- Email: [email protected]
High galactosylation of oligosaccharides in umbilical cord blood IgG, and its relationship to placental function
Clinica Chimica Acta 299 (2000) 169–177 www.elsevier.com / locate / clinchim
High galactosylation of oligosaccharides in umbilical cord blood IgG, and its relationship to placental function a, b c Satoshi Kimura *, Masahide Numaguchi , Tokio Kaizu , c a a Donghyun Kim , Yasushi Takagi , Kunihide Gomi a
Department of Clinical Pathology, Showa University School of Medicine, Shinagawa-ku, Tokyo 142 -8666, Japan b Department of Gynecology and Obstetrics, Dokkyo University Koshigaya, Hospital, Koshigaya-shi, Saitama 343 -8555, Japan c Genzyme Japan K.K., 333 Yabukicho, Shinjuku-ku, Tokyo 162 -0801, Japan Received 17 December 1999; received in revised form 3 April 2000; accepted 12 April 2000
Abstract N-linked oligosaccharides on human serum IgGs have been reported to modulate IgG function. We studied umbilical cord blood to determine whether neonatal IgGs have characteristic structures related to developmental and pathological status. Oligosaccharide patterns of serum IgG from 45 umbilical cord blood samples were characterized by HPLC, and compared with those of serum IgG from 11 normal adults. Oligosaccharyl amines from purified IgG were released by recombinant N-glycanase, labeled with fluorescence reagent FMOC (9-fluorenylmethyl chloroformate), and analyzed quantitatively by high-pressure liquid chromatography (HPLC). Increased galactosylation was observed in cord blood. The ratio of galactosylated to non-galactosylated oligosaccharides on IgG was 7.9063.92 (mean6S.D.) in cord blood, significantly higher than the ratio in adults (1.6060.62, P , 0.0001). There were weak but not significant correlations between the ratio and birth weight, gestation period, mother’s age, and no correlation with serum IgG concentration. The ratio was lower for premature or intra-uterine growth retarded neonates. Our results, in conjunction with previous reports that galactosylated IgG stimulates Fc-mediated phagocytosis of monocytes, suggest that increased galactosylation of IgG enhances neonatal immunity. 2000 Elsevier Science B.V. All rights reserved. Keywords: Cord blood; IgG; Carbohydrate; Galactosylation; Placental function
*Corresponding author. Tel.: 1 81-3-3784-8577; fax.: 1 81-3-3788-4927. E-mail address: [email protected] (S. Kimura) 0009-8981 / 00 / $ – see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0009-8981( 00 )00289-8
170
S. Kimura et al. / Clinica Chimica Acta 299 (2000) 169 – 177
1. Introduction N-linked oligosaccharides in human IgG are known to modulate immunological function. Complex oligosaccharide moieties (biantennary mannose structure, with or without galactose on N-acetylglucosamine outer arms) are attached to asparagine-297 of gamma heavy chains in human IgG molecules, and help maintain the three-dimensional structure [1]. Carbohydrate moieties help stabilize the IgG molecule [2], and prolong its half-life in serum [3]. Carbohydrates modulate antigen binding of IgG [4], and Fc-mediated phagocytosis [5]. Human neonates suffer fewer infectious diseases than do infants older than 3 months, because they receive maternal immunoglobulins (particularly IgG) via the placenta [6]. However, the average half-life of adult serum IgG is less than a month, which does not account for the duration of strong immunity in neonates. It is possible that IgGs present in neonates have distinctive structures which prolong their half-life. We characterized, for the first time, carbohydrate structures on neonatal serum IgGs in comparison to adult serum IgGs.
2. Materials and methods
2.1. Sample collection Umbilical cord blood provides a convenient, non-invasive means of studying neonatal serum IgG. Cord blood samples (n 5 45; 24 males and 21 females) were collected during deliveries at Dokkyo University Koshigaya Hospital, with the parents’ written consent. Ranges and mean6S.D. values were 25–41 weeks (38.963.0) for gestation period, 860–3824 g (29146641) for birth weight, 495–2040 mg / dl (13376346) for cord blood IgG level, and 21–39 years (30.064.4) for age of mother. Blood samples were collected at Showa University Hospital from 11 healthy adult male volunteers (age range 23–40; 26.965.6), whose serum aminotransferases, Na, K, Cl, total protein, direct bilirubin, total cholesterol, and triglyceride levels were all within normal range.
2.2. HPLC analysis FMOC-Cl, urea, and 1,4-dioxane were from Sigma Chemical, St. Louis, MO, USA. Recombinant N-glycanase (E.C.3.5.1.52) was from Genzyme, Cambridge, MA, USA. Orcinol, dithiothreitol, ethylenediaminetetraacetate, iodoacetic acid, acetonitrile (HPLC grade), sodium monobasic phosphate, sodium dibasic phosphate, and dichloromethane were from Wako Pure Chemical, Osaka, Japan.
