• β1,3-galactosyltransferase;
  • C1GALT1;
  • Cosmc;
  • immunoglobulin A glycosylation;
  • immunoglobulin A nephropathy;
  • real-time reverse transcriptase polymerase chain reaction;
  • vicia villosa lectin


  1. Top of page
  2. Abstract.
  3. Introduction
  4. Design, setting and methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References

Purpose.  Aberrant O-glycosylation of serum IgA1 is presumed to be one of the main pathogenesis of immunoglobulin A nephropathy (IgAN). β1,3-galactosyltransferase (β1,3GT), whose activity requires coexistence of a specific chaperone, is the main enzyme which participate in the glycosylation process. The current study was carried out to elucidate the expression level of β1,3GT (C1GALT1) and its chaperone (Cosmc) in IgAN, and their relationships with clinical features as well as IgA glycosylation level.

Design, setting and subjects.  Forty-one patients with IgAN, 21 patients with non-IgAN glomerulonephritis and 26 normal controls were included in the present study. Peripheral B lymphocytes were isolated, and then expression level of C1GALT1 and Cosmc were quantitatively measured by real-time reverse transcriptase polymerase chain reaction (RT-PCR). Serum IgA level and glycosylation level were determined by enzyme-linked immunosorbent assay (ELISA) and VV lectin-binding method. Correlation analysis was performed between C1GALT1/Cosmc expression levels and clinical manifestations (severe proteinuria, renal dysfunction, gross haematuria).

Results.  B-lymphocyte Cosmc gene expression level was significantly lower in IgAN patients than that of normal control and non-IgAN patients (P < 0.05), whilst no apparent disparity was observed in C1GALT1 expression level. Cosmc expression showed a negative correlation with IgA O-glycosylation level indicated by VV lectin-binding assay. Statistical analysis also indicated that the level of Cosmc expression was negatively correlated with severe proteinuria (P < 0.05) instead of gross haematuria (P > 0.05).

Conclusion.  These data suggested that the aberrant IgA O-glycosylation in IgAN was resulted from a downregulation of β1,3GT chaperone (Cosmc) expression in B lymphocyte, which is closely associated with clinical characteristics of the disease. This downregulation might be one of the fundamental pathogenic abnormalities in IgAN.


  1. Top of page
  2. Abstract.
  3. Introduction
  4. Design, setting and methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References

Immunoglobulin A nephropathy (IgAN) is characterized by proliferative changes in the glomerular mesangial cells and matrix associated with marked deposition of IgA, mainly IgA1 [1]. IgAN is the most common glomerulonephritis in the world, accounting for more than 50% biopsy-proved primary glomerulonephritis in Asia. Approximately 20% patients progress to end-stage renal disease (ESRD) within 20 years after renal biopsy. The pathogenesis and mechanism of IgAN remains to be determined. Recent investigations indicated that abnormalities of IgA1 O-glycosylation might be one of the key pathogenesis factors of IgAN. It was reported that O-linked Gal, GalNAc were significantly decreased in sera, mesangial deposited and tonsil secreted IgA1 molecules in IgAN patients [2–4]. Allen et al. had reported that the β1,3-galactosyltransferase (β1,3GT) synthesis activity was remarkably lower in peripheral B lymphocyte of IgAN patients when compared with that of controls, but it was not clarified whether this activity reduction was a result from downregulation of gene expression or postexpressional inhibition [5]. Regarding that the β1,3GT (C1GALT1) [6] activity needs the coexistence of a specific chaperone, Cosmc (core 1 β3-Gal-T-specific molecular chaperone) [7], also known as C1GALT2 [8], the present study was designed to determine the β1,3GT and its chaperone expression level in IgAN patients as well as relationships with clinical manifestations. Although more than 30 years have passed since the first report of IgAN, there is still a lack of appropriate diagnosis, monitoring and prognosis judgement index of IgAN. Conventional indices such as severity of proteinuria, haematuria, hypertension and oedema are low in sensitivity and specificity, and renal biopsy is unfit for continuous monitoring of diseases progress. Therefore, in the present study, we also try to elucidate whether the expression of β1,3GT and Cosmc could be used as specific index of IgAN.

