Search

Background: Metabolomics, defined as the comprehensive identification and quantification of low-molecular-weight metabolites to be found in a biological sample, has been put forward as a potential tool for classifying individuals according to their risk of coronary heart disease (CHD). Here, we investigated whether a single-point blood measurement of the metabolome is associated with and predictive for the risk of CHD., Methods and Results: We obtained proton nuclear magnetic resonance spectra in 79 cases who developed CHD during follow-up (median 8.1 years) and in 565 randomly selected individuals. In these spectra, 100 signals representing 36 metabolites were identified. Applying least absolute shrinkage and selection operator regression, we defined a weighted metabolite score consisting of 13 proton nuclear magnetic resonance signals that optimally predicted CHD. This metabolite score, including signals representing a lipid fraction, glucose, valine, ornithine, glutamate, creatinine, glycoproteins, citrate, and 1.5-anhydrosorbitol, was associated with the incidence of CHD independent of traditional risk factors (TRFs) (hazard ratio 1.50, 95% CI 1.12-2.01). Predictive performance of this metabolite score on its own was moderate (C-index 0.75, 95% CI 0.70-0.80), but after adding age and sex, the C-index was only modestly lower than that of TRFs (C-index 0.81, 95% CI 0.77-0.85 and C-index 0.82, 95% CI 0.78-0.87, respectively). The metabolite score was also associated with prevalent CHD independent of TRFs (odds ratio 1.59, 95% CI 1.19-2.13)., Conclusion: A metabolite score derived from a single-point metabolome measurement is associated with CHD, and metabolomics may be a promising tool for refining and improving the prediction of CHD., (Copyright © 2014 Mosby, Inc. All rights reserved.)

Aims: To re-estimate the SCORE risk function using individual data on risk factors and coronary heart disease (CHD) incidence from the Dutch Cardiovascular Registry Maastricht (CAREMA) population-based cohort study; to evaluate changes that may improve risk prediction after re-estimation; and to compare the performance of the resulting CAREMA risk function with that of existing risk scores., Methods and Results: The cohort consisted of 21,148 participants, born in 1927-1977 and randomly sampled from the Maastricht region in 1987-1997. After follow-up (median 10.9 years), 783 incident CHD cases occurred. Model performance was assessed by discrimination and calibration. The additional value of including other risk factors or current risk factors in a different manner was evaluated using the net reclassification index (NRI). The c statistic of the re-estimated SCORE model was 0.799 (95% CI 0.782-0.816). Separating the total/high-density lipoprotein (HDL) cholesterol ratio into total and HDL cholesterol levels did not improve the c statistic (p = 0.22), but reclassified 6.0% of the participants into a more appropriate risk category (p < 0.001) compared with the re-estimated model. The resulting CAREMA function reclassified 28% of the participants into a more appropriate risk category than the Framingham score. Compared with the SCORE functions for high- and low-risk regions, the NRIs were 28% and 35%, respectively, which can largely be explained by the difference in outcome definition (CHD incidence vs. CHD mortality)., Conclusion: In this Dutch population, a re-estimated SCORE function with total and HDL cholesterol levels instead of the cholesterol ratio can be used for the risk prediction of CHD incidence.

Background: Genome-wide association studies (GWAS) have identified many single-nucleotide polymorphisms (SNPs) associated with coronary heart disease (CHD) or CHD risk factors (RF). Using a case-cohort study within the prospective Cardiovascular Registry Maastricht (CAREMA) cohort, we tested if genetic risk scores (GRS) based on GWAS-identified SNPs are associated with and predictive for future CHD., Methods and Results: Incident cases (n=742), that is, participants who developed CHD during a median follow-up of 12.1 years (range, 0.0-16.9 years), were compared with a randomly selected subcohort of 2221 participants selected from the total cohort (n=21 148). We genotyped 179 SNPs previously associated with CHD or CHD RF in GWAS as published up to May 2, 2011. The allele-count GRS, composed of all SNPs, the 153 RF SNPs, or the 29 CHD SNPs were not associated with CHD independent of CHD RF. The weighted 29 CHD SNP GRS, with weights obtained from GWAS for every SNP, were associated with CHD independent of CHD RF (hazard ratio, 1.12 per weighted risk allele; 95% confidence interval, 1.04-1.21) and improved risk reclassification with 2.8% (P=0.031). As an exploratory approach to achieve weighting, we performed least absolute shrinkage and selection operator (LASSO) regression analysis on all SNPs and the CHD SNPs. The CHD LASSO GRS performed equal to the weighted CHD GRS, whereas the Overall LASSO GRS performed slightly better than the weighted CHD GRS., Conclusions: A GRS composed of CHD SNPs improves risk prediction when adjusted for the effect sizes of the SNPs. Alternatively LASSO regression analysis may be used to achieve weighting; however, validation in independent populations is required.

Background: Prevalence of metabolic factors such as diabetes, hypertension, obesity, HDL and total cholesterol that are associated with an increased risk of coronary heart disease (CHD) is increasing worldwide. However, less is known about combinations of these factors that are associated with the highest CHD risk. Therefore, the associations between combinations of these metabolic factors and the incidence of CHD, acute myocardial infarction (AMI), and unstable angina pectoris (UAP) were studied in the Cardiovascular Registry Maastricht (CAREMA) cohort study., Methods: The CAREMA study consists of 21,148 participants, born in 1927-1977 and randomly sampled from Maastricht and surrounding communities in 1987-1997. At baseline, all participants completed a self-administered questionnaire. Height, weight, blood pressure, total and HDL cholesterol levels were measured during a physical examination. After follow-up of maximally 16.9 years, 780 CHD, 437 AMI, and 286 UAP cases of first occurrence were registered. Incidence rate ratios (RRs) were estimated using Cox proportional hazards models adjusted for age, sex, smoking, and alcohol consumption., Results: Compared with subjects without any of the metabolic factors, the RRs of CHD were 2.37, 4.34, and 7.36 for subjects with 1, 2, or ≥ 3 metabolic factors, respectively. These RRs were higher for AMI but lower for UAP. Especially combinations of metabolic factors that included diabetes or both a low HDL (≤ 0.9 mmol/L in men; ≤ 1.0 mmol/L in women) and high total cholesterol (≥ 6.21 mmol/L) were associated with increased risks., Conclusion: The risk of total CHD, AMI, and UAP varies considerably between different combinations of metabolic factors., (Copyright © 2011 Elsevier Ireland Ltd. All rights reserved.)