S. Kimura et al. / Clinica Chimica Acta 299 (2000) 169 – 177
171
Neuraminidase (E.C.3.2.1.18) was from Nacalai Tesque, Kyoto, Japan. TSKGEL Amide 80 column was from Tosoh K.K., Tokyo, Japan. To isolate IgG proteins, 500 ml of serum was passed through a Protein A-conjugated sepharose CL-4B column equilibrated with 0.05 mol / l Tris–HCl, pH 8.2, containing 0.15 mol / l NaCl and 0.02% NaN 3 . The column was rinsed with the same buffer, and eluted with 50 mmol / l sodium citrate buffer containing 0.15 mol / l NaCl and 0.02% NaN 3 , pH 3.5. Purified IgG was dissolved in 1.0 ml of 20 mmol / l acetic acid pH 5.0, and incubated with 250 mU of neuraminidase at 378C for 24 h. The neuraminidase treatment, although not absolutely necessary, gave simpler peak patterns for HPLC analysis. Preparation of reduced, carboxymethylated IgG proteins, and release of (oligosaccharyl) amines from these proteins by recombinant N-glycanase treatment, were performed as described previously [7]. To derive oligosaccharyl amines, a portion ( | 50 mg) of each of the samples was dissolved in 800 ml of 20 mmol / l sodium phosphate buffer, pH 8.6, containing 50% ethanol. To 800 ml of this solution, 80 ml of FMOC (49 mg of FMOC-Cl in 4 ml of 1,4-dioxane) was added, and incubated at room temperature for 1 h. The solution was added to 800 ml of dichloromethane, mixed well, and centrifuged. The supernatant was applied to the HPLC. FMOC-amino-oligosaccharide derivatives were separated on Amide 80 column (4.6 3 250 mm, particle size 5 mm) by HPLC with a linear elution gradient (25 to 49%, v / v) of aqueous acetonitrile within 40 min, flow-rate 1.0 ml / min. Peaks were recorded and integrated by a Hitachi D-2500 Chromato-integrator. The HPLC system was a Waters 600E system controller, with JASCO 821-FP Intelligent Spectrofluorometer. FMOC-amino-oligosaccharide derivatives (1 nmol) from human serum IgG were digested with 10 mU of beta-galactosidase (Seikagaku, Tokyo, Japan) in 100 ml of 50 mmol / l sodium acetate buffer, pH 6.0, at room temperature for 10 min. Statistical analysis was by Student’s t-test.
3. Results Statistical data (sex, birth weight, gestation period, mother’s age) and analytical results for the 45 cord blood samples are summarized in Table 1.
3.1. Galactose content between neonates and adults FMOC-amino-oligosaccharide derivatives from two different serum IgG samples were separated quantitatively by HPLC (Figs. 1A and B). Peak 3
S. Kimura et al. / Clinica Chimica Acta 299 (2000) 169 – 177
172
Table 1 Statistical data and analytical results for 45 umbilical cord blood samples
Mean median S.D.
Birth weight (g)
Gestation period (weeks)
Mother’s age
IgG (mg / dl)
%G0
%G1
%G2
G2 / G0 ratio
2914 3122 641
38.9 39.6 3.0
30.0 31.0 4.4
1337 1360 346
9.0 7.8 4.2
33.5 33.6 6.5
57.5 59.5 8.6
7.90 7.46 3.92
isolated from cord blood was digested with beta-galactosidase to remove non-reducing terminal galactose residues, and the reaction mixture was reexamined by HPLC under the same conditions (Fig. 1C). As shown in Figs. 1B and C, peak 3 shifted to peak 1 after beta-galactosidase digestion, indicating that galactose residues in the non-reducing terminal of the oligosaccharide molecule were eluted in peak 3. In adults, peak 1 (non-galactosyl oligosaccharides; G0) represented 25.466.2% (mean6S.D.) of oligosaccharides from the IgG molecule, while peak 2 (G1) and peak 3 (G2) represented 36.567.0 and 37.867.8%, respectively. In neonates, on the other hand, G0, G1 and G2 took 9.064.2, 33.566.5 and 57.568.6% of the IgG molecule, respectively. According to these results, %G0 was significantly higher in adults, whereas %G2 was higher in neonates (P , 0.001 in both cases). The ratio of G2 to G0 (G2 / G0 ratio) was significantly higher in neonates (7.9063.92) than in adults (1.6060.62; P , 0.0001) (Fig. 2A). Similarly, G1 / G0 ratio in neonates (4.3861.94) was higher than in adults (1.5360.53; P , 0.0001) (Fig. 2B).