Design, setting and methods

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Design, setting and methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References


Forty-one biopsy-proved IgAN patients were included in the present study. Diagnosis criterion of IgAN was based on the manifestation of generalized glomerular mesangial proliferation with presence of IgA as the sole or predominant immunoglobulin deposition in the mesangial area of glomeruli. Patients with systemic disease such as Schonlei-Henoch purpura, systemic lupus erythematosus, Sjögren syndrome, rheumatoid arthritis, diabetes mellitus or liver cirrhosis were excluded.

Twenty-one biopsy-proved non-IgAN glomerulonephritis patients were include as controls. Of them, there were 11 cases of minimal change disease, five cases of focal segmental glomerulosclerosis (FSGS), two cases of membranous nephropathy, one case of membranoproliferative glomerulonephritis, two cases of lupus nephritis. The 1995 WHO classification criterions are applied in the diagnosis.

The severity of renal lesions was graded according to Haas and Lee's classifications [9, 10], grades I and II were considered as mild lesions; grade III as moderated lesions; grades IV and V as severe lesions. Severity of proteinuria was classified according to 24-h quantification: <3.5 g per 24 h (or <35 mg kg−1 per 24 h) was defined as mild/moderate and >3.5 g day−1 (or >35 mg kg−1 per 24 h) as severe. The WHO definition of hypertension was applied in hypertension classification [11]. Judgement of renal dysfunction was based on serum creatinine (sCr) concentration: 133 μmol L−1 was applied as the upper normal limit. Blood pressure (BP), urine routine, sCr and 24 h urinary protein quantification were measured in all patients (Table 1).

Table 1.  Baseline clinical characters of immunoglobulin A nephropathy (IgAN) and non-IgAN patients
  1. aExpressed as mean ± SD.

Age (years)27.53 ± 9.38a29.4 ± 10.03a
Disease duration (months)14.12 ± 23.28a18.3 ± 40.87a
Blood pressure systolic (mmHg)117.6 ± 15.8a116.9 ± 16.4a
Blood pressure diastolic (mmHg)75.5 ± 11.6a73.6 ± 13.0a
Patients with renal hypertension55
Proteinuria quantification (g per 24 h)2.84.94
Severe proteinuria rate (%)3238
Serum creatinine (μmol L−1)89.3 ± 35.6a90.0 ± 29.5a
Renal dysfunction rate (%)1519
Gross haematuria rate (%)4929
Pathological severity (mild : moderate : severe)18 : 20 : 3 

Normal controls

Twenty-six sex- (nine of 17) and age-matched (18–50 years old, mean 25) unrelated normal controls without known diseases were selected. BP, urine routine and sCr were measured to exclude those who has abnormal findings.

Serum samples and B-lymphocyte isolation

About 10 mL venous blood sample was taken into ethylenediaminetetraacetic acid (EDTA) anticoagulated tubes. Peripheral blood mononuclear cells (PBMCs) were separated by density-gradient centrifugation on Ficoll (density 1.077, Sangon, China). PBMCs were washed three times with phosphate-buffered saline (PBS) and resuspended in PBS + 1% bovine serum albumin (BSA). Peripheral B lymphocytes were then isolated using Dynabeads M450 CD-19 (PanB) magnetic beads (Dynal Biotech ASA, Oslo, Norway) according to manufacturer's protocol [12]. The purity of harvested cells was measured by flow cytometry, which indicate that B-cell population was >90%, about 1.5 mL of sera was also collected.

IgA ELISA analysis

Serum IgA concentration was determined by sandwich enzyme-linked immunosorbent assay (ELISA). Briefly, 96-well immunoplates were coated with 100 μL per well goat antihuman IgA (Southern Biotechnology Associates, Birmingham, AL, USA) in 0.05 mol L−1 carbonate-bicarbonate buffer (pH 9.6), overnight at 4 °C. The plates were washed five times with washing buffer (50 mmol L−1 Tris/0.14 mol L−1 NaCl/0.05% Tween 20, pH 8.0) by automated plate washer (Bio-Rad model 1575 Immunowash; Bio-Rad, Hercules, CA, USA), and excess protein-binding sites were blocked with 200 μL per well (50 mmol L−1 Tris/0.14 mol L−1 NaCl/1% BSA) for 2 h at room temperature. After further washing, samples diluted 50 000 : 1 with dilution solution (50 mmol L−1 Tris/0.14 mol L−1 NaCl/0.05% Tween 20/1% BSA) were added to the plates at 100 μL per well in duplicate. The plates were incubated overnight at 4 °C, washed again, and then incubated with 100 μL per well horseradish peroxidase-conjugated goat antihuman IgA (Southern Biotechnology Associates) diluted in dilution solution (6000 : 1) for 2 h at ambient temperature. After final washing, the colour was developed using 50 μL per well O-Phenylenediamine (OPD) solution with 0.5% (v/v) 30% H2O2, and the reaction was stopped after 20 min with 100 μL per well 1 mol L−1 H2SO4. The absorbance was detected with Bio-Rad 550 at 490 nm. A standard curve was constructed with a serial dilution of Human Reference Serum (Bethyl Laboratories, Montgomery, USA). Serum IgA concentration was then read from the standard curve of each plate [13].