Background: Intakes of n-3 polyunsaturated fatty acids (PUFAs), especially EPA (C20:5n-3) and DHA (C22:6n-3), are known to prevent fatal coronary heart disease (CHD). The effects of n-6 PUFAs including arachidonic acid (C20:4n-6), however, remain unclear. δ-5 and δ-6 desaturases are rate-limiting enzymes for synthesizing long-chain n-3 and n-6 PUFAs. C20:4n-6 to C20:3n-6 and C18:3n-6 to C18:2n-6 ratios are markers of endogenous δ-5 and δ-6 desaturase activities, but have never been studied in relation to incident CHD. Therefore, the aim of this study was to investigate the relation between these ratios as well as genotypes of FADS1 rs174547 and CHD incidence., Methods: We applied a case-cohort design within the CAREMA cohort, a large prospective study among the general Dutch population followed up for a median of 12.1 years. Fatty acid profile in plasma cholesteryl esters and FADS1 genotype at baseline were measured in a random subcohort (n = 1323) and incident CHD cases (n = 537). Main outcome measures were hazard ratios (HRs) of incident CHD adjusted for major CHD risk factors., Results: The AA genotype of rs174547 was associated with increased plasma levels of C204n-6, C20:5n-3 and C22:6n-3 and increased δ-5 and δ-6 desaturase activities, but not with CHD risk. In multivariable adjusted models, high baseline δ-5 desaturase activity was associated with reduced CHD risk (P for trend = 0.02), especially among those carrying the high desaturase activity genotype (AA): HR (95% CI) = 0.35 (0.15-0.81) for comparing the extreme quintiles. High plasma DHA levels were also associated with reduced CHD risk., Conclusion: In this prospective cohort study, we observed a reduced CHD risk with an increased C20:4n-6 to C20:3n-6 ratio, suggesting that δ-5 desaturase activity plays a role in CHD etiology. This should be investigated further in other independent studies.

This Viewpoint article addresses comments made on our original article describing a symbolic system for the depiction of N- and O-linked carbohydrate structures and proposes a method for extending the symbol set to include monosaccharides commonly found in carbohydrates present in bacteria and plants. As before, basic monosaccharides are shown by shape with one or more additions such as solid fill or additions of lines, crosses or dots to represent functional groups. The use of colour to differentiate constituent monosaccharides is avoided, thus enabling the system to be used in a variety of formats. Linkage and anomericity are shown by the angle and type of line connecting the symbols. In this extended version, new symbols are proposed for additional hexoses and it is proposed that pyranose and furanose forms of the monosaccharides could be shown by solid or broken outlines to the symbols. Conventions for depicting the presence of multiple functional groups such as deoxy-(NH(2))(2) are also discussed. It is hoped that these proposals will stimulate discussion so that a consensus can be reached as to how the glycobiology community can best convey complex information in a simple manner., (Copyright © 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.)

Background: Few studies investigated the association between smoking, alcohol consumption, or physical activity and the risk of unstable angina pectoris (UAP), while the strength of these associations may differ compared to other coronary diseases such as acute myocardial infarction (AMI). Therefore, we investigated whether the associations of these lifestyle factors with UAP differed from those with AMI. Additionally, we investigated whether these effects differed between subjects with and without a family history of myocardial infarction (MI)., Methods: The CAREMA study consists of 21,148 persons, aged 20-59 years at baseline and randomly sampled from the Maastricht region in 1987-1997. At baseline, all participants completed a self-administered questionnaire. After follow-up of maximally 16.9 years, 420 AMI and 274 UAP incident cases were registered. Incidence rate ratios (RRs) were estimated using Cox proportional hazards models., Results: For both diseases, smoking increased the risk while alcohol consumption was associated with a protective effect. Associations with both risk factors were stronger for AMI than UAP, although this difference was only statistically significant for smoking. In men, an inverse association was found with physical activity during leisure time which seemed to be stronger for the risk of UAP than of AMI. On the contrary, physical activity during leisure time was associated with an increased risk of both AMI and UAP in women which seemed to be weaker for UAP than for AMI. Except for occupational physical activity in women, no significant interactions on a multiplicative scale were found between the lifestyle factors and family history of MI. Nevertheless, the highest risks were found in subjects with both a positive family history and the most unfavorable level of the lifestyle factors., Conclusions: The strength of the associations with the lifestyle factors did not differ between AMI and UAP, except for smoking. Furthermore, the effects of the lifestyle factors on the risk of both coronary diseases were similar for subjects with and without a positive family history., (© 2011 Merry et al; licensee BioMed Central Ltd.)

Symbolic diagrams are commonly used to depict N- and O-linked glycans but there is no general consensus as to how individual constituent monosaccharides or linkages are shown. This article proposes a system that avoids ambiguities inherent in most other systems and is appropriate for both hand drawing and computer applications. Constituent monosaccharides are depicted by shapes modified to show OAc, deoxy, etc. Linkage is indicated by the bond angle and anomericity by solid (beta) or dashed (alpha) lines.