3.2. Galactose content among neonates A high G2 / G0 ratio was observed even in a premature case (gestation period 25 weeks, birth weight 860 g, G2 / G0 ratio 5.91), although this neonate’s IgG content was only 532 mg / dl, | 40% of the typical level for full-term deliveries. There were weak but not significant correlations between the G2 / G0 ratio and birth weight (r 5 0.219), gestation period (r 5 0.207), mother’s age (r 5 2 0.272), and no correlation with serum IgG concentration (r 5 0.063) (Fig. 3). The G2 / G0 ratio was 4.9662.57 for intra-uterine growth retardation (IUGR) babies (n 5 6), lower than for non-IUGR babies (n 5 39) (8.3663.91; P , 0.05). G2 / G0 ratio was 8.0664.06 for full-term (n 5 41) vs. 6.3361.11 for premature infants (n 5 4), although this difference was not statistically significant because of small sample size.
S. Kimura et al. / Clinica Chimica Acta 299 (2000) 169 – 177
173
Fig. 1. HPLC elution patterns of FMOC-labeled oligosaccharides from a healthy adult (A), a neonate (B), and neonatal peak 3 after galactosidase digestion (C). Elution times of oligosaccharides ranged from 24.1 to 36.5 min. (A) Adult. The three major peaks represent nongalactosyl (peak 1; G0), and galactosyl structures (peaks 2 and 3, G1 and G2, respectively) attached to a mannose core biantennary structure. The G2 / G0 ratio is 1.09, and G1 / G0 ratio is 0.73. (B) Neonate. Peak 3 is much larger and peak 1 smaller in comparison to the adult pattern. This neonate had a gestation period of 37.6 weeks and birth weight of 2450 g. The G2 / G0 ratio is 10.74, and the G1 / G0 ratio is 6.34. (C) Peak 3 from the neonate, after purification and beta-galactosidase digestion, shifted to the location of peak 1 in panel B.
174
S. Kimura et al. / Clinica Chimica Acta 299 (2000) 169 – 177
Fig. 2. Comparison of the G2 / G0 and G1 / G0 ratio between neonates and adults. The ratio of peak 3 to peak 1 (G2 / G0, see Fig. 1) ranged from 1.58 to 20.56 in neonates, and 0.84–2.79 in adults. Mean value for neonates (7.9063.92) was significantly higher than for adults (1.6060.62, P , 0.0001). The ratio of peak 2 to peak 1 (G1 / G0) was also higher in neonates than in adults.
S. Kimura et al. / Clinica Chimica Acta 299 (2000) 169 – 177
175
Fig. 3. Regression of G2 / G0 ratio against serum IgG concentration.
4. Discussion This is the first report on oligosaccharide profiles of neonatal serum IgGs. Parekh et al. [8] and Shikata et al. [9] investigated age-related galactosylation patterns of N-linked oligosaccharides in human serum IgGs in subjects ranging from one to 73 years old. Parekh et al. found that the proportion of nongalactosylated oligosaccharide was highest (27.3% of total galactosyl structures) in a 1-year-old infant and lowest (21.2%) in a 25-year-old subject. In our study, non-galactosylated oligosaccharide accounted for 9.064.2% (mean6S.D.) of total oligosaccharides in neonates, compared to 25.466.2% in adults (P , 0.0001). The G2 / G0 ratio was significantly higher in cord blood IgGs than in adult IgGs. Our results are consistent with those of Parekh et al., and there appears to be little difference between infants and adults. We therefore conclude that cord blood IgGs are more highly galactosylated than infant or adult IgGs. Increased galactosyltransferase activity has been reported in pregnant women [10], consistent with our finding of increased galactosylation of cord blood IgG. In neonates, we must also consider the half-lives of galactosylated and sialylated
176
S. Kimura et al. / Clinica Chimica Acta 299 (2000) 169 – 177
IgGs. In a rat model, clearance of IgG proteins from the bloodstream was prolonged for animals with N-linked oligosaccharides [3]. There are many reports on changes of carbohydrate structures in IgG molecules as correlated with disease status [1,11,12]. Glycosylation status in relation to IgG subclass was studied by Keusch et al. [13]. Sugar chains, particularly glycosylated N-linked oligosaccharides, may hinder the degradation of IgG proteins, and play a role in maintaining steric structure [13,14]. This preserves the active sites present in these domains, and allows more effective binding to C1q and Fc-receptors [15,16]. Fc receptor-mediated phagocytosis of macrophages, and feedback immunosuppression, are modified by the IgG carbohydrate structure [5,17,18]. These previous findings, in conjunction with ours, suggest that increased galactosylation of IgG proteins prolongs their half-life in neonatal blood, and strengthens the neonatal immune system by maintaining immunity of maternal origin. Maternal IgG is transferred to the fetus via active transport through the placenta [6]. Decreased galactosylation of IgG in IUGR neonates may reflect placental dysfunction. Neonates and mothers are often exposed to infectious agents during delivery. Higher galactosylation and sialylation of carbohydrate structures on serum IgGs may reduce the risk of infection, or enhance transport of IgG through the placenta. Further studies are needed to assess these possibilities.