Vicia villosa lectin-binding assay

Immunoplates were coated with anti-IgA antibody and blocked with BSA as above. About 100 μL per well sample at 50 μg mL−1 were added to the plates in duplicate, and incubated at 4 °C overnight. After washing, the plates were incubated at ambient temperature for 2 h with 50 μL per well biotinylated VV lectin (Vector Laboratories, Peterborough, UK) at 1 μg mL−1 in PBS, washed again, and incubated for a further 2 h at ambient temperature with 50 μL per well peroxidase-avidin D (Vector Laboratories) at 2.5 μg mL−1 in PBS. Colour was developed as above [13]. As results obtained from different assays run at different times could not be compared directly, we included a same control serum as internal calibrator in each assay. For each sample, the observed optical density (OD) was calibrated according to the OD of calibrator sample, allowing for further comparisons.

RT-PCR and real-time RT-PCR

Total RNA was extracted using Trizol reagent (Invitrogen, Carlsbad, CA, USA), concentration and purity were determined by Beckman DU-600 UV spectrometer, OD ratio of 260/280 nm (Beckman, Fullerton, CA, USA). cDNA was synthesized using 500 ng total RNA with revert first-strand cDNA kit (MBI Fermentas, Vilnius, Lithuania) according to manufacturer's protocol. In order to accurately quantify the expression level of β1,3GT and its chaperone, real-time reverse transcriptase polymerase chain reaction (RT-PCR) with TaqMan probe technique was performed. Primers and probes of C1GALT1, Cosmc and the internal calibrator GAPDH (synthesized by Sangon, Shanghai, China) were listed in Table 2.

Table 2.  Primers and probes of C1GALT1, Cosmc and the internal calibrator GAPDH

A 30 μL PCR solution (10X Taq buffer 3 μL; 25 mmol L−1 MgCl2μL; 10 mmol L−1 dNTPs 1 μL; 10 μmol L−1 forward primer 1 μL; 10 μmol L−1 reverse primer 1 μL; 10 μmol L−1 TaqMan probe 1 μL; Taq DNA Polymerase 1 U; cDNA 2 μL; DEPC-H2O 17.8 μL) was amplified in ABI PRISM 7700 Sequence Detection System (Applied Biosystems, Foster, CA, USA), Taq DNA polymerase, dNTPs were purchased from MBI Fermentas. The thermocycles of C1GALT1 PCR was as follows: after an initial denaturation at 94 °C for 10 min, 45 cycles of denaturation 94 °C for 30 s, annealing 55.5 °C for 30 s, polymerization 72 °C for 60 s were performed, then the final polymerization was carried out at 72 °C for 10 min. Thermocycles of Cosmc was similar to that of C1GALT1 only the annealing temperature was 60 °C. PCR products were agarose gel electrophoretic analysed. During the real-time PCR process, fluorescence sign was collected during annealing steps. Cycle threshold (Ct) was calculated for further statistical analysis. In order to examine the efficiency of real-time PCR, standard curves were established with serial dilution of sample RNA (500 ng, C1GALT1 10 times dilution, Cosmc five times dilution, GAPDH 10 times dilution).

The PCR products of three randomly selected samples of each group were purified on a 1% agarose gel using Agarose Gel DNA Purification Kit (Takara, Shiga, Japan) and then directly sequenced.

Statistical analysis

The real-time RT-PCR results were analysed using Pfaffl's method [14]; anova analyses were carried out using SPSs 12.0 software. Two-sided P-value of 0.05 was taken as the level of statistical significance.