Incidence rates of cardiovascular diseases are often estimated by linkage to hospital discharge and mortality registries. The validity depends on the quality of the registries and the linkage. Therefore, we validated incidence rates of coronary heart disease (CHD), acute myocardial infarction, unstable angina pectoris, and heart failure, estimated by this method, against the disease registry of the cardiovascular registry Maastricht cohort study. The cohort consists of 21,148 persons, born between 1927 and 1977, who were randomly sampled from Maastricht and surrounding communities in 1987-1997. Incident cases were identified by linkage to the Netherlands causes of death registry and either the hospital discharge registry (HDR) or the cardiology information system (CIS) of the University Hospital Maastricht. Sensitivities and positive predictive values were calculated using the CIS-based registry as gold standard. Relatively high sensitivities and positive predictive values were found for CHD (72 and 91%, respectively) and acute myocardial infarction (84 and 97%, respectively). These values were considerably lower for unstable angina pectoris (53 and 78%, respectively) and heart failure (43 and 80%, respectively). A substantial number of cases (14-47%) were found only in the CIS-based registry, because they were missed or miscoded in the HDR-based registry. As a consequence, the incidence rates in the HDR-based registry were considerably lower than in the CIS-based registry, especially for unstable angina pectoris and heart failure. Incidence rates based on hospital discharge and mortality data may underestimate the true incidence rates, especially for unstable angina pectoris and heart failure.

Background: In the last decades, the incidence of oesophageal and gastric cardia adenocarcinoma has increased rapidly in the Western world. We investigated the association between body mass index (BMI), height and risk of oesophageal and gastric cardia adenocarcinoma., Methods: The Netherlands Cohort Study was initiated in 1986. All participants (n = 120,852), aged 55-69 years, completed a self administered questionnaire. Cases were identified through annual record linkage with the Netherlands Cancer Registry. After 13.3 years of follow-up, excluding the first follow-up year, complete data from 4552 subcohort members, 133 oesophageal and 163 gastric cardia adenocarcinomas were available for case-cohort analyses. Incidence rate ratios (RRs) and corresponding 95% confidence intervals were estimated using Cox proportional hazard models., Results: The RRs (95% CI) of oesophageal adenocarcinoma were 1.40 (0.95 to 2.04) and 3.96 (2.27 to 6.88) for overweight (BMI 25.0-29.9 kg/m(2)) and obese subjects (BMI >or=30.0 kg/m(2)), respectively, compared to subjects with normal weight (BMI 20.0-24.9 kg/m(2)). For gastric cardia adenocarcinoma, these RRs were 1.32 (0.94 to 1.85) and 2.73 (1.56 to 4.79). Also change in BMI during adulthood was positively associated with the risk of oesophageal and gastric cardia adenocarcinoma (p trend 0.001 and 0.02, respectively), while no association was found with BMI in early adulthood (p trend 0.17 and 0.17, respectively). None of the tumour types investigated was significantly associated with height., Conclusions: These results confirm higher risks of oesophageal and gastric cardia adenocarcinoma with increasing BMI. This implies that the increasing prevalence of obesity may be one of the explanations for the rising incidence of oesophageal and gastric cardia adenocarcinoma in the Western world.

Urinary glycoproteins are important inhibitors of calcium oxalate crystallization and adhesion of crystals to renal cells, both of which are key mechanisms in kidney stone formation. This has been attributed to glycosylation of the proteins. In South Africa, the black population rarely form stones (incidence < 1%) compared with the white population (incidence 12-15%). A previous study involving urinary prothrombin fragment 1 from both populations demonstrated superior inhibitory activity associated with the protein from the black group. In the present study, we compared N-linked and O-linked oligosaccharides released from urinary prothrombin fragment 1 isolated from the urine of healthy and stone-forming subjects in both populations to elucidate the relationship between glycosylation and calcium oxalate stone pathogenesis. The O-glycans of both control groups and the N-glycans of the black control samples were significantly more sialylated than those of the white stone-formers. This demonstrates a possible association between low-percentage sialylation and kidney stone disease and provides a potential diagnostic method for a predisposition to kidney stones that could lead to the implementation of a preventative regimen. These results indicate that sialylated glycoforms of urinary prothrombin fragment 1 afford protection against calcium oxalate stone formation, possibly by coating the surface of calcium oxalate crystals. This provides a rationale for the established roles of urinary prothrombin fragment 1, namely reducing the potential for crystal aggregation and inhibiting crystal-cell adhesion by masking the interaction of the calcium ions on the crystal surface with the renal cell surface along the nephron.

mRNA, and latent and active levels MMP-2 and -9 were higher in tumor tissue compared to normal tissue from 63 patients with colorectal cancer, whereas RECK and EMMPRIN levels were lower. Correlations between mRNA, latent, and active MMP were particular high for MMP-2 in tumor tissue (R(s)=0.6-0.8, P<0.001). For active MMP-2, but not for MMP-9, a significant negative partial correlation (R(p)=-0.440, P<0.001) for RECK was found in tumor tissue, which was confirmed by linear regression analysis. In exploratory survival analyses we found that in patients with localized disease the RECK level in normal or tumor tissue had a significant (P=0.017) association with overall survival.

Studies of mucins suggest that the structural effects of O-glycans are restricted to steric interactions between peptide-linked GalNAc residues and adjacent polypeptide residues. It has been proposed, however, that differential O-glycan sialylation alters the structure of the stalk-like region of the T cell co-receptor, CD8, and that this, in turn, modulates ligand binding (Daniels, M. A., Devine, L., Miller, J. D., Moser, J. M., Lukacher, A. E., Altman, J. D., Kavathas, P., Hogquist, K. A., and Jameson, S. C. (2001) Immunity 15, 1051-1061; Moody, A. M., Chui, D., Reche, P. A., Priatel, J. J., Marth, J. D., and Reinherz, E. L. (2001) Cell 107, 501-512). We characterize the glycosylation of soluble, chimeric forms of the alphaalpha- and alphabeta-isoforms of murine CD8 containing the O-glycosylated stalk of rat CD8alphaalpha, and we show that the stalk O-glycans are differentially sialylated in CHO K1 versus Lec3.2.8.1 cells (82 versus approximately 6%, respectively). Sedimentation analysis indicates that the Perrin functions, Pexp, which reflect overall molecular shape, are very similar (1.61 versus 1.54), whereas the sedimentation coefficients (s) of the CHO K1- and Lec3.2.8.1-derived proteins differ considerably (3.73 versus 3.13 S). The hydrodynamic properties of molecular models also strongly imply that the sialylated and non-sialylated forms of the chimera have parallel, equally highly extended stalks ( approximately 2.6 A/residue). Our analysis indicates that, as in the case of mucins, the overall structure of O-glycosylated stalk-like peptides is sialylation-independent and that the functional effects of differential CD8 O-glycan sialylation need careful interpretation.