Acknowledgements This work was supported by Grants-in-aid for Scientific Research (A) 07772258, (A) 08772145, and (A) 09772071 (to S.K.) from the Ministry of Education, Science, Culture, and Sports of Japan. The authors thank Dr. Stephen Anderson for English assistance.
References [1] Kobata A. Function and pathology of the sugar chains of human immunoglobulin G. Glycobiology 1990;1:5–8. [2] Hobbs SM, Jackson LE, Hoadley J. Interaction of aglycosyl immunoglobulins with the IgG Fc transport receptor from neonatal rat gut:comparison of deglycosylation by tunicamycin treatment and genetic engineering. Molec Immunol 1992;29:949–56. [3] Wawrzynczak E, Cumber A, Parnell G, Jones P, Winter G. Blood clearance in the rat of a recombinant mouse monoclonal antibody lacking the N-linked oligosaccharide side chains of the CH2 domains. Molec Immunol 1992;29:213–20. [4] Fishman AJ, Fucello AJ, Pellegrino-Gensey JL. Effect of carbohydrate modification on the localization of human polyclonal IgG at focal sites of bacterial infection. J Nucl Med 1992;33:1378–82.
S. Kimura et al. / Clinica Chimica Acta 299 (2000) 169 – 177
177
[5] Fukushima K, Takasaki S. Processing inhibition of N-linked sugar chains associated with induction of Fc receptor-mediated phagocytosis in the mouse monocytoid cells. Glycobiology 1993;3:15–22. [6] Gitlin D, Kumate J, Urrusti J, Morales C. The selectivity of the human placenta in transfer of plasma proteins from mother to fetus. J Clin Invest 1964;43:938–51. [7] Kim D, Rasmussen J, Kimura S, Kaizu T. Application of FMOC-Cl for the quantitative determination of 1-amino-oligosaccharides in the low range. Glycoconjugate J 1995;12:408. [8] Parekh R, Roitt I, Isenberg D, Dwek R, Radenmacher T. Age-related galactosylation of the N-linked oligosaccharides of human serum IgG. J Exp Med 1988;167:1731–6. [9] Shikata K, Yasuda T, Takeuchi F, Konishi T, Nakata M, Mizuochi T. Structural changes in the oligosaccharide moiety of human IgG with aging. Glycoconjugate J 1998;15:683–9. [10] Lacord-Bonneau M, Khalfoun B, Weill JD, Bardos P. Long-term study of serum glycosyltransferase levels in pregnant women and their newborn infants. Int J Biochem 1988;20:997– 1000. [11] Bond A, Alavi A, Axford JS et al. A detailed lectin analysis of IgG glycosylation, demonstrating disease specific changes in terminal galactose and N-acetylglucosamine. J Autoimmun 1997;10:77–85. [12] Malhotra R, Wormald MR, Rudd PM, Fischer PB, Dwek RA, Sim RB. Glycosylation changes of IgG associated with rheumatoid arthritis can activate complement via the mannose-binding protein. Nature Med 1995;1:237–43. [13] Keusch J, Levy Y, Shoenfeldt Y, Youinou P. Analysis of different glycosylation states in IgG subclasses. Clin Chim Acta 1996;252:147–58. [14] Nardella F, Teller D, Mannik M. Studies on the antigenic determinants in the self-association of IgG rheumatoid factor. J Exp Med 1981;154:112–25. [15] Henny C, Stanworth D. Reaction of rheumatoid factor with the isolated polypeptide chains of human 7s gamma-globulin. Nature 1964;201:511–2. [16] Koide N, Nose M, Muramatsu T. Recognition of IgG by Fc receptor and complement: effects of glycosidase digestion. Biochem Biophys Res Commun 1977;75:838–44. [17] Nose M, Wigzell H. Biological significance of carbohydrate chains on monoclonal antibodies. Proc Natl Acad Sci USA 1983;80:6632–6. [18] Heyman B, Nose M, Weigle W. Carbohydrate chains on IgG2b: a requirement for efficient feedback immunosuppression. J Immunol 1985;134:4018–23.
