Review_glyco-diagnostics - Current Molecular Medicine 2007...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Current Molecular Medicine 2007, 7, 397-416 397 Clinical Laboratory Testing in Human Medicine Based on the Detection of Glycoconjugates Benjamin L. Schulz#,1 , Wouter Laroy#,2 , $ and Nico Callewaert*,#,2,3 1GlycoINIT and Institute of Microbiology, Swiss Federal Paulistrasse 10, CH-8093 Zürich, Switzerland 2Unit for Molecular Glycobiology, Department Technologiepark 927, B-9052 Ghent, Belgium for Institute Molecular of Technology, Biomedical Wolfgang- Research, VIB, 3Unit for Molecular Glycobiology, Department for Biochemistry, Physiology and Microbiology, Ghent University, K.L.-Ledeganckstraat 35, B-9000 Ghent, Belgium Abstract: The purpose of this review is to provide a concise overview of developments over the last 15 years in the field of laboratory tests in human medicine that are based on the detection of alterations in the glycan part of glycoconjugates. We show how glycosylation-based diagnostic testing is widespread in the current clinical practice, in different formats. To provide the necessary focus in this extremely broad field, we have only included assays that are either in actual clinical use or that are under active development towards clinical use, with some bias towards assays that were recently developed. The fields included are: cancer, infectious disease, genetic defects of glycoconjugate biosynthesis and catabolism, auto-immunity, drug abuse and liver disease. n tio u rib To conclude this review, we provide a viewpoint on the future of the glyco-diagnostics field in terms of novel technologies, especially with regard to the discovery and clinical implementation of biomarkers that are based on pathologically altered endogenous glycotopes. t is D r o F ot GLYCOCONJUGATE-BASED TESTING IN CURRENT CLINICAL PRACTICE OR ADVANCED RESEARCH STAGE Diagnostic and Prognostic Tests for Malignant Transformation (Cancer) Introduction Changes in glycosylation have long been associated with cancers. Similar to all but a few biochemical markers for cancer, while none of these glycosylation-based markers is 100% accurate, they are still clinically very useful, as will be pointed out below. Secreted and surface glycoproteins of a broad variety of adenocarcinomas commonly express Tn, T, sialyl-Tn glycotopes and increased levels of the different Lewis blood group glycotopes [1] (see Fig. 2 for structures of these glycotopes). Glycobiological cancer diagnosis and prognosis relies on lectin or antibody-based glycotope detection, performed either on tumor biopsy material using histochemistry, or in biofluids using solid-phase or insolution immunoassays. The detection of cancerassociated glycotopes is currently the most widely studied subfield of glycomics-based diagnosis. The activity in this field over the last 10 years has mainly been in trying to establish the utility of these N *Address correspondence to this author at the Unit for Molecular Glycobiology, Department for Molecular Biomedical Research, VIB, Technologiepark 927, B-9052 Ghent, Belgium; Tel: +32 9 331 3620; Fax: +32 9 331 3609; E-mail: [email protected] $Current address: Pronota NV, VIB Bio-incubator, Technologiepark 4, B9052 Ghent, Belgium #All these authors have contributed equally. 1566-5240/07 $50.00+.00 markers in a very broad range of carcinoma types, for a variety of clinical purposes. A second focus has been the elucidation of the genetic and biochemical mechanisms behind these glycobiological alterations. And, third, a range of studies has aimed at explaining how these glycan changes mechanistically contribute to metastasis phenotype. The latter hypothesis arose from prognostic studies in which several of the glycobiological alterations were found to be clearly predictive of a worse patient outcome. In the sections that follow, we have only included those tumor antigens with relatively well-established validity for diagnostic purposes (i.e. in current clinical use or well on the way to that status). Many more might be therapeutically relevant as tumor-specific antigens or might have been described in the literature, but they have either not proven to yield more accurate diagnostic information than the more established carbohydrate-dependent tumor markers, or are still in a very preliminary stage of validation. Carbohydrate-Dependent Tumor Markers in Current Clinical Use, Mainly Detected Via Histology on Tumor Tissue T-Glycotope Probably the best-known family of tumor antigens of carbohydrate nature are based on the T-antigen (Galβ-1,3-GalNAc- α -Ser/Thr) [2]. This structure is also known as the Thomsen-Friedenreich antigen. The detection of the T-antigen is performed using the lectin PNA (peanut agglutinin) or with specific monoclonal antibodies. PNA-based detection has © 2007 Bentham Science Publishers Ltd. 398 Current Molecular Medicine, 2007, Vol. 7, No. 4 Schulz et al. n tio Glycolipids u rib Polysaccharides Glycosylation Machinery t is D r o F ot Glycoproteins N O O-glycans N-glycans N Fig. (1). Overview of the diagnostic testing procedures discussed in this review. Glycoconjugate-based testing is widespread in the current clinical practice, and several assays are in advanced stages of clinical development. the advantage that it is more protein-carrier independent than the monoclonal antibodies. An enzyme-linked PNA assay has also been developed to measure the T-antigen in serum in the context of cervical carcinoma (utility in other adenocarcinomas has not been reported) [3]. The detection of the Tantigen is currently most often used in histological diagnosis of colorectal carcinoma [4], where it has a very good specificity (i.e. ratio of number of true positives to the number of all positive test results) as the normal colonic mucosa is T-antigen negative. The marker has been assessed in a large range of other carcinomas as well, with varying degrees of clinical utility concluded (down to none in lung cancer, where the T antigen is expressed on terminally differentiated pneumocytes and is thus a differentiation antigen) [5]. Tn-Glycotope A related tumor-associated glycotope is known as Tn (GalNAc-α -Ser/Thr). The Tn antigen itself is detected using monoclonal antibodies, whereas the lectin from Helix pomatia (HPA) recognizes a broad array of GalNAc-modified glycoconjugates on cancer cells (and some GlcNAc binding is also very likely [6]). Remarkably, HPA lectin histochemistry results Clinical Laboratory Testing in Human Medicine Based depend very much on the technical details of the method [7, 8]. Given the low affinity of lectins for their glycotopes, and the fact that lectins, unlike antibodies, are a diverse family of proteins with different folds, care should always be taken to not alter the lectin's glycotope binding during modifications (such as conjugation to peroxidase for direct detection). To allow detection of Tn in serum, a heterogenous immuno-lectin sandwich assay was developed [9], which captures Tn-bearing glycoproteins on an anti-Tn surface, followed by glycoprotein detection using biotinylated Vicia villosa isolectin B4-streptavidin. The assay appeared to have good analytical characteristics and very good specificity for the indiscriminate detection of a range of adenocarcinomas (test negative in all healthy controls and in patients with non-malignant diseases). However, we are not aware of it being used in routine clinical practice. Increased levels of this glycotope are good predictors for the metastatic potential of adenocarcinomas of the breast, the colon and the stomach, the skin [10], the ovary, oesophagus, thyroid and prostate [11]. Human adenocarcinoma cell lines that are HPA positive metastasize in immunodeficient mice models, while their HPA negative counterparts generally do not [12, 13]. In in vitro studies with breast cancer-derived cell lines, the level of Tn expression correlated with the capability of the cell lines to invade through Matrigel, but the Tn glycotope itself appeared not to be directly mechanistically involved in this phenotype, as the glycotope blocking with HPA lectin did not affect the invasivity of the cell lines [14]. In vitro, cell lines have been derived from a common progenitor colon carcinoma line, which are either Tn+ or Tn-. Tn positivity was associated with a loss of core 1 β-1,3galactosyltransferase activity [15]. It is unsure whether this is also the mechanism behind increased Tn abundance during natural tumor progression. Current Molecular Medicine, 2007, Vol. 7, No. 4 the metastatic competence of adenocarcinomas. This would explain why Tn/HPA positivity correlates closely with increased metastasis, while it seems not to be directly involved in this behaviour. In the case of colon carcinoma, a very nice study [17] used quantitative T/sia-T and Tn/Sia-Tn histology in a comparison of paired primary tumor and metastatic tissues. The metastases had a decrease in the Tn and T structures versus the primary tumour and a reciprocal increase in Sia-Tn and Sia-T. The authors went on to show that the sialylated glycotopes were important in weakening the adhesion of the metastasizing cells to basement membranes, which is compatible with the finding on β1-integrin modfication. It appears that, at least in colon carcinoma, deregulation of gene expression leads to some cells acquiring a hyper-α -2,6-sialylation phenotype [18], which helps these cells to detach from the tissue and at the same time increases the likelihood for these cells to re-attach at endothelial sites elsewhere in the body. Of course, metastasis formation is a multifactorial process which might follow different paths in different patients, but changes in cancer cell glycosylation certainly is an important contributing factor. In breast carcinoma, the detection of Sia-Tn correlates with other predictors of poor prognosis (high nuclear grade, aneuploidy and high S-phase fraction of epithelial cells) [19], and Sia-Tn is already detectable in a significant fraction of in situ carcinomas, indicating that it is often expressed early on in the process of cancer formation. N The α -2,6-sialylated derivative of Tn (Sia-Tn) is also a cancer-associated glycotope that is used in clinical histology, and its increased abundance on cancer cells is often correlated with an increased activity of the ST6 GalNAc 1 sialyltransferase. When the latter enzyme is overexpressed in epithelial cell lines, the cells dramatically change in morphology and migration properties, in a phenotype that is reminiscent of epithelial-to-mesenchymal transition (EMT). An important pathway in this phenotype is the modification of the β1- integrin subunit with Oglycans that are sialylated by the over-expressed enzyme [16]. This appears to significantly modify integrin-initiated signaling, and might be associated with cancer progression. In this sense, and