Calreticulin is a ubiquitous and highly conserved Ca(2+)-binding protein that is involved in intracellular Ca(2+) homeostasis and molecular chaperoning in the endoplasmic reticulum (ER). Plant calreticulin, in contrast to its animal counterpart, is often glycosylated: its N-glycans have been shown so far to be of the high-mannose type, typical of ER-resident glycoproteins. During the characterization of calreticulin from vegetative and reproductive tissues of Liriodendron tulipifera L., we gained some biochemical evidence that prompted us to investigate the monosaccharide composition and primary structure of the calreticulin N-glycans isolated from the ovary of this dicotyledon tree. The structures of the components of the N-glycan pool were elucidated by HPLC analysis and exoglycosidase sequencing, and further confirmed by matrix-assisted laser desorption/ionization mass spectrometry. The 16 identified oligosaccharide structures, which consisted of both the high-mannose and complex type, are indicative of calreticulin glycan processing through the ER-to-Golgi pathway up to the medial and trans Golgi stacks. Approximately 45% of calreticulin glycan chains are of the complex type, always containing beta(1,2)-xylose, and approximately a third of these also contain alpha(1,3)-fucose in the core. The most complex glycoform harbors the Lewis-a epitope Gal(beta)1-3[Fuc(alpha)1-4]GlcNAc. Immunolocalization of calreticulin with anti-calreticulin antibodies was consistent with protein transit through the Golgi. Thus, although it contains the tetrapeptide HDEL ER retention signal, the reticuloplasmin calreticulin possesses the competence to transit from the ER compartment to the distal Golgi stacks. The final fate of the protein after its complete maturation is still obscure.

Recent advances have elucidated the detailed glycosylation of the prion protein and highlighted the size of the sugars, which shield large areas of the protein and confer some conformational stability on the normal cellular form. The reliability of SDS-PAGE banding patterns of different "glycoforms" as a diagnostics tool has been discussed. The possibility exists that the glycans may play a role in the location of the prion protein on the neuronal cell surface. Alternative topologies and tethering of the prion glycoprotein on the cell membrane affect glycan site occupancy and may play a role in disease pathogenesis.

A sensitive, rapid, quantitative strategy has been developed for O-glycan analysis. A structural database has been constructed that currently contains analytical parameters for more than 50 glycans, enabling identification of O-glycans at the subpicomole level. The database contains the structure, molecular weight, and both normal and reversed-phase HPLC elution positions for each glycan. These observed parameters reflect the mass, three-dimensional shape, and hydrophobicity of the glycans and, therefore, provide information relating to linkage and arm specificity as well as monosaccharide composition. Initially the database was constructed by analyzing glycans released by mild hydrazinolysis from bovine serum fetuin, synthetic glycopeptides, human glycophorin A, and serum IgA1. The structures of the fluorescently labeled sugars were determined from a combination of HPLC data, mass spectrometric composition and mass fragmentation data, and exoglycosidase digestions. This approach was then applied to human neutrophil gelatinase B and secretory IgA, where 18 and 25 O-glycans were identified, respectively, and the parameters of these glycans were added to the database. This approach provides a basis for the analysis of subpicomole quantities of O-glycans from normal levels of natural glycoproteins., ((c)2002 Elsevier Science (USA).)

The initial step in quantitative analysis of O-linked glycans of glycoproteins is to release them in high yield, nonselectively, unmodified, and with a free reducing terminus. In contrast to other techniques, hydrazinolysis can meet these criteria. However, when analyzing pools of O-linked glycans as described in the accompanying article by L. Royle et al. (2002, Anal. Biochem. 304), some peeling of the glycans was observed. Critical steps in the sample preparation and glycan recovery were therefore evaluated by analyzing and identifying both intact O-glycans and degraded products. Synthetic O-glycopeptides were characterized by mass spectrometry. Released glycans were identical to those on the glycopeptide. O-Linked glycans from a range of glycoproteins of increasing complexity, namely, bovine serum fetuin, glycophorin A, and previously uncharacterized glycopeptides isolated from human salivary mucin Muc5B, were also analyzed. Quantitative analysis of the glycan profile confirmed that there was <2% peeling of O-glycans released by hydrazinolysis conditions of 60 degrees C for 6 h, and recovered using the optimised procedure now described. This demonstrated that O-glycans can be prepared by hydrazinolysis without degradation and, as part of an analytical strategy, makes the analysis of O-glycans attached to low-microgram levels of naturally occurring glycoproteins feasible., ((c)2002 Elsevier Science (USA).)

Apolipoprotein(a) (apo(a)) is a multikringle domain glycoprotein that exists covalently linked to apolipoprotein B100 of low density lipoprotein, to form the lipoprotein(a) (Lp(a)) particle, or as proteolytic fragments. Elevated plasma concentrations of apo(a) and its fragments may promote atherosclerosis, but the underlying mechanisms are incompletely understood. The factors influencing apo(a) proteolysis are also uncertain. Here we have used exoglycosidase digestion and mass spectrometry to sequence the Asn (N)-linked and Ser/Thr (O)-linked oligosaccharides of human apo(a). We also assessed the potential role of apo(a) O-glycans in protecting thermolysin-sensitive regions of the polypeptide. Apo(a) contained two major N-glycans that accounted for 17% of the total oligosaccharide structures. The N-glycans were complex biantennary structures present in either a mono- or disialylated state. The O-glycans were mostly (80%) represented by the monosialylated core type 1 structure, NeuNAcalpha2-3Galbeta1-3GalNAc, with smaller amounts of disialylated and non-sialylated O-glycans also detected. Removal of apo(a) O-glycans by sialidase and O-glycosidase treatment dramatically increased the sensitivity of the polypeptide to thermolysin digestion. These studies provide the first direct sequencing data for apo(a) glycans and indicate a novel function for apo(a) O-glycans that is potentially related to the atherogenicity of Lp(a).

