Eur. 3. Biochem. 93,431 -435 (1979)

Determination of Leb-Active Oligosaccharides in Urine of Pregnant and Lactating Women by Radioimmunoassay David A. ZOPF, Victor GINSBURG, Peter HALLGREN, Anne-Charlotte JONSSON, Bo S. LINDBERG, and Arne LUNDBLAD National Cancer Institute and National Institute of Arthritis, Metabolism and Digestive Diseases National Institutes of Health, Bethesda, Maryland; Department of Clinical Chemistry, University Hospital, Lund; and Department of Obstetrics and Gynecology, Central Hospital, Vaster% (Received August 28, 1978)

Antiserum obtained by immunizing a goat with lacto-N-difucohexaose I, a Leb blood-group hapten, coupled to poly-L-lysine was used in a radioimmunoassay to detect Leb-active oligosaccharides (chiefly lacto-N-difucohexaose I) in urine of 138 pregnant and lactating women of different ABO and Lewis blood groups. Specificity of the method was determined by comparing inhibitory activities of 18 oligosaccharides. Only women who belonged to the Leb blood group excreted Leb-active oligosaccharides in urine. Leb-active oligosaccharides increase during pregnancy, reaching levels up: to approximately 70 pmol lacto-N-difucohexaose I equivalents per pmol creatinine in the third trimester and early post-partum period. Excretion varies considerably but tends to be highest in those individuals who strongly express the Leb antigen on their red blood cells or who belong to blood group 0. Leb-active oligosaccharides were detected in plasma of a few individuals but at concentrations 1000-fold lower than in urine.

In previous studies the urinary oligosaccharides excreted by two women during pregnancy and lactation were identified by chemical methods and shown to increase during pregnancy, reaching maximum levels during the last trimester [I -31. These studies have now been extended to include more individuals by using a rapid and sensitive radioimmunoassay [4] that detects specific oligosaccharides in urine without prior purification. In the present report this assay is used to measure the level of Leb-active oligosaccharides, chiefly lacto-N-difucohexaose I, in the urine and plasma during pregnancy and lactation of 138 women belonging to different ABO and Lewis blood groups. MATERIALS AND METHODS Oligosaccharides '

Oligosaccharides from human urine were isolated as previously described [l, 2,5,6]. Oligosaccharides VIH, VIIa, and VIIs were gifts from Dr Gerhard Strecker (Lille, France) [7].Lacto-N-difucohexaose I was radioactively labelled by reduction with sodium borotritide [8] (specific activity = 15.6 niCi/pmol; New England Nuclear Corporation, Boston, Massachusetts). For structure of oligosaccharides see Table 1.

Antiserum

Goat antiserum that binds [3H]lacto-N-difucohexaitol I was prepared as described previously [4]. Urine and Plasma

Urine and plasma were collected from 10 nonpregnant women of Le(a- b + ) blood type and from 130 women at different stages of pregnancy without selection for Lewis blood type. From four of these women, three Le(a - b +) and one Le(a - b -), 24-h urine specimens were collected at biweekly intervals beginning in the first trimester and continuing through the post-partum period until lactation ceased. All samples were stored at - 10 "C. Urinary creatinine was determined by the Jaffe reaction modified by L$ken [9], but adsorption to Lloyd's reagent was omitted. Blood Typing

ABO and Lewis blood type were determined using blood samples taken at the time of urine collection. Typing was performed by the Blood Center (University Hospital, Lund) using commercial antisera. During typing it was noted that red blood cells

432

Leb-Active Oligosaccharides in Urine during Pregnancy and Lactation

Trivial Name (abbreviation)

Structure Gal p I-myo-inositol 2

difucogalactosyl inositol

I

from some individuals were agglutinated weakly by anti Leh serum; these individuals are designated in this paper as Le(a-bb) (weak).

I

Fucal Fucal Gal p I-3GlcNAc p I4Gal 2 4 6

VIH

I

I

Fucal VllA

1

I

I

Fucal

Fucal

Galal3Galp IJGlcNAc p 14Gal 2 4 6 I I I Fucal Fucal Fucal

2'-fucosyllactose

Galp 14Glc 2

I

Fucal Gal p IMGlc 3

3-fucosyllactose

I

Fucal

Gal p 14Glc 2 3

lactodifucotetraose (LDFT)