taking into account that the Tn epitope is the substrate for the ST6 GalNAc 1 sialyltransferase, one would be induced to think that Tn/HPA positivity is merely a pre-condition for Sia-Tn formation, which then would be the functionally important glycotope in increasing n tio u rib t is D r o F ot Sia-Tn Glycotope 399 Le Blood Group Determinants The glycan structures that determine the Le blood group have been implicated in many cancers as risk factors for a bad prognosis [20, 21], but it seems very difficult to draw general conclusions for different cancer types: different cancers seem to disturb the flux through the complex network of O-glycosylation biosynthesis reactions in quite different ways. For example, in pancreas cancer, Sia-Lea detection is used clinically for diagnosis, whereas in prostate cancer, it is Sia-Lex and Ley which are the only blood-group related antigens that are minimal or absent in benign epithelial cells and that are more highly expressed in malignant tissue. It is likely that the expression pattern of the O-glycosylation genes in the benign tissue (which can be very different from tissue to tissue) greatly influences the effect that upor downregulation of some of these genes has on the O-glycosylation phenotype of the cancer. And of course, the 'ground state' of O-glycosylation determines whether a specific glycotope is practically useful or not in detecting a change in that tissue: it is much easier to detect the appearance of a normally absent glycotope than to reliably measure -for example- a 25% increase in the abundance of a glycotope. As tissues differ in their 'ground state' Oglycosylation, the type of glycotopes that are useful to detect alterations in these tissues are obviously also different. If one general conclusion can be drawn, it would be that if a tumor displays the altered 400 Current Molecular Medicine, 2007, Vol. 7, No. 4 glycosylation which is cancer-specific for that tissue, then the clinical course of the cancer is most often significantly worse than for tumors in that tissue that do not diplay the glycosylation change. Whether all of these different glycosylation changes contribute causally to this more agressive behaviour or rather merely reflect a more advanced tumor stage has not been firmly established in all cases. In colon carcinoma, Sia-Lex detection is predictive for disease recurrence, whereas Sia-Lea detection is not [22, 23]. In primary lung cancer, a high level of Sia-Lex predicts short survival time of the patient [24]. Sia-Lex In gastric cancer, expression is an independent risk factor for liver metastasis [25] and Sia-Lea expression is a predictor for worse outcome [26]. A monoclonal antibody (designated as FH6) is available which specifically recognizes dimeric SiaLe x glycotopes. The detection of such a dimeric glycotope would indicate a high density of Sia-Lex and thus be particularly suitable for use in cancer diagnostics based on Sia-Lex. Indeed, in gastric carcinoma patients it was found that detection of this complex glycotope was significantly associated with venous invasion of the tumour and with a more advanced histological classification [27]. Moreover, patients with a high level of this glycotope had a shorter survival time than patients with none or low levels. In patients with advanced stage gastric carcinoma, dimeric Sia-Lex was the only independent prognostic factor for patient outcome in this study. Schulz et al. increased adhesion to umbilical vein endothelial cells. This phenotype was correlated with an increased abundance of the Sia-Lex glycotope, which is a ligand for the endothelial E-selectin [32]. In hepatocellular carcinoma and in colon carcinoma [33], Mgat5 mRNA is strongly upregulated. Non-Invasive Carbohydrate-Dependent Tumor Markers in Current Clinical Use, Mainly Detected Via Immuno-Assay on Body Fluids CA 19-9 Glycotope Detection in Gastro-Intestinal Cancer: Sia-Lea The epitope of the monoclonal antibody CA 19-9 encompasses the Sia-Lea glycotope. The detection of this epitope in serum is currently part of routine diagnostic testing for pancreatic cancer [34, 35], for cholangiocarcinoma (bile duct cancer), and for colorectal [36-38] and gastric cancer [39] (and the marker has sporadically been studied for a large range of other cancers). Approximately 80% of all pancreatic cancer cases are positive for CA19-9, with a false positive rate of approx. 20% in benign pancreatic and hepatic conditions. Postoperatively high CA19-9 or re-appearing CA19-9 upon therapy are useful predictors of therapy failure and poor survival [40]. Remarkably, the CA19-9 marker, in combination with CEA, can usefully detect even imaging-occult cholangiocarcinoma on a background of primary sclerosing cholangitis (the most important risk factor for cholangiocarcinoma development) [41]. This finding was recently validated and screening of primary sclerosing cholangitis patients with CA19-9 and CEA was found useful [42]. CA19-9 is thus one of the more succesful glycosylation-based biomarkers. Because its use is widespread, there is also quite some information on non-cancer pathologies that can cause false positive CA19-9 test results. As with most cancer biomarkers, these pathologies often are benign alterations of the target tissue (such as pancreatitis [43], gallstone-induced bile duct inflammation [44], or acute hepatitis [45]), but other diseases can also cause false positives in clinically distinct settings (splenic cysts [45-47], endometriosis [49], or pulmonary alveolar proteinosis [50, 51]). Also, in the rare cases where these cancers occur at young age, the marker has to be used with utmost caution [52]. The levels of CA19-9 are also influenced by the Lewis secretor phenotype of the individual under investigation [53, 54], and negative CA19-9 test results can be obtained in secretor negative pancreatic cancer patients. This rather broad understanding of the factors that can adversely influence CA19-9 test results is telltale of a mature biomarker, which has stood the reality test of clinical implementation. u rib t is D r o F ot α -1,2-Fucosylated Le Glycotopes The mAb MIA-15-5 recognizes the H/Le y/Le b glycotopes, which have a Fuc-α -1,2-Gal determinant in common. Patients with MIA-15-5 positive primary lung cancers have a 3-to 6-fold reduced chance on 5y survival than those with MIA-15-5 negative tumors [28]. Most carcinomas express LeY glycotopes on their plasma membrane glycoproteins, including epidermal growth factor receptors. As was recently demonstrated, anti-LeY antibodies can potently interfere with EGF-R signaling and this might be a novel target for tumor therapy [29]. N Increased L-PHA Binding (β-1,6-GlcNAc Branching) N-acetylglucosaminyltransferase V is the enzyme that catalyzes β-1,6-GlcNAc branching of N-glycans in the mammalian Golgi apparatus. The coding Mgat5 gene is under control of Ras signaling, which is very commonly upregulated in tumor cells. Glycans which are modified with β-1,6-GlcNAc branches can be detected by the lectin L-PHA (leukoagglutinating phytohemagglutinin)[30]. Increased levels of L-PHA binding are especially well documented in human breast and colon carcinoma, and these increased levels are associated with progressively advanced stages of the diseases [31]. In an in vitro study, over-expression of GnT V in colon carcinoma cells leads to decreased fibronectin adhesion, but n tio On the mechanistic side, a recent study has shown that the β3GalT5 β-1,3-galactosyltransferase is important in biasing the biosynthesis of O-glycans in pancreas cancer cells towards Sia-Lea -capped type 1 chain O-glycans, as its silencing by antisense RNA resulted in poly-N-acetyllactosamine chain elongation and capping by Sia-Le x structures [55]. Clinical Laboratory Testing in Human Medicine Based CA125 in Ovarian Carcinoma The protein backbone of the CA125 antigen is encoded by the Muc16 gene [56, 57]. Detection of the CA125 glycoform of MUC16 is commonly used in the diagnosis of epithelial ovarian cancer [58], using the OC125 monoclonal antibody [59]. As with most cancer markers, false positives can occur in patients with non-malignant disorders of the target tissue or of other tissues [60, 61]. The O- and N-glycosylation of CA125 has recently been comprehensively characterised [62] (although we still do not know which of the observed glycan structures form part of the CA125 epitope), potentially opening the way to yet more specific measurement, and to an understanding of its role in ovarian cancer progression. Glycolipid Alterations in Cancer: Limited Established Non-Invasive Diagnostic Utility Next to mucinous O-glycosylation and Nglycosylation, glycolipids are also known to be altered in malignant transformation. Especially in brain tumors, ganglioside fingerprinting has met with some success in classifying different tumor types [63]. Several gangliosides are or have been the focus for development of tumor-targeting therapies [64]. In a non-invasive diagnostics context, glycolipids have a more limited role than glycoproteins, as they are mainly cell-associated. In those cases where cellular material is readily sampled, detection of altered glycolipids can be useful, as in the case of endometrial carcinoma [65]. Some studies have also quantified gangliosides in serum of cancer patients, and found ganglioside levels to be upregulated in cancer patients as compared to healthy patients and those patients with benign tumors [66]. With the development of improved glycolipid analytical technology, we might as yet see more studies appearing that explore the diagnostic potential of biofluid glycolipid composition. Current Molecular Medicine, 2007, Vol. 7, No. 4 tumor marker protein. Glycans add a layer of structural diversity to the underlying molecule (especially protein), the quality and quantity of which can be sensitive to the physiology of the cell producing the molecule, often independent of the mere abundance of this molecule. Glycosylation is a very frequent post-translational modification of secreted proteins (i.e. those proteins that are the most likely ones to be found in the easily accessible patient biofluids), in contrast to other posttranslational modifications that can be modulated by malignant transformation, such as protein phosphorylation. Highly Fucosylated α -Fetoprotein (AFP) Glycoforms for the Detection of Hepatocellular Carcinoma The best-established example of the utility of measuring glycoform ratios of tumor marker proteins rather than the protein levels per se, is to be found in the diagnosis of hepatocellular carcinoma in patients with underlying liver cirrhosis. Clinically, an α -fetoprotein level in serum of >20 ng/ml detects patients with hepatocellular carcinoma with a useful sensitivity (>60%; sensitivity is defined as the number of afflicted individuals scoring positive on the test to the total number