The majority of biologically active proteins are glycosylated, therefore any approach to proteomics which fails to address the analysis of oligosaccharides is necessarily incomplete. To appreciate the structure of a glycoprotein fully, to understand the roles for the attached oligosaccharides and to monitor disease associated changes it is necessary to visualise the sugars as well as the protein. To achieve this aim when biological samples are available at the low microgram level or less has involved increasing the sensitivity of the technology for glycan analysis. Since one protein may have many different oligosaccharides attached to it (glycoforms) this is a major technical challenge. CD59, for example, has over 100 different sugars at one N-linked glycosylation site. Applications of recently developed technology suggest that it is now becoming realistic to extend the proteomics analysis of glycoproteins to include details of glycosylation. This is achieved by releasing the N-glycans from the protein in a gel by optimised peptide-N-glycosidase F digestion. The released glycans are then tagged with the fluorophore, 2-amino benzamide. The labelled glycan pools (containing 50-100 femtomoles of glycans) are resolved by predictive normal phase high performance liquid chromatography (HPLC) on an amide based column or by reverse phase HPLC on a C18 column. Preliminary structural assignments are confirmed by exoglycosidase array digestions of the entire glycan pool. Complementary matrix-assisted laser desorption/ionization-mass spectrometry, which requires 10-20 times as much sugar for a single run, can be used where there is sufficient material. This provides a composition analysis but not linkage information.

The external envelope glycoprotein (gp160) of HIV-1 is a candidate for vaccines against AIDS. Most of the surface of the molecule is shielded by carbohydrate and the structures and locations of these glycans may be important in defining the immunogenicity of the viral coat. Here we report a sensitive mapping strategy for profiling and analysing the N-glycosylation of gp160, based on chemical release of glycans, fluorescent labelling and HPLC analysis. This approach has been validated in terms of establishing the reproducibility of all steps in the analytical procedure and on overall reproducibility on a run-to-run and day-to-day basis. The validated analysis technique was used to monitor the consistency of N-glycosylation of one rgp 160 vaccine candidate produced in baby hamster kidney (BHK) cell culture. It was demonstrated that the variation in the glycan profiles of 6 different lots was not statistically significant.

IgG carries bi-antennary N-linked glycans which differ in degrees of galactosylation, core fucosylation and bisecting N-acetyl glucosamine. The majority of these are non-sialyated closely related neutral structures which can be resolved by HPLC analysis, but which are difficult to separate in techniques such as fluorophore-coupled carbohydrate electrophoresis. Derivatisation with the singly charged fluorophore, 2-amino benzoic acid and separation in gels with a 30% monomer content in tris/glycine buffer enabled separation of neutral glycans. In particular, agalactosyl glycans with either a core fucose substitution or bisecting N-acetyl galactosamine could be resolved. Good separation of mono- and di-galactosylated glycans was also achieved with this system. It was shown that IgG can be separated from serum by size-exclusion and anion exchange chromatography with minimal contamination, with complete glycan release accomplished by the enzyme peptide-N-glycosidase F (F. meningosepticum). This method of resolving IgG glycans could be used to monitor patients in which glycosylation changes may have a diagnostic value, as in rheumatoid arthritis. It could also be used to monitor recombinant IgG glycosylation where routine screening is required in the biotechnology industry.

Lysis of papain-treated group A and B erythrocytes by human complement was studied by an anti-A (BRIC. 131) and an anti-B (BRIC. 30) IgM monoclonal antibody in 51Cr release assays. The indirect effect of membrane-bound antibody, i.e. its influence on complement binding to sensitized surrounding cells, was examined in a cold target competition test in which sensitized, non-labelled cells are present along with sensitized labelled cells and complement. The mode by which anti-A antibodies indirectly suppressed lysis of sensitized B cells up to 20-fold was studied by following C1q and C3b binding. C1q binding to both types of erythrocytes was not altered in mixed populations of erythrocytes in the presence of both antibodies. Binding of C3b to a mixture of both cell types was, however, suppressed, when both antibodies were present. C3b deposition in mixed cell populations did not reach a significantly higher extent than deposited to one type of erythrocyte alone. This was consistent with the results from competitive lysis and suggests that the anti-A captured most C3b at high anti-A concentrations and deprived the similarly sensitized B erythrocytes of complement. We think that this phenomenon is not due to an uneven removal of complement regulatory proteins from A and B erythrocytes by papain. Instead, the phenomenon might be due to an inherent property of anti-A mAb to better produce nucleation sites for C3 convertases which, upon binding factor B, better compete for the limiting factor D. A mathematical analysis of cold target competition experiment (containing 2430 individual measurements) also shows that the distribution of complement between the competing A and B erythrocyte population is uneven, since it predicts that in any given antibody combination the majority of complement is bound to A erythrocytes. This is consistent with the measured average percentage of lysis.

Lysis of group A and B erythrocytes by human complement was studied by an anti-A (BRIC.131) and an anti-B (BRIC.30) IgM monoclonal antibody in a 51Cr-release assay. The relative concentration of membrane-bound immunoglobulins was detected by flow cytometric analysis, and the amount of C1q and C3 bound to the sensitized red cells was measured by using purified, 125I-labelled molecules. The direct haemolysis was identical with both reagents in the presence of excess and suboptimal complement over a wide range of antibody concentration (between 50 and 7000 ng/ml). The indirect effect of membrane-bound antibody, i.e. its influence on complement binding by sensitized bystander cells, was examined in a cold target competition assay in which sensitized, non-labelled cells are present when complement is incubated with sensitized labelled cells. We have found that the competitive capacity of sensitized erythrocytes correlated with the amount of membrane-bound immunoglobulins. In accordance with our earlier findings, an equal level of target and competitor cell lysis was obtained only if the fluid phase anti-B antibody concentration was 2 to 4 times higher than that of the anti-A antibodies. We demonstrate in this paper that the different competitive activity of IgM anti-A and anti-B monoclonal antibodies might be accounted for by differences in their C1q and C3 binding capacities.