High galactosylation of oligosaccharides in umbilical cord blood IgG, and its relationship to placental function a, b c Satoshi Kimura *, Masahide Numaguchi , Tokio Kaizu , c a a Donghyun Kim , Yasushi Takagi , Kunihide Gomi a
Department of Clinical Pathology, Showa University School of Medicine, Shinagawa-ku, Tokyo 142 -8666, Japan b Department of Gynecology and Obstetrics, Dokkyo University Koshigaya, Hospital, Koshigaya-shi, Saitama 343 -8555, Japan c Genzyme Japan K.K., 333 Yabukicho, Shinjuku-ku, Tokyo 162 -0801, Japan Received 17 December 1999; received in revised form 3 April 2000; accepted 12 April 2000
Abstract N-linked oligosaccharides on human serum IgGs have been reported to modulate IgG function. We studied umbilical cord blood to determine whether neonatal IgGs have characteristic structures related to developmental and pathological status. Oligosaccharide patterns of serum IgG from 45 umbilical cord blood samples were characterized by HPLC, and compared with those of serum IgG from 11 normal adults. Oligosaccharyl amines from purified IgG were released by recombinant N-glycanase, labeled with fluorescence reagent FMOC (9-fluorenylmethyl chloroformate), and analyzed quantitatively by high-pressure liquid chromatography (HPLC). Increased galactosylation was observed in cord blood. The ratio of galactosylated to non-galactosylated oligosaccharides on IgG was 7.9063.92 (mean6S.D.) in cord blood, significantly higher than the ratio in adults (1.6060.62, P , 0.0001). There were weak but not significant correlations between the ratio and birth weight, gestation period, mother’s age, and no correlation with serum IgG concentration. The ratio was lower for premature or intra-uterine growth retarded neonates. Our results, in conjunction with previous reports that galactosylated IgG stimulates Fc-mediated phagocytosis of monocytes, suggest that increased galactosylation of IgG enhances neonatal immunity. 2000 Elsevier Science B.V. All rights reserved. Keywords: Cord blood; IgG; Carbohydrate; Galactosylation; Placental function
*Corresponding author. Tel.: 1 81-3-3784-8577; fax.: 1 81-3-3788-4927. E-mail address: [email protected] (S. Kimura) 0009-8981 / 00 / $ – see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0009-8981( 00 )00289-8
170
S. Kimura et al. / Clinica Chimica Acta 299 (2000) 169 – 177
1. Introduction N-linked oligosaccharides in human IgG are known to modulate immunological function. Complex oligosaccharide moieties (biantennary mannose structure, with or without galactose on N-acetylglucosamine outer arms) are attached to asparagine-297 of gamma heavy chains in human IgG molecules, and help maintain the three-dimensional structure [1]. Carbohydrate moieties help stabilize the IgG molecule [2], and prolong its half-life in serum [3]. Carbohydrates modulate antigen binding of IgG [4], and Fc-mediated phagocytosis [5]. Human neonates suffer fewer infectious diseases than do infants older than 3 months, because they receive maternal immunoglobulins (particularly IgG) via the placenta [6]. However, the average half-life of adult serum IgG is less than a month, which does not account for the duration of strong immunity in neonates. It is possible that IgGs present in neonates have distinctive structures which prolong their half-life. We characterized, for the first time, carbohydrate structures on neonatal serum IgGs in comparison to adult serum IgGs.
2. Materials and methods
2.1. Sample collection Umbilical cord blood provides a convenient, non-invasive means of studying neonatal serum IgG. Cord blood samples (n 5 45; 24 males and 21 females) were collected during deliveries at Dokkyo University Koshigaya Hospital, with the parents’ written consent. Ranges and mean6S.D. values were 25–41 weeks (38.963.0) for gestation period, 860–3824 g (29146641) for birth weight, 495–2040 mg / dl (13376346) for cord blood IgG level, and 21–39 years (30.064.4) for age of mother. Blood samples were collected at Showa University Hospital from 11 healthy adult male volunteers (age range 23–40; 26.965.6), whose serum aminotransferases, Na, K, Cl, total protein, direct bilirubin, total cholesterol, and triglyceride levels were all within normal range.
2.2. HPLC analysis FMOC-Cl, urea, and 1,4-dioxane were from Sigma Chemical, St. Louis, MO, USA. Recombinant N-glycanase (E.C.3.5.1.52) was from Genzyme, Cambridge, MA, USA. Orcinol, dithiothreitol, ethylenediaminetetraacetate, iodoacetic acid, acetonitrile (HPLC grade), sodium monobasic phosphate, sodium dibasic phosphate, and dichloromethane were from Wako Pure Chemical, Osaka, Japan.