of afflicted individuals in the study group). However, the large majority of HCC arises on a background of liver cirrhosis, and AFP levels are often strongly increased in cirrhosis. Thus, between 20 and 500 ng/ml, there is a 'grey zone' for HCC detection, which severly impairs the utility of AFP as an early marker. Early observations using L ens culinaris lectin crossed affinoimmunophoresis on AFP demonstrated that a highly core-α -1,6-fucosylated glycoform of AFP (designated as AFP-L3) was specifically more abundant in hepatocellular carcinoma than in 'benign' liver disease, including liver cirrhosis [67-74]. The AFP 'fucosylation index' (normalizing the highly fucosylated forms to the amount of AFP) is known as AFP-L3% and kits were developed to perform this analysis [75, 76]. The cutoff value for HCC detection that is reported varies somewhat between the studies, but appears to lie between 10 and 15%. Soon, immunoblotting replaced the second dimension gel in the lectin crossed affinoimmunophoresis procedure [77], a development which afforded improved sensitivity (down to about 1 ng/ml per detected band). Using this improved sensitivity, it was found that even when AFP levels were below the 20 ng/ml cutoff, AFP-L3% was still diagnostic in many HCC cases (especially those with small tumors, i.e. <20 mm diameter). The sensitivity of the assay for small tumors is about 35%, with a specificity of >90%. In longitudinal studies of patients with cirrhosis, AFP-L3% has a lead time of 9-12 months in HCC detection over currently used imaging techniques [78]. This offers a time window for resection of HCC in a small (often still localized) stage. AFP-L3% positivity of small HCC is a risk factor for recurrence of the disease upon treatment, so especially in these cases, waiting until the tumor becomes discernable by imaging techniques may prove lethal for the patient. Such a Glycoform Detection of Known Tumor Marker Proteins or Other Proteins Produced by the Affected Tissue Several currently used serological tumor markers are glycoproteins that can also be produced at a certain level by non-cancerous tissues, and this often limits their clinical utility. Quite often potential pre-cancerous lesions are discovered through imaging or clinical examination screens (of the breast, the prostate gland, the liver,...) of relatively large population groups, based on their sex and age, and based on known risk factors (such as chronic viral hepatitis, smoking habit,...). Those patients in which suspected lesions are found could benefit the most from high-frequency highly specific non-invasive tumor marker testing (to yield early warning of conversion to cancer). Recent work shows that measuring tumor-specific glycoforms of these tumor marker proteins can help in making a clearer distinction between benign hyperplasia and cancer than the measurement of the total amount of the n tio u rib t is D r o F ot N 401 402 Current Molecular Medicine, 2007, Vol. 7, No. 4 strategy of repeated AFP-L3% screening of cirrhosis patients obviously presumes that good assays to pre-symptomatically detect liver cirrhosis are available, and also here, glycomics-based testing provides a solution (see below). Post-operative AFPL3% positivity is the most significant independent factor for predicting a decreased survival chance for the patient [79]. The AFP-L3% assay has recently been automated [80] in a heterogenous liquid-phase lectin-antibody sandwich assay, by making use of LCA and 2 anti-AFP antibodies. The assay cycle is now less than 4 minutes, with almost perfect correlation of the data with the previous lectin affinoelectrophoresis technique. This technological feat proves that, given sufficiently convincing clinical utility and given favorable market forces, glycoformbased cancer diagnostics are realistic and can be implemented in formats that are almost indistinguishable from routine clinical assays. Along similar lines, but less well-established, it has been found that the sialylation level of AFP is lower in HCC than in benign liver disease. Specifically, the proportion of monosialylated AFP can be measured upon isoelectric focusing of serum proteins, followed by anti-AFP western blotting. This assay has a reported accuracy of 89% for distinguishing HCC from liver cirrhosis in cases with nondiagnostic total AFP (<500 ng/ml) [81], but this has not (yet) been validated by different investigators. Schulz et al. normal prostate cells synthesize sialylated and less fucosylated structures. More recently, similar differences have been shown for urinary PSA in a comparison of malignant versus benign tumors [88]. Another recent study determined that the level of α 2,3-sialylation of PSA N-glycans is significantly increased in prostate cancer vs. benign prostate hyperplasia, both in seminal fluid (where it was detected through direct glycan structural analysis) and in serum (determined through increased binding to the Maackia amurensis-derived lectin (MAA) [90]). A determination of the clinical utility of these glycosylation differences awaits larger-scale clinical experimentation and the development of a simple, standardized assay method, along the development path that has been succesful for AFP-L3% in HCC. CEA (Carcinoembryonic Detection Antigen) N Prostate-Specific Antigen (PSA) Glycoforms Prostate cancer is one of the leading causes of cancer-related deaths worldwide. However, it can be treated effectively if caught in its early stages. To do so, markers for early diagnosis are needed. Currently, rectal examination and prostate specific antigen (PSA) test are widely used for this purpose. PSA, a serine protease of the kallikrein gene family [83], is secreted almost exclusively by epithelial prostate cells. In cases where disruption of the prostate basement membrane occurs, elevated serum PSA levels are observed. However, increased PSA levels can also be the consequence of more benign prostate disorders, and this yields many false positive results in risk-population screening (positive predictive value is about 30%) [84]. However, recent studies indicate that the type of N-glycosylation of PSA is indicative for the type of prostate disease [86-90]. Prostate cancer cells produce neutral, nonsialylated and highly fucosylated N-glycans, whereas n tio The carcinoembryonic antigen is the longest studied [91, 92] and still one of the most frequently used glycosylation-dependent serum tumor markers (originally for colon carcinoma, but now in a very wide spectrum of carcinomas). In the current clinical CEAmeasurement practice, antibodies are used of which the binding to CEA is carbohydrate-independent [93, 94], despite the massive glycosylation of the CEA protein (40-65% carbohydrate). However, these CEA measurements suffer from rather low specificity, especially in patients with non-malignant diseases of the intestine, the pancreas, the liver and the lung. Therefore, the main indication for the use of the CEA test is in the follow-up of cancer patients posttreatment: succesful therapy is hallmarked by a drop of the CEA levels to reference values, while remaining CEA most often indicates that the tumor has not been completely removed or has metastasized. The re-appearance of CEA indicates relapse or outgrowth of previously covert micrometastases. The present CEA tests are useless for wide at-risk population screening because of their high false positive rates. Therefore, attempts have been made to make use of cancer-associated CEA carbohydrate structural alterations (presence of Siaα 2,6-Gal- β1,4-GlcNAc glycotopes on cancer patients CEA and not on the cross-reactive molecules from non-cancerous patients) [95-97] to improve the specificity for cancer [98]. However, these assays await further validation to determine their clinical utility. u rib t is D r o F ot Another study has shown that gammaglutamyltransferase (an enzyme of which the activity is elevated in active liver disease) is differentially glycosylated in hepatocellular carcinoma versus liver cirrhosis, with useful clinical properties of the ensuing assay [82]. This difference was detected through separation of GGT isoforms over boronate affinity columns. The boronate groups interact with vicinal cis-hydroxyl groups, found mainly on proteins in their sugar moieties. An elucidation of the glycanstructural underpinning of this observation is not available. Glycoform T-Antigen on MUC1 (in Histology) The BW835 monoclonal antibody binds the Thomsen-Friedenreich antigen on a motif within the MUC1 repeat region [99]. Thus, quantitative measurement of this epitope in patient material would elegantly make use of the combined knowledge that MUC1 is frequently upregulated in human cancers, and that Thomsen-Friedenreich antigen is also upregulated in (some of) these conditions. This strategy has been tested in gastric carcinoma [100], and it was found that the BW835 immunoreactivity correlated with the presence of Clinical Laboratory Testing in Human Medicine Based lymph node metastases and was a marker of unfavorable prognosis (although the latter analysis was univariate). In colorectal carcinoma, BW835 staining was significantly correlated with more advanced tumor stage [101]. The antibody binds to about 75% of oesophageal squamous cell carcinomas [102], whereas antibodies raised against the T-glycotope alone detected only 40% of these carcinoma cases. Altered N-Glycan Branching of Urinary Fibronectin in Bladder Cancer In bladder cancer patients, the lectin binding behaviour of the common extracellular matrix protein fibronectin, shed in urine from the bladder epithelium, has been shown to be dramatically altered [103]. The lectin binding pattern suggested increases in the activity of the Nacetylglucosaminyltransferases that are involved in determining the branching pattern of N-glycans, and these were subsequently assayed in patient tissues. N-acetylglucosaminyltransferase III was especially strongly upregulated, explaining the 5-fold reduction in ConA binding of the fibronectin (ConA does not bind N-glycans with a bisecting GlcNAc-modified trimannosyl core structure). Proper clinical validation of this finding has not yet been reported. Current Molecular Medicine, 2007, Vol. 7, No. 4 403 Diagnostic Tests to Identify the Agent of Infectious Disease Mycology Periodic Acid Schiff staining in histology for fungus detection is based on the presence in the yeast/fungal cell wall of copious quantities of polysaccharides. Combined with the morphology of the PAS-stained entity, this can yield conclusive evidence of fungal infection [104] (in human medicine, candidiasis, cryptococcidiosis and aspergillosis are the most frequent). Candida albicans serotyping is based on agglutination assays [105] or flow cytometry [106] using specific antibodies against the yeast’s cell wall mannan epitopes, and such a serotyping strategy is competitive with other clinical yeast classification assays that are based mainly on differential carbohydrate utilitzation [107]. n tio Aspergillus fumigatus is the most frequent causative agent of human aspergillosis, and its mannoproteins contain α -galactofuranose determinants that are immunogenic. The detection of these carbohydrate epitopes via immunoassay has a role in the diagnosis and monitoring of aspergillosis patients [108]. u rib t is D r o F ot N Fig. (2). Some of the glycotope structures measured in the discussed diagnostic tests. 