  1. Top of page
  2. Abstract.
  3. Introduction
  4. Design, setting and methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References

IgA ELISA and VV lectin binding

Serum IgA concentration of IgAN patients, non-IgAN patients and the normal control were measured by ELISA. We found that mean serum IgA levels were remarkable higher in IgAN patients (2.12 ± 0.95 mg mL−1) compared with non-IgAN patients (1.65 ± 0.35 mg mL−1) and normal control (1.64 ± 0.12 mg mL−1; P < 0.05). In VV lectin-binding assay, the intra-plate coefficient of variation (CV) of assay was 3.2%, and the inter-plate CV was 6.1%. As expected, VV lectin binding was remarkably higher in IgAN patients (OD: 0.43 ± 0.07) compared with non-IgAN patients (OD: 0.36 ± 0.07) and normal controls (OD: 0.32 ± 0.09; P < 0.001). However, no apparent difference was noticed between non-IgAN patients and normal control in serum IgA concentration and VV lectin-binding results (Fig. 1).


Figure 1. Results of immunoglobulin A (IgA) enzyme-linked immunosorbent assay (ELISA) measuring and VV lectin-binding assay.

Download figure to PowerPoint


Clear fragments of PCR products could be seen after RT-PCR agarose gel electrophoresis. PCR product of C1GALT1 was 164 bp; Cosmc was 90 bp and GAPDH was 143 bp, as illustrated in Fig. 2. Sequencing result was analysed using ncbi blast software at, the result showed that the amplified fragment was in accordance with the GenBank recorded data.


Figure 2. Electrophoretic analysis result of reverse transcriptase polymerase chain reaction (RT-PCR) products (M = marker).

Download figure to PowerPoint

Real-time RT-PCR analysis

Standard curves were established to examine the efficiency of real-time PCR. Figure 3 indicated that the efficiency and correlation index (r2) were >0.95.


Figure 3. Standard curves of real-time reverse transcriptase polymerase chain reaction (RT-PCR) analysis of C1GALT1, Cosmc and GAPDH.

Download figure to PowerPoint

The Ct ratio (Ct target/Ct internal calibrator) of IgAN patients, non-IgAN patients and normal control were calculated for statistical analysis in order to exclude influence of loading volume difference. anova analysis indicated that significant differences could be observed when comparing IgAN patients with non-IgAN patients and normal controls in the expression level of Cosmc, but not in C1GALT1 (Table 3).

Table 3. β1,3GT and Cosmc expression level in IgAN patients and controls
 IgAN patients (n = 41)Non-IgAN patients (n = 21)Normal control (n = 26)All controla (n = 47)
  1. aNon-IgAN patients and normal controls added together to be all controls.

  2. *Significantly different (P < 0.05).

  3. No significant difference was observed in the C1GALT1 expression (P > 0.05).

  4. IgAN, immunoglobulin A nephropathy; β1,3GT, β1,3-galactosyltransferase.

Ct ratio target/GAPDH (mean ± SD)
 C1GALT11.22 ± 0.361.11 ± 0.131.11 ± 0.081.11 ± 0.1
 Cosmc1.36 ± 0.31*1.22 ± 0.18*1.24 ± 0.10*1.23 ± 0.14*

Ratios of target gene expression between IgAN patients and controls were calculated according to the mathematical model of relative quantification in real-time RT-PCR developed by Pfaffl [14]. Using this method relative quantification in real-time RT-PCR of a target gene transcript was measured in comparison with a reference gene transcript. The relative expression ratio is calculated only from the real-time PCR efficiencies and Ct deviation of an unknown sample versus a control. Eqn 1 shows the mathematical model of relative expression ratio in real-time PCR.

  • image(1)

Etarget and Ereference are the real-time PCR efficiencies of target gene (C1GALT1 and Cosmc) and reference gene (GAPDH). Real-timePCR efficiency was calculated according to eqn 2, the slope was measured as illustrated in Fig. 3.

  • image(2)

As anova analysis showed no difference in the target gene expression level between non-IgAN patients and normal controls, therefore they were treated as one control group in the following analysis. The relative quantification results of C1GALT1 and Cosmc gene expression were listed in Table 4. As could be seen, expression levels of these two genes were remarkably downregulated in IgAN patients.

Table 4.  The C1GALT1 and Cosmc expression ratio between immunoglobulin A nephropathy (IgAN) patients and controls
Control (mean)22.7725.1527.64
Patient (mean)23.3327.3230.81
E(target)ΔCT 0.1600.091
E(reference)ΔCT 0.6880.688
Absolute gene regulation 0.2330.132

Correlation between VV lectin-binding OD and target gene expression

Pearson correlation analysis between OD results of VV lectin-binding assay and Cosmc expression level (GAPDH/Cosmc Ct ratio) showed a significant (P =0.026) but weak negative correlation (r = −0.24; Fig. 4). However, no correlation with C1GALT1 was noticed.