I

I

Fucal Fucal urine A pentasaccharide

GalNAcal-3Galp14Glc 2 3

I

I

Fucal Fucol urine 6 pentasaccharide

GalalJGalp 14Glc 2 3

I

I

Fucal Fucal lacto-N-fucopentaose I

Gal p IJGlcNAc p I-3Gal p I4Glc 2 Fuc a 1

lacto-N-fucopentaose II

Gal p 13GlcNAc p I-3Gal p I4Glc 4 I Fuc a 1

lacto-N-fucopentaose 111

Gal B IMGlcNAc R I-3GalB 14Glc 3

I

Fucal lacto-N-difucohexaose I (LND I)

Gal p I-3GlcNAcp I-3Gal pI4Glc 4 2

I

Fucal lacro-N-neo difucohexaose I

I

lacto-N-difucohexaose II

lacto-N-difucohexaose IV

lacto-N-trifucoheptaose I1

Gal p I-3GlcNAcp I-3Gal p I4Glc 2 2

I

I

Fucal

Fucal

Gal pI-3GlcNAc fi I-3GaI p14Glc 2 2 4

I

The chemical specificity of a goat antiserum that binds [3H]lacto-N-difucohexaitolI has been studied previously by comparing some milk oligosaccharides as inhibitors of binding [4].Additional studies to determine which urinary oligosaccharides are Lebactive were performed as follows: 10 p1 goat antiserum was mixed with varying amounts of inhibitor in 0.01 M Tris buffer, pH 7.5, containing 0.14 M NaCl, 0.5 mM MgS04, and 0.15 mM CaC12 in a final volume of 300 pl. After incubation for 60 min at room temperature, 6.7 pinol of [3H]lacto-N-difucohexaitol I (100000 counts/min) in 100 pl of buffer were added and incubation was continued for 30 min. The mixture was then passed through a nitrocellulose filter and the filter was washed with 10 ml of buffer and counted by liquid scintillation. Total radioactivity was corrected by substracting radioactivity unspecifically retained by the filter (generally 30 to 40 counts/min) when non-immune goat serum was used in place of the immune serum. Of the 18 oligosaccharides tested (Fig.1 and Table 1) lacto-N-difucohexaose I is the best inhibitor, 0.1 nmol giving 50 % inhibition. Lacto-N-trifucoheptaose 11, whose level in urine is much lower than lactoN-difucohexaose I, and oligosaccharide VIH, which has not been found in urine of pregnant women, are about 10 % as active as lacto-N-difucohexaose I. Other oligosaccharides are less than 1 as active as lacto-Ndifucohexaose I on a molar basis. Previously reported levels of most urinary oligosaccharides [ 3 ] are far below the level required for detection in this assay. Therefore, the Leb activity exhibited by urinary oligosaccharides is mainly due to lacto-N-difucohexaose I.

I

Fucal

Gal p I-3GlcNAc p I-3Gal p 14Glc 4 3 I I Fuc a 1 Fucal

Fucal lacto-N-neo trifucoheptaose I1

I

Fucal

Gal p I4GlcNAc p I-3Gal p I4Glc 2 3 Fucal

Specificity of the Radioimmunoassay

Fucol

GalNAcal3Gal p I-3GlcNAcp I4Gal 4 6 2 Fucal

Vllg

I

Fucal

RESULTS AND DISCUSSION

I

Fucal

I

Fucal

Gal p IMGlcNAc p I-3Galp I4Glc 2 3 2

I

I

I

Fucal

Fucal

Fucal

.' Fucose has the I> configuration, all other monosaccharidcs shown have the D configuration

Radioimmunoassay of Leb-Active Oligosaccharides in Urine and Plasma 100-pl aliquots of urine cleared by centrifugation or plasma ultrafiltered through an Amicon PM-10 membrane, were added to the assay mixture described above. Levels of Leb-active oligosaccharides, calculated as lacto-N-difucohexaose I, were estimated from the inhibition obtained using a standard curve for authentic lacto-N-difucohexaose I. Coefficients of variation of duplicate determinations were less than 10 % when inhibition was 30 to 60 % (Fig. 1). Thus, 30 % inhibition (equivalent to 0.03 nmol 1acto-Ndifucohexaose I) was defined as the lower limit of detection. When inhibitions above 70 % were obtained