404 Current Molecular Medicine, 2007, Vol. 7, No. 4 Bacteriology The bacterial cell wall contains a range of immunogenic glycoconjugates. O-antigen and capsular polysaccharide-based bacterial typing using specific antisera is still a mainstay in clinical bacteriology, although these assays are gradually being replaced by nucleic-acid detection tests (PCRbased) that genotype the genes that biosynthesize the strain-specific components of the cell wall polysaccharide biosynthesis machinery (especially the O-antigen polymerases and the genes that are postulated to be the O-antigen lipid precursor flippases). However, with the advent of protein microarray technology for the fast detection of specific pathogens (both clinically and in bio-defense settings), the use of anti-cell wall lectins and serotyping antisera might enjoy a revival as specific capture reagents for the bacteria [109]. In settings where rapid and simple testing for specific pathogens is required or cost-efficient, immunological assays, often detecting pathogenspecific carbohydrate antigens, are still very competitive, and we discuss a number of cases that were reported during the review period (many more were commercially developed into point-of-care tests and are not necessarily formally reported in the peerreviewed literature). Schulz et al. patients. The latter might be very valuable, especially in low-resource medical settings: urine is generally sterile and its testing does not require BSL3 containment laboratories [116]. When 3 serological tests (including anti-PGL-1 and anti-LAM) are combined, the sensitivity for M. tuberculosis can be as high as 85% even in acid fast smear negative/ culture negative patients, with about 10-15% false positives in a healthy control population [117]. Parasitology Also for several 'tropical' diseases, glycoconjugate-detecting assays are under development (the whole field was reviewed recently: [118]). A remarkable case is the use of keyhole limpet hemocyanin (KLH) in the diagnosis of acute schistosomiasis: this protein, derived from a completely different organism than the parasite (Schistosoma mansoni), has immunogenic carbohydrate epitopes (LDN-DF: [GalNAcβ1,4 (Fucα 1,2Fucα 1,3)GlcNAc β1-] and variants thereof [119, 120] in common with the surface of schistosomula. Thus, KLH can be used as standard antigen in ELISA assays to detect the active stage of schistosomiasis with high accuracy [121]. It is even so that the shared glycotope between KLH and Schistosoma is the relevant pathogenetic factor in the induction of hepatic granulomas by S. mansoni eggs that are trapped in the liver (this is the main pathology of the disease) [122]. Schistosoma mansoni also sheds antigens of glycoconjugate nature that are filtered into the urine and can be detected there using specific monoclonal antibodies, with an efficiency which approaches diagnosis using serum [123]. u rib t is D r o F ot Rapid immunoassays for the detection of streptococcal infection from throat swabs, based on detection of the Group A streptococcal carbohydrate antigen, surpass traditional culture-based diagnostics in sensitivity (and, obviously, speed) [110] and can be formulated in point-of-care diagnostic formats. Bacterial endocarditis caused by gram-positive bacterial infection can be detected using an ELISA for lipid S (a glycolipid which is related to lipoteichoic acid). This assay has sensitivity and specificity of close to 90%, is rapid, and complements the results from traditional culture microbiology and from endocardial imaging [111]. N A field that is still under active development is the detection of mycobacterial infection using blood or urine as the matrix. Several of these assays use antibodies that specifically recognize unique mycobacterial glycolipids (amongst others: phenolic glycolipid I [112]; lipoarabinomannan (commercial name of the assay: MycoDot TM [113]). For M. leprae infection, the anti-PGL-I IgM ELISA has been widely used for the testing of high-risk groups in countries where leprosy is endemic, in an effort to detect subclinical infections. However, the assay has a low predictive value in this setting of early leprosy detection [114]. A more recent study demonstrated that the levels of PGL-I and anti-PGL-I IgM in the serum of leprosy patients correlates with the bacterial load in the patients [115]. For M. tuberculosis, lipoarabinomannan assays are commercially available with serum as the matrix, and this antigen can also be detected in the urine of tuberculosis n tio In the case of leishmaniasis (Leishmania donovani ), surface expression of 9-O-acetylated sialic acid glycotopes on infected hematopoietic cells and erythrocytes is elevated (produced by promastigotes of the parasite) [124]. Patients have increased titers of anti-9-O-acetylsialic acid IgM antibodies in their serum, which forms the basis for a novel diagnostic test with good clinical characteristics [125, 126]. The standard antigen in this ELISA test is bovine submaxillary mucin, which contains the same 9-O-acetylsialic acid glycotope as Leishmania. The parasite also appears to shed low-molecular weight glycans that are filtered by the kidney into the urine, where they can be detected using antiLeishmania glycoconjugate antibodies [127]. In infections caused by Trichinella spiralis (trichinellosis), the diagnostic antigen in use contains terminal tyvelose residues (3,6-dideoxy-D-arabinose). The tyvelose-GalNAc determinant was chemically synthesized and coupled to BSA to serve as standard antigen in an ELISA with 100% sensitivity [128]. Tapeworm (Taenia solium)-caused neurocysticercosis can be quite reliably diagnosed by measuring Lens culinaris-reactive parasite proteins via ELISA or immunoblot [129, 130]. The diagnostic Clinical Laboratory Testing in Human Medicine Based C-antigens for Echinococcal infections in humans are also of glycoconjugate nature [131, 132], though the detailed structure of these remains unknown. Airway and Salivary Mucus Glycosylation as Predisposing Factors for Infectious Disease Glycoconjugates are abundant at mucosal surfaces, which are often the initial sites of infection. Changes in mucosal glycosylation as a consequence of an underlying disease are under investigation in mucosa that are accessible for clinical diagnostic sampling, as risk factors for infectious disease in compromised patients, especially in cystic fibrosis and asthma. Cystic fibrosis (CF) is caused by lack of function of cystic fibrosis transmembrane conductance regulator protein (CFTR) [133], which results in increased viscosity [134], and reduced clearance [134, 135], of pulmonary mucus. At present, CF diagnosis is specifically and accurately performed with genetic screening [136]. Nevertheless, the disease burden generally comes in bouts, and there are currently no biochemical predictors for when a new bout is upcoming in a patient. The severity of airway infection in CF lung disease and in chronic bronchitis is correlated with high sialylation of airway mucus glycoconjugates [137]. Moreover, increased sulfation of respiratory mucins in CF was observed relative to other respiratory diseases including chronic bronchitis [138] and asthma [139]. It is still a matter of debate whether these glycosylation changes are a consequence of altered conditions in the mucin-secreting cells' secretory pathway due to the primary CFTR defect, or whether secondary effects such as chronic airway inflammation and infection lead to these mucin glycosylation changes that have been observed in vivo. Some in vitro analyses of secreted glycoconjugates in CF have reported increased sulfation [140-143], which would be in support of the first hypothesis, whereas others have failed to detect in vitro CF-specific differences [144, 145]. Methodological differences in these studies are a likely cause of this discrepancy. Whatever may be the underlying cause, the correlation of mucin sialylation with infection severity [137] suggests that prognosis of the progression of CF lung disease may be possible by measuring changes in airway mucus glycosylation. Such prognostic biochemical marker would complement the currently available prognostic analyses such as forced expiratory volume in 1 second (FEV1) and high-resolution computer aided tomography (HRCT) scanning [146, 147]. Accurate biochemical markers that provide this prognostic information would be clinically useful in CF management, as they could trigger prophylactic antibiotic treatment to stem the severe airway infection that often accompanies an active CF period. Current Molecular Medicine, 2007, Vol. 7, No. 4 (MG1) and soluble monomeric MUC7 (MG2) [148]. These mucins are heavily glycosylated with different glycans: MUC7 contains sTn, Le x, sLex and binds LSelectin [149] and various streptococci through these epitopes [150]. On the other hand, MUC5B contains blood group epitopes (A, B, H/O, Lea, Lex, Le b, Le y), sialylated epitopes and many sulfated structures [151-153]. As a consequence the glycosylation of MUC5B is heavily dependent on the subject's blood group status: MUC5B carries blood group antigens corresponding to the subject's blood groups in secretors, but increased sialyl-Le a in nonsecretors [154]. Using lectins, salivary mucous glycosylation differences have been noted between individuals with high vs. low susceptibility to dental caries [155]. It would be interesting to correlate the lectin data with direct structural analysis of salivary mucin glycosylation, to perhaps find out which glycotopes confer increased resistance to the dental caries pathogens (especially Streptococcus mutans). u rib Mucins are also a major component of saliva, and are present in two main fractions, gel-forming MUC5B n tio Diagnostic Tests for Inherited Glycoconjugate Anabolic and Catabolic Defects: Congenital Disorders of Glycosylation and Lysosomal Storage Diseases t is D r o F ot N 405 Lysosomal Storage Diseases The lysosomal storage diseases are a family of rare genetic diseases of which the causative mutation(s) result in a defective lysosomal catabolism of cellular constituents. This defective catabolism causes these cellular constituents to accumulate in the lysosomal compartment of the cell, and the family of diseases owes its name to this lysosomal ‘storage’ phenotype. There are several ways in which storage diseases can be diagnosed. The profiling of glycoconjugates in patient urine plays an important role in newborn screening for lysosomal storage diseases and in the diagnostic workup of identified cases, as most of these defects are in the catabolism of glycoconjugates. As the lysosomal load becomes too large, cells die and release the storage product in the interstitial medium, and at least some of this storage product is filtered out in the kidney and is excreted in urine. The analysis is classically performed using Thin Layer Chromatography for urinary lipids and oligosaccharides [156-158]. For specific subcategories, especially the mucopolysaccharidosis group of diseases (caused by defects in the catabolism of glycosaminoglycans), Fluorophore Assisted Carbohydrate Electrophoresis (FACE) on high-density polyacrylamide gels [159], or electrospray mass spectrometry [160] can be used. The diagnosis of a lysosomal storage disorder is complemented by histology on patient cells (morphology and general chemical reactivity of the storage product). Quantitative assaying of the suspected enzymatic activity (or activities) is also a mainstay for these diagnoses, nowadays often accompanied by genetic analysis of the locus or loci of interest for mutations (such as for Tay-Sachs disease in known risk populations). 