Skip Navigation


NDT Advance Access originally published online on January 4, 2008
Nephrology Dialysis Transplantation 2008 23(6):1931-1939; doi:10.1093/ndt/gfm913
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
23/6/1931    most recent
gfm913v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Shimozato, S.
Right arrow Articles by Sugiyama, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Shimozato, S.
Right arrow Articles by Sugiyama, S.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author [2008]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org


Serum under-galactosylated IgA1 is increased in Japanese patients with IgA nephropathy

Sachiko Shimozato, Yoshiyuki Hiki, Hiroko Odani, Kazuo Takahashi, Kouichiro Yamamoto and Satoshi Sugiyama

Department of Nephrology, Fujita Health University, School of Medicine, 1-98 Dengakugakubo, Kutsukake-cho, Toyoake, Aichi 470-1192, Japan

Correspondence and offprint requests to: Yoshiyuki Hiki, Department of Nephrology, Fujita Health University, School of Medicine, 1-98, Dengakugakubo, Kutsukake-cho, Toyoake, Aichi 470-1192, Japan. Tel: +81-562-93-9245; Fax: +81-562-93-1830; E-mail: yyhiki{at}fujita-hu.ac.jp



   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 References
 
Background. Immunoglobulin A nephropathy (IgAN) is characterized by an aberrant structure of O-glycans in the IgA1 hinge region. Recently, under-galactosylated IgA1 has been found to be increased in Caucasian IgAN patients. Thus, we examined this in Japanese IgAN patients.

Methods. An enzyme-linked immunosorbent assay of binding between Helix aspersa (HAA) and serum IgA was performed in Japanese IgAN patients and the HAA–IgA binding levels were compared among IgAN patients (n = 41), patients with other forms of kidney disease (OKD, n = 43) and healthy controls (n = 38). The clinicopathological severity of IgAN was then analysed between patients with high and low HAA–IgA binding levels. The levels were also compared in 11 patients before and after the combination of tonsillectomy and steroid pulse therapy. Furthermore, we examined the O-glycan structure of IgA1 hinge glycopeptides by mass spectrometry (MS).

Results. The HAA–IgA binding levels were significantly higher in IgAN patients compared with either healthy controls (P = 0.0025) or those with OKD (P = 0.016). To reflect the absolute level of under-galactosylated IgA, we multiplied the HAA–IgA binding level by the serum IgA concentration to produce an indicative value. The specificity and sensitivity of this value were 89% and 66%, respectively. MS showed that peak distribution of IgA1 hinge glycopeptides was shifted to smaller molecular weights in high HAA–IgA-binding IgAN patients. There was no correlation between the HAA–IgA binding level and either disease severity or the use of combination therapy.

Conclusions. HAA–IgA binding is significantly increased in Japanese IgAN patients. This potential IgAN marker is not affected by disease severity or therapeutic intervention.

Keywords: IgA1; HAA; GalNAc; galactose; neuraminic acid



   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 References
 
IgA nephropathy (IgAN) is the most common type of primary glomerulonephritis in the world and is a major cause of end-stage kidney disease [1–3].

The disease is characterized by predominant deposits of the IgA1 subclass in the kidney mesangium [4]. The human IgA1 molecule is quite exceptional among the serum proteins because it contains O-linked oligosaccharides (O-glycans) in its hinge region [5]. The structure of O-glycans in IgA1 has been well studied over the past 10 years, and there is increasing evidence for the involvement of aberrantly glycosylated IgA1 in the pathogenesis of IgAN [6–10]. A variant with terminal GalNAc or sialylated GalNAc is rare in normal serum IgA1, but previous studies suggest that the presence of the truncated O-glycan with an exposed N-acetylgalactosamine (GalNAc) residue is more common in the IgA1 of IgAN patients.

In previous studies, we analysed the O-glycan structure in the IgA1 hinge using mass spectrometry (MS) including matrix-assisted laser desorption ionization time-of-flight MS (MALDI-TOFMS) and liquid chromatography/electro- spray ionization/MS (LC/ESI/MS). These techniques enabled us to investigate the heterogeneity of O-glycan structures in the IgA1 hinge region. However, these techniques are not suitable for clinical application because sample preparation (isolation of the IgA1 hinge region) is complicated.