433

D. A. Zopf, V. Ginsburg, P. Hallgren, A.-C. Jonsson, Bo S. Lindberg. and A. Lundblad

the samples were diluted appropriately with the Tris assay buffer and the determination was repeated. Lacto-N-difucohexaose I added to 100 pl of urine from a nonpregnant woman of blood type Le(a - b -) gave an inhibition curve identical to that obtained with lacto-N-difucohexaose I alone. Essentially all Leb activity of urine is present in the neutral oligosaccharide fraction as it was quantitatively recovered from urine of a pregnant woman of blood type Le(a - b +) following ultrafiltration through an Amicon PM-10 membrane and passage through a mixed-bed ion-exchange column containing Bio Rad AG3-X4A, OH- form plus AG-50W-X8, H’ form. Leb-active oligosaccharides in urine of four women were determined at approximately biweekly intervals from early in gestation through the post-partum period (Fig.2). Three of these were of blood type Le(a- b + ) and excreted increasing amounts of Lebactive oligosaccharides beginning in the first trimester with a maximum during the first week post partum. Thereafter the amount of Leb-active oligosaccharides gradually decreased and finally fell to non-pregnancy levels when lactation ceased. One Le(a - b -) woman followed at biweekly intervals showed no significant increase in urinary Leb-activity during the whole period. These results agree with previous data on urinary oligosaccharides obtained by chemical methods [3]. Table 2 shows the level of Leb-active oligosaccharides in urine of 138 women determined by radioimmunoassay. In all women of blood group Le(a - b +) levels of Leb-active oligosaccharides in urine increased during pregnancy, reaching the highest level during the third trimester and the first weeks post partum. In general, women with weak expression of Leb antigen on their red blood cells [Lewis blood type Le(a- b + ) (weak)] had less Leb activity in their urine. As expected from previous studies [3] Le” activity was

not detected in urine from nonpregnant Le(a- b + ) women. Decreased expression of Lewis antigens on erythrocytes during pregnancy has been reported [l 1,121. The incidence of Le(a - b -) blood type among women in this study was greater than expected (31 % as compared with 6 % for the general population in Sweden) and some of the women who typed as Le(a-b-) during pregnancy were actually of blood groups Le(a + b -) or Le(a - b +), explaining the small amount of Le” activity in urine during pregnancy and the post-partum period in some women with redblood-cell phenotype Le(a-b-) (Table 2). Of the 44 women typed during pregnancy as Le(a - b -), 22 were available for retyping after delivery. Two of these changed Lewis blood type to Le(a b -) and five to Le(a- b+). The five latter individuals all had relatively high levels of Leb activity in urine during pregnancy and lactation.

+

100 r

10-2

lo-’

1 I n h i b i t o r added

10

100

(nrnol)

Fig. 1. ltihibition of’untihodj binditig o f f ’H]lucto-N-difircolleraitol I by oligusuccharides. Lacto-N-difucohexaose I (I), oligosaccharide VIrr (2), lacto-N-trifucoheptaose 11 (3), lacto-hr-neotrifucoheptaose I1 (4), lacto-N-fucopentaose 111 (9,lacto-N-difucohexaose IV (6), lacto-N-fucopentaose I1 (7), urine-B-pentasaccharide (8), oligosaccharide VIIA (Y), oligosaccharide VIIB (lo), lacto-N-neodifucohexaose I (1 I), lactodifucotetraose (12), lacto-N-fucopentaose I (1 3), urine-A-pentasaccharide (14), lacto-N-difucohexaose I1 (15), 3-fucosyllactose (16), Z’-fucosyllactose (1 7), difucogdlactosylinositol (18). See Table 1 for structures

k G e s t a t i o n ( w e e k s ) 4 T i r n e postpart urn-I-Tirne (days)

postpart urn (rnont hs)

Fig. 2. Lacto-N-dfucohexaose I in 24-h urine specimens collected approximute1.v biweekly fiom Jbur women during pregnancy and subsequent Blood group 0 Le(a - b +) lactating after 5 months; ( t - m ) blood group A Le(a - b +) lactating after 2 months; lactation. (0-0) (---O) blood group B Le(a - b +) lactation ceased after 4 months; (-0) blood group A Le(a - b -) lactating after 1 month; (A) partus

1000

434

Leh-Active Oligosaccharides in Urine during Pregnancy and Lactation

Table 2. Leb-active oligosaccharides in urine of non-pregnant, pregnant and lactating women with different Lewis blood groups The number of subjects in each group is given in square brackets. Results are a mean value with the range in round brackets Leb-active oligosaccharides as lacto-N-difucohexaose I in urine specimen

Lewis blood group

-

1st trimester

Nonpregnant

2nd trimester

3rd trimester

post partum

6.3 [13] (0.3 -20)

12.2 [16] (0.7 - 32.9)

29.8 [I41 (5.2-69.7)

30.8 [I41 (16.7-49.1)

1.5 [4] (0.6-3.1)

9.7 [5] (3.6 - 20.5)

20.3 [ I l l (2.7 - 66.6)