S. Kimura et al. / Clinica Chimica Acta 299 (2000) 169 – 177
171
Neuraminidase (E.C.3.2.1.18) was from Nacalai Tesque, Kyoto, Japan. TSKGEL Amide 80 column was from Tosoh K.K., Tokyo, Japan. To isolate IgG proteins, 500 ml of serum was passed through a Protein A-conjugated sepharose CL-4B column equilibrated with 0.05 mol / l Tris–HCl, pH 8.2, containing 0.15 mol / l NaCl and 0.02% NaN 3 . The column was rinsed with the same buffer, and eluted with 50 mmol / l sodium citrate buffer containing 0.15 mol / l NaCl and 0.02% NaN 3 , pH 3.5. Purified IgG was dissolved in 1.0 ml of 20 mmol / l acetic acid pH 5.0, and incubated with 250 mU of neuraminidase at 378C for 24 h. The neuraminidase treatment, although not absolutely necessary, gave simpler peak patterns for HPLC analysis. Preparation of reduced, carboxymethylated IgG proteins, and release of (oligosaccharyl) amines from these proteins by recombinant N-glycanase treatment, were performed as described previously [7]. To derive oligosaccharyl amines, a portion ( | 50 mg) of each of the samples was dissolved in 800 ml of 20 mmol / l sodium phosphate buffer, pH 8.6, containing 50% ethanol. To 800 ml of this solution, 80 ml of FMOC (49 mg of FMOC-Cl in 4 ml of 1,4-dioxane) was added, and incubated at room temperature for 1 h. The solution was added to 800 ml of dichloromethane, mixed well, and centrifuged. The supernatant was applied to the HPLC. FMOC-amino-oligosaccharide derivatives were separated on Amide 80 column (4.6 3 250 mm, particle size 5 mm) by HPLC with a linear elution gradient (25 to 49%, v / v) of aqueous acetonitrile within 40 min, flow-rate 1.0 ml / min. Peaks were recorded and integrated by a Hitachi D-2500 Chromato-integrator. The HPLC system was a Waters 600E system controller, with JASCO 821-FP Intelligent Spectrofluorometer. FMOC-amino-oligosaccharide derivatives (1 nmol) from human serum IgG were digested with 10 mU of beta-galactosidase (Seikagaku, Tokyo, Japan) in 100 ml of 50 mmol / l sodium acetate buffer, pH 6.0, at room temperature for 10 min. Statistical analysis was by Student’s t-test.
3. Results Statistical data (sex, birth weight, gestation period, mother’s age) and analytical results for the 45 cord blood samples are summarized in Table 1.
3.1. Galactose content between neonates and adults FMOC-amino-oligosaccharide derivatives from two different serum IgG samples were separated quantitatively by HPLC (Figs. 1A and B). Peak 3
S. Kimura et al. / Clinica Chimica Acta 299 (2000) 169 – 177
172
Table 1 Statistical data and analytical results for 45 umbilical cord blood samples
Mean median S.D.
Birth weight (g)
Gestation period (weeks)
Mother’s age
IgG (mg / dl)
%G0
%G1
%G2
G2 / G0 ratio
2914 3122 641
38.9 39.6 3.0
30.0 31.0 4.4
1337 1360 346
9.0 7.8 4.2
33.5 33.6 6.5
57.5 59.5 8.6
7.90 7.46 3.92
isolated from cord blood was digested with beta-galactosidase to remove non-reducing terminal galactose residues, and the reaction mixture was reexamined by HPLC under the same conditions (Fig. 1C). As shown in Figs. 1B and C, peak 3 shifted to peak 1 after beta-galactosidase digestion, indicating that galactose residues in the non-reducing terminal of the oligosaccharide molecule were eluted in peak 3. In adults, peak 1 (non-galactosyl oligosaccharides; G0) represented 25.466.2% (mean6S.D.) of oligosaccharides from the IgG molecule, while peak 2 (G1) and peak 3 (G2) represented 36.567.0 and 37.867.8%, respectively. In neonates, on the other hand, G0, G1 and G2 took 9.064.2, 33.566.5 and 57.568.6% of the IgG molecule, respectively. According to these results, %G0 was significantly higher in adults, whereas %G2 was higher in neonates (P , 0.001 in both cases). The ratio of G2 to G0 (G2 / G0 ratio) was significantly higher in neonates (7.9063.92) than in adults (1.6060.62; P , 0.0001) (Fig. 2A). Similarly, G1 / G0 ratio in neonates (4.3861.94) was higher than in adults (1.5360.53; P , 0.0001) (Fig. 2B).