At other institutes, O-glycans in the hinge region of IgA1 have been evaluated by enzyme-linked immunosorbent assay (ELISA) with lectin-specific binding [6,10,11]. Xu et al. [11] reported that aberrantly glycosylated serum IgA1 was closely associated with the pathologic phenotype of IgAN. Their analyses used elderberry bark lectin (SNA), peanut agglutinin (PNA) and Vicia villosa lectin (VVL), which recognize the {alpha}2,6-linked neuraminic acid (NeuAC), galactose (Gal) and GalNAc residues, respectively. This report stimulated our interest, but our efforts with VVL were disappointing, in that reproducible dose–response curves could not be obtained.

Moore et al. [12] recently evaluated the binding characteristics of several commercial preparations of GalNAc-specific lectins, and found that lectins from Helix aspersa (HAA) and Helix pomatia bound exclusively to IgA1 containing Gal-deficient O-linked glycans. On the other hand, V. villosa reacted not only with IgA1 but also with IgA2 and IgG, both of which have N-glycans but no O-glycans. This indicated that VVL recognized not only GalNAc in O-glycans but N-linked glycans as well.

Most recently, Moldoveanu et al. [13] reported the increased binding of HAA to serum IgA1 in Caucasian patients with IgAN with high specificity and sensitivity. In a pilot study using a similar experimental procedure, we also obtained a reproducible dose–response binding curve of IgA to HAA. Therefore, we performed the HAA–IgA binding assay to determine its specificity and sensitivity in Japanese patients with IgAN and assessed the potential association between the altered O-glycans of IgA1 molecules and clinicopathological severity and therapeutic intervention. Furthermore, to analyse the heterogeneity of the O-glycan structures in the IgA1 hinge region, we used MS (LC/ESI/MS) analyses of IgA1 purified from IgAN patients with high and low HAA–IgA binding levels and from healthy individuals.



   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 References
 
Patients and test sera
Forty-one patients diagnosed with IgAN were examined in the present study. The diagnostic criteria used for IgAN consisted of the following: predominant deposits of IgA and C3 within mesangium on renal biopsy, absence of liver disease, purpura or arthritis suspecting Henoch–Schönlein nephritis and collagen disease. Serum samples were obtained at the time of renal biopsy in 39 IgAN patients. In two additional patients, serum samples were taken ~2 years after the renal biopsy. These two subjects did not undergo tonsillectomy nor did they receive steroid therapy. In 11 of the IgAN patients, serum was taken before and after the combination therapy. The patient and control profiles are summarized in Table 1.


View this table:
[in this window]
[in a new window]

  		Table 1 Clinical characteristics of the patients with IgAN and other forms of kidney disease (OKD) and healthy controls

 
An additional 43 patients with other forms of kidney disease (OKD) were also examined. The details are shown in Table 2. Renal biopsy had confirmed the absence of glomerular IgA deposits in all OKD patients. Thirty-eight healthy control subjects were selected from healthy individuals matched for gender and age to the IgAN patient group.


View this table:
[in this window]
[in a new window]

  		Table 2 Itemizing other forms of kidney disease

 
Pathologic severity
The severities of each biopsy specimen were evaluated by the semi-quantitative method of Kobayashi et al. [14]. The seven parameters, namely mesangial hypercellularity, crescent, tuft adhesion, mesangial sclerosis, global sclerosis, interstitial fibrosis and tubular atrophy, were graded from 0 to 3. The overall pathologic severity of each biopsy specimen was defined as the sum of the scores of the seven parameters; the total pathologic score ranged between 0 and 21.

Combination therapy
Palatine tonsillectomy was performed prior to steroid pulse therapy. Steroid pulse therapy was performed according to the protocol of Pozzi et al. [15] with a slight modification. A 500 mg amount of methylprednisolone was administeredintravenously for three consecutive days and this course was repeated 2, 4 and 6 months later. Oral prednisone was given at an initial dose of 20 mg daily for 6 months with a gradual decrease in dosage over 1 year. Patients were given diuretics, antihypertensive drugs including ACEI and/or ARB and antiplatelet agents, as needed.

ELISA
As a pilot study, we examined the dose–response curve of IgA bound to HAA using serially diluted (1/10–1/1000) serum samples, to ascertain the specificity of HAA. The results show the expected dose–response curve (results not shown).