22.0 [9] (6.8 -45.1)

0.8 [S] (0.4- 2.2)

2.5 [ l l ] (0.1 -8.5)

2.8 [21] (0.1 - 8.6)

5.1 [9] (0.7-15.7)

0.4 [6] (0-1.0)

0.4 [2] (0.3 -0.4)

0.6 [3] (0.2-1.0)

0.5 [2] (0.1-0.9)

pmol/pmol creatinine (a-b+)

1.7 [lo] (0.4-4.9) -

(a - b +) (weak)” (a-b-)b

0.3 [I] (0.3)

(d+ b - )

-

~

a

See blood typing in Materials and Methods Some of these on subsequent retyping were (Le (a - b +). See text

Table 3. Leb-active oligosaccharides in urine of pregnant and lactating Le(a - b +) women with different ABO blood groups The number of subjects in each group is given in square brackets. Results are a mean value with the range in round brackets ABO blood group

Leb-active oligosdccharides as lacto-N-difucohexaose I in urine specimen _.

1st trimester

2nd trimester

3rd tiimester

post partum

pmol/pmol creatinine _-

A

2.9 [lo] (0.3 - 11.4)

11.1 [S] (0.9-30.3)

16.5 [ l l ] (2.7- 31.4)

24.6 [I41 (6.8- 45.1)

B

3.2 [3] (1 .5- 5 .O)

12.5 [5] (3.5 - 32.9)

9.9 [4] (7.6-13.3)

22.4 [3] (17.2 - 28.4)

AB

-

0.7 [I] (0.7)

5.4 [I] (5.4)

0

12.3 [4] (4.9 - 20)

12.5 [7] (6.5-20.5)

Decreased expression of Lewis antigens during pregnancy may be a reflection of the usual variability in expression of the Leh antigen on red blood cells such that some Le(a- b + ) individuals whose cells normally contain large amounts of Leh antigen retain their usual Lewis blood group phenotype whereas others whose cells normally contain smaller amounts of Lehantigen become weakly agglutinable or completely lose agglutinability by standard test sera. The correlation observed between agglutinability of red blood cells and levels of urinary Leb-active oligosaccharides during pregnancy (Table 2) probably is due in part to individual differences in the rate of synthesis of all Leh-active structures. The biosynthesis of Lebactive structures requires the concerted action of several independently inherited glycosyltransferases and the biochemical basis for variation in the overall

44.6 [9] (19.6-69.7)

36.2 [6] (27.0-49.1)

rate of synthesis among individuals of blood group Le(a - b +) has been reviewed [13]. Precursors of Lebactive oligosaccharides are also intermediates in the synthesis of structures active in blood groups A and B : glycosyltransferases coded by blood group A and B genes utilize UDP-N-acetylgalactosamine or UDPgalactose, respectively, as well as H-active precursors to form blood-group-A-active or B-active products [13]. Thus, some of the observed variation in Leh activity in urine and on red blood cells may be due to competition for precursors among enzymes that form Leb, A, and B antigens. This idea is supported by data shown in Table 3 where the level of Leh-active oligosaccharides (as lacto-N-difucohexaose I) in urine from 81 pregnant or lactating women of Le(a-b+) blood type is tabulated according to the four main ABO blood groups of the donors. During the third

D. A. Zopf, V. Ginsburg, P. Hallgren, A.-C. Jonsson, Bo S. Lindberg, and A. Lundblad

trimester and post-partum period the largest increase of Leb activity occurs in members of blood group 0. Similarly, Tilley et al. [14] described variations in A and B blood-group-active glycolipids in plasma dependent upon Lewis blood group and secretor status. The origin of Leb-active oligosaccharides in urine during pregnancy and lactation is unknown but fetal or placental origin seems unlikely for three reasons. First, no Leb-active oligosaccharides were detected in any of the 13 women with blood type Le(a-b-) although some of these are expected to have been carrying Le(a - b +) fetuses. Lewis antigens have been detected in fetal tissues by immunofluorescence [15] but they are weakly expressed and are undetectable in fetal erythrocytes [16]. Second, the fact that the strength of expression of Le” antigen on erythrocytes correlates with the amount of Leb-active oligosaccharides in urine suggests common genetic regulation of the levels of both antigens. Third, the pattern of urinary excretion, which increases during pregnancy, peaks in the early post-partum period, and continues at high levels during lactation, would not be expected if the fetus were the source of Leb-active oligosaccharides. Clearly the prolonged post-partum urinary excretion during lactation requires some maternal source, possibly lactating mammary gland as many urinary oligosaccharides are identical to oligosaccharides isolated from human milk[17]. However, an additional source seems likely since some oligosaccharides present in urine (e.g. difucogalactosylinositol) have not been detected in milk. Most of the plasma samples contained no detectable Leb-active oligosaccharides. In a few cases in which Leb activity was markedly elevated in urine, Le” activity was detected in plasma ultrafiltrates but the concentrations were approximately 1000-fold lower than in urine obtained at the same time. These results suggest that oligosaccharides circulate in blood at very low concentration and are concentrated in the kidney. The physiologic basis for the decreased expression of Lewis antigens on red blood cells during pregnancy is unknown. The Lea and Leb antigens of red blood