Figure 4. Correlation between VV lectin-binding assay and Cosmc expression.

Download figure to PowerPoint

Relationship between β1,3GT and Cosmc expression and clinical manifestations

We analysed the relationship between target gene expression level and clinical manifestations of IgAN patients, such as severe proteinuria, hypertension, renal dysfunction and gross haematuria. As these manifestations have already been reported to be independent prognostic indicators, we expected to find the evidence to support using β1,3GT and its chaperone expression levels as indices of IgAN severity and prognosis.

β1,3GT and Cosmc expression with severe proteinuria

Thirteen IgAN patients included had severe proteinuria (>3.5 g per 24 h) and were diagnosed clinically as Nephropathy Syndrome. Real-time RT-PCR suggested that compared to patients without severe proteinuria expression levels of C1GALT1 and Cosmc in these patients were dramatically lower (Table 5).

Table 5.  The β1,3GT and Cosmc expression in IgAN patients with/without severe proteinuria
 Patients with severe proteinuria (n = 13)
C1GALT1/GAPDH (1.41 ± 0.37)Cosmc/GAPDH (1.51 ± 0.52)
  1. IgAN, immunoglobulin A nephropathy; β1,3GT, β1,3-galactosyltransferase.

Patients without severe proteinuria (n = 28)1.12 ± 0.08 (P = 0.01; ratio: 0.05)1.29 ± 0.08 (P = 0.03; ratio: 0.20)
Controls (n = 47)1.11 ± 0.08 (P = 0.001; ratio: 0.04)1.22 ± 0.10 (P = 0.001; ratio: 0.05)

Pearson correlation assay between 24 h urine protein amount (UPA) and target gene expression level (GAPDH/target gene Ct ratio) showed significant but weak correlations with C1GALT1 (r =−0.375, P = 0.016) and Cosmc (r = −0.378, P =0.015; Fig. 5). These results showed a relationship between severe proteinuria (mainly manifested as Nephropathy Syndrome) and expression levels of C1GALT1 as well as Cosmc.


Figure 5. Correlation between 24 h urine protein amount (UPA) and C1GALT1, Cosmc expression.

Download figure to PowerPoint

β1,3GT and Cosmc expression with gross haematuria

In the 41 IgAN patients included, repeated gross haematuria was reported in 20 cases. No difference in C1GALT1 (1.30 ± 0.5 vs. 1.13 ± 0.09, P = 0.14) as well as Cosmc (1.42 ± 0.44 vs. 1.31 ± 0.09, P = 0.27) expression levels between patients with or without repeated gross haematuria was noticed. These results indicated that β1,3GT and its chaperone expression were not related to repeated gross haematuria in IgAN patients.

β1,3GT and Cosmc expression with renal dysfunction

Only six IgAN patients included in the present study had renal dysfunction before biopsy. Serum creatinine levels were between 135 and 249 μmol L−1, which indicated early stage of renal dysfunction. Because of the limited sample size, no effective statistical analysis could be performed. But Pfaffl's [14] analysis suggested that the Cosmc expression level was remarkably lower in these patients, just 3.7% and 22% of controls and patients with normal renal functions respectively. Increased Scr levels were noticed in four non-IgAN patients (140–160 μmol L−1); Pfaffl's [14] quantification found that Cosmc expression was 20% of controls. Pearson correlation analysis showed no significant correlation between Scr level and C1GALT1 or Cosmc expression level.


  1. Top of page
  2. Abstract.
  3. Introduction
  4. Design, setting and methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References

O-glycans, an important motif of many glycoproteins, can be classified into several different groups according to core structures [15]. Core 1 Galβ1–3GalNAcα1-serine/threonine, is the major constituent of O-glycan in IgA1 molecules. Core 1 β1,3-galactosyltransferases (C1Gal-Ts) transfer a galactose from UDP-Gal to a GalNAc residue on proteins with a β1,3-linkage to synthesize the core 1 structure. Recently, C1Gal-T gene had been cloned and named as C1GALT1, which locates on chromosome 7p14-p13 and has a 1794 bp cDNA sequence encoding a protein of 363 amino acids [6]. A special protein associated with C1GALT1 and is required for its activity had been cloned and termed as Cosmc (core 1 β3-Gal-T-specific molecular chaperone), which is mapped to chromosome Xq23 and has a 1471 bp cDNA sequence encoding a protein of 318 amino acids [7, 8]. Cosmc may relate with the folding intermediates of C1GALT1. They may form complexes such as Cosmc(C)-inactive C1GALT1 (U)-active C1GALT1 (A) and CUUC. After potential rounds of binding and dissociation between Cosmc and enzyme, stable active forms of the enzyme, either dimeric (AA) or monomeric (A) may be generated and perform its glycosylation function.