A 100 µl aliquot of 0.05 M bicarbonate buffer (pH 9.6), containing the F(ab')2 fragment of goat anti-human serum IgA (Jackson Immuno Research Labs, West Grove, PA, USA, 15 µg/ml), was added into the wells of a polystyrene microtiter plate (Corning, NY, USA) and incubated overnight at 4°C. The plates were washed with 0.01 M phosphate-buffered saline (PBS)–0.05% Tween-20 (PBST) three times for each step. Coated plates were blocked with 1% bovine serum albumin (BSA; Sigma Chemical Company, St Louis, MO, USA) in PBST for 1 h at room temperature (RT). Beforehand, the serum IgA levels of each sample had been measured by conventional immunonephelometry and adjusted to 1 mg/ml. A 100 µl/well aliquot of the diluted sera (100 µg/well of IgA) in blocking solution was added to the coated wells, in triplicate, and incubated overnight at RT. For quantitation of Gal-deficient IgA1, the terminal neuraminic acid was removed by incubation with 100 µl of 20 mU/ml neuraminidase (Roche Diagnostic Corp., Indianapolis, IN, USA) in acetate buffer (pH 5) for 3 h at 37°C. After washing, 100 µl of biotin-labelled GalNAc-specific lectin HAA (2 µg/ml, Sigma) was added to each well. Following a further 3 h of incubation at 37°C, the plates were washed and the avidin–biotin complex (ABC) peroxidase (Pierce, Rockford, IL, USA) was added and incubated for 1 h at 37°C. The wells were then developed with O-phenylenediamine (OPD)–H2O2 (Pierce) in 0.1 M citrate phosphate buffer. The colour reaction was stopped with 1 M sulphuric acid and absorbance at 490 nm was measured.

Each sample was tested in triplicate and the median values were regarded as the absorbance levels of each sample. The HAA–IgA-binding ELISA was performed on two different days using a total of five ELISA plates. To minimize variation due to experimental conditions, a quality control sample (serum sample from an IgAN patient) was included in each ELISA plate. We described the HAA–IgA binding level in each sample as a relative value calculated by the absorbance level being divided by the absorbance level of the quality control in the same plate. The corrected mean and SD values in healthy controls in the two experiments (first experiment, n = 14; second experiment, n = 24) were 0.90 ± 0.26 and 1.05 ± 0.23, respectively, showing minimal variability.

HAA–IgA binding levels without treatment with neuraminidase were similarly measured in 15 IgAN patients, 15 patients with OKD and 14 healthy controls, all of whom had been randomly selected. As the data were obtained in one experiment using one ELISA plate, the results were expressed as native 490-nm absorbance levels.

LC/ESI/ITMS
To investigate the association of the increased HAA–IgA binding with the actual structure of O-glycans in the IgA1 hinge, the peak distribution of IgA1 hinge glycopeptides was analysed for IgA1 isolated from four IgAN patients: two with high HAA–IgA binding and two with low HAA–IgA binding. Two healthy controls were also included.

The analyses were conducted using the method of Odani et al. [7] with some modifications. Each serum sample was mixed with an equal volume of saturated ammonium sulphate, and the resulting precipitate fraction was applied to an anti-IgA antibody-coupled sepharose column. The bound IgA was eluted with 0.1 M glycine–HCl (pH 2.5).

The IgA preparations were diluted with 0.4 M Tris–HCl buffer (pH 8.6) containing 6 M guanidine–HCl and 0.2 Methylenediaminetetraacetic acid. To dissociate the disulphide linkage, 5 µl of dithiothreitol solution (200 mg/ml) was added to 1 mg IgA with stirring. After heating at 50°C for 4 h, 1.6-µl 4-vinylpyridine /mg IgA was added, and the reaction mixture was allowed to stand for 90 min at RT. The reaction was terminated by the addition of 50 µl of 2 M formic acid.

The S-pyridylethylated IgA1 fraction was dissolved in 50 mM Tris–HCl buffer (pH 8.0) containing 2.0 M urea. To this solution, 20 µl of trypsin solution (10-µg trypsin/20 µl of the above buffer) and 20 µl of 0.1 M CaCl2were added, and then the reaction mixture was incubated overnight at 37°C. The volume of the sample was adjusted to 2 ml with 0.01 M PBS (pH 7.8) and applied to a Jacalin–agarose column. The column was then washed with PBS. After washing the column with 6 ml of 0.8 M glucose in the buffer, the hinge glycopeptide fraction was eluted with 6 ml of 0.1 M melibiose in the buffer. Purification of the glycopeptides by high performance liquid chromatography (HPLC) (Alliance 2690, Waters, Milford, MA, USA) was carried out on a Cosmosil 5C18-300 column (Nacalai Tesque; 4.6 x 150 mm).