3.2. Galactose content among neonates A high G2 / G0 ratio was observed even in a premature case (gestation period 25 weeks, birth weight 860 g, G2 / G0 ratio 5.91), although this neonate’s IgG content was only 532 mg / dl, | 40% of the typical level for full-term deliveries. There were weak but not significant correlations between the G2 / G0 ratio and birth weight (r 5 0.219), gestation period (r 5 0.207), mother’s age (r 5 2 0.272), and no correlation with serum IgG concentration (r 5 0.063) (Fig. 3). The G2 / G0 ratio was 4.9662.57 for intra-uterine growth retardation (IUGR) babies (n 5 6), lower than for non-IUGR babies (n 5 39) (8.3663.91; P , 0.05). G2 / G0 ratio was 8.0664.06 for full-term (n 5 41) vs. 6.3361.11 for premature infants (n 5 4), although this difference was not statistically significant because of small sample size.
S. Kimura et al. / Clinica Chimica Acta 299 (2000) 169 – 177
173
Fig. 1. HPLC elution patterns of FMOC-labeled oligosaccharides from a healthy adult (A), a neonate (B), and neonatal peak 3 after galactosidase digestion (C). Elution times of oligosaccharides ranged from 24.1 to 36.5 min. (A) Adult. The three major peaks represent nongalactosyl (peak 1; G0), and galactosyl structures (peaks 2 and 3, G1 and G2, respectively) attached to a mannose core biantennary structure. The G2 / G0 ratio is 1.09, and G1 / G0 ratio is 0.73. (B) Neonate. Peak 3 is much larger and peak 1 smaller in comparison to the adult pattern. This neonate had a gestation period of 37.6 weeks and birth weight of 2450 g. The G2 / G0 ratio is 10.74, and the G1 / G0 ratio is 6.34. (C) Peak 3 from the neonate, after purification and beta-galactosidase digestion, shifted to the location of peak 1 in panel B.
174
S. Kimura et al. / Clinica Chimica Acta 299 (2000) 169 – 177
Fig. 2. Comparison of the G2 / G0 and G1 / G0 ratio between neonates and adults. The ratio of peak 3 to peak 1 (G2 / G0, see Fig. 1) ranged from 1.58 to 20.56 in neonates, and 0.84–2.79 in adults. Mean value for neonates (7.9063.92) was significantly higher than for adults (1.6060.62, P , 0.0001). The ratio of peak 2 to peak 1 (G1 / G0) was also higher in neonates than in adults.
S. Kimura et al. / Clinica Chimica Acta 299 (2000) 169 – 177
175
Fig. 3. Regression of G2 / G0 ratio against serum IgG concentration.
4. Discussion This is the first report on oligosaccharide profiles of neonatal serum IgGs. Parekh et al. [8] and Shikata et al. [9] investigated age-related galactosylation patterns of N-linked oligosaccharides in human serum IgGs in subjects ranging from one to 73 years old. Parekh et al. found that the proportion of nongalactosylated oligosaccharide was highest (27.3% of total galactosyl structures) in a 1-year-old infant and lowest (21.2%) in a 25-year-old subject. In our study, non-galactosylated oligosaccharide accounted for 9.064.2% (mean6S.D.) of total oligosaccharides in neonates, compared to 25.466.2% in adults (P , 0.0001). The G2 / G0 ratio was significantly higher in cord blood IgGs than in adult IgGs. Our results are consistent with those of Parekh et al., and there appears to be little difference between infants and adults. We therefore conclude that cord blood IgGs are more highly galactosylated than infant or adult IgGs. Increased galactosyltransferase activity has been reported in pregnant women [10], consistent with our finding of increased galactosylation of cord blood IgG. In neonates, we must also consider the half-lives of galactosylated and sialylated
176
S. Kimura et al. / Clinica Chimica Acta 299 (2000) 169 – 177
IgGs. In a rat model, clearance of IgG proteins from the bloodstream was prolonged for animals with N-linked oligosaccharides [3]. There are many reports on changes of carbohydrate structures in IgG molecules as correlated with disease status [1,11,12]. Glycosylation status in relation to IgG subclass was studied by Keusch et al. [13]. Sugar chains, particularly glycosylated N-linked oligosaccharides, may hinder the degradation of IgG proteins, and play a role in maintaining steric structure [13,14]. This preserves the active sites present in these domains, and allows more effective binding to C1q and Fc-receptors [15,16]. Fc receptor-mediated phagocytosis of macrophages, and feedback immunosuppression, are modified by the IgG carbohydrate structure [5,17,18]. These previous findings, in conjunction with ours, suggest that increased galactosylation of IgG proteins prolongs their half-life in neonatal blood, and strengthens the neonatal immune system by maintaining immunity of maternal origin. Maternal IgG is transferred to the fetus via active transport through the placenta [6]. Decreased galactosylation of IgG in IUGR neonates may reflect placental dysfunction. Neonates and mothers are often exposed to infectious agents during delivery. Higher galactosylation and sialylation of carbohydrate structures on serum IgGs may reduce the risk of infection, or enhance transport of IgG through the placenta. Further studies are needed to assess these possibilities.
