was a lack of formal information on whether these patients had Turner syndrome, and therefore only one copy of PHKA2. Caution is required when using information from public databases and it is necessary to clarify the status of these patients.
Our review of the literature on PHKA2 mutations in Asian populations with GSD IX is summarized in Table 4. Among 18 different PHKA2 mutations reported in Asian populations, nine (50 %) were splicing or deletion mutations. The overall PHKA2 mutation spectrum derived from a comprehensive literature review is summarized in Fig. 2. The most common PHKA2 mutation type among other ethnic groups and Asian populations Total cholesterol > 240 mg/dL and/or triglyceride >200 mg/dL for at least one measurement during follow-up period f Serum creatine kinase >204 IU/L for at least one measurement during follow-up period g Glucose monitoring in which blood glucose was measured upon waking in the morning in a fasting state identified one episode of asymptomatic fasting hypoglycemia (<70 mg/dL) during the follow-up period. The patient did not experience symptomatic hypoglycemia h X-linked inheritance was identified by maternal PHKA2 mutation analysis i Mild improvement from two fingerbreadths to one fingerbreadth during follow-up period other than Koreans was sequence variants, such as missense, nonsense, or frameshift mutations.

Discussion
In this study we examined the PHKA2 mutation spectrum in Korean patients with GSD IX. GSD IX is associated with a deficiency in liver PhK (caused by mutations in PHKA2, PHKB and PHKG2) or the muscle form of PhK (caused by mutations in PHKA1). Mutations in PHKA2 have been reported to be the most common cause of GSD IX (responsible for approximately 75 % of cases) [20,21]. The second most common cause of liver PhK deficiency is mutations in PHKG2 (autosomal recessive), followed by PHKB (also autosomal recessive) [3]. X-linked liver glycogenosis (XLG) caused by PHKA2 mutations can be divided into two subtypes: XLG I with no detectable activity of phK in liver and peripheral blood cells, and XLG II with normal activity in peripheral blood cells and deficiency in the liver [8]. Although several female GSD IX patients with PHKA2 mutations have been reported in other populations, including Western and other Asian populations [18,[22][23][24], only male patients were identified among our Korean patient population. The PHKA2 mutation spectrum is known to be distributed across the entirety of PHKA2 exons, with the exception of exon 13. Except for amino acids 420,  423, and 432-434, the protein sequence of PHKA2 is strictly conserved among various mammals and zebrafish (Fig. 3). Exon 13, which corresponds to amino acids 416 to 442, is in a six-hairpin glycosidase domain (amino acids 9 -453) that contains pathogenic mutations including several missense mutations [2]. Although several missense variant alleles have been identified in exon 13 (ExAC), hemizygosity and lack of clinical information make interpretation of their clinical significance difficult. Further studies are needed to clarify the clinical significance of mutations in exon 13. No recurrent mutations were observed among Korean GSD IX patients, although a missense PHKA2 mutation of p.Pro1205Leu suggestive of a founder mutation has been reported among Dutch patients [22]. This mutation has also been reported in Japanese and Chinese patients, but has not been observed in the Korean population [22,24,25].
In this study, we identified one novel and five known PHKA2 mutations. The missense mutation of c.884G > A (p.Arg295His) has been reported in a patient with unusually severe clinical manifestation with marked ketosis and hyperlipidemia [8,26] In our study, patient 2 carrying the same mutation showed a relatively mild clinical manifestation, with asymptomatic hepatomegaly and asymptomatic hypoglycemia and height and weight within the 25th to 50th percentiles, although he had experienced hyperlactic acidemia. This mutation has been reported previously in a patient with XLG I who had a pronounced PhK deficiency in both liver and erythrocytes (PhK activity was 10.0 % of the control mean in both liver and erythrocytes) and in a patient with XLG II who had normal erythrocyte PhK activity (83.0 % of the control mean), therefore the significance of this mutation for biochemical function in specific tissues and clinical severity remains speculative [8]. This mutation was predicted to affect protein function by in silico analyses (SIFT and PolyPhen-2) and was predicted to alter splicing through activation of an exonic cryptic donor site by both Human Splice Finder software and by a machine-learning technique that scores This mutation has been reported previously in two patients with GSD IX and is predicted to affect protein function by in silico analyses (SIFT and PolyPhen-2) and to affect splicing, potentially through activation of an exonic cryptic donor site, by both Human Splice Finder software and by a machine-learning technique that scores how strongly genetic variants affect RNA splicing [27] b This mutation has been reported as a founder mutation in the Dutch population c Although in vitro analysis for the splicing effect of c.537 + 5G > A was not performed, in vivo results confirming phosphorylase b kinase deficiency have been reported to constitute a pathogenic mutation in a patient with GSD IX in previous literature. This mutation was predicted to affect splicing, potentially through alteration of the wild-type donor site how strongly genetic variants affect RNA splicing [27]. The c.537 + 5G > A mutation has been reported in a male GSD IX patient of European ancestry who presented with hepatomegaly and growth delay that presented from 9 months of age [6]. This mutation was predicted to affect splicing through alteration of the wild-type donor site. Although in vitro analysis of the splicing effect of c.537 + 5G > A was not performed, in vivo results confirmed phosphorylase b kinase deficiency in the patient's lymphocytes (32 U/g, normal range 100-240 U/g) and erythrocytes (0.7 U/g Hb, normal range 4-10 U/g Hb) [6]. Clinical manifestations of the patient carrying this mutation were reported to include growth delay, mild fasting hypoglycemia, post-prandial lactatemia, and elevated serum triglycerides, with age at onset of 9 months [6]. In our study, patient 1 carrying the same mutation had post-prandial lactatemia and elevated serum triglycerides in common, but no fasting hypoglycemia or growth delay. A small deletion mutation of c.3210_3212delGAG (p.Arg1072del) has been reported in a Finnish male GSD IX patient with hepatomegaly without short stature or hyperlactic acidemia [18] whose age at onset was 16 months. That patient was diagnosed with GSD IX by molecular diagnosis, but a PhK enzyme activity test was not performed [18]. In this study, the patient with the same mutation was of short stature and his liver glycogen concentrations increased over time (8.8 %/g wet liver weight at initial diagnosis and 22.5 %/g wet liver weight at 2.5 year follow-up; reference range 1-6 %/g wet liver weight). His age of onset was 4 years and he experienced hyperlactic academia, which was different from the Finnish male. A gross deletion mutation of exons 27-33 has been reported in a male GSD IX patient of European ancestry with PhK deficiency diagnosed by markedly decreased phosphorylase activities in his lymphocytes and erythrocytes, whose age at onset was 8 months [6]. Because PhK (defective PhK activities cause GSD IX) activates glycogen phosphorylase b (defective phosphorylase activities cause GSD VI) by phosphorylation, patients with GSD IX can have decreased phosphorylase activities [6]. Molecular studies and analysis of PhK are needed to make the correct diagnosis and avoid misdiagnosis of GSD VI [28]. The male patient in the previous literature was reported to have post-prandial lactatemia with elevated serum triglycerides [6]. In this study, patient 6 carrying the same large deletion had asymptomatic fasting hypoglycemia and post-prandial lactatemia but no elevated serum triglycerides. His age at first diagnostic workup was 4 years. Of note, most mutations in Korean GSD IX patients were deletion or splicing mutations, except for one known missense mutation of c.884G > A (p.Arg295His) [8]. Fifty percent of Korean GSD IX patients had gross deletion mutations, and 83.3 % had splicing or deletion mutations that were different from those reported in other ethnic populations. These results suggest that laboratory tests for large deletions in PHKA2 should be included in the variety of methods that may be used such as quantitative PCR, long-range PCR, multiplex ligation-dependent probe amplification, and that chromosomal microarrays including this gene/chromosome segment should be used as the first approach for Korean patients suspected to have GSD IX [9]. In this study, the presence of deletions in male GSD IX patients was initially implied by the lack of amplification of exons by PCR and additional multiplex PCR was then was employed to better define the deletion. Although only male patients had PHKA2 mutations in this study, multiplex PCR would be necessary to detect deletions in females with suspected GSD IX [9]. Methods to detect deletions would be important for patients of all ethnicities [6,9,29].
Historically, diagnosis of GSD was based on enzymatic defects of each type of GSD or on liver histopathology compatible with GSD [6]. Recent molecular diagnostic approaches based on mutation analysis for diseasecausing genes associated with each type of GSD provide the advantages of avoiding invasive liver biopsy and allowing differentiation between several types of GSD with similar clinical findings [6].
Because of its variable clinical manifestations with mild symptoms, underestimation of GSD IX is possible.
For example, short stature was observed in 33.3 % of patients in this study, which was comparable to previous studies in European and Argentinian populations [18] but different from a Canadian study that reported no patients with short stature [3]. Furthermore, the effect of PHKA2 mutations on tissues other than liver should be clarified through future studies. A recent study reported that genetic diagnosis for X-linked mental retardation revealed several PHKA2 variants [30]. Although the natural history of GSD IX in Canadian patients has been reported, there are no reliable data on long-term outcomes for Korean GSD IX patients [3]. The study on 11 Canadian GSD IX patient with PHKA2 mutations with a follow-up period ranging from 1 to 16 years reported improved, normalized, or stable liver enzymes in all patients, and development of likely liver adenoma in one patient at the 5-year follow-up [3]. Detailed long-term natural history studies of Korean patients with GSD IX caused by PHKA2 mutation will be helpful to understand whether these patients are at increased risk of developing additional complications such as liver cirrhosis or hepatocellular carcinoma, or other conditions such as reproductive or mental disorders, later in life.
PhK deficiency is very complex and many genes play a key role in the GSD IX phenotype; the spectrum includes autosomal recessive forms of GSD IX (caused by mutations in PHKB, PHKG1, PHKG2, CALM1, CALM2, and CALM3) as well as the X-linked liver form (caused by mutations in PHKA2) and X-linked muscle form (caused by mutations in PHKA1). Only PHKA2 mutations were studied in this study. Future studies involving comprehensive mutational analyses of multiple causative genes in the Korean population are needed.

Conclusions
In this study we summarized the PHKA2 mutation spectrum in Korean GSD IX patients and found that the most common mutation type was gross deletion. The present study expands our knowledge of the mutational spectrum in Korean GSD IX patients, which differs from that in other ethnic populations. Considering the PHKA2 mutation spectrum in Korean patients with GSD IX, molecular diagnostic methods for deletions should be combined with direct sequence analysis to provide accurate molecular diagnosis of this disease.