This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by O’Donnell, J. Articles by Laffan, M. A. Search for Related Content PubMed PubMed Citation Articles by O’Donnell, J. Articles by Laffan, M. A. Related Collections Deep vein thrombosis Coagulation Thrombophilia Thrombosis risk factors

(Arteriosclerosis, Thrombosis, and Vascular Biology. 2002;22:335.)
© 2002 American Heart Association, Inc.

Thrombosis

Amount of H Antigen Expressed on Circulating von Willebrand Factor Is Modified by ABO Blood Group Genotype and Is a Major Determinant of Plasma von Willebrand Factor Antigen Levels

James O’Donnell; Frank E. Boulton; Richard A. Manning; Michael A. Laffan

From the Department of Haematology, Imperial College School of Medicine, Hammersmith Hospital, London, and the National Blood Service (F.E.B.), Southampton, UK.

Correspondence to Dr James O’Donnell, Department of Haematology, Hammersmith Hospital/ICSM, DuCane Road, East Acton, London W12 ONN, UK. E mail james.o'donnell{at}virgin.net

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
To investigate whether the effect of ABO blood group on plasma von Willebrand factor (vWF) levels is mediated by the ABH antigenic determinants carried on N-linked glycans of vWF, we studied 158 group A and group O healthy volunteers. vWF antigen (vWF:Ag) and factor VIII antigen (FVIII:Ag) levels were highest in A¹A¹ individuals and higher in A¹O¹ than in A²O¹ or O¹O¹ individuals. Plasma A transferase activity and the amount of A antigen expressed per unit vWF (AvWF) were significantly higher in A¹A¹ than in A¹O¹ individuals and higher in A¹O¹ than in A²O¹ individuals. AvWF was correlated strongly with plasma levels of A transferase activity. Thus, we have clearly demonstrated a direct relationship between ABO genotype, A transferase expression, and the amount of A antigen expressed on circulating vWF. H antigen expression per unit vWF (HvWF) was highest in group O individuals. Among group A individuals, the pattern of HvWF expression was A²O¹>A¹O¹>A¹A¹. In group O and group A²O¹ individuals, HvWF was inversely correlated with plasma vWF levels. In contrast, among group A¹A¹ and A¹O¹ individuals, there was no relationship between AvWF and plasma vWF levels. These findings suggest that it is H antigen expression that mediates the ABO effect on plasma vWF concentration.

Key Words: ABO blood group • von Willebrand factor • factor VIII

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
A large adhesive glycoprotein, von Willebrand factor (vWF) circulates in the plasma as a series of heterogeneous multimers.¹ It plays an important role in the interaction between platelets and the blood vessel wall and also serves as the carrier molecule for procoagulant factor VIII (FVIII).² Both functions are essential for normal hemostasis.³ In vivo biosynthesis of vWF is limited to endothelial cells and megakaryocytes,⁴ but the vWF synthesized within megakaryocytes is stored within -granules, so that the majority of plasma vWF is derived from endothelial cells.⁵ The normal population distribution of plasma vWF shows a wide range, with skewing toward higher levels. Deficiency of vWF (von Willebrand disease) has long been established as a cause of excessive bleeding.⁶ Conversely, there is accumulating evidence that elevated vWF-FVIII levels represent an important risk factor for myocardial infarction and venous thromboembolic disease.^7–10 Besides the vWF gene (12p12) and the FVIII gene (Xq28), it is well established that other gene loci exert major quantitative effects on plasma levels of the vWF-FVIII complex. The most important of these loci has been shown to be the ABO blood group locus on chromosome 9q34.

The antigens of the ABO system, (A, B, and H determinants, respectively) consist of complex carbohydrate molecules.¹¹ The A and B alleles of the ABO locus encode A and B glycosyltransferase activities, which convert precursor H antigen into either A or B determinants. These A and B structures differ only with respect to their terminal sugar moieties (N-acetylgalactosamine and D-galactose, respectively). Group O individuals lack such transferase enzymes and, consequently, continue to express the basic H structure constituting a solitary terminal fucose moiety.

A major influence of the ABO blood group on plasma vWF and FVIII levels was described by Preston¹² in 1964 and has been confirmed in many subsequent studies.^13–15 Normal group O individuals have significantly lower levels of vWF and FVIII than do non-O individuals.^16,17 Furthermore, 2 recent studies have demonstrated that normal individuals carrying 1 O allele (genotype AO and BO) have significantly lower plasma levels of vWF antigen (vWF:Ag) than do those carrying no O allele (genotypes AA, BB, and AB, respectively).^17,18 The mechanism underlying this association between the ABO blood group and vWF-FVIII levels remains unknown. However, linkage analysis has shown that the effect of the ABO group on plasma vWF levels is due to a direct functional effect of the ABO locus rather than a linkage disequilibrium between the ABO locus and another unidentified vWF regulatory locus.¹⁸

The effect of ABO blood group on plasma vWF:Ag levels may be due to altered rates of vWF synthesis or secretion by endothelial cells of a different ABO genotype. Alternatively, the effect of ABO on plasma vWF:Ag levels may be due to different rates of vWF clearance from the plasma in vivo. Thus, the lower vWF:Ag levels found in group O individuals may be the result of more rapid vWF clearance, with consequent reduction in vWF half-life. Matsui et al¹⁹ have demonstrated that ABH structures are carried on the N-linked oligosaccharide chains of vWF according to the blood type of the individual. Although it is not known whether the effect of ABO blood group on plasma vWF is directly mediated by these ABH structures on vWF, a recent report has shown that other carbohydrate antigenic determinants on murine vWF can significantly influence the plasma clearance rate.²⁰