406 Current Molecular Medicine, 2007, Vol. 7, No. 4 Congenital Disorders of Glycosylation Congenital Disorders of Glycosylation (CDGs) are a rapidly growing chapter in pediatrics, after the first description of such disease [161, 162]. The original diagnostic test is probably still the best: iso-electric focusing of serum proteins, followed by immunodetection of the transferrin isoforms [163]. Transferrin is a major iron transporting glycoprotein synthesized by the liver. The two N-glycans present can have two to four antennae, each of which can be sialylated or not. Differential sialylation (either because entire N-glycan chains or lost or because the structure of the N-glycans contains fewer sialic acid residues) results in several transferrrin isoforms with different isoelectric points that can be separated using iso-electric focusing. CDG type I is defined as those CDGs with defects that affect the biosynthesis of the N-glycan precursor Glc3Man 9GlcNAc2PPDol, or the transfer of the Glc3Man 9GlcNAc2 moiety to N-glycosylation sites on proteins. These defects generally cause a lower efficiency of N-glycosylation site occupancy, which is detected in transferrin IEF as increased abundance of the disialo-and asialo-isoforms of transferrin. On the other hand, defects in the processing of the precursor glycan after transfer to the protein are classified as CDG type II. The underlying functions that are affected are very diverse: they can be in glycosidases, glycosyltransferases, sugar-nucleotide transporters, and also in components that are involved in secretory pathway homeostasis. Defects in these biosynthetic steps towards fully processed complex-type N-glycans most often result in a lower degree of sialylation of glycoproteins (because of incomplete processing towards the galactosylated glycans that are the substrate of sialyltransferases, because of altered branching, or because of defects in the sialylation process itself). Thus, type II CDGs have a transferrin IEF pattern that also tends to have increased abundance of the monosialylated and trisialylated isoforms, or more generally, their transferrin IEF is ‘not normal and not type I’. The transferrin IEF is a rapid, cheap, quite reliable assay for the biosynthetic pathway that assembles the bulk of serum N-glycans. It is suitable for screening purposes, and this is its current role in the clinic. Nevertheless, the assay’s utility only stretches as far as its molecular basis allows: it will not detect defects in O-glycan biosynthesis, and also not in the biosynthesis of N-glycan modifications that are not present on transferrin (for example brain-specific modifications). It also does not detect defects in noncharged substituents on transferrin glycans (such as fucosylation and bisecting GlcNAc). For fine-typing the N-glycan structural differences in type II CDGs, one can either resort to direct serum protein N-glycan analysis using mass spectrometry, HPLC or capillary electrophoresis, or to mass spectrometrical analysis of transferrin using MALDI or ESI technology [164,165]. Immunoaffinity-LC-ESI [166] is now routinely used in many centers in the US for neonate screening for CDG. Schulz et al. A range of congenital muscular dystrophies (Walker-Warburg syndrome; Muscle-Eye-Brain disease and Fukuyama muscular dystrophy) are caused by defects in O-mannosylation [167]. MEB is caused by a defect in the gene coding for POMGnTI (protein O-mannosylation-N-acetylglucosaminyltransferase I). An enzymatic assay that should be useful in screening for this disorder amongst muscular dystrophy patients has recently been described [168]. Further, Ehler Danlos syndrome is caused by glycosylation defects in a small proteoglycan [169]. Hereditary multiple exostoses are probably caused by defects in heparan sulfate proteoglycan biosynthetic glycosyltransferases [170, 171]. Although technically not a 'congenital' disorder in the true sense of the word, paroxysmal nocturnal hemoglobinuria (PNH) [172] somehow belongs to the same category of genetic disorders in glycosylation biosynthesis. This is a rare hemolytic anemia characterized by the increased sensitivity of cells of the myeloid lineage to complement. This leads to periodic intravascular hemolysis and hemoglobinuria, accompanied with an increased risk on thrombosis. The disease is caused by acquired mutations in the PIG-A gene in haematogenic progenitor cells (stem cells), followed by periodical clonal expansion of these stem cells. PIG-A is involved in glycosylphosphatidylinositol anchor biosynthesis, and thus, PIG-A deficient cells are devoid of or have a strongly reduced abundance of GPI-anchored cell surface proteins (amongst others: those that protect cells against complement lysis: CD55, CD59 and C8BP). This forms the basis of the current diagnostic methods for the disease [173]. In one method, flow cytometry is used on blood cells to detect GPIanchored proteins (CD55 and CD59), whereas in a second method, fluorescently labeled, non-toxic mutants of bacterial GPI-binding toxins (aerolysine or mutants of Clostridium septicum U-toxin [174]) are used for blood cell staining and flow cytometry. u rib t is D r o F ot N n tio Diagnostic Tests for Drugs of Abuse and Illicit Performance-Enhancing Drugs Carbohydrate Deficient Transferrin (CDT) for the Detection of Alcohol Abuse Alcohol is considered the major drug of abuse worldwide, with important social and economical consequences. Therefore, there is a need for the diagnosis and follow-up of alcohol abuse. Since the discovery that alcohol abuse leads to increased levels of under-sialylated transferrin isoforms as compared to subjects with moderate or no alcohol use, carbohydrate-deficient transferrin (CDT) has become the standard marker to evaluate these patients [175-177]. The procedure is the same as for transferrin IEF in the diagnosis of Congenital Disorders of Glycosylation (see above), although the phenotype due to alcohol abuse is usually less severe than the one observed in CDG patients. However, overlaps do exist and there are cases Clinical Laboratory Testing in Human Medicine Based known of heterozygous carriers of 'severe' CDG-I causing mutations that score positive on the CDT test for alcoholism. Tetrasialylated transferrin is the most predominant glycoform in healthy controls (70-80% of total transferrin). The analysis of serum from alcoholic patients typically reveals elevated levels of di- and asialo forms that are formed due to impaired glycosylation: one or two entire N-glycan chains are missing [178]. The percentage of these two glycoforms found in serum is used to demonstrate alcohol abuse (definitions of abuse vary, but >50g/day for more than 5 days is about the average of the definitions). As transferrin has a half-life of ten to fifteen days, a change of this value will be seen when a patient entered a withdrawal treatment, or when a patient returns to alcohol abuse while under treatment. Although CDT is currently the best available marker to detect recent alcohol abuse in at risk populations, the positive predictive value that can be attained in the general population is less than 50% (when compared to self-declared alcohol consumption as 'gold standard', which is certainly not perfect) [179]. It is the higher prevalence of alcohol abuse in risk groups that gives the CDT test a sufficient positive predictive value to be useful. Liver disease appears to only interfere substantially with CDT measurements in a very severe stage (advanced cirrhosis), where a general undersialylation of liver-produced glycoproteins (including transferrin) is observed. Current Molecular Medicine, 2007, Vol. 7, No. 4 polymorphisms often yield inaccurate results in immunoassay-based CDT tests, and IEF, CE or HPLC-based [188] CDT testing is recommended to confirm immunoassay results in critical cases [189]. Recombinant EPO Detection Erythropoietin (EPO) is a glycosylated hormone produced by the kidney in human adults. It induces increased red blood cell mass, haemoglobin concentration and aerobic power. For those reasons, the protein is used therapeutically in the treatment of certain forms of anemia [190]. As the administration of this hormone and its analogues also leads to a substantial improvement of exercise performance, it has been abused as a doping agent by endurance athletes [191]. Both the International Olympic Committee (IOC) and the World Anti-Doping Agency (WADA) have banned the use of recombinant human EPO (rhEPO). Currently, the use of rhEPO is tested using a direct identification method based on isoelectric focusing and double blotting of the EPO present in urine [192]. As in the case of transferrin (described above), differences in the degree and type of sialylation of EPO between endogenous EPO and exogenously administered rhEPO lead to diffent isoelectric points. Human EPO has 3 N- and one O-glycosylation site [193, 194]. Glycosylation is important for both EPO function and pharmacokinetic properties [194, 195]. Most commonly, rhEPO is produced in Chinese hamster ovary cells (CHO) or human kidney cells. In both cases, glycosylation is different from human endogenous EPO [193, 196]. n tio u rib t is D r o F ot CDT measurements can currently be performed using a choice of commercially available and homebrew test formats, which all differ somewhat from each other in their analytical properties [180]. This does not make it easier to compare the data from one study to another, but most studies seem to result in a sensitivity of 40-70% for a specificity of >80% (reviewed