Acknowledgements This work was supported by Grants-in-aid for Scientific Research (A) 07772258, (A) 08772145, and (A) 09772071 (to S.K.) from the Ministry of Education, Science, Culture, and Sports of Japan. The authors thank Dr. Stephen Anderson for English assistance.
References [1] Kobata A. Function and pathology of the sugar chains of human immunoglobulin G. Glycobiology 1990;1:5–8. [2] Hobbs SM, Jackson LE, Hoadley J. Interaction of aglycosyl immunoglobulins with the IgG Fc transport receptor from neonatal rat gut:comparison of deglycosylation by tunicamycin treatment and genetic engineering. Molec Immunol 1992;29:949–56. [3] Wawrzynczak E, Cumber A, Parnell G, Jones P, Winter G. Blood clearance in the rat of a recombinant mouse monoclonal antibody lacking the N-linked oligosaccharide side chains of the CH2 domains. Molec Immunol 1992;29:213–20. [4] Fishman AJ, Fucello AJ, Pellegrino-Gensey JL. Effect of carbohydrate modification on the localization of human polyclonal IgG at focal sites of bacterial infection. J Nucl Med 1992;33:1378–82.
S. Kimura et al. / Clinica Chimica Acta 299 (2000) 169 – 177
177
[5] Fukushima K, Takasaki S. Processing inhibition of N-linked sugar chains associated with induction of Fc receptor-mediated phagocytosis in the mouse monocytoid cells. Glycobiology 1993;3:15–22. [6] Gitlin D, Kumate J, Urrusti J, Morales C. The selectivity of the human placenta in transfer of plasma proteins from mother to fetus. J Clin Invest 1964;43:938–51. [7] Kim D, Rasmussen J, Kimura S, Kaizu T. Application of FMOC-Cl for the quantitative determination of 1-amino-oligosaccharides in the low range. Glycoconjugate J 1995;12:408. [8] Parekh R, Roitt I, Isenberg D, Dwek R, Radenmacher T. Age-related galactosylation of the N-linked oligosaccharides of human serum IgG. J Exp Med 1988;167:1731–6. [9] Shikata K, Yasuda T, Takeuchi F, Konishi T, Nakata M, Mizuochi T. Structural changes in the oligosaccharide moiety of human IgG with aging. Glycoconjugate J 1998;15:683–9. [10] Lacord-Bonneau M, Khalfoun B, Weill JD, Bardos P. Long-term study of serum glycosyltransferase levels in pregnant women and their newborn infants. Int J Biochem 1988;20:997– 1000. [11] Bond A, Alavi A, Axford JS et al. A detailed lectin analysis of IgG glycosylation, demonstrating disease specific changes in terminal galactose and N-acetylglucosamine. J Autoimmun 1997;10:77–85. [12] Malhotra R, Wormald MR, Rudd PM, Fischer PB, Dwek RA, Sim RB. Glycosylation changes of IgG associated with rheumatoid arthritis can activate complement via the mannose-binding protein. Nature Med 1995;1:237–43. [13] Keusch J, Levy Y, Shoenfeldt Y, Youinou P. Analysis of different glycosylation states in IgG subclasses. Clin Chim Acta 1996;252:147–58. [14] Nardella F, Teller D, Mannik M. Studies on the antigenic determinants in the self-association of IgG rheumatoid factor. J Exp Med 1981;154:112–25. [15] Henny C, Stanworth D. Reaction of rheumatoid factor with the isolated polypeptide chains of human 7s gamma-globulin. Nature 1964;201:511–2. [16] Koide N, Nose M, Muramatsu T. Recognition of IgG by Fc receptor and complement: effects of glycosidase digestion. Biochem Biophys Res Commun 1977;75:838–44. [17] Nose M, Wigzell H. Biological significance of carbohydrate chains on monoclonal antibodies. Proc Natl Acad Sci USA 1983;80:6632–6. [18] Heyman B, Nose M, Weigle W. Carbohydrate chains on IgG2b: a requirement for efficient feedback immunosuppression. J Immunol 1985;134:4018–23.