ABH structures are expressed on a wide variety of human tissues, including red blood cells, platelets, epithelium, and vascular endothelial cells.²¹ However, the only circulating plasma glycoproteins expressing N-linked ABH antigens are vWF, FVIII, and ₂-macroglobulin.^22–24 The effect of the ABO blood group on plasma FVIII coagulant activity levels is similar to that previously described for plasma vWF:Ag levels, with significantly lower levels in group O individuals.^13,18 However, no quantitative data exist comparing ₂-macroglobulin levels in the different ABO blood groups.

We hypothesized that the quantitative effects of ABO blood group on plasma vWF levels are mediated by variation in the quantity of ABH antigenic determinants carried on circulating vWF. To test this hypothesis, we began by investigating the effect of ABO genotype on glycosyltransferase enzyme expression and plasma vWF levels in a series of healthy volunteers. We then quantified the amount of A and H antigenic determinants expressed per unit vWF and looked for the hypothesized correlation between these variables.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Sample Collection
One hundred fifty-eight healthy volunteers were recruited from routine blood donors at the Wessex Regional Transfusion Center, Southampton, UK. All donors were either blood group A or O and were aged between 18 and 70 years. The study was approved (South and West Research Ethics Committee), and each donor provided written informed consent. Blood was collected from the antecubital vein into Becton-Dickinson Vacutainer tubes containing 0.105 mol/L trisodium citrate. Platelet-poor plasma was obtained by centrifugation at 2000g for 20 minutes, within 90 minutes of collection. The plasma was then divided into aliquots and stored at -70°C.

ABO Genotyping
For each individual, the red blood cell ABO phenotype was determined by routine serological testing with the use of monoclonal anti-A and -B, respectively (Biotest AG). The ABO genotype was also established. Genomic DNA was extracted from 1 mL of citrated whole blood by using a commercial kit (Scotlab Ltd UK). ABO genotyping was performed by polymerase chain reaction amplification of exons 6 and 7 of the ABO gene, followed by diagnostic restriction enzyme digestion. Eight different primers were used to amplify 4 fragments, each spanning a different polymorphic site of the ABO gene. Primers ABO-1 and ABO-2 in conjunction with the restriction enzyme KpnI were used to differentiate O¹ alleles from A¹, A², B, and O² alleles²⁵; primers ABO-3 and ABO-4 in conjunction with the restriction enzyme PvuII were used to differentiate A² alleles from A¹, B, O¹, and O² alleles²⁶; primers ABO-5 and ABO-6 in conjunction with the restriction enzyme AluI were used to differentiate B alleles from A¹, A², O¹, and O² alleles²⁵; primers ABO-7 and ABO-8 in conjunction with the restriction enzyme MnII were used to differentiate O² alleles from A¹, A², B, and O¹ alleles.²⁷

vWF:Ag
Plasma vWF:Ag levels were determined by using a standard sandwich ELISA technique. Ninety-six–well plates were coated with rabbit polyclonal anti-human vWF antibodies (Dako) diluted 1:500 in 0.05 mol/L (pH 9.6) carbonate buffer overnight at room temperature. After a washing with Tris-buffered saline (TBS) containing 0.05% Tween, the plates were blocked for nonspecific binding with TBS containing 1% BSA (Sigma Chemical Co) for 1 hour. After 3 further washings, the plasma samples were added to the wells and incubated for 2 hours at room temperature. All samples were tested in duplicate at 3 different dilutions. The plates were washed with TBS/Tween and then incubated with a rabbit polyclonal anti-human vWF peroxidase conjugate (Dako) diluted 1:500 in TBS/Tween for 1 hour. After 3 further washings, the plates were incubated with a substrate solution. The reaction was stopped with 1 mol/L H₂SO₄, and the optical density was measured at a wavelength of 492 nm. Dilutions of 100% reference plasma (vWF:Ag 1.05 IU/mL, Immuno) were used to construct standard curves for calibration. The intra-assay and interassay coefficients of variation were <5%.

FVIII:Ag
Plasma FVIII antigen (FVIII:Ag) levels were measured by using a commercial ELISA kit (Immunozym, Technoclone) according to the manufacturer’s recommendations. All samples were measured against a reference plasma calibrated in international units per milliliter (Immunozym).

A Transferase Enzyme Expression
To measure -1,3-N-acetylgalactosaminyltransferase (A transferase) activity, 20 µL of each plasma sample was incubated in a total volume of 100 µL, together with 5 mmol/L ATP, 20 mmol/L MnCl₂, 50 mmol/L sodium cacodylate buffer (pH 7.4), 2.5 mmol/L 2'-fucosyllactose acceptor, and radiolabeled donor sugar in the form of 6.5 mmol/L UDP-[³H]N-acetylgalactosamine (50 000 cpm, Amersham UK). The reaction mixtures were incubated at 37°C for 16 hours, and the radioactive product was separated on a Dowex-1 column (0.8 mL) in the formate form. The columns were eluted with 1.6 mL water, and the eluate was mixed with 4 mL of water-miscible scintillant (Pico-Fluor 40, Canberra Packard) and counted on a Beckman LS 6800 scintillation counter. Control mixtures not containing acceptor substrates were performed for each plasma sample, and any eluted counts representing breakdown of the labeled nucleotide sugar were subtracted from the product counts. Pooled A plasma (40 blood group A donors) was tested at 4 dilutions to check that the relationship between A transferase and percent UDP-[³H]N-acetylgalactosamine incorporation was linear over the assayed range.