in [181]). Because of the higher test volume than in the setting of CDG screening, classical (rather laborious) IEF is not much used anymore, although it remains the gold standard. The commercial CDTect test uses anion exchange microcolumns to separate CDT from non-CDT isoforms, and quantifies transferrin in both fractions using antitransferrin enzyme-linked immunoassay [182]. The %CDT-TIA (turbidimetric immunoassay) and its 2nd generation %CDT use rate nephelometry to quantify total transferrin and CDT upon anion exchange separation. Capillary electrophoresis with UV absorbance detection of total serum proteins can be used to determine CDT: although all serum proteins are detected, immunosubtraction experiments were used to pinpoint the transferrin isoforms amongst the observed peaks. The assay is available commercially with excellent analytical characteristics (CeoFix CDT) [183-185], and some laboratories have developed a homebrew variant [186, 187]. As for IEF-based CDT tests, transferrin genetic variations that result in a more anionic or cationic transferrin protein molecule shift all transferrin glycoform peaks in the IEF pattern or the electropherogram. Such transferrin N 407 There are some disadvantages related to the analysis of urinary EPO. The major drawback is the quick disappearance of measurable rhEPO levels soon after administration. After about four days [197, 198], levels have dropped back to baseline. In contrast, the athlete retains the benefits associated to its use for a longer time. Moreover, conditions like proteinurea influence urinary EPO levels [199] and endurance sports influence renal functions [200]. Therefore, alternative tests are under investigation, which analyze total serum EPO in combination with reticulocytes, heamoglobulin and soluble transferrin receptor concentrations [201]. Diagnostic Tests in Auto-Immune Disorders and Other Chronic Inflammatory States Auto-Antibodies Reactivity with Glycoconjugate (Cross) Anti- S. Cerevisiae Mannan Antibodies (ASCAs) in Crohn's Disease A subgroup of Crohn's disease patients develops an adaptive immune response against S. cerevisiae mannan (detected as Anti- S. cerevisiae Antibodies or ASCAs) [202, 203]. When used in a clinical context, the ASCA assay has a high specificity in the important differential diagnosis from ulcerative colitis [204]. It was recently found that low-expressing 408 Current Molecular Medicine, 2007, Vol. 7, No. 4 alleles of Mannose Binding Lectin (MBL) are very significantly more frequent in ASCA-positive than in ASCA-negative Crohn's disease patients [205]. ASCA-positive patients also had a significantly lower MBL protein level than ASCA-negative patients. This would indicate a mechanism in which the increased intestinal permeability in Crohn's disease causes an increased S. cerevisiae burden, which would normally be handled by the MBL-mediated branch of the alternative complement activation pathway. This would not be the case in MBL-deficient patients, who hence develop an anti- S. cerevisiae humoral and cellular immune response. Anti-Ganglioside Antibodies in Several Neuropathies A large body of data (about 500 studies, curated bibliography available upon request) is available in the literature on disorders of the Guillain-Barre syndrome family and the role that anti-ganglioside antibodies play in the pathogenesis of these diseases. These syndromes are strongly associated with previous infection with different pathogens, but most notably with Campylobacter jejuni and related species. Conclusive evidence has recently been obtained to prove that the Lipid-linked Oligosaccharide in the cell wall of these pathogens contains an epitope which induces antibodies that are cross-reactive with gangliosides that are prevalent in the myelin sheath of peripheral nerves [206]. These antibodies block neuromuscular conduction and may attract phagocytic cells to attack the antibody-decorated Schwann cells. Schulz et al. cannot be established by a single test, but is accomplished by a combination of laboratory test results and clinical scoring systems. Now that new therapies are starting to make an impact, more biomarkers are needed to follow-up the condition of these patients (as objective measures of response to these expensive therapies that are not devoid of serious side-effects such as emergence of tuberculosis [217, 218]). One of the possible bases for such a therapymonitoring marker is the well-established observation that the N-glycans of immunoglobulin-G molecules are hypo-galactosylated in RA-patients [219-221]. However, a lower galactosylation degree of IgGs was later also observed in other inflammatory diseases like lupus, tuberculosis, Crohn’s disease, systemic vasculitis and in some cancers [222] (possibly reflecting inflammatory aspects of the disease), whereas an increase in galactosylation is associated with pregnancy [223]. Because of this lack of specificity, this marker is not in widespread clinical use for diagnostic purposes. With the advent of DMARDs, however, there is currently more need for non-invasive markers to assess chronic rheumatoid disease activity, progression rate and response to therapy. We believe that IgG galactosylation status could play a much more useful role here than for the differential diagnosis purposes for which it was originally conceived. We recently completed a study to explore the utility of this marker in a number of these clinical contexts [Laroy and Callewaert, manuscript submitted]. The technologies used are most often micro-purification of serum immunoglobulins, followed by N-glycan preparation and fluorescent labeling and profiling of these glycans using HPLC or CE. The entire procedure can now be performed with clinically relevant throughput. u rib t is D r o F ot Because of the pathophysiological role of ganglioside-crossreactive antibodies in the Guillain Barre syndrome and related disorders such as Fisher Syndrome and Bickerstaff's brainstem encephalitis, the laboratory detection of such antibodies (using immunoassays with the gangliosides as antigens) aids in the diagnosis of the disease [207-210]. N Under-Galactosylation of Serum IgG in a Range of Chronic Inflammatory Disorders Rheumatoid arthritis (RA) is a major cause of disability, morbidity and mortality [211, 212]. It is a chronic, systemic inflammatory autoimmune disease with as the primary target the synovium. It affects almost 1% of the adults worldwide and is three times more common in women, in whom it also has an earlier onset. Although the exact mechanisms behind the disease are yet unknown, the understanding of cytokine networks responsible for the ongoing inflammatory response and of the pathophysiology of the disease have led to several therapies that modifiy the disease process. Methotrexate is currently the most prescribed of these Disease Modifying anti-Arthritis Drugs (DMARDs), and new therapies directly target the proinflammatory cytokines TNFα and IL1. An early diagnosis is critical for successful treatment, as a delay of treatment of only three months results in substantially more bone damage after 5 years [213216]. At present, however, a definite diagnosis n tio Recently, it was indicated that differences in IgG galactosylation may be due to increased sialylation of the β-1,4-galactosyltransferase enzyme in rheumatoid arthritis, which inhibits its enzymatic activity [224]. However, this finding was based on serum β-1,4-galactosyltransferase isoform profiling and should be confirmed on plasma cell β-1,4galacytosyltransferase. Cholinesterase Glycoforms in the Diagnosis of Alzheimer's Disease One laboratory has reported that an increased level of non-concanavalin A (ConA)-binding isoforms of acetylcholinesterase in lumbar cerobrospinal fluid is diagnostic for Alzheimer's disease and was not detected in other illnesses that cause dementia [225]. This increase in non-ConA binding was also found in a mouse model of AD (APP Tg2576 transgenic mice) [226]. In a more recent study, the level of non-ConA binding CSF acetylcholinesterase was found to positively correlate with the duration of Alzheimer's disease, but to be not suitable for earlier diagnosis of AD than is currently clinically possible. Clinical Laboratory Testing in Human Medicine Based These findings are very interesting, but there are as yet no reports available from other laboratories to validate this assay. In our own studies involving CSF glycomics (unpublished), we have found substantial differences in glycosylation profiles with samples from different clinics. This most likely reflects different sampling and pre-analytical procedures (centrifugation, storage,...) possibly compounded by different average levels of blood serum contamination of these samples. CSF sampling and handling requires substantially more skill and is unfortunately much less standardized in the medical practice than blood serum sampling and handling. Diagnostic Tests to Monitor Status of the Liver Chronic liver pathology is a complex syndrome with different stages. As a result of the exposure to hepatotoxic agents (alcohol, viruses), liver tissue undergoes inflammatory necrosis. As a reaction, activated hepatic stellate cells in this tissue will secrete extracellular matrix components. Due to the chronic character of the necrosis, this process (which is a component of normal healing) overshoots, leading to excessive deposition of extracellular matrix and a disturbance in liver architecture. This is called fibrosis and different stages are histologically distinguishable. The most severe form of fibrosis is liver cirrhosis and is characterized by regenerative nodules: clusters of replicating hepatocytes in an abnormal architecture. This chronic regenerative activity is a major risk factor for the development of hepatocellular carcinoma. Up till now, the clinical assessment of a patient's liver condition is achieved by the histological examination of liver tissue. The invasive percutanous biopsy which is needed to obtain this tissue involves serious discomfort to the patient, and it is a costly procedure (estimated at 1500 USD). Moreover, the procedure is not entirely risk-free, especially in patients with rather advanced liver disease. All disadvantages taken together, liver biopsy is not a suitable procedure for regular followup of patients who have been diagnosed with a chronic liver disorder, and non-invasive alternatives are needed to this end. Current Molecular Medicine, 2007, Vol. 7, No. 4 liver cirrhosis strongly pre-disposes patients for the development of liver cirrhosis (about 40-fold increased risk), another important question is: does the patient have the chronic regenerative activity associated with liver cirrhosis [227]? This entails detection of the regenerative activity before clinical signs of cirrhosis appear, to be able to put these patients on an intensified