LC/ESI/ITMS measurement
The purified glycosylated peptides were analysed by LC/ESI/ITMS (LCQ DECA XP plus, Thermo electron, Waltham, MA, USA). The glycopeptides samples were resolved by reverse phase HPLC (RP-HPLC, SI-2, Shiseido, Tokyo, Japan) and analysed by LC/ESI/ITMS using a LCQ DECA XP plus ion-trap mass spectrometer (Thermo electron, Waltham, MA, USA). RP-HPLC was conducted on a Monitor C-18M column (5 mm, 150 x 4.6 mm ID, Column Engineering, USA) equilibrated with solvent A (0.03% TFA in H2O), and eluted with a linear gradient of 0–50% solvent B (40% acetonitrile, 60% H2O, 0.02% TFA) during the first 30 min. The column was then washed with 50% B for 20 min, with 100% B for 5 min and then re-equilibrated in 100% solvent A for 20 min with a flow rate of 0.3 ml/min. The elution time for the above glycopeptide was ~35 min. For ESI, the ionizing energy, current of spray and voltage of spray were 72 eV, 1.5 mA and 4.5 kV, respectively. Mass spectrometric analyses were performed in the ESI positive-ion mode. To analyse and identify the multiply protonated ion mass spectrum, the BIOMASS deconvolution program (Bio WorksTM 3.0, Thermo Fisher Scientific, USA) was used. The ESI/MS spectrum of multiply charged ions was transformed into a deconvoluted spectrum containing a single peak at the MW of the sample component.

Statistical analyses
The Kruskal–Wallis H-test was used to compare HAA–IgA binding among patients with IgAN and OKD, and healthy controls. To investigate the association of HAA–IgA binding with the clinicopathological severity, 41 IgAN patients were divided into two groups: a high HAA–IgA binding group (more than the 90th percentile of the healthy group value, 1.25) and a low HAA–IgA binding group (<1.25). The chi-square test with Yates’ correction was used to compare the male/female ratio of IgAN patients between high and low HAA–IgA binding groups. Other clinicopathological parameters (age, pathological scores, serum IgA, serum creatinine, urinary protein and urinary occult blood) of the IgAN patients were also compared between high and low HAA–IgA binding groups using the Mann–Whitney U-test. The Wilcoxon t-test was used to compare the clinical parameters and HAA–IgA binding in 11 patients before and after combination therapy. A correlation coefficient between neuraminidase-treated and non-treated HAA–IgA levels was calculated using Spearman's correlation. The 90th percentile for healthy controls was the cutoff point for calculation of sensitivity and specificity. A P-value <0.05 was considered statistically significant.



   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 References
 
HAA–IgA binding levels in patients with IgAN and OKD, and healthy controls
The levels of HAA–IgA binding were higher in IgAN patients (mean ± SD 1.33 ± 0.40) compared to patients with OKD (1.06 ± 0.26) (P = 0.016) and to healthy controls (1.00 ± 0.25) (P = 0.0025) (Figure 1).


Figure 1
View larger version (16K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1 HAA–IgA binding levels in patients with IgAN and other forms of kidney disease (OKD) and healthy controls. The levels of HAA–IgA binding were higher in IgAN patients (mean ± SD; 1.33 ± 0.40) compared to patients with OKD (1.06 ± 0.26) (P = 0.016) and to healthy controls (1.00 ± 0.25) (P = 0.0025).

 
We also calculated a value corresponding to the total serum level of HAA–IgA binding by multiplying the HAA–IgA binding level by the serum IgA level. There were obvious significant differences between IgAN (mean ± SD 430.4 ± 166.9) and the healthy controls (222.3 ± 113.5) (P = 0.000001), and between IgAN and the OKD patients (282.2 ± 120.3) (P = 0.0001) (Figure 2).


Figure 2
View larger version (18K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2 HAA–IgA binding levels multiplied by serum IgA levels in patients with IgAN and OKD and healthy controls. We also calculated a value corresponding to the total serum level of HAA–IgA binding calculated by multiplying the HAA–IgA binding level by the serum IgA level. There were obvious significant differences between IgAN (mean ± SD; 430.39 ± 166.85) and the healthy controls (222.28 ± 113.45) (P = 0.000001), and between IgAN and the OKD patients (282.20 ± 120.34) (P = 0.0001).