435

cells are glycosphingolipids that are carried on high density and low density lipoproteins in plasma and they adsorb reversibly to red cells [18,19]. Possibly the marked increase in both plasma volume [20] and concentration of plasma lipoproteins [21] that occurs during pregnancy alters the distribution betweeen erythrocytes and plasma of glycosphingolipids active in the Lewis blood group. This work was aided by grants from the Swedish Medical Council (03X-2) and Stiffelsen Expressens pretalforskningsnumnd.

REFERENCES 1. Hallgren, P. & Lundblad, A. (1977) J . Bid. Chem. 252, 10141022. 2. Hallgren, P. & Lundblad, A . (1977) J . B i d . Chem. 252, 10231033. 3. Hallgren, P., Lindberg, B. S. & Lundblad, A. (1977) J . B i d Chem. 252, 1034-1040. 4. Zopf, D. A,, Ginsburg, A. & Ginsburg, V. (1975) J . Immunol. 115, 1525 - 1529. 5. Lundblad, A. (1967) Biochim. Biophys. Acta, 148, 151 -157. 6. Lundblad, A. (1968) Biochim. Biophys. Acta, 165, 202-207. 7. Strecker, G., Trentesaux-Chauvet, C., Riazi-Farzad, B., Bouquelet, S. & Montreuil, J. (1973) C. R. Hehd. Seances Acad. Sci. 277, IS69 - 1572. 8. Kobata, A. & Ginsburg, V. (1969) J . Biol. Chem. 224, 54965502. 9. LBken, F. (1954) Scand. J . Clin.Lab. Invest. 6, 325-334. 10. Lundblad, A. (1978) Methods Enzymol. 50, 226-235. 11. Brendemoen, 0. J. (1952) Acta Pathol. Microhiol. Scand. 31, 579 - 582. 12. Seki, K. (1968) Jpn. J . Legal. Med. 22, 487-498. 13. Ginsburg, V. (1972) Adv. Enzymol. 36, 131-149. 14. Tilley, C. A., Crookston, M. C., Brown, B. L. & Wherrett, J. R. (1975) Vox Sang. 28, 25-33. 15. Szulnian, A. E. & Marcus, D. M. (1973) Lab. Invest 28, 565 - 574. 16. Jordal, K. (1956) Acta Path. Microbiol. Scand. 39, 399-406. 17. Kobata, A. (1972) Methods Enzymol. 28, 262-271. 18. Marcus, P. M. & Cass, L. E. (1969) Science (Wash. D.C.) 164, 553 -555. 19. Sneath, J. S. &Sneath, P. H. A.(1955)Nature (Lond.) 176,172. 20. Rovinsky, J. J. & Jaffin, H. (1965) Am. J . Ohstet. Gynecol. 93, 1-15. 21. Warth, A. R., Arky, R. A. & Knopp, R. H. (1975) J . Clin. Endocrinol. Metab. 41, 649-655.

D. A. Zopf, National Cancer Institute, National Institutes of Health, 9000 Rockville Pike, Bethesda, Maryland, U.S.A. 20014 V. Ginsburg, National Institute of Arthritis, Metabolism, and Digestive Diseases, National Institutes of Health.

9000 Rockville Pike, Bethesda, Maryland, U.S.A. 20014 P. Hallgren, A,-C. Jonsson, and A. Lundblad, Institutionen for KIinisk Kemi, Lunds Universitet, Lasarettet, S-221 85 Lund, Sweden

Bo S. Lindberg, Kvinnoklininken, Centrallasarettet, S-721 89 VasterPs, Sweden

Determination of Leb-active oligosaccharides in urine of pregnant and lactating women by radioimmunoassay.

Eur. 3. Biochem. 93,431 -435 (1979) Determination of Leb-Active Oligosaccharides in Urine of Pregnant and Lactating Women by Radioimmunoassay David A...
429KB Sizes 0 Downloads 0 Views