The IgAN is one of the diseases reported to be related to C1GALTs. Allen et al. studied the re-glycosylation capacity of leucocytes lysate on degalactosylated sera protein fragment, and found a decreased synthesis activity of β1,3GT, which positively correlated with glycosylation aberration degree, in the B-lymphocytes lysate of IgAN patients [5]. They suspected that the decrease may result from lower gene expression or post-translational modifications. However, considering that there is no significant post-translational modification of the C1GALTs [6], the expression level abnormality was highly suspected to be the underlying mechanism of decreased synthesis activity was observed. Unfortunately, their study was just a preliminary investigation in the β1,3GT function in IgAN. As the only integrated β1,3GT glycosylation ability reflected by cytolysate re-glycosylation capacity, which was indirect and could be affected by many factors other than expression levels, was measured, they could not tell specifically whether the decreased activity was due to C1GALTs or Cosmc downregulation and the degree of it. In order to obtain a clear picture of β1,3GT and its chaperone expression, we determined the level of C1GALT1 and Cosmc expression respectively.

In the present study, we found that the β1,3GT chaperone Cosmc expression level in peripheral B lymphocytes of IgAN patients (Ct ratio: 1.36 ± 0.31) was significantly lower than that of non-IgAN glomerulonephritis patients (Ct ratio: 1.22 ± 0.18) and normal controls (Ct ratio: 1.24 ± 0.10; 87% decrease compared with controls), and the expression level was negatively correlated with the level of glycosylation aberrance (P = 0.026, r = −0.239). Whilst there is no apparent discrepancy of C1GALT1 expression level between IgAN patients and controls. With respect to the findings of Ju and Kudo that in cell lines which express C1GALT1 only (LSC-G1Gal-T1 and Hi-5+C1β3Gal-T respectively) the β1,3GT activity was very low (only 2% to that of cell line express Cosmc in Kudo et al.'s study) [7, 8], our results suggest that it is the decreased expression of Cosmc that causes the lower of β1,3GT activity and IgA1 glycosylation disorder in IgAN patients. Although the C1GALT1 expression level is normal in these patients, yet without the coexistence of Cosmc the C1GALT1 cannot exert its glycosylation ability normally. It was reported that the lack of C1GALT activity in some cell lines was due to mutations in Cosmc gene: in LSC line there is a T insertion leads to termination of translation; and in Jurkat cells there are a missense C-T mutation and a T deletion leads to termination of translation. It is postulated that these mutations genetically inherited or occurring randomly in precursor stem cell might be associated with decreased β1,3GT activity in B cells responsible for IgA1 production. And the location of Cosmc on X chromosome is also in accordance with the finding that IgAN exhibits a 2 : 1 to 6 : 1 male predominance [1], which supports mutation hypothesis. We have already started to sequence the genes in order to address this issue. But, till now, no mutation has been identified in human being.

Our data showed that the expression level of C1GALT1 and Cosmc genes were remarkably decreased in patients with severe proteinuria when compared to patients without severe proteinuria and controls, and there was a significant correlation between 24 h UPA and the expression level of C1GAL1 as well as Cosmc. Considering that severe proteinuria is a poor prognosis indicator of IgAN, the downregulated gene expression may also be a prognosis index. Although many efforts have been exerted to establish a good evaluation criterion of IgAN, there still lacks a proper index to diagnosis, monitoring and prognosis judgement of IgAN. Conventional indices such as severity of proteinuria, haematuria, hypertension, oedema and renal dysfunction are low in sensitivity and specificity. And regarding the limitation and proper risk, pathological biopsy is also unfit for the continuous monitoring of diseases progress. Our results suggested a potentiality of using Cosmc expression level as such an index to evaluate IgAN patients, whilst a large size and long-term clinical study are needed to confirm this assumption.