A and H Blood Group Antigenic Determinants on vWF
A (N-acetylgalactosaminyl -13[fucosyl -12] galactose) antigenic determinants on plasma vWF were measured by using a modified sandwich ELISA.^28,29 ELISA plates were coated with rabbit anti-human vWF antibody (Dako), washed, and blocked with TBS containing 1% BSA as per the vWF:Ag ELISA described previously. After 3 further washings, the plasma samples were added to the wells and incubated for 2 hours at room temperature. Each plasma was tested in duplicate at 3 dilutions. The plates were washed with TBS/Tween and incubated with murine anti-A monoclonal antibody (Ortho Diagnostics), which was diluted 1:10 in TBS, for 1 hour. After 3 further washings, the plates were incubated with goat anti-mouse IgM peroxidase conjugate (Sigma), which was diluted 1:1000 in TBS, for 1 hour. After another TBS/Tween wash, peroxidase substrate solution was added and incubated in the dark for 20 minutes. The reaction was stopped with 1 mol/L H₂SO₄ after 3 minutes, and the optical density was measured at a wavelength of 492 nm by use of an ELISA reader. Pooled group A plasma was assayed to produce a standard curve for each ELISA.

By use of the standard curve, a value for A antigen on vWF expression was determined for each plasma sample. Plasma vWF:Ag concentration strongly influenced the amount of A antigen detected in each ELISA well (P<0.0001 and =0.74, Spearman rank correlation). To determine the amount of A antigen expressed per unit vWF, the amount of A antigen detected was divided by the amount of vWF:Ag (see vWF:Ag ELISA) present in the ELISA well. These values of A or H per unit vWF are denoted AvWF and HvWF, respectively. The pooled normal A plasma was assigned a value of 1 U/mL for the amount of AvWF.

H (fucosyl -12 galactose) antigenic determinants on plasma vWF were measured by using a similar modified ELISA methodology. After incubation with the plasma samples, the plates were washed and incubated with biotin-conjugated Ulex europaeus (Vector Laboratories UK) diluted 1:500 for 1 hour. After a further washing, the plates were then incubated with streptavidin conjugated to peroxidase (Vector Laboratories UK) diluted 1:1000 for 45 minutes. Pooled group O plasma was assayed to produce a standard curve. The pooled normal O plasma was assigned a value of 1 U/mL for the amount of HvWF.

Statistical Analysis
All statistical analyses were performed by using the SPSS statistics package (version 10.0, SPSS Inc), and statistical significance was assigned at a value of P<0.05. Data were not normally distributed. Consequently, nonparametric statistical analyses (Mann-Whitney and Spearman rank correlation) have been used.

	Results

Top Abstract Introduction Methods Results Discussion References

 
We investigated a total of 158 normal blood donors (82 male and 76 female). All individuals recruited into the present study were repeat donors. Previous serological testing had shown that all were either blood group A or O. The ABO genotype distribution of these individuals was as follows: 13 were A¹A¹, 59 were A¹O¹, 1 was A¹A², 13 were A²O¹, and 72 were O¹O¹. All donors were white, and the blood group frequencies are in keeping with this.

ABO Genotype and Plasma vWF:Ag/FVIII:Ag Levels
In keeping with previous reports, vWF:Ag levels were significantly higher in A¹A¹ individuals (mean 111.8 IU/dL) than in A¹O¹ individuals (mean 96.1 IU/dL, P=0.03) and higher in A¹O¹ individuals than in A²O¹ or O¹O¹ individuals (means 84.1 and 77.4 IU/dL, respectively; P=0.03 and P=0.033, Mann Whitney; Figure 1). There was no significant difference in vWF:Ag level between A²O¹ and O¹O¹ individuals. Similarly, FVIII:Ag levels were highest in A¹A¹ individuals (mean 105.2 IU/dL) and significantly higher in A¹O¹ individuals (mean 97.9 IU/dL) than in A²O¹ or O¹O¹ individuals (means 85.3 and 76.2 IU/dL, respectively; P=0.01 and P=0.016, Mann-Whitney test). Overall, there was a strong correlation between vWF:Ag and FVIII:Ag levels (P<0.001 and =0.70, Spearman rank correlation).

View larger version (20K): [in this window] [in a new window]   Figure 1. Plasma vWF:Ag levels against ABO genotype. Plasma vWF:Ag levels were measured by ELISA in 158 healthy volunteers, and the ABO genotype of each subject was determined by polymerase chain reaction amplification of genomic DNA. The mean vWF:Ag level for each genotype is shown. vWF:Ag levels were highest in A1A1 individuals and significantly higher in A1O1 than in A2O1 or O1O1 individuals (P=0.03 and P=0.033, respectively; Mann-Whitney).