 
Specificity and sensitivity of HAA–IgA binding, serum IgA levels and the product of the HAA–IgA binding and serum IgA level
The values of specificity and sensitivity of HAA binding were 89% and 49%, respectively. Median serum IgA concentration was 325.3 mg/dl (range 157.8–477.5 mg/dl) for IgAN patients compared with 217.3 mg/dl (range 116.7–411.5 mg/dl) for healthy controls (P < 0.000005). Eighteen IgAN patients had a serum IgA concentration >339.3 mg/dl (the 90th percentile for controls) indicating that the specificity and sensitivity of the serum IgA levels were 89% and 44%, respectively.

The specificity and sensitivity of the product values calculated by multiplying the HAA–IgA binding level by the serum IgA level increased to 89% and 66%, respectively.

Association of clinicopathological severity with HAA–IgA binding in IgAN patients
None of the clinicopathological parameters differed significantly between the high and low HAA–IgA binding groups (Table 3).


View this table:
[in this window]
[in a new window]

  		Table 3 Association of clinicopathological severity with HAA–IgA binding in IgAN patients

 
The effect of combination therapy on serum HAA–IgA binding levels
Serum IgA, urinary protein and urinary occult blood obviously improved after the therapy. The renal function was kept constant in all of the patients, showing no significant elevation of serum creatinine after the therapy. However, the HAA–IgA binding levels before and after the therapy showed no significant difference in any of the patients (Table 4).


View this table:
[in this window]
[in a new window]

  		Table 4 Comparison of clinical parameters in the IgAN patients before and after the combination therapy

 
Analysis of carbohydrate structure of the IgA1 hinge region with LC/ESI/ITMS in IgAN patients with high or low HAA–IgA binding and healthy controls
Two IgAN patients with a high HAA–IgA binding level, two IgAN patients with a low HAA–IgA binding level and two healthy controls (Figures 3c and 4c) were used in this part of the study.


Figure 3
View larger version (28K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 3 LC/ESI/ITMS analysis of the O-glycan structure of the IgA1 hinge region in IgAN patients with high or low HAA–IgA binding and in the healthy controls. The peaks obviously assigned as 33mer IgA1 hinge glycopeptides based on their m/z (Table 5) were named using capital letters A, B and C and Arabic numbers 1–5 as shown in this figure and Figure 4. As described in Table 5, the numbers of GalNAc, Gal and neuraminic acid (NeuAc) of each peak were expressed as x, y and z, respectively. Figures 3a and 4a represent the mass spectra of two IgAN patients with high-IgA binding levels, Figures 3b and 4b represent the spectra of two patients with low HAA-IgA binding, and Figures 3c and 4c represent those of healthy controls. The HAA-IgA binding levels of the IgAN patients as shown in Figures 3a, 4a, 3b and 4b were 2.02, 2.38, 1.00 and 1.08, respectively. As shown in Figures 3 and 4, peaks 1–5 were mainly observed in IgAN patients with low HAA–IgA binding (Figures 3b and 4b) and healthy controls (Figures 3c and 4c). On the other hand, peaks A, B and C, located to the left of peaks 1–5, were predominantly observed in IgAN patients with high HAA–IgA binding (Figures 3a and 4a).

 

Figure 4
View larger version (30K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 4 The description of Fig. 4 is included in the legend of Fig. 3.

 
The peaks obviously assigned as 33mer IgA1 hinge glycopeptides based on their m/z (Table 5) were named using capital letters A, B and C and Arabic numbers 1–5, as shown in Figures 3 and 4. As described in Table 5, the numbers of GalNAc, Gal and NeuAc of each peak were expressed as x, y and z, respectively. Figures 3a and 4a represent the mass spectra of two IgAN patients with high IgA binding levels; Figures 3b and 4b represent the spectra of two patients with low HAA–IgA binding and Figures 3c and 4c represent those of healthy controls. The HAA–IgA binding levels of the IgAN patients as shown in Figures 3a, 4a, 3b and 4b were 2.02, 2.38, 1.00 and 1.08, respectively.


View this table:
[in this window]
[in a new window]

  		Table 5 Distribution of assigned peaks of IgA1 hinge glycopeptides

 
As shown in Figures 3 and 4, peaks 1–5 were mainly observed in IgAN patients with low HAA–IgA binding (Figures 3b and 4b) and healthy controls (Figures 3c and 4c). On the other hand, peaks A, B and C, located to the left of peaks 1–5, were predominantly observed in IgAN patients with high HAA–IgA binding Figures 3a and 4a).