Besides IgAN, idiopathic Tn syndrome is also characterized by aberrant expression of bare GalNAc O-linked glycans on red blood cell (RBC) and platelet, which was speculated to be result from stem cell deficiency of β1,3GT activity. Considering that treatment with 5-azacytidine (a potent inhibitor of DNA methylation) and NaB (an inhibitor of histone deacetylation) could reactive the defected β1,3GT activity in Tn syndrome, it was speculated that repression of both or one of the β1,3GT gene copies may be the underlying mechanism of Tn phenotype [16]. As our findings are consistent with their speculations, treatment using 5-azacytidine and NaB or drugs with similar pharmacological actions could be a new therapeutical method to IgAN.

We suspected that decrease expression of Cosmc, caused by mutations or external depression, will affect the activity of C1GALT1 and leads to insufficiency of glycosylation activity. Under-glycosylation may (i) cause IgA1‘self-aggregate’ and increase the affinity towards mesangial cells and extracellular matrix (ECM) of IgA molecules [17]; (ii) change receptor-binding affinity (decrease in ASGPR [18, 19] increase in transferrin receptor CD71 [20]); (iii) expose pathogenic domains and forming IgA–IgG complex [21–23]; (iv) cause increasing of C3 activation capability [24]; (v) affect the proliferation, apoptosis of mesangial cells [25]; as well as cytokines and inflammation processes [26–28], and all above-mentioned factors eventually contribute to the pathological changes of IgAN.

We found a significant but weak negative correlation between Cosmc mRNA expression level and IgA glycosylation abnormality. This low correlation may be due to the following causes: (i) because the antibodies needed to study the protein levels are unavailable at present, only mRNA levels were measured in the present study. Although normally there is a good correlation between protein and mRNA levels, further studies were needed to elucidate the protein expression level when conditions allowed; (ii) researchers thought there might be more than just two members in the family of β1,3GT [15], therefore C1GALT1 and Cosmc may only play a partial role in the IgA glycosylation process; (iii) the majority of human IgA is produced by plasma cells, which are rarely found in peripheral blood and technically inaccessible. As Allen et al.'s study indicated a strong correlation between peripheral B lymphocyte β1,3GT activity and VV lectin-binding level [5], therefore it was reasonable to use peripheral B lymphocytes, precursor of plasma cells, in the present study. However, although a close relationship could be expected between these two kinds of cells, further studies were needed to clarify the potential error brought forward by this.

In conclusion, Cosmc expression level in peripheral B lymphocytes of IgAN patients is significantly decreased when compared with non-IgAN patients and normal controls, which is also negatively correlated with the IgA glycosylation degree. Decrease of C1GALT1 and Cosmc gene expression levels are related to the severity of proteinuria in IgAN patients, which suggested a potential prognosis indication property. Genetic mutation may be the cause of downregulated expression, which needs further study. Investigation of the actual reason underlying the expression deficiency may shed light on the mechanism of IgAN and help to find new therapy of this disease.


  1. Top of page
  2. Abstract.
  3. Introduction
  4. Design, setting and methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References

Sincere thanks are due to Dr Qingjie Xia, Yingjie Wang and Xiaohong Ma for their help. This work was partly supported by Scientific Research Fund of Sichuan Traditional Chinese Medicine Administration Bureau, 2004A63.