The lysis of group AB erythrocytes by human complement was studied by different anti-A and anti-B IgM monoclonal antibodies (mabs) in a 51Cr-release assay. The concentration of membrane-bound immunoglobulin was detected by ELISA, and the amount of C1q and C3 bound to sensitized red cells was measured by using purified, 125I-labelled molecules. We have demonstrated that there is an exponential relationship between the concentration of the sensitizing IgM mabs and C1q binding to the sensitized AB cell. The efficiency of binding was related to the number of antibodies bound; thus, anti-A sensitized cells bound 3-6 times more C1q than anti-B sensitized cells did. AB cells, on the other hand, bound similar amounts of C3 whether anti-A or anti-B was present. The lytic efficiencies of the various IgM mabs during short incubation times were different, suggesting that the complement activation rates vary widely with different antibodies on the AB cell membrane. The binding of C1q to an antibody-sensitized target activates a cascade, whose components may migrate away from the sensitizing antibody; interactions between the activation processes generated by the anti-A and anti-B antibodies may thus occur. Choosing appropriate pairs of anti-A and anti-B mabs for the simultaneous sensitization of AB cells has indeed resulted in stimulation in some and inhibition in other combinations of mabs. It is suggested that stimulation is observed when the activated intermediates are produced in excess, whereas inhibition occurs when a shortage of activated intermediates prevents mutual utilization.

We have attempted to determine the cellular mechanism by which alpha-thalassaemia may protect against Plasmodium falciparum malaria. Invasion and development of P. falciparum in the microcytic red cells of two-gene deletion forms of alpha-thalassaemia when measured morphologically or by [3H]hypoxanthine incorporation were normal compared to controls. Normal invasion rates were also observed following schizogony in thalassaemic red cells. Neither the addition of the oxidant menadione, 30% oxygen, nor modified medium, produced differential damage to parasites within thalassaemic cells. Furthermore, there were no significant differences in the binding of P. falciparum-parasitized alpha-thalassaemic and normal cells to C32 melanoma cells in vitro. However, when neoantigen expression on the surface of infected thalassaemic cells was estimated using a quantitative radiometric antiglobulin assay, clear differences were observed. It was found that alpha-thalassaemic cells bound higher levels of antibody from serum obtained from individuals living in a malaria endemic area than control normal red cells. The binding ratio for thalassaemic compared with controls was 1.69 on a cell-for-cell basis, and 1.97 when related to surface area. The binding of antibody from immune serum increased exponentially during parasite maturation. We also found increased binding of naturally occurring antibody present in non-immune serum to parasitized thalassaemic red cells which also increased during parasite maturation. We conclude that the protection afforded by thalassaemia against malaria may not reside in the ability of parasites to enter, grow or cytoadhere to endothelium in such cells, but may be related to immune recognition and subsequent clearance of parasitized red cells.

In an attempt to determine the mechanism whereby thalassemia in its milder forms may protect against malaria, we have examined the expression of neoantigen at the surface of Plasmodium falciparum-parasitized thalassemic red cells. Neoantigen expression was estimated by measurement of antibody bound after incubation in serum from adults living in a malaria-endemic area, using a quantitative radiometric antiglobulin assay. We found that P. falciparum-parasitized alpha- and beta-thalassemic red cells bind greater levels of antibody from endemic serum than controls: mean binding ratios (+/- SE), respectively, for alpha- and beta-thalassemia compared with controls were 1.69 +/- 0.12 and 1.23 +/- 0.06 on a cell for cell basis, and 1.97 +/- 0.11 and 1.47 +/- 0.08 after a correction for surface area differences. Binding of antibody increased exponentially during parasite maturation. In addition, we found a small but significant degree of binding of naturally occurring antibody to parasitized red cells, the extent of which was also greater in thalassemia. The apparent protective effect of thalassemia against malaria may be related to enhanced immune recognition and hence clearance of parasitized erythrocytes.

B-lymphoblastoid cell lines transformed by Epstein-Barr virus were produced from cells obtained from a hyperimmunised donor with serum anti-D activity against category DVI red cells and enriched for this activity by rosetting with category DVI red cells. Three clones produced IgG1 anti-D and had stable cell growth and continuous secretion of antibody in prolonged culture. The monoclonal antibodies reacted with category DVI red cells, when assessed manually and in an automated blood grouping system, and are useful blood grouping reagents for the detection of the category DVI phenotype. Using a radiometric technique, the number of antibody molecules bound to category DVI red cells from 5 individuals was estimated to range from 2,800 to 11,200 per cell. Five percent of blood donors classed as Du in the south western region were found to have the category DVI phenotype.

The application of monoclonal antibodies to the quantification of blood group antigen sites on erythrocytes was examined. A second antibody technique using labelled anti-mouse IgG could not be used as it was not possible to predict the binding ratio between this and the monoclonal antibody. A series of monoclonal antibodies (R10, R18, BRIC 13, BRIC 14) to the erythrocyte sialoglycoprotein alpha (syn: glycophorin A) showed the number of antigen sites to be from 0.3 X 10(6) to 1.2 X 10(6) per erythrocyte and supported the conclusion that the Wrb antigen is located on this protein. An antibody with a specificity related to the Rh blood group system (R6A) showed 4.6 - 10.4 X 10(4) binding sites to be present on cells of phenotype cCDEe. On cells of phenotype -D- 1.24 X 10(4) binding sites were present but protease treatment increased the number of available sites to 1.3 X 10(5). An antibody to a Kell-related antigen (BRIC 18) recognized 2.5 - 5.9 X 10(3) sites per erythrocyte on cells of phenotype Kk. However, a similar number also appeared to be present on cells of the McLeod and Ko phenotypes although the affinity for the antigen on these cells was very much reduced. The potential of using monoclonal antibodies for this purpose and the value of this in the study of blood group systems has been demonstrated.

65 normal, healthy people with a positive direct antiglobulin test (DAT) have been identified in a population of blood donors over a period of 14 years. 32 of them have been recalled for detailed study. A strong positive correlation with increasing age was noted, comparable to that seen in hospital patients with a positive DAT. No feature known to cause a positive DAT was identified in the healthy individuals, only 1 of whom went on to develop autoimmune haemolytic anaemia.

The number of Lub antigen sites on the human erythrocyte membrane, as recognised by the murine monoclonal antibody BRIC-108, has been determined. The number of antibody molecules bound per cell on one example of cells of the phenotype Lu(a-b-) with recessive inheritance was an average of around 200 in replicate determinations, which probably represents non-specific antibody binding. A similar number of antibody molecules to this bound to trypsin- or pronase-treated normal cells or cells of the Lu(a-b-) phenotype associated with the inheritance of the X-borne gene, XS2. The number of binding sites on three examples of cells of the phenotype Lu(a-b-) with dominant inheritance was from 440 to 690. A variation in the number of binding sites per cell from 1,640 to 4,070 was found in five individuals with the phenotype Lu(a-b+) and from 850 to 1,820 in four individuals with the phenotype Lu(a+b+). Four individuals with the Lu(a+b-) phenotype had an average of 480 binding sites per cell. The Lub antigen therefore appears to a have low-site density and a variable level of expression on the erythrocyte surface.

A method is described for the production of an anti-C3+d specific antiserum in rabbits. The method does not require purification of C3 but relies on the binding of human C3 to the rabbit erythrocytes following activation of the alternative pathway. The antiserum produced is suitable for use as either a serological reagent or for immunoprecipitation.

We have studied the erythrocytes of two individuals (P. L. and K. W.) who lack the Gerbich (Ge) blood-group antigen. The erythrocytes of P. L. and K. W. were not reactive with two monoclonal antibodies (NBTS/BRIC 4 and NBTS/BRIC 10) which reacted with normal erythrocytes. The membranes of P. L. and K. W. erythrocytes appeared to lack three minor sialoglycoproteins (beta, beta 1 and gamma). These three minor sialoglycoproteins were found to be associated with the cytoskeletons of normal erythrocytes. Approx. 10% of the erythrocytes of P. L. and K. W. were frankly elliptocytic. We suggest that one or more of the minor sialoglycoproteins may play a part in maintaining the discoid shape of the human erythrocyte.

The amount of antibody bound to cells in a low ionic strength solution (LISS) has been quantitated for several antibodies including anti-D, anti-c, anti-Kell, anti-Fya, and anti-Jka. With the exception of the Kell antibodies there was an enhancement of the rate of antibody uptake in LISS. For Rh antibodies the amount bound after a 5-min LISS incubation is comparable to that bound after 45 min in saline. For Kell antibodies a smaller amount was bound in LISS than in saline. The effect of the ratio of serum to cells was also studied, and with several antibodies there was an increase in the amount of antibody bound with a higher serum to cell ratio irrespective of suspending medium. For Kell antibodies this ratio appears to be of greater importance than the ionic strength for antibody detection. A modification to the LISS method to increase the serum to cell ratio is, therefore, presented.

Rhnull human erythrocytes lack the antigens of the Rhesus blood-group system, have an abnormal shape, have an increased osmotic fragility, and are associated with mild chronic haemolytic anaemia. Rhnull erythrocytes also lack all antigens of the LW blood-group system, but the functional significance of this deficiency is unknown. We have identified, by immunoblotting with two mouse monoclonal antibodies (BS46 and BS56), the LW-active component(s) in normal human erythrocytes as a broad band of Mr 37 000-47 000 on SDS/polyacrylamide-gel electrophoresis. Treatment of intact human erythrocytes with endoglycosidase F preparation destroyed the epitopes recognized by antibodies BS46 and BS56, suggesting that one or more N-glycosidically linked oligosaccharides are required for the formation of the LW antigens. Estimation of the number of LW antigen sites per erythrocyte by using radioiodinated purified antibody BS46 gave average values of 4400 molecules/cell for Rh(D)-positive adult erythrocytes and 2835 molecules/cell for Rh(D)-negative adult erythrocytes. Like the Rh(D) polypeptide, the LW polypeptide(s) is (are) associated with the cytoskeleton of normal erythrocytes. These results suggest the possibility that the absence of the LW polypeptide may also contribute to the functional and/or morphological abnormalities of Rhnull erythrocytes.

A sensitive method has been developed using 125I-labelled anti-IgG which allows detection of IgG present on the red cell surface. By suitable absorption and correction for non-specific binding to trypsinized red cells, a reproducible, accurate method of quantitating IgG absorbed onto red cells is obtained. The number of IgG molecules present on normal cells was found to range from 5 to 90 with an average of 39. Studies of correlation of agglutination in the antiglobulin test with numbers of IgG molecules on the cell were also undertaken.

1. The molecular weights of chick tendon and cartilage procollagens, and their constituent polypeptides, were determined by gel filtration and gel electrophoresis. The values obtained are in good agreement and indicate that the mol.wts. of the secreted procollagens (types I and II) and their individual pro-alpha-chains are of the order of 405 000-445 000 and 137 000-145 000 respectively.2. Digestion of tendon procollagen with human rheumatoid synovial collagenase gave products consistent with the presence of large non-helical peptide extensions at both N-and C-termini. Electrophoretic analysis gave apparent mol.wts. of 17 500 and 36 000 for the respective N- and C-terminal extensions of pro-alpha1(I)-and pro-alpha2-chains, and inter-chain disulphide bonds were restricted to the C-terminal location. 3. During the biosynthesis of procollagen by tendon and cartilage cells a close correlation was observed between the extent of inter-chain disulphide bonding and the proportion of procollagen polypeptides having a triple-helical conformation. These processes appeared to commence in the rough endoplasmic reticulum and be completed in the smooth endoplasmic reticulum, but the rate at which they occur in cartilage cells is markedly slower than that found in tendon cells. 4. When the intracellular [14C]procollagen polypeptides present in the rough-endoplasmic-reticulum fractions of tendon and cartilage cells were analysed under non-reducing conditions on agarose/polyacrylamide composite gels, no significant pools of dimeric intermediates were detected. 5. In both cell types, inter-chain disulphide-bond formation occurred even when hydroxylation, and hence triple-helix formation, was inhibited. The presence of pro-alpha1- and pro-alpha2-components in a ratio of 2:1 in the disulphide-linked unhydroxylated procollagen isolated from tendon cells demonstrated that correct chain association occurs in the absence of hydroxylation. This observation is consistent with a model for the assembly of pro-gamma112-chains in which the recognition and selection of pro-alpha1-and pro-alpha2-chains in a 2:1 ratio are directed by the non-helical C-terminal extension peptides of tendon procollagen.

The preparation of test cells coated with specific IgG of known subclass is described. Such cells are required in the standardization of IgG subtyping reagents. At present, these test cells usually are prepared by coating cells with IgG myeloma paraproteins. However, these paraproteins may not be generally available and an alternative method is presented using more readily obtainable materials. Quantification of cell-bound IgG showed that the subclass did not affect the sensitivity of the antiglobulin test when using broad-spectrum anti-IgG and that the test cells produced had an optimal IgG coating. Reactions with subclass-specific antisera were however, considerably weaker that those obtained with broad-spectrum anti-IgG. A modified spin-layering technique for use with subclass-specific antisera is described.

Evidence of immune mediated haemolysis was sought in 83 patients with P. falciparum malaria in eastern Thailand. Amongst 73 patients with uncomplicated infection 12 (16.4%) had a weakly positive direct antiglobulin test (DAT). The incidence in 32 children aged 8-16 years was similar to that in adults. Of 10 patients with cerebral malaria, six adults, all of whom were in unrousable coma, had a positive DAT. Erythrocyte-bound IgG1 accounted for the positive DAT in all cases; sensitization with complement or other IgG subclasses was not found. Patients with uncomplicated malaria had a median value of 70 IgG molecules per erythrocyte compared with 65 molecules per cell in 67 healthy controls. This difference was not statistically significant but could account for the lower incidence of a positive DAT in control subjects (4.5%). There was no correlation between the number of IgG molecules per cell and the degree of anaemia during the acute or convalescent phases of the infection. There is no evidence from this study that an immunohaemolytic process contributes to the anaemia of falciparum malaria in eastern Thailand.

The binding of 15 125I-labelled mouse monoclonal antibodies to cell-surface sialoglycoprotein alpha (SGP alpha: synonym Glycophorin A) was studied using intact IgG and Fab fragments. It was estimated that the number of sialoglycoprotein alpha (SGP alpha) molecules per red cell is of the order of 1 x 10(6) and the number of sialoglycoprotein delta (SGP delta; synonym Glycophorin B) molecules per red cell is of the order of 1.7-2.5 x 10(5). Competitive binding assays showed that antibodies of the same blood group specificity (four anti-Ns reacting with SGP alpha and SGP delta (BRIC 33, BRIC 115, BRIC 120, BRIC 123) and four anti-Wrbs reacting with SGP alpha (R7, BRIC 14, BRIC 89, BRIC 93) inhibited binding of each other to red cells. Two antibodies (R1.3 reacting with SGP alpha and SGP delta and R18 reacting with SGP alpha) recognized distinct epitopes, and the remaining five antibodies (BRIC 116, BRIC 117, BRIC 119, BRIC 127, R10 reacting with SGP alpha) partially inhibited binding of each other to red cells. This latter observation and the finding that four of these antibodies (BRIC 116, BRIC 117, BRIC 119, BRIC 127) bind to a considerably smaller number of antigen sites (1.69-2.71 x 10(5) for intact IgG) than the maximum value obtained, suggests heterogeneity of glycosylation within SGP alpha molecules. The functional affinities of the IgG antibodies ranged from 1 x 10(5) to 4 x 10(7)M-1.

Monocyte-erythrocyte interaction in patients with autoimmune haemolytic anaemia (AIHA) was assessed by phagocytosis and rosette assays. In most patients, a relationship was observed between haemolysis and the phagocytosis of their own erythrocytes by allogenic peripheral monocytes. An evaluation of the number of immunoglobulins on patient erythrocytes and IgG subclasses of autoantibodies shows that in patients with only IgG1 antibody or with additional IgG2 or/and IgG4, phagocytosis was always observed when the number of erythrocyte-bound IgG molecules was above 2,000. On the other hand, in all patients where IgG3 was detectable, phagocytosis was observed even if the amount of IgG was as low as 230 molecules per erythrocyte. Similar observations were made in the rosette assay. Generally, the number of erythrocyte-bound IgG and the presence of phagocytosis were correlated with the degree of haemolysis, but there were exceptions, i.e. the amount of IgG and phagocytosis were high but there was no evidence of haemolysis, or where there was little IgG, no phagocytosis but haemolysis was present. Our data do not indicate that erythrocytes from AIHA are preferentially bound to autologous monocytes.

Thirty-four IgG anti-D human monoclonal antibodies (mAb) derived from 18 donor were assessed for their ability to mediate lysis of D+ red cells by lymphocytes in antibody-dependent cell-mediated cytotoxicity assays. Cell-bound antibody was quantified and the mAb were compared at similar levels of sensitization. The majority (23/31) of IgG1 and all (3/3) IgG3 mAb were ineffective; two donors produced both lytic and non-lytic anti-D mAb. Greater sensitivity was achieved using fluid-phase antibody (as culture supernatants) in the assay than was obtained with pre-sensitized red cells. Minimum levels of 2000 anti-D molecules per cell were required for lysis using pre-sensitized cells. Partial D red cells (DIVa, DVa and DVI) were lysed by three mAb that were lytic with normal D+ cells. There was no relationship between lytic ability and Gm allotype or D epitope specificity of the antibodies. Four mAb to other blood group specificities were tested: two (anti-E and anti-G) were lytic and two (anti-c and anti-Kell) were not lytic. Possible reasons for the heterogeneity of the lytic activity by the mAb are discussed.