Although FVIII:Ag and vWF:Ag levels were significantly different in individuals of different ABO groups, the ratio of FVIII:vWF was not significantly different between blood group A and blood group O subjects. Moreover, the FVIII:vWF ratios for each of the 4 genotypes studied demonstrated no significant differences (A¹A¹ 0.95, A¹O¹ 1.01, A²O¹ 1.01, and O¹O¹ 0.99).

A Transferase Enzyme Expression
From Figure 1, it is clear that the ABO genotype exerts a dosage effect on plasma vWF:Ag and FVIII:Ag levels, such that AO heterozygotes have levels intermediate between AA and O. To determine whether the ABO genotype mediates this by exerting a dosage effect on A transferase enzyme expression, we measured plasma A transferase activities. A transferase levels were highest in homozygous A¹A¹ individuals and significantly higher in A¹O¹ individuals than in A²O¹ individuals (P=0.002 and P<0.0001, Mann Whitney; Figure 2). The level of A transferase activity detected in A²O¹ individuals was minimal. The A² allele differs from the A¹ allele in having deletion of the normal stop codon. Consequently, the A transferase protein encoded is longer and less efficient in converting H structures into A antigens. No A transferase activity was detectable in group O plasmas.

View larger version (16K): [in this window] [in a new window]   Figure 2. Plasma A transferase enzyme activity against ABO genotype. Plasma A transferase activity was measured in each plasma sample by using UDP-[3H]N-acetylgalactosamine (3H GalNAC) as donor sugar at pH 7.4. The mean A transferase level for each genotype is shown. A transferase activity was greatest in A1A1 individuals and significantly higher in A1O1 than in A2O1 individuals (P=0.002 and P<0.0001, respectively; Mann-Whitney).

Blood Group A Genotype and vWF A Antigenic Determinant Expression
To determine whether genotype at the ABO locus exerts a quantitative effect on vWF ABH antigen expression, the amount of AvWF was measured in the 158 plasma samples. The amount of AvWF was significantly higher in A¹A¹ than in A¹O¹ individuals and higher in A¹O¹ than in A²O¹ individuals (P=0.001 and P<0.001, Mann-Whitney; Figure 3). Furthermore, AvWF levels demonstrated marked variation even among individuals of the same genotype. AvWF in A²O¹ individuals was low, in keeping with the low levels of A transferase activity previously demonstrated in this group.

View larger version (16K): [in this window] [in a new window]   Figure 3. Amount of A antigenic determinant expressed on circulating vWF against ABO genotype. The amount of AvWF was measured by ELISA. A reference pooled group A plasma was assayed to produce a standard curve for each ELISA. Plasma vWF had significantly more AvWF in homozygous A1A1 individuals than in A1O1 or A2O1 individuals (P=0.001 and P<0.001, respectively; Mann-Whitney).

Among A¹A¹ and A¹O¹ individuals, in whom accurate quantification of AvWF was possible, AvWF levels were strongly correlated with plasma levels of A transferase activity (P<0.001 and =0.54, Spearman rank correlation; Figure 4). AvWF levels demonstrated the same trend as plasma vWF:Ag levels (A¹A¹>A¹O¹>A²O¹>O¹O¹). However, among A¹A¹ and A¹O¹ individuals, there was no correlation between AvWF and the plasma vWF:Ag level (P=0.24 and =0.15, Spearman rank correlation).

View larger version (15K): [in this window] [in a new window]   Figure 4. Amount of AvWF against plasma A transferase activity. For each individual, plasma A transferase activity was measured by using 3H GalNAC as donor sugar, and the amount of AvWF was measured by ELISA. A1A1 individuals are represented by open circles, and A1O1 individuals are represented by filled circles. The amount of AvWF clearly demonstrates a strong correlation with the amount of A transferase activity present in plasma (P<0.001 and =0.54, Spearman rank correlation).

ABO Genotype and vWF H Antigenic Determinant Expression
The amount of HvWF was highest in group O individuals (Figure 5). Among blood group A individuals, the pattern of HvWF expression was A²O¹>A¹O¹>A¹A¹. There was a strong inverse correlation between HvWF and A transferase activity (P<0.001 and =-0.59, Spearman rank correlation) and between HvWF and AvWF levels (P=0.0017 and =-0.36, Spearman rank correlation).

View larger version (19K): [in this window] [in a new window]   Figure 5. Amount of HvWF against ABO genotype. The amount of HvWF was measured by ELISA. A reference pooled plasma was assayed at 4 dilutions to produce a standard curve for each ELISA. Plasma vWF in homozygous O1O1 individuals had significantly more HvWF. The amount of HvWF was inversely correlated with the amount of AvWF (data not shown).

Among A²O¹ and O¹O¹ individuals, HvWF was inversely correlated with plasma vWF levels (P=0.015 and =-0.26, Spearman rank correlation). Moreover, within the group O individuals, HvWF also demonstrated a statistically significant inverse correlation with plasma vWF level (P=0.021 and =-0.27, Spearman rank correlation; Figure 6).

View larger version (13K): [in this window] [in a new window]   Figure 6. Amount of HvWF against plasma vWF:Ag level in group O individuals. Plasma vWF:Ag levels were determined by ELISA. The amount of HvWF was measured by modified sandwich ELISA. Among group O1O1 individuals, the plasma vWF level demonstrated a significant correlation with the amount of HvWF (P=0.021 and =-0.27, Spearman rank correlation).

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
The ABO blood group exerts a major quantitative effect on plasma FVIII-vWF complex levels. Evidence from twin studies has shown that 66% of the variation observed in plasma vWF levels is genetically determined and that 30% of the total variation is explained by the effect of the ABO blood group. Orstavik et al¹³ have clearly demonstrated that the effect on plasma FVIII concentration is the result of a primary effect of the ABO locus on plasma vWF levels. The mechanism underlying the relationship between the ABO blood group and vWF levels remains unknown. Theoretically, the ABO blood group may influence the rate of vWF synthesis or secretion by endothelial cells or the rate of plasma clearance.

Our results are consistent with previous reports in demonstrating that both vWF:Ag and FVIII:Ag are highest in A¹A¹ individuals and are significantly higher in A¹O¹ than in A²O¹ or O¹O¹ individuals.^17,18 As previously described, the normal range of vWF:Ag found in group O individuals clearly extends below 50 IU/dL (mean±2 SD, 36 to 157 IU/dL). Unless ABO group–specific vWF:Ag reference ranges are used, normal group O individuals will be diagnosed as type I von Willebrand disease.

Although the ABO blood group exerts a marked effect on plasma vWF:Ag and FVIII:Ag levels, there is no significant difference in the FVIII:vWF ratio across the ABO genotypes studied. FVIII and vWF circulate in plasma as a noncovalent 1:1 complex.^30,31 If the effect of ABO on plasma FVIII-vWF levels is mediated by the rate of clearance of this complex, then the FVIII:vWF ratio should be the same in group O and non-O individuals. Conversely, if group O individuals have reduced endothelial cell vWF synthesis/secretion, then the ratio of FVIII:vWF in these individuals would be higher than that observed in non-O groups. Eikenboom et al³² have previously established the validity of this argument by showing that the FVIII:vWF ratio is increased in heterozygous carriers of vWF null alleles, in whom low vWF is due to a synthetic defect. Thus, concordance in this study of the FVIII:vWF ratio in different ABO blood group phenotypes and genotypes supports the hypothesis that the effect of the ABO locus on plasma FVIII and vWF levels is mediated by altering plasma clearance rates.

Although ABH antigenic determinants have been demonstrated on the N-linked glycans of vWF,^19,22 there is no previous evidence to suggest that these determinants are responsible for mediating the effect of ABO blood group on plasma vWF levels. However, it is interesting that platelet-derived vWF lacks ABH determinants and that platelet vWF concentration is not influenced by the ABO group.³³ Our results demonstrate that the quantity of A antigenic determinant present on vWF (AvWF) is related to the ABO genotype of the individual. AvWF is highest in A¹A¹ individuals and is significantly higher in A¹O¹ than in A²O¹ individuals. Moreover, the amount of A antigen expressed on vWF demonstrates a significant correlation with the plasma level of A transferase activity. Curtis et al³⁴ have recently shown that plasma levels of A and B glycosyltransferases are also correlated with the amount of A and B antigens expressed on platelets and various glycoproteins, including glycoprotein IIb. We have demonstrated that the highest levels of plasma A transferase activity are found in homozygous A¹A¹ individuals, with significantly lower levels in A¹O¹ individuals and barely detectable levels in A²O¹ individuals. This dosage effect of the ABO locus on glycosyltransferase expression levels has not been previously described, although it is well recognized for the Lewis blood group locus.³⁵

The strong correlation that we have demonstrated between plasma A transferase activity and AvWF levels could be explained by A antigen addition to vWF after secretion from endothelial cells. However, this is unlikely because the level of appropriate sugar nucleotide donors and of other cofactors in normal plasma is low.³⁴ Moreover, the majority of human endothelial cells express A and B transferases appropriately, and renal vein endothelial extracts contain vWF with ABH blood group determinants.²⁹ Last, vWF secreted by endothelial cells in response to 1-deamino-D-argnine vasopressin has been reported to have normal A antigenic determinant expression.²⁸ Therefore, it seems likely that the explanation for the correlation is that A transferase expression levels within endothelial cells parallel the levels observed in plasma.

H antigen is converted into A antigen by the addition of N-acetylgalactosamine.³⁶ Consequently, one would expect the amount of H and A antigenic determinants expressed on N-linked oligosaccharide chains to be inversely related. Indeed, our results demonstrate that the amount of HvWF is highest in O¹O¹ individuals and lowest in A¹A¹ homozygotes. Overall, HvWF is inversely related to plasma A transferase activity and AvWF levels. Thus, the ABO blood group locus determines A transferase expression, which subsequently determines the amount of H antigen on vWF that is converted into A.

Distinct hepatic receptors with specificity for fucose and for N-acetylgalactosamine have been characterized.^37,38 Our observation that the amount of H antigen expressed on vWF is inversely correlated with vWF levels in group O individuals (in whom there is no confounding A antigen expression) and in group A²O¹ individuals suggests that accelerated clearance via a fucose-mediated mechanism may be the underlying explanation. Thus, in group A and B individuals, conversion of N-linked H antigen into A and B structures reduces H antigenic expression, resulting in delayed vWF clearance and higher plasma levels. Although the amount of H antigen expressed on vWF in A¹O¹ individuals is extremely low, the difference in vWF levels from other groups suggests that it is still significant. The correlation coefficient for the relationship between HvWF and vWF:Ag level is relatively low, indicating that only a small proportion of the total variation in vWF:Ag levels in these individuals can be attributed to the amount of H antigen expression. This finding is consistent with the estimation of Orstavik et al,¹³ who reported that only 30% of the total genetic variation in plasma vWF:Ag levels is explained by the ABO locus. In view of the other factors known to influence plasma vWF levels (eg, the vWF gene,³⁹ 5' single nucleotide polymorphisms of the vWF gene,⁴⁰ Secretor blood group locus,⁴¹ and acute-phase reaction⁴²), one would be surprised to find a stronger correlation coefficient. Addition of the A²O¹ individuals results in a reduction in the P value with no change in the value, adding weight to this conclusion. The virtual absence of H from A¹A¹ vWF suggests that other mechanisms of clearance exist, although the lack of correlation between AvWF and plasma vWF concentration indicates that it is unlikely to be via an N-acetylgalactosamine receptor.

A marked variation in the amount of A antigen expressed on the surface of platelets of normal group A individuals has recently been reported.³⁴ Our data confirm that there is also a large variation in the amount of A antigen on plasma vWF and indicate that this variation has at least 2 origins. First, we have shown that the amount of H antigen expressed on group O vWF can vary over a wide range. Because it appears that most H structures are converted into A structures in A¹A¹ and A¹O¹ individuals, a variation in the amount of H availability is likely to be a major determinant of A antigen expression. The origin of the variation in H antigen expression may lie among the many glycosyltransferases and glycosidases required for synthesis of complex N-linked glycan structures.⁴³ Second, we have shown that A transferase enzyme levels can vary significantly even within group A genotypes. The regulatory elements controlling A transferase expression have not been fully characterized. Moreover, the wide variation that we have observed in plasma A transferase activity levels could also result from differences in the rates of cellular secretion or plasma clearance of the transferase enzyme.

	Acknowledgments

 
Dr O’Donnell was sponsored by a Medical Research Council Training Fellowship award. The authors would like to thank the staff of the Wessex Regional Transfusion Center, Southampton, for their help.

Received October 8, 2001; accepted October 8, 2001.

	References

Top Abstract Introduction Methods Results Discussion References

 

1. Ruggeri ZM, Zimmerman TS. The complex multimeric composition of factor VIII/vWF. Blood. 1981; 57: 1140–1143.[Abstract/Free Full Text]
2. Sadler JE. von Willebrand factor. Ann Rev Biochem. 1998; 67: 395–424.[CrossRef][Medline] [Order article via Infotrieve]
3. Ruggeri ZM, Zimmerman TS. von Willebrand factor and von Willebrand disease. Blood. 1987; 70: 895–904.[Abstract/Free Full Text]
4. Mayadas TN, Wagner DD. von Willebrand factor biosynthesis and processing. Ann N Y Acad Sci. 1991; 614: 153–166.[Medline] [Order article via Infotrieve]
5. Nichols TC, Samama CM, Bellinger DA, Roussi J, Reddick RL, Bonneau M, Read MS, Bailliart O, Koch GG, Vaiman M, et al. Function of von Willebrand factor after crossed bone marrow transplantation between normal and von Willebrand disease pigs: effect on arterial thrombosis in chimeras. Proc Natl Acad Sci U S A. 1995; 92: 2455–2459.[Abstract/Free Full Text]
6. Nichols WC, Ginsberg D. von Willebrand disease. Rev Mol Med. 1997; 76: 1–20.
7. Meade TW, Cooper JA, Stirling Y, Howarth DJ, Ruddock V. Factor VIII, ABO blood group and the incidence of ischaemic heart disease. Br J Haematol. 1994; 88: 601–607.[Medline] [Order article via Infotrieve]
8. Medalie JH, Levene C, Papier C, Goldbourt U, Dreyfuss F, Oron D, Neufeld H, Riss E. Blood groups, myocardial infarction and angina pectoris among 10,000 adult males. N Engl J Med. 1971; 285: 1348–1353.[Medline] [Order article via Infotrieve]
9. Koster T, Blann AD, Briet E, Vandenbroucke JP, Rosendaal FR. Role of clotting factor VIII in effect of von Willebrand factor on occurrence of deep-vein thrombosis. Lancet. 1995; 345: 152–155.[CrossRef][Medline] [Order article via Infotrieve]
10. O’Donnell J, Mumford AD, Manning RA, Laffan M. Elevation of FVIII:C in venous thromboembolism is persistent and independent of the acute phase response. Thromb Haemost. 2000; 83: 10–13.[Medline] [Order article via Infotrieve]
11. Lowe J. The blood group-specific human glycosyltransferases. Baillieres Clin Haematol. 1993; 6: 465–490.[Medline] [Order article via Infotrieve]
12. Preston AE, Barr A. The plasma concentration of factor VIII in the normal population. Br J Haematol. 1964; 10: 238–245.[Medline] [Order article via Infotrieve]
13. Orstavik K, Magnus P, Reisner H, Berg K, Graham J, Nance W. Factor VIII and factor IX in a twin population: evidence for a major effect of ABO locus on factor VIII level. Am J Hum Genet. 1985; 37: 89–101.[Medline] [Order article via Infotrieve]
14. McCallum CJ, Peake IR, Newcombe RG, Bloom AL. Factor VIII levels and blood group antigens. Thromb Haemost. 1983; 50: 757.[Medline] [Order article via Infotrieve]
15. Mohanty D, Ghosh K, Marwaha N, Kaur S, Chauhan AP, Das KC. Major blood group antigens: a determinant of factor VIII levels in blood? Thromb Haemost. 1984; 51: 414.[Medline] [Order article via Infotrieve]
16. Cox Gill JC, Endres-Brooks J, Bauer PJ, Marks WJ, Montgomery RR. The effect of ABO blood group on the diagnosis of von Willebrand disease. Blood. 1987; 69: 1691–1695.[Abstract/Free Full Text]
17. Shima M, Fujimura Y, Nishiyama T, Tsujiucchi T, Narita N, Matsui T, Titani K, Katayama M, Yammamoto F, Yoshioka A. ABO blood group genotype and plasma von Willebrand factor in normal individuals. Vox Sang. 1995; 68: 236–240.[Medline] [Order article via Infotrieve]
18. Souto JC, Almsay L, Muniz-Diaz E, Soria JM, Borrell M, Bayern L, Mateo J, Madoz P, Stone W, Blangero J, et al. Functional effects of the ABO locus polymorphism on plasma levels of von Willebrand factor, factor VIII, and activated partial thromboplastin time. Arterioscler Thromb Vasc Biol. 2000; 20: 2024–2028.[Abstract/Free Full Text]
19. Matsui T, Titani K, Mizuochi T. Structures of the asparagine-linked oligosaccharide chains of human von Willebrand factor. J Biol Chem. 1992; 267: 8723–8731.[Abstract/Free Full Text]
20. Mohlke KL, Purkayastha AA, Westrik RJ, Smith PL, Petryniak B, Lowe JB, Ginsberg D. Mvwf, a dominant modifier of murine von Willebrand factor, results from altered lineage-specific expression of a glycosyltransferase. Cell. 1998; 96: 111–120.
21. Eastlund T. The histo-blood group ABO system and tissue transplantation. Transfusion. 1998; 38: 975–988.[CrossRef][Medline] [Order article via Infotrieve]
22. Matsui T, Fujimura Y, Nishida S, Titani K. Human plasma ²-macroglobulin and von Willebrand factor possess covalently linked ABO(H) blood group antigens in subjects with corresponding ABO phenotype. Blood. 1993; 82: 663–668.[Abstract/Free Full Text]
23. Sodetz J, Paulson J, McKee P. Carbohydrate composition and identification of blood group A, B, and H oligosaccharide structures on human factor VIII/von Willebrand factor. J Biol Chem. 1979; 254: 10754–10760.[Free Full Text]
24. Hironaka T, Furukawa K, Esmon PC, Fournel MA, Sawada S, Kato M, Minaga T, Kobata A. Comparative study of the sugar chains of factor VIII purified from human plasma and from the culture media of recombinant baby hamster kidney cells. J Biol Chem. 1992; 267: 8012–8020.[Abstract/Free Full Text]
25. Ogasawara K, Bannai M, Saitou N, Bannai M, Nakata K, Takkenaka M, Fujisawa K. Extensive polymorphism of ABO blood group gene: three major lineages of the alleles for the common ABO phenotypes. Hum Genet. 1996; 97: 777–783.[Medline] [Order article via Infotrieve]
26. O’Keefe DS, Dobrovic A. A rapid and reliable PCR method for genotyping the ABO blood group: A² and O² alleles. Hum Mutat. 1996; 8: 358–361.[CrossRef][Medline] [Order article via Infotrieve]
27. Nishimukai H, Fukumori Y, Okiura I, Shinomiya T, Ohnoki S, Shibata H, Vogt U. Genotyping of the ABO blood group system: analysis of nucleotide position 802 by PCR-RFLP and the distribution of ABO genotypes in a German population. Int J Legal Med. 1996; 109: 90–93.[CrossRef][Medline] [Order article via Infotrieve]
28. Brown SA, Bowen DJ, Ellis D, Collins PW. von Willebrand factor blood group-associated antigens: immunological detection and distribution. Blood. 1997; 90: 466a.Abstract.
29. Matsui T, Shimoyama T, Matsumoto M, Fujimura Y, Takemoto T, Sako M, Hamako J, Titani K. ABO blood group antigens on human plasma von Willebrand factor after ABO-mismatched bone marrow transplantation. Blood. 1999; 94: 2895–2900.[Abstract/Free Full Text]
30. Vlot AJ, Koppelman SJ, Bouma BN, Sixma JJ. Factor VIII and von Willebrand factor. Thromb Haemost. 1998; 79: 456–465.[Medline] [Order article via Infotrieve]
31. Vlot AJ, Koppelman SJ, Meijers JC, Dama C, van den Berg HM, Bouma BN, Sixma JJ, Willems GM. Kinetics of factor VIII–von Willebrand factor association. Blood. 1996; 87: 1809–1816.[Abstract/Free Full Text]
32. Eikenboom JC, Castaman G, Kamphuisen PW, Bertina RM. Factor VIII/von Willebrand factor in healthy controls and in heterozygous carriers of vWF null alleles and von Willebrand disease type 2N mutations. Thromb Haemost. 1999; 83: 686aAbstract.
33. Sweeney JD, Hoernig LA. Intraplatelet von Willebrand factor and ABO blood group. Thromb Res. 1992; 68: 393–398.[CrossRef][Medline] [Order article via Infotrieve]
34. Curtis BR, Edwards JT, Hessner MJ, Klein JP, Aster RH. Blood group A and B antigens are strongly expressed on platelets of some individuals. Blood. 2000; 96: 1574–1581.[Abstract/Free Full Text]
35. Orntoft TF, Vestergaard EM, Holmes EH, Sinkbaek J, Wolf H. Influence of Lewis 1-3/4-L fucosyltransferase (FUT3) gene mutations on enzyme activity, erythrocyte phenotyping and circulating tumor marker Sialyl-Lewis a levels. J Biol Chem. 1997; 271: 32260–32268.
36. Lowe J. The blood group-specific human glycosyltransferases. Baillieres Clin Haematol. 1993; 6: 465–489.[Medline] [Order article via Infotrieve]
37. Ashwell G, Harford J. Carbohydrate-specific receptors of the liver. Annu Rev Biochem. 1982; 51: 531–554.[CrossRef][Medline] [Order article via Infotrieve]
38. Biesson EA, Bakkeren HF, Beuting DM, Kuiper J, Van Berkel TJ. Ligand size is a major determinant of high affinity binding of fucose- and galactose-exposing lipoproteins by the hepatic fucose receptor. Biochem J. 1994; 299: 291–296.[Medline] [Order article via Infotrieve]
39. Eikenboom JC, Ploos van Amstel HK, Reitsma PH, Briet E. Mutations in severe, type III von Willebrand disease in the Dutch population: candidate missense and nonsense mutations associated with reduced levels of von Willebrand factor messenger RNA. Thromb Haemost. 1992; 68: 448–454.[Medline] [Order article via Infotrieve]
40. Keightley AM, Lam YM, Brady JN, Cameron CL, Lillicrap D. Variation at the von Willebrand factor gene locus is associated with plasma vWF:Ag levels: identification of three novel single nucleotide polymorphisms in the vWF gene promoter. Blood. 1999; 93: 4277–4283.[Abstract/Free Full Text]
41. O’Donnell J, Boulton FE, Manning R, Laffan M. Genotype at the Secretor blood group locus is a determinant of plasma von Willebrand factor level. Br J Haematol. In press.
42. Conlan MG, Folsom AR, Finch A, Davis CE, Sorlie P, Marcucci G, Wu KK. Associations of factor VIII and von Willebrand factor with age, race, sex, and risk factors for atherosclerosis: the Atherosclerosis Risk in Communities (ARIC) Study. Thromb Haemost. 1993; 70: 380–385.[Medline] [Order article via Infotrieve]
43. Varki A. Biological roles of oligosaccharide: all of the theories are correct. Glycobiology. 1993; 3: 97–130.[Abstract/Free Full Text]

This article has been cited by other articles:

				L. Gallinaro, M. G. Cattini, M. Sztukowska, R. Padrini, F. Sartorello, E. Pontara, A. Bertomoro, V. Daidone, A. Pagnan, and A. Casonato A shorter von Willebrand factor survival in O blood group subjects explains how ABO determinants influence plasma von Willebrand factor Blood, April 1, 2008; 111(7): 3540 - 3545. [Abstract] [Full Text] [PDF]

				T. A. J. McKinnon, A. C. K. Chion, A. J. Millington, D. A. Lane, and M. A. Laffan N-linked glycosylation of VWF modulates its interaction with ADAMTS13 Blood, March 15, 2008; 111(6): 3042 - 3049. [Abstract] [Full Text] [PDF]

				N. C. Sousa, J. M. Anicchino-Bizzacchi, M. F. Locatelli, V. Castro, and M. L. Barjas-Castro The relationship between ABO groups and subgroups, factor VIII and von Willebrand factor Haematologica, February 1, 2007; 92(2): 236 - 239. [Abstract] [Full Text] [PDF]

				J. G. Kang, H. J. Ahn, G. S. Kim, T. S. Hahm, J. J. Lee, M. S. Gwak, and S. J. Choi The Hemostatic Profiles of Patients with Type O and Non-O Blood After Acute Normovolemic Hemodilution with 6% Hydroxyethyl Starch (130/0.4) Anesth. Analg., December 1, 2006; 103(6): 1543 - 1548. [Abstract] [Full Text] [PDF]

				J. S. O'Donnell, T. A. J. McKinnon, J. T. B. Crawley, D. A. Lane, and M. A. Laffan Bombay phenotype is associated with reduced plasma-VWF levels and an increased susceptibility to ADAMTS13 proteolysis Blood, September 15, 2005; 106(6): 1988 - 1991. [Abstract] [Full Text] [PDF]

				J. T. B. Crawley, J. K. Lam, J. B. Rance, L. R. Mollica, J. S. O'Donnell, and D. A. Lane Proteolytic inactivation of ADAMTS13 by thrombin and plasmin Blood, February 1, 2005; 105(3): 1085 - 1093. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by O’Donnell, J. Articles by Laffan, M. A. Search for Related Content PubMed PubMed Citation Articles by O’Donnell, J. Articles by Laffan, M. A. Related Collections Deep vein thrombosis Coagulation Thrombophilia Thrombosis risk factors