  1. Top of page
  2. Abstract.
  3. Introduction
  4. Design, setting and methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References
  • 1
    Floege J , Feehally J . IgA nephropathy: recent developments. J Am Soc Nephrol 2000; 11: 2395403.
  • 2
    Hiki Y , Tanaka A , Kokubo T et al. Analyses of IgA1 hinge glycopeptides in IgA nephropathy by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. J Am Soc Nephrol 1998; 9: 57782.
  • 3
    Allen AC , Bailey EM , Brenchley PE et al. Mesangial IgA1 in IgA nephropathy exhibits aberrant O-glycosylation: observations in three patients. Kidney Int 2001; 60: 96973.
  • 4
    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: 48696.
  • 5
    Allen AC , Topham PS , Harper SJ et al. Leucocyte beta 1,3-galactosyltransferase activity in IgA nephropathy. Nephrol Dial Transplant 1997; 12: 7016.
  • 6
    Ju T , Brewer K , D'Souza A et al. Cloning and expression of human core 1 beta1,3-galactosyltransferase. J Biol Chem 2002; 277: 17886.
  • 7
    Ju T , Cummings RD . A unique molecular chaperone Cosmc required for activity of the mammalian core 1 beta 3-galactosyltransferase. Proc Natl Acad Sci U S A 2002; 99: 166138.
  • 8
    Kudo T , Iwai T , Kubota T et al. Molecular cloning and characterization of a novel UDP-Gal:GalNAc(alpha) peptide beta 1,3-galactosyltransferase (C1Gal-T2), an enzyme synthesizing a core 1 structure of O-glycan. J Biol Chem 2002; 277: 4772431.
  • 9
    Lee SM . Prognostic indicators of progressive renal disease in IgA nephropathy: emergence of a new histologic grading system. Am J Kidney Dis 1997; 29: 9538.
  • 10
    Haas M . Histologic subclassification of IgA nephropathy: a clinicopathologic study of 244 cases. Am J Kidney Dis 1997; 29: 82942.
  • 11
    Whitworth JA . 2003 World Health Organization (WHO)/International Society of Hypertension (ISH) statement on management of hypertension. J Hypertens 2003; 21: 198392.
  • 12
    Norrback KF , Dahlenborg K , Carlsson R et al. Telomerase activation in normal B lymphocytes and non-Hodgkin's lymphomas. Blood 1996; 88: 2229.
  • 13
    Allen AC , Willis FR , Beattie TJ et al. Abnormal IgA glycosylation in Henoch-Schonlein purpura restricted to patients with clinical nephritis. Nephrol Dial Transplant 1998; 13: 9304.
  • 14
    Pfaffl MW . A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 2001; 29: e45.
  • 15
    Hennet T . The galactosyltransferase family. Cell Mol Life Sci 2002; 59: 108195.
  • 16
    Thurnher M , Rusconi S , Berger EG . Persistent repression of a functional allele can be responsible for galactosyltransferase deficiency in Tn syndrome. J Clin Invest 1993; 91: 210310.
  • 17
    Kokubo T , Hiki Y , Iwase H et al. Protective role of IgA1 glycans against IgA1 self-aggregation and adhesion to extracellular matrix proteins. J Am Soc Nephrol 1998; 9: 204854.
  • 18
    Roccatello D , Picciotto G , Torchio M et al. Removal systems of immunoglobulin A and immunoglobulin A containing complexes in IgA nephropathy and cirrhosis patients. The role of asialoglycoprotein receptors. Lab Invest 1993; 69: 71423.
  • 19
    Basset C , Devauchelle V , Durand V et al. Glycosylation of immunoglobulin A influences its receptor binding. Scand J Immunol 1999; 50: 5729.
  • 20
    Moura IC , Centelles MN , Arcos-Fajardo M et al. Identification of the transferrin receptor as a novel immunoglobulin (Ig)A1 receptor and its enhanced expression on mesangial cells in IgA nephropathy. J Exp Med 2001; 194: 41725.
  • 21
    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: 50916.
  • 22
    Tomana M , Julian BA , Waldo FB et al. IgA nephropathy. A disease of incomplete IgA 1 glycosylation? Adv Exp Med Biol 1995; 376: 221.
  • 23
    Kokubo T , Hashizume K , Iwase H et al. Humoral immunity against the proline-rich peptide epitope of the IgA1 hinge region in IgA nephropathy. Nephrol Dial Transplant 2000; 15: 2833.
  • 24
    Nikolova EB , Tomana M , Russell MW . The role of the carbohydrate chains in complement (C3) fixation by solid-phase-bound human IgA. Immunology 1994; 82: 3217.
  • 25
    Amore A , Cirina P , Conti G et al. Glycosylation of circulating IgA in patients with IgA nephropathy modulates proliferation and apoptosis of mesangial cells. J Am Soc Nephrol 2001; 12: 186271.
  • 26
    Gomez-Guerrero C , Gonzalez E , Hernando P et al. Interaction of mesangial cells with IgA and IgG immune complexes: a possible mechanism of glomerular injury in IgA nephropathy. Contrib Nephrol 1993; 104: 12737.
  • 27
    Peruzzi L , Amore A , Cirina P et al. Integrin expression and IgA nephropathy: in vitro modulation by IgA with altered glycosylation and macromolecular IgA. Kidney Int 2000; 58: 233140.
  • 28
    Amore A , Conti G , Cirina P et al. Aberrantly glycosylated IgA molecules downregulate the synthesis and secretion of vascular endothelial growth factor in human mesangial cells. Am J Kidney Dis 2000; 36: 124252.