screening regime for HCC. Glycosylation-based assays can help in answering these questions, based on the fact that the major fraction of serum glycoprotein is synthesized by hepatocytes, while the liver also resorbs a large amount of serum glycoproteins. Therefore, changes in the whole serum protein glycosylation profile reflect the liver condition. A very basic approach has been the monosaccharide composition measurement of serum proteins [228], with conclusions that fucose and GlcNAc are elevated when normalized per 3 mol mannose (3 mannose residues are the common element of every abundant serum N-glycan) in severe liver disease. The increase in GlcNAc was interpreted as an increased degree of branching, but taking the result of recent studies [229] on the structures of serum protein N-glycans in liver disease into account, one has to revise this interpretation: the extra GlcNAc in the early studies most likely ensues from an increase in 'bisecting' GlcNAc rather than from an increase in branching of the N-glycans. A typical setting in which such repetitive assessment is needed, is in determining whether a patient responds to therapy with a stabilisation or improvement of the fibrosis (therapies are either antiviral as in the case of hepatitis C infection, or antifibrotic (the latter still experimental)). Changes in a number of serological parameters (most notably hyaluronic acid concentration, α 2macroglobulin concentration, aminotransferase activities, bilirubin concentration, etc.) correlate somewhat with the stage of liver fibrosis. These changes have been mathematically combined in a number of regression models to obtain scores that better answer the clinical questions at hand than the single parameters. One important question is: is there 'clinically significant' fibrosis in the patient's liver or not (important in making treatment decisions)? As n tio u rib t is D r o F ot N 409 The quantitation of Aleuria aurantia lectin-binding (i.e. fucosylated) serum cholinesterase isoforms, relative to the total amount of serum cholinesterase, can distinguish compensated liver cirrhosis (Child's stage A) from chronic active hepatitis with an accuracy of 70% [230]. Along similar lines, the fucosylation of α 1-acid glycoprotein was found to be higher in patients with liver cirrhosis than in patients with non-cirrhotic liver disease [231], and the diagnostic accuracy of the lectin immunoassay that was developed for this purpose was very similar to that of hyaluronic acid (about 75-80%), although the cirrhosis patients in this study were not stratified according to clinical severity. As mentioned above, the core-fucosylation of α -fetoprotein is higher in patients with hepatocellular carcinoma than in those with non-HCC liver disease, and this is useful diagnostically. However, AFP levels and AFP fucosylation are also increased in conditions associated with massive hepatic regeneration, such as acute hepatitis. As a consequence, the AFP-L3 assay should be used with caution and probably only in patients with chronic liver disease (and not in the acute phase of the disease). Curiously, AFP fucosylation does not correlate with the fucosylation of other, abundant serum glycoproteins (tested were transferrin and α -1-antitrypsin) [232], which probably helps to explain why AFP seems only hyperfucosylated in HCC, whereas hyperfucosylation of the abundant proteins can be used diagnostically for liver cirrhosis (which normally precedes HCC in 95% of the cases). 410 Current Molecular Medicine, 2007, Vol. 7, No. 4 Schulz et al. Some studies have explored imaging of the biodistribution of technetium 99m-labeled, galactosesubstituted human serum albumin to measure the activity of the liver asialoglycoprotein-mediated clearance (and thereby of the functional reserve of the liver for this function), and one generally finds good correlations with other such assays of liver capacity [233]. In our own work [229], we have developed a semi-automated capillary electrophoresis-based profiling tool for the total serum protein N-glycome. This technology is very rapid and robust and strikes a good balance between a low technical complexity (as is desirable for clinical implementation) on the one hand and yielding sufficient analyte structural information on the other hand (as is desirable to link the diagnostic proflie changes to known aspects of disease pathology). This glycomics diagnostic technology was explored for the diagnosis of liver cirrhosis, and we found that about 75-80% of compensated liver cirrhosis cases can be detected by measuring changes in the serum protein N-glycan profile, with very high specificity. We also found evidence to show that the serum protein N-glycan profile contains information that can be used in monitoring the progression of liver fibrosis from early stages onwards. Some of the structural alterations that were detected are compatible with the lectinbased studies that have shown increased fucosylation of serum proteins in liver cirrhosis, but the best diagnostic parameter that was derived from the serum N-glycome profiles was not dependent on fucosylation, but rather on an increased abundance of glycans that are substituted with a bisecting GlcNAc residue and a decreased abundance of the triantennary non-fucosylated structure. Based on lectins, it would be very difficult to find and reliably measure this combination of glycan structural alterations. stringency of the clinical laboratory in terms of limitations in assay complexity, validation requirements, and not in the least: cost pressure. Therefore, a realistic perspective on the booming diagnostic 'omics' efforts must be that the technologies used for discovery of novel markers will generally not be the ones that will be used for the final clinical implementation of these markers. DNA arrays with tens of thousands of features will hardly be used in the routine clinical lab, shotgun proteomics measurements also not, and highresolution mass spectrometrical glycoconjugate profiling neither. The goal of the discovery stage with high-complexity, comprehensive profiling technology must be to identify as small a set of analytes as possible that, when measured together, contain as much of the diagnostic information contained in the entire dataset as possible. Then, focused test methodology needs to be devised to measure this limited set of analytes in a format that allows thorough validation and easy implementation in a moderate complexity clinical laboratory environment. In practice, such tests are almost invariably based on specific binding with the analytes of interest, be they nucleic acids (hybridization), proteins (antibodies) or glycans (lectins). In rare cases, robust separation technology and direct non-analyte specific detection can be clinically implemented. This is mainly the case when the diagnostic information is contained in the major components of a mixture. This is the case for serum protein electrophoresis, for total serum protein N-glycome profiling, and for capillary electrophoresisbased Carbohydrate Deficient Transferrin detection, all of which have diagnostic utility. u rib t is D r o F ot N All of these findings are currently being validated in larger studies of different designs, to fully assess the cllinical utility of these novel assays. As the assays are based on analytes which have not been used before in a clinical context, we are investing significant efforts at present in the analysis of factors that may influence the serum protein N-glycome, in order to establish sound guidelines for physicians and clinical laboratories that will perform the assay (patient pre-conditioning, choice of serum or plasma as the matrix, way of preparing the biofluid, potential influence of different storage conditions etc.). GLYCOCONJUGATE BIOMARKER DISCOVERY/TESTING TECHNOLOGIES CURRENTLY USEDANDUNDERDEVELOPMENT:THE FUTURE OF GLYCOCONJUGATE-BASED TESTING General Remark: Development Reality of Clinical Assay All 'omics' technologies that move out of the academic research laboratory soon encounter the n tio Glycoconjugate Mass Spectrometry Since the introduction of soft ionisation techniques, mass spectrometry (MS) has increased in importance and success as a carbohydrate analytical technique [234]. The benefits of carbohydrate analysis using MS include remarkable sensitivity (sub-picomolar to femtomolar), the possibility for a large degree of automation, high analysis speed (minutes to seconds per analysis depending on the specific MS technique), the ability to couple on-line separation of isomeric oligosaccharides with MS detection, and most importantly: more or less unbiased detection of a wide variety of molecular species in the same experiment. However, oligosaccharide characterisation by MS relies heavily on biological rules determined previously or in combination with other methods, such as NMR or glycosidase digestion. For instance, oligosaccharide linkage anomericity is very difficult to determine by MS alone. However, once these general rules have been outlined for a given biological system, MS provides a very useful compromise between speed and extent of characterisation. Because it lends itself to robust medium- to high-throughput analyses, MS shows great promise as a discovery tool in diagnostic glycomics. Clinical Laboratory Testing in Human Medicine Based Essentially all modern mass spectrometers can be used successfully for oligosaccharide analysis and characterisation [235]. These different techniques have various advantages. For instance, MALDI analysis is typically faster, while on-line LC and GC allow separation and detection of isomeric oligosaccharides, and ion trap instruments typically allow easier structural analysis through ion fragmentation. MS-based approaches have certainly shown their merit for sensitive and detailed characterisation of carbohydrates, but the robust performance required for routine clinical application is often lacking. Therefore, the challenge is now for instrument manufacturers, researchers, and clinicians to develop MS as a robust and cost-effective technique for general clinical application. The recent success of clinical applications of MS (particularly ESI-MS/MS) for inherited metabolic disease screening (including amino acid, acylcarnitine, steroid and lysosomal enzymes analysis) (reviewed in [236] and [237]) bodes well for the future. Glycodiagnostic markers must not only be well characterised, but also relatively or absolutely quantifiable. Standard MS techniques provide only limited quantitation, in terms of total ion current measurements, which can vary considerably between measurements. A rapidly expanding technique for relative quantitation of analyte abundances from different samples is labelling with stable isotope-containing reagents [238]. The central premise of quantitation through stable