| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Molecular Diagnostics and Genetics |
Departments of1
Laboratory of Pediatrics and Neurology, 3
Pediatrics, and 5
Neurology, Radboud University Nijmegen Medical Center, Nijmegen, The Netherlands.
2 Department of Pediatric Metabolic Medicine and Institute for Child Health, Great Ormond Street Hospital, London, United Kingdom.
4 Institut National de la Santé et de la Recherche Médicale (INSERM U602), Groupement de Recherche, Paul Brousse Hospital, University of Paris Villejuif, France.
aAddress correspondence to this author at: Laboratory of Pediatrics and Neurology (830), Radboud University Nijmegen Medical Center, Geert Grooteplein 10, 6525 GA Nijmegen, The Netherlands. Fax 31-24-3668754; e-mail r.wevers{at}cukz.umcn.nl.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: We studied plasma samples from 55 patients with various primary defects in N- and/or O-glycosylation, 21 patients with secondary N-glycosylation defects, and 6 patients with possible glycosylation abnormalities. Furthermore, we analyzed 500 plasma samples that were sent to our laboratory for selective screening for inborn errors of metabolism.
Results: Plasma samples from patients with congenital disorders of glycosylation (CDG) types –IIe and –IIf showed a hypoglycosylated apoC-III isoform profile, as did plasma samples from 75% of the patients with an unspecified CDG type II. Hyposialylated O-glycan profiles were also seen in plasma from 2 patients with hemolytic-uremic syndrome. In the 500 plasma samples from the selective screening, 3 patients were identified with a possible isolated defect in the biosynthesis of core 1 mucin-type O-glycans.
Conclusions: To our knowledge this is the first study in which use of a plasma marker protein has identified patients in whom only O-glycan biosynthesis might be affected. The primary defect(s) remain as yet unknown. Plasma apoC-III IEF is complementary to transferrin isofocusing. In conjunction both tests identify biosynthesis defects in N-glycan and mucin-type core 1 O-glycan biosynthesis. The apoC-III IEF assay is likely to help metabolic laboratories to identify and unravel further subtypes of inborn errors of glycan biosynthesis.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In mammals, the most common form of O-glycosylation is mucin-type O-glycosylation in which glycans are attached to the protein via GalNAc. This mucin-type O-glycosylation can be subdivided into 8 core structures. The core 1 O-glycan, with Galß1–3GalNAc-(Ser/Thr) as the core structure, is the most common subtype and occurs on many membrane and secreted proteins (4)(5). Until now, only 1 disorder with a genetic defect in the biosynthesis of mucin-type O-glycans has been described: familial tumoral calcinosis (FTC) associated with mutations in the UDP- N-acetyl-
-D-galactosamine:polypeptide-N-acetylgalactosaminyltransferase 3 gene (GALNT3)2
(6). O-glycosylation biosynthesis routing and different O-glycan biosynthesis defects have recently been reviewed by Wopereis et al. (7).
Three years ago, our group developed apoC-III IEF as a means to study biosynthesis defects in mucin-type core 1 O-glycans and addressed technical aspects such as linearity, reproducibility, and reference intervals (1). This study is an exploratory retrospective study into the diagnostic significance of apoC-III isoform analysis. We investigated abnormalities in apoC-III isoform profiles in samples from 4 patient groups. Among these are samples from patients with (genetically) confirmed primary and secondary glycosylation defects.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The 2 unreported CDG-IIe cases (1 female patient, 1 month old; 1 male, 7 months old) presented with growth retardation; progressive, severe microcephaly; hypotonia; adducted thumbs; feeding problems from gastrointestinal pseudoobstruction; failure to thrive; cardiac anomalies, wrinkled skin; and episodes of extreme hyperthermia. Both patients were found to have the same homozygous, intronic splice site mutation (c.169 + 4A>C) of the COG7 gene identified in the patients described by Wu et al. (11). Clinical signs and symptoms of these 2 CDG-IIe patients will be described in more detail separately by Morava et al.
patient group 2; secondary defects in the n-glycosylation
Group 2 patients had secondary defects in N-glycosylation. These patients had genetically confirmed secondary diseases or confirmed syndromes with abnormal TIEF profiles. We obtained 21 plasma samples from patients with secondary N-glycosylation alterations [fructosemia due to aldolase B deficiency, n = 2; galactosemia due to galactose-1-phosphate uridyltransferase deficiency, n = 5; chronic alcohol abuse, n = 12; and hemolytic-uremic syndrome (HUS), n = 2]. The plasma samples from the patients with fructosemia and galactosemia were obtained before dietary treatment. The plasma samples from the HUS cases were obtained in the acute phase of the disease before or soon after the start of the appropriate treatment.
patient group 3; possible glycosylation abnormalities
Group 3 patients had confirmed genetic defects suspected to influence the biosynthesis of glycans or a syndrome with abnormal glycosylation patterns described in the literature. We studied 6 plasma samples from patients with disorders suspected to cause abnormal glycosylation [Hutchinson Gilford progeria syndrome (HGPS), n = 2; and a combined deficiency of factor V and factor VIII (F5F8D), n = 4]. In HGPS, abnormal N-glycosylation has been reported (14). Patients with F5F8D have a defect in component 53 of the endoplasmic reticulum (ER)-Golgi intermediate compartment, which could influence the biosynthesis and trafficking of glycoproteins.
patient group 4; selective screening for inborn errors of metabolism
Group 4 patients had plasma samples sent to our laboratory for selective metabolic screening. In a prospective study, we analyzed 500 different plasma samples collected during the last 3 years from patients presenting with psychomotor retardation that was isolated or with additional symptoms, muscle diseases, encephalopathy, and other symptoms compatible with metabolic disorders.
samples and sample preparation
Blood samples were obtained from patients with informed parental consent. Plasma was prepared by centrifugation and stored immediately at –80 °C until required for analysis.
ief of apoc-iii
IEF of apoC-III was carried out as described by Wopereis et al. (7). The apoC-III IEF profile was defined as abnormal when the ratio of the 3 apoC-III isoforms was outside the described reference intervals in at least 2 different plasma samples from 1 patient sample in duplicate (1). In general, 2 abnormal apoC-III IEF profile types can be differentiated; the apoC-III0 IEF profile characterized by increased concentrations of apoC-III0, and the apoC-III1 profile characterized by increased concentrations of the monosialo apoC-III form (8). The day-to-day (total) imprecision (CV), assessed by performing the test on the same plasma sample on 6 different days, was <2% for each of the 3 isoforms (1). The apoC-III IEF assays were performed and interpreted by 2 technicians with 3 years of expertise.
ief of transferrin
TIEF was carried out as described by Wopereis et al. (7).
neuraminidase treatment
Human plasma was incubated with neuraminidase (5 kU/L) from Clostridium perfringens (cat. no. 1585886; Roche; 5U in 0.5 mL 0.1 mol/L Tris, pH 7.0) overnight at room temperature. Samples were analyzed for transferrin and apoC-III IEF as described above.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
patient group 1; primary cdg
All plasma samples from 28 patients with CDG type I (CDG-Ia, -1b, -Ic, -Ie, -If, and CDG-Ix) showed a normal apoC-III isoform distribution. The plasma sample from 3 CDG-IIe patients showed an apoC-III0 IEF profile with increased amounts of the apoC-III0 isoform (35%, 27%, and 23% respectively; reference interval 0%–8%) and decreased amounts of the apoC-III2 isoform (15%, 20%, and 22% respectively; reference interval 40%–62%; Fig. 1
, lane 2). The plasma sample from the CDG-IIf patient showed a different isoform profile, an apoC-III1 IEF profile with increased amounts of the apoC-III1 isoform (78%, reference interval 33%–67%) and decreased amounts of the apoC-III2 isoform (16%, reference interval 27%–60%; Fig. 1
, lane 3). Of the 12 CDG-IIx patients, 3 had a normal apoC-III IEF profile, whereas 9 had an abnormal apoC-III profile. Of the 9 patients with abnormal apoC-III IEF results, 5 had an apoC-III0 IEF profile and 4 had the apoC-III1 profile. The relative amounts of the apoC-III isoforms in the patients with abnormal apoC-III IEF profiles are summarized in Table 1
. Clinical signs and biochemical results of the 12 CDG-IIx patients have been described in more detail (8). The plasma samples from a CDG-IIa, CDG-IId, HIBM, MEB, and 6 FTC patients showed a normal apoC-III isoform distribution. The apoC-III IEF and TIEF results from patients with primary CDG are summarized in Table 2
.
|
|
|
patient group 2; secondary n-glycosylation alterations
Several disorders and/or conditions are known to cause N-glycan abnormalities, such as galactosemia (16), fructosemia (17), chronic alcohol abuse (18), and HUS due to a Streptococcus pneumoniae infection (19). Plasma samples from 2 patients in the acute phase of HUS resulting from a S. pneumoniae infection showed an apoC-III0 profile with increased relative amounts of apoC-III0 (18% and 23%, respectively, reference interval 0%–12%) and slightly decreased amounts of apoC-III2 (26% and 24% respectively, reference interval 27%–60%; Fig. 1
, lane 4). Relative amounts of the apoC-III isoforms in the HUS patients are summarized in Table 1
. The profiles normalized soon after the start of appropriate treatment. The plasma samples from 12 chronic alcohol abuse, 2 fructosemia, and 5 galactosemia patients all had normal plasma apoC-III isoform profiles (Table 2
).
patient group 3; possible glycosylation abnormalities
In patient group 3, we investigated 2 different disorders that were suspected to have abnormal glycan biosynthesis, namely HGPS and F5F8D. Robinson et al. (14) reported that proteome analysis in HGPS showed abnormal glycosylation in patients with and without mutations in the lamin A/C gene. Our HGPS patient with a mutation in the LMNA gene had an apoC-III1 profile with increased amounts of apoC-III1 (76%, reference interval 33%–67%) and decreased amounts of apoC-III2 (20%, reference interval 27%–60%; Fig. 1
, lane 5). The apoC-III isoform distribution in the HGPS patient without a mutation in the LMNA gene (the underlying defect is still unknown) was slightly but not convincingly abnormal. The ApoC-III1 isoform was just above the reference interval (70%, reference interval 33%–67%). The results for TIEF were normal in both HGPS patients. The relative amounts of the apoC-III isoforms in the HGPS patients are summarized in Table 1
.
F5F8D also might lead to glycosylation abnormalities. As is seen in patients with a defect in subunit 7 of the conserved oligomeric Golgi complex (COG7) or CDG-IIe, defects in Golgi trafficking can influence the biosynthesis of glycoproteins. Patients with F5F8D have a defect in component 53 of the ER–Golgi intermediate compartment, thought to function as a molecular chaperone for transport of a specific subset of secreted proteins, including factor V and VIII, from the ER to the Golgi (20). Therefore, this defect might lead to a disturbance of glycosylation. The plasma samples from 4 patients with F5F8D, however, gave normal results for both apoC-III IEF and TIEF.
patient group 4; selective screening for inborn errors of metabolism
ApoC-III IEF was performed in 500 plasma samples sent to our laboratory for selective screening for inborn errors of metabolism. Four samples were found to have an abnormal apoC-III IEF profile (Table 1
). All patients had normal results for TIEF.
Patient 1.
In this 9-year-old boy with pronounced developmental regression and deterioration of verbal communication at age 2 years, ApoC-III IEF repeatedly showed an apoC-III1 profile with increased amounts of apoC-III1 (80%, reference interval 33%–67%) and decreased amounts of apoC-III2 (13%, reference interval 27%–60%).
Patient 2.
In this 18-year-old woman with psychomotor retardation, rhabdomyolysis, and kidney insufficiency, ApoC-III IEF of a plasma sample obtained in the acute phase of rhabdomyolysis showed a hypersialylated apoC-III profile with decreased amounts of apoC-III1 (17%, reference interval 43%–69%) and increased amounts of apoC-III2 (81%, reference interval 23%–50%).
Patient 3.
This 5-year-old boy with psychomotor retardation and dysmorphic features had a repeatedly abnormal apoC-III IEF profile; an apoC-III0 profile with increased amounts of apoC-III0 (18%, reference interval 0%–12%), slightly increased amounts of apoC-III1 (70%, reference interval 33%–67%), and decreased amounts of apoC-III2 (11%, reference interval 27%–60%).
Patient 4.
In this 16-year-old woman with muscle cramps of unknown cause and excessive sweating, ApoC-III IEF repeatedly showed an apoC-III1 profile with increased amounts of apoC-III1 (79%, reference interval 33%–67%), and decreased amounts of apoC-III2 (19%, reference interval 27%–60%). Interestingly, her mother presented similar clinical symptoms and also an apoC-III1 profile with increased amounts of apoC-III1 (75%, reference interval 43%–69%) and decreased amounts of apoC-III2 (22%, reference interval 23%–50%). Serum creatine kinase was normal and a quadriceps muscle biopsy showed no morphological or enzyme histochemical abnormalities. Enzyme histochemistry showed no abnormal mitochondrial staining and also Periodic Acid Schiff-staining was normal. Biochemical investigations of the muscle biopsy showed normal overall oxidation rates and normal respiratory chain complex activity.
apoc-iii pattern in premature neonates and apoc-iii polymorphisms
We found that apoC-III isoforms in premature neonates were hyperglycosylated in 6 of 13 cases, suggesting abnormal O-glycan biosynthesis possibly due to liver immaturity. Furthermore, in the 500 plasma samples we found 2 from individuals who had a polymorphism in the protein backbone of apoC-III. These results have been described in the Supplemental data [see the Data Supplement that accompanies the online version of this article at http://www.clinchem.org/content/vol53/issue2].
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The patient with CDG-IIf had a deficient cytidine 5'monophospho-N-acetylneuraminic acid (CMP-NeuAc) transporter (12). Martinez-Duncker et al. (12) found that the sialylation pattern of several N-glycosylated plasma proteins was normal, but lectin staining revealed defective sialylation of core 1 mucin-type O-glycans. ApoC-III IEF showed an apoC-III1 profile (8) with increased amounts of the apoC-III1 isoform and decreased amounts of apoC-III2. The patient with CDG-IIf was identified by apoC-III IEF but had normal TIEF results. A comparison with other sialylation defects shows that in all cases mainly mucin-type O-glycosylation is affected, rather than N-glycosylation. Three defects affecting sialylation are known to date: CDG-IIf, HIBM, and sialuria. Sialuria patients have a defect in uridine-5'-diphosphate-N-acetylglucosamine-2-epimerase/N-acetylmannosamine kinase (GNE/MNK) leading to an overproduction of CMP-NeuAc. The patients have a clear hypersialylation of plasma core 1 O-glycans as shown in the apoC-III IEF assay but have only minor changes in the sialylation pattern of the N-glycans (22). Patients with HIBM also have a defect in the gene coding for the enzyme GNE/MNK, which is a bifunctional enzyme that catalyzes the first 2 steps in the biosynthesis of CMP-NeuAc. The theoretical decrease in CMP-NeuAc biosynthesis leads to hyposialylated muscle core 1 O-glycans shown by lectin staining (23) and to a decrease of muscle -dystroglycan staining suggestive for hyposialylation of O-mannosylglycans (13). The sialylation of N-glycans, however, seems to be unaffected (13)(23). ApoC-III IEF resulted in a normal sialylation pattern in plasma of a HIBM patient and did not confirm the hyposialylation of core 1 O-glycans found in muscle tissue. The disturbance in mucin-type core 1 O-glycan biosynthesis may be tissue specific, leading to abnormal O-glycans in muscle and possibly not affecting the hepatic synthesis of apoC-III.
The primary defect in 12 CDG-IIx patients, most probably presenting a genetic heterogeneous group, is probably localized in the processing of N-glycans, which mainly occurs in the Golgi compartment. The 3 CDG-IIx patients with normal apoC-III IEF likely had the primary defect in an N-glycan specific processing step, whereas the 9 CDG-IIx patients with abnormal apoC-III IEF likely had the defect in an enzyme or protein that plays a role in the biosynthesis of both N- and core 1 O-glycans.
In patients with other glycosylation disorders, patients with FTC have the only genetically defined defect in the biosynthesis of mucin-type O-glycosylation identified so far. The defective GALNT3 gene encodes UDP-GalNAc transferase 3 (GALNT3), responsible for the transfer of UDP-GalNAc to Thr/Ser to the protein backbone. GALNT3 is highly expressed in human pancreas, skin, kidney, and testis and weakly expressed in prostate, ovary, intestine, and colon (24). The results of a normal apoC-III IEF profile reflected this expression pattern because GALNT3 is not expressed in liver tissue where apoC-III is synthesized (6). The normal results in the CDG-I, CDG-IIa, -IId, and MEB patients are explained by the fact that the defective enzymes in these disorders are not involved in the biosynthesis of apoC-III O-glycan.
In patient group 2, where we tested for abnormalities in the biosynthesis of mucin-type core 1 O-glycans, we found diseases/conditions leading to secondary N-glycan abnormalities. Among these are galactosemia, fructosemia, alcohol abuse, and HUS. Plasma apoC-III was hyposialylated in the acute phase of S. pneumoniae–associated HUS of 2 patients. S. pneumoniae excrete neuraminidase, which catalyzes the hydrolysis of NeuAc residues from glycoproteins, explaining the apoC-III hyposialylation profile. Plasma from patients with galactosemia, fructosemia (all before dietary treatment), and alcohol abuse showed normal results for apoC-III IEF. The 3 conditions result in a transferrin type 1 profile (17)(18)(25), which suggests that these conditions have an influence on the early N-glycan biosynthesis pathway, localized in the ER or cytoplasm. Because the biosynthesis of mucin-type core 1 O-glycans is situated in the Golgi, these results were as expected.
Finally, we analyzed 500 plasma samples that were sent to our laboratory for the selective screening of inborn errors of metabolism. We found 4 patients with an abnormal apoC-III IEF profile. Three of the 4 patients had a hypoglycosylated apoC-III IEF profile, whereas 1 patient had a hypersialylated apoC-III IEF profile. The sample from patient 2 was taken in the acute phase of rhabdomyolysis, and apoC-III isoforms showed a hypersialylated profile. The rhabdomyolysis in our patient caused secondary renal failure. Six additional patients with kidney failure had normal apoC-III IEF profiles (data not shown), so it is unlikely that the kidney failure in the rhabdomyolysis patient caused the hypersialylation. Unfortunately, the patient was lost for follow-up so it remains unclear whether this patient also had an abnormal apoC-III IEF profile after the rhabdomyelytic phase. The mechanism behind the hypersialylation remains unknown.
The other 3 patients displayed a hypoglycosylated apoC-III isoform distribution on several occasions (patients 1, 3 and 4). Patients 1 and 4 had an apoC-III1 profile with increased amounts of apoC-III1 and decreased amounts of apoC-III2. The mother of patient 4 had symptoms similar to those of her daughter and turned out to have an abnormal apoC-III1 IEF pattern as well, suggesting dominant or X-linked inheritance. Patient 3 had an apoC-III0 profile with increased apoC-III0 and apoC-III1 and decreased apoC-III2. The results for TIEF were normal in all 3 patients. We were unable to reach a classifying diagnosis in any of the 3 patients. They are likely to have the primary defect situated in one of the steps involved in the biosynthesis of mucin-type core 1 type O-glycans. To our knowledge this is the first time that patients have been identified in which only the O-glycan biosynthesis might be affected while N-glycan biosynthesis if unaffected.
In conclusion, the apoC-III IEF test is helpful, especially in patients with CDG type II, and should be considered complementary to selective screening with TIEF. The apoC-III isoforms have a broad reference interval in all age cohorts. In some patients only slightly abnormal values are found for apoC-III isoforms (as in case CDG-IIx*1 in Table 1
). When such slight abnormalities are confirmed in a 2nd sample they may be meaningful, but as long as the primary defect in these patients has not been determined the significance of slight abnormalities remains uncertain. No alternative techniques for the screening of O-glycan biosynthesis disorders are available, although newly developed mass spectrometric approaches are promising (27)(28). The apoC-III IEF test is pivotal to pick up inborn errors in mucin-type core 1 O-glycan biosynthesis. The assay is likely to help metabolic screening laboratories to identify and unravel a further series of inborn errors of glycan biosynthesis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Human genes: GALNT3, UDP-N-acetyl--D-galactosamine:polypeptide-N-acetylgalactosaminyltransferase 3; COG7, component of oligomeric golgi complex 7; LMNA, lamin A/C (previous symbols: LMN1, CMD1A); MGAT2, mannosyl (alpha-1,6-)-glycoprotein beta-1,2-N-acetylglucosaminyltransferase; GCS1, glucosidase; SLC35C1, solute carrier family 35, member C1; B4GALT1, UDP-Gal:betaGlcNAc beta 1,4-galactosyltransferase, polypeptide 1; SLC35A1, solute carrier family 35 (CMP-sialic acid transporter), member A1; POMGNT1, protein O-linked mannose beta1,2-N-acetylglucosaminyltransferase; GNE, glucosamine (UDP-N-acetyl)-2-epimerase/N-acetylmannosamine kinase; GALT, galactose-1-phosphate uridylyltransferase; ALDOB, aldolase B, fructose-bisphosphate.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-dystroglycan in patients with hereditary IBM due to GNE mutations. Mol Genet Metab 2004;81:196-202.[CrossRef][ISI][Medline]
[Order article via Infotrieve]-D-galactosamine. Polypeptide N-acetylgalactosaminyltransferase, GalNAc-t3. J Biol Chem 1996;271:17006-17012.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