Correlation between neuraminidase-treated and non-treated HAA–IgA levels
The levels of HAA–IgA binding without neuraminidase treatment were generally much lower for all samples, although still significantly elevated in the IgAN patients (mean ± SD absorbance 0.12 ± 0.06) compared to the healthy controls (0.05 ± 0.04 versus IgAN; P = 0.007). There were no significant differences between the levels in IgAN patients and OKD patients (0.08 ± 0.05 versus IgAN; P = 0.39). The levels of each individual were significantly associated with neuraminidase treatment levels (correlation coefficient =0.81, P = 3.0 x 10–11) (Figure 5).


Figure 5
View larger version (16K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 5 Correlation between neuraminidase-treated and non-treated HAA–IgA levels. The levels of each individual were significantly associated with neuraminidase treatment levels (correlation coefficient =0.81, P = 3.0 x 10–11). Figure 5, and {blacktriangleup} indicate IgAN, OKD and healthy individuals, respectively.

 


   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
References
 
The serum levels of HAA–IgA binding were significantly higher in the IgAN patients compared to patients with OKD and healthy controls. The HAA ELISA was performed using serum samples adjusted to 1 mg/ml IgA levels, and thus the results were not influenced by IgA serum levels, but by the proportion of Gal-deficient IgA1 molecules and/or the degree of under-galactosylation (number of exposed GalNAc residues).

The levels of HAA–IgA binding did not differ between OKD patients and healthy controls (P = 1.99). However, the product values of HAA–IgA binding and serum IgA levels did show significant differences between the two groups (P = 0.036). The cause of this phenomenon is most likely an increase in serum IgA levels. Increases in serum IgA levels due to the original disease were observed in some OKD patients as follows: interstitial nephritis due to Sjögren syndrome (serum IgA level 813.2 mg/dl); four patients with ANCA-related nephritis (554.0, 443.5, 391.3 and 376.6 mg/dl) and acute glomerulonephritis (467.0 mg/dl). Consequently, we adjusted serum IgA levels among the three groups beforehand to avoid false positive data not due to the proportional increase in HAA-responsive IgA but actually due to an increase in serum IgA.

To evaluate the actual under-glycosylation, MS was performed on IgA1 hinge glycopeptides isolated from IgAN patients with high and low HAA–IgA binding levels and from healthy controls. Previously, we analysed the O-glycan structure in the IgA1 hinge region using MS [7, 8]. In that study, the O-glycan structure of pooled serum IgA1 and IgA1 extracted from glomeruli in renal biopsy specimens appeared to be under-glycosylated in IgAN patients.

In the present study, the distribution of the peaks of IgA1 hinge glycopeptides in IgAN patients with high HAA–IgA binding was shifted to a lower molecular weight range, similar to our previous study using pooled sera [7]. On the other hand, the peak distributions in IgAN patients with low HAA binding were comparable with those of the healthy controls. To elucidate the precise structure of peaks A (numbers of GalNAc:Gal:NeuAc, x:y:z = 4:1:1), B (3:0:3) and C (6:3:0) found in IgAN patients with high HAA–IgA binding, the numbers of truncated GalNAc residues, known as Tn antigens, of these peaks were calculated as x – y, found to be all three (4 – 1, 3 – 0 and 6 – 3, respectively). However, the Tn antigens of peaks 1–5 found among patients with low HAA and healthy controls were 0 or 1. Although the number of subjects examined was limited, the Tn antigen seemed to be increased specifically in IgAN patients with high HAA. This result is reasonably consistent with the HAA–IgA binding levels because HAA recognizes the Tn antigen. However, to confirm the phenomenon, further MS with a greater number of subjects is required.

We also measured HAA–IgA levels without the treatment of neuraminidase in 15 IgAN patients, 15 patients with OKD and 14 healthy controls. The levels were generally much lower for all samples compared with those of the neuraminidase-treated samples. However, the untreated values were still higher in IgAN patients compared with the other groups, significantly so when compared with the healthy controls. Further, there was a close correlation between neuraminidase-treated and non-treated values in each individual. We used commercially available neuraminidase that had been isolated from Arthrobacter ureafaciens. According to the manufacturer's instructions, neuraminidase preferentially hydrolyzes terminal {alpha}2,6-linked NeuAc rather than {alpha}2,3- or {alpha}2,8-linked NeuAc. These results therefore suggest that the GalNAc residue is uniformly sialylated among the three groups. These results are interesting because the formation of the NeuAc {alpha}2,6–GalNAc–Ser/Thr structure is known to interfere with subsequent galactosylation of GalNAc in O-glycans [16]. However, we cannot completely rule out the possibility that the dramatic increase in HAA–IgA binding after neuraminidase treatment was caused not only by the desialylation of GalNAc but also by the elimination of {alpha}2,3-linked NeuAc from galactose, which sterically hinders the access of HAA to neighbouring GalNAc.

Moldoveanu et al. [13] reported that the specificity and sensitivity of HAA–IgA binding for the diagnosis of IgAN were 94% and 76.5%, respectively, which would give it the potential as a diagnostic marker for IgAN. In the present study, the specificity and sensitivity of HAA–IgA binding adjusted to IgA levels of 1 mg/ml were 89% and 49%, respectively. We also compared the values relevant to the total levels of HAA-binding IgA among the groups, calculating a product value of the HAA–IgA binding level and serum IgA concentration. The specificity and sensitivity of the products were 89% and 66%, respectively. The cutoff point of specificity (90th percentile for healthy controls) was the same as in [13]. However, the sensitivity of our data for both HAA–IgA binding and HAA–IgA x IgA levels was inferior. Although the present results still provide ancillary support for diagnosis, our fundamental aim in this study was to confirm the qualitative difference (under-galactosylation) of IgA1 molecules in IgAN not reflected by serum IgA levels. Superior results could be obtained by excluding the error of dilution procedure if we directly evaluated HAA–IgA binding without adjustment for the IgA level.

Our results indicated that the HAA–IgA binding capacities were not associated with the clinicopathological parameters of IgAN. In this study, we found that the therapy drastically improved the clinical parameters, including a decrease in serum IgA levels. However, the HAA–IgA levels were not affected by the aggressive combination therapy, indicating that under-galactosylation itself was not affected by the change in the immunological conditions caused by the therapy. In contrast, HAA-IgA x IgA levels were significantly decreased after the therapy, most likely caused by the drastic decrease in serum IgA. From a clinical point of view, the decreased amount of aberrantly glycosylated IgA1 due to the decrease in total IgA may play a role in the effectiveness of the therapy. We frequently observed decreased serum IgA levels after tonsillectomy without steroid therapy in some IgAN patients. However, we suspect that the main effectiveness of this therapy is due to the influence of the anti-inflammatory effect on the injured glomerulus and not of the under-glycosylation of IgA1 molecules.

Moldoveanu et al. also observed that the HAA–IgA binding levels did not correlate with glomerular filtration rate, histological abnormality or degree of urinary proteins among the IgAN patients. These results suggest that the under-glycosylation of IgA1 plays a significant role for the glomerular deposition of IgA, but not for the glomerular tissue damage.

In an animal model, we recently observed that enzymatically desialylated and degalactosylated IgA1 tended to deposit in glomeruli of mice known as KM mice [17]. These mice lack endogenous genes for immunoglobulins and carry the entire human IgH locus and the IgLk transgene [18]. However, there was an absence of tissue injury. Therefore, there is no evidence that the mesangial deposition of under-glycosylated IgA1 directly provokes glomerular injury. It is also well known that mesangial IgA deposition is incidentally encountered in asymptomatic individuals. From these observations, it was suspected that there may be other factors responsible for the occurrence and progression of glomerular damage.

In conclusion, the present study found that binding of HAA to serum IgA1 is significantly higher in Japanese patients with IgAN. These increases were independent of the serum IgA concentrations. The levels of HAA–IgA binding in IgAN did not correlate with disease severity and the combined therapy. Further studies are required to examine the induction of tissue damage and progression to end-stage renal disease in IgAN.



   Acknowledgments
 
The authors appreciate the valuable assistance of Professor Jan Novak (University of Alabama, Birmingham, USA) in the development of the lectin assay and critical reading of the manuscript. The authors are also grateful to Dr Yutaka Kobayashi (Akebono Hospital, Tokyo, Japan) for supplying materials indispensable to this study. This study was supported by the New Energy and Industrial Technology Development Organization (NEDO) as a part of the Developing Technology Project for implementing sugar chain functions in Japan. This study was also supported in part by the following grants: a grant-in-aid for the 21st Century Center of Excellence Program of Fujita Health University, the Ministry of Education, Science and Culture of Japan (no 1659804) and Fujita Health University Research Foundation.

Conflict of interest statement. None declared.



   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 

  1. D'Amico G. The commonest glomerulonephritis in the world: IgA nephropathy. Q J Med (1987) 64:709–727.[Web of Science][Medline]
  2. Clarkson AR, Woodroffe AJ, Aarons I, et al. IgA nephropathy. Annu Rev Med (1987) 38:157–168.[CrossRef][Web of Science][Medline]
  3. Julian BA, Waldo FB, Rifai A, et al. IgA nephropathy, the most common glomerulonephritis worldwide. A neglected disease in the United States? Am J Med (1988) 84:129–132.[Web of Science][Medline]
  4. Donadio JV Jr, Grande JP. Immunoglobulin A nephropathy: a clinical perspective. J Am Soc Nephrol (1997) 8:1324–1332.[Web of Science][Medline]
  5. Baenziger J, Kornfeld S. Structure of the carbohydrate units of IgA1 immunoglobulin. I: composition, glycopeptide isolation, and structure of the asparagine-linked oligosaccharide units. J Biol Chem (1974) 249:7260–7269.[Abstract/Free Full Text]
  6. Tomana M, Matousovic K, Julian BA, et al. Galactose-deficient IgA1 in sera of IgA nephropathy patients is present in complexes with IgG. Kidney Int (1997) 52:509–516.[Web of Science][Medline]
  7. Odani H, Hiki Y, Takahashi M, et al. Direct evidence for decreased sialylation and galactosylation of human serum IgA1 Fc O-glycosylated hinge peptides in IgA nephropathy by mass spectrometry. Biochem Biophys Res Commun (2000) 271:268–274.[CrossRef][Web of Science][Medline]
  8. Hiki Y, Odani H, Takahashi M, et al. Mass spectrometry proves under-O-glycosylation of glomerular IgA1 in IgA nephropathy. Kidney Int (2001) 59:1077–1085.[CrossRef][Web of Science][Medline]
  9. Horie A, Hiki Y, Odani H, et al. IgA1 molecules produced by tonsillar lymphocytes are under-O-glycosylated in IgA nephropathy. Am J Kidney Dis (2003) 42:486–496.[CrossRef][Web of Science][Medline]
  10. Allen AC, Harper SJ, Feehally J. Galactosylation of N- and O-linked carbohydrate moieties of IgA1 and IgG in IgA nephropathy. Clin Exp Immunol (1995) 100:470–474.[Web of Science][Medline]
  11. Xu LX, Zhao MH. Aberrantly glycosylated serum IgA1 are closely associated with pathologic phenotypes of IgA nephropathy. Kidney Int (2005) 68:167–172.[CrossRef][Web of Science][Medline]
  12. Moore JS, Kulhavy R, Tomana M, et al. Reactivities of N-acetylgalactosamine-specific lectins with human IgA1 proteins. Mol Immunol (2007) 44:2598–2604.[CrossRef][Web of Science][Medline]
  13. Moldoveanu Z, Wyatt RJ, Lee JY, et al. Patients with IgA nephropathy have increased serum galactose-deficient IgA1 levels. Kidney Int (2007) 71:1148–1154.[CrossRef][Web of Science][Medline]
  14. Kobayashi Y, Tateno S, Hiki Y, et al. IgA nephropathy: prognostic significance of proteinuria and histological alterations. Nephron (1983) 34:146–153.[Web of Science][Medline]
  15. Pozzi C, Bolasco PG, Fogazzi GB, et al. Corticosteroids in IgA nephropathy: a randomised controlled trial. Lancet (1999) 353:883–887.[CrossRef][Web of Science][Medline]
  16. Schachter H, McGuire EJ, Roseman S. Sialic acids. XIII: a uridine diphosphate D-galactose: mucin galactosyltransferase from porcine submaxillary gland. J Biol Chem (1971) 246:5321–5328.[Abstract/Free Full Text]
  17. Hiki Y, Takahashi K, Itoh M, et al. Continuous administration of human desial-degalacto IgA1 causes mesangial IgA deposition in KM mouse having entire Ig loci. Nephrology (2005) 10:A430.[CrossRef]
  18. Ishida I, Tomizuka K, Yoshida H, et al. Production of human monoclonal and polyclonal antibodies in TransChromo animals. Cloning Stem Cells (2002) 4:91–102.[CrossRef][Medline]
Received for publication: 21. 8.07
Accepted in revised form: 3.12.07


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
23/6/1931    most recent
gfm913v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Shimozato, S.
Right arrow Articles by Sugiyama, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Shimozato, S.
Right arrow Articles by Sugiyama, S.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?