isotope coding is to label analytes from different samples with the same chemical derivative, but containing different amounts of heavy isotope atoms (typically deuterium, 13C or 15N). The samples are then mixed, and detected in a single MS experiment. Peptides are detected as doublets with molecular weights differing by the difference in mass of the heavy and unlabelled reagents, and quantiation is performed by comparing the relative intensities of peaks from each sample. The most commonly used reactive groups on proteinaceous analytes are cysteine residues, the peptide amino and carboxy termini, and lysine ε amino groups [238]. It is conceivable that oligosaccharides could be similarly derivatised with heavy and light reagents (for example through reducing terminal reductive amination) for relative or absolute quantitation in MS. This would enable more reliable quantitative approaches to disease diagnosis and prognosis based on glycan biomarkers. Current Molecular Medicine, 2007, Vol. 7, No. 4 been developed to provide equivalents to peptide mass fingerprinting [241] of peptides, or SEQUEST pattern matching of peptide fragmentation. After glycan characterisation, the large amounts of processed information also require consistent bioinformatic and database implementation (including carbohydrate structure, disease state and analytical methodology). Such databases are now available: GlycoSuite ( [242] and SweetDB ( [243]. Central data repositories enable the accumulation of many researchers’ work, and become acutely necessary as the amount of data generated increases. The value of genomic and proteomic databases is apparent, and the same will become true of glycomic databases, enabling efficient data mining and critical assessment of the usefulness and validity of potential glycodiagnostics. MS analytical techniques are a very useful tool in glycome characterisation by enabling the collection of vast amounts of data describing the structures of large numbers of glycans in a single experiment. In these circumstances, structural and statistic data analysis is often much more time consuming and difficult than the wet-lab data collection itself. These large data sets necessitate efficient bioinformatic tools to assist data handling and processing. Tools such as the glycan-fragment mass-fingerprinting tool GlycosidIQ [237] or glyco-search-ms [240] have n tio Improved Sample Preparation: Glycoprotein Enrichment Technologies Selective A consistent problem in (glyco)proteomic analyses is the orders of magnitude difference in protein abundances in a single sample. This causes difficulties because analytical techniques have inherently limited dynamic detection ranges, and also because more abundant components will tend to mask lower abundant components. A solution to these problems is to prefractionate proteins in a sample before analysis [244]. Similar approaches are also possible for glycomic and glycoproteomic analysis. For instance, affinity chromatography enrichment of glycoproteins for glycosylation or proteomic analysis [245] has been performed with lectins [246, 247] and boronate [248], while hydrazide chemistry allows the covalent capture of glycopeptides [249]. These various approaches allow a focused investigation of glycosylation and glycosylated proteins and can improve the chances to detect also the low-abundance glycoproteins which often contain the most useful diagnostic information (as is the case for tumor marker glycoproteins in early tumor stages). However, it must be said that virtually all of these enrichment methodologies are, to a different extent, selective for certain glycan structures. u rib t is D r o F ot N 411 Capillary Array Analyzers as a Platform for Diagnostic Glycomics Analyzers that use an array of 96 parallel capillaries for capillary electrophoresis were the workhorse platforms in the sequencing of the human and many other genomes. At present, such instruments are available from several manufacturers at scales of 1 up to 384 capillaries, thus fitting the most diverse needs of genetic analysis laboratories. For DNA-analysis, these analyzers are equipped with 4-or 5-colour fluorescence detection, most often excited by an argon laser at 488 nm, but also from red diode lasers or from very bright green or blue light emitting diodes. The technology is now also being implemented in small, less versatile, cheaper, 412 Current Molecular Medicine, 2007, Vol. 7, No. 4 dedicated analyzers that are more suitable for the clinical laboratory, to cope with the increasing number of DNA-based diagnostic tests, especially in such high-throughput fields as food and clinical microbiology [250, 251]. We implemented carbohydrate analysis on the existing DNA-analyzers to benefit from the technological developments in the higher-volume DNA-diagnostics market. Virtually no modification is needed to the buffer systems, capillary dynamic coatings and separation polymers that were optimized for DNA separation to obtain state-of-theart resolution for 8-amino-1,3,6-pyrenetrisulfonic acid labeled glycans on the most widely used DNAanalysis systems ([229]; Laroy et al. manuscript in preparation). Indeed, DNA fragment analysis and glycan analysis can effectively be performed in parallel, if needed even in the same electrophoresis run. This makes for very cost-effective glycan analysis in molecular biology laboratories, for which glycan analysis is often an infrequent need which would not justify the major expenses needed for dedicated glycan analytical equipment. We are currently also starting to explore capillary array DNA-analysers for different glycomics purposes besides fluorophore-labeled glycan analysis. Schulz et al. field of glycosylation-based diagnostics would greatly benefit from a set of lectins with simple and strict binding specificity and high binding affinity to common, small substructures of human O-and Nglycans. However, as most of these human glycotopes are 'self' to the species in general use for monoclonal antibody generation (mouse, rat), classical immunisation strategies of these species generally fail in yielding these desired antibodies. One recent spin-off of the generation of glycosyltransferase knock-outs in mice has been that these mice can be used for the generation of antibodies against the glycotope which is missing as a consequence of the biosynthetic defect [253]. However, with many glycosyltransferase activities being encoded by multigene families, it is unlikely that this will become a generic strategy for the production of better glycotope-binding proteins of desired specificity. What remains is the everexpanding range of in vitro panning strategies of huge libraries of sequence-varied protein scaffolds to select those few sequence variants that bind the glycotope of interest (and not other, structurally related ones in the mixture in which the glycotope has to be specifically detected to be of diagnostic utility). This can be achieved by repetitive positive and negative selection rounds, followed by more directed in vitro affinity maturation of the few clones with promising characteristics. In the hands of careful and skilled specialists, these strategies can be succesful in rapidly generating a binding protein with the desired characteristics. Nevertheless, these techniques currently seem under-used for the generation of glycoconjugate-recognizing proteins. The most important problem is that most display technologies are geared towards monomeric binding proteins, whereas high-affinity in carbohydrate recognition (at least in nature) is generally achieved through avidity effects based on oligomeric binding proteins. Even in phage display experiments of monomeric carbohydrate-recognizing modules, a strong selection pressure towards mutants that oligomerise the modules instead of mutants with higher intrinsic affinity of the binding site has been observed [254]. This causes problems in then going from the phage-displayed protein to recombinant protein production, as the protein might not oligomerize as it did in the phage context, unless the adaptation to oligomerisation is very strong. A very recent study has indeed built upon that logic in the case of generating single-chain antibodies against the T-antigen that are not protein-dependent, by keeping the linker between the phage protein and the single chain scaffold very short (0 or 1 amino acids). This yielded oligomeric single chains which were also oligomerizing when expressed in E. coli out of the phage context [254]. The future will tell whether this strategy is of general applicability in generating glycan-binding proteins with the desired properties. u rib t is D r o F ot Upon detection of diagnostically relevant differences amongst the many samples that need to be analyzed in clinical discovery programs, the analytes of interest can then be identified using focused, low-throughput CE-MS, which is now coming of age as a powerful analytical technique [252]. Development of Carbohydrate-Binding Proteins with Higher Specificity and Binding Affinity N Much of contemporary diagnostic glyco(proteo)mics research depends on complex analytical methodologies for the profiling of glycoconjugate mixtures. Many of these methods rely on multi-dimensional separations and sophisticated mass spectrometry techniques, as described above. However, once a glycosylation-based biomarker is discovered using these technologies, a clinically useful assay then has to be developed which can be performed in an economical way in the high qualitystandard analytical environment of a routine clinical laboratory. Most successful such assays rely on specific binding proteins to either capture the molecules that contain the biomarker information, to detect the glycotope(s) of interest on these molecules, or both. Most of the currently available lectins have complex specificities, which can cause difficulties in developing an assay for the glycosylation-based marker that was found in the discovery phase of a project. Moreover, the binding affinity generally is orders of magnitude lower than one typically can count on for antibodies (micromolar instead of nanomolar), which can severely affect the robustness of lectin-based assays. Therefore, the n tio We have recently explored the Yeast Surface Display system for the purpose of expression cloning Clinical Laboratory Testing in Human Medicine Based of glycan-binding proteins from complex DNA libraries (Ryckaert, S., Callewaert, N. et al. manuscript submitted). The salient features of this system are cell-surface wide expression of thousands of molecules consisting of a fusion between the protein of interest and a yeast cell-wall associated protein, very much mimicking the cell surface of a lectinexpressing cell. By using multivalent glycan molecules or high-density glycan surfaces as the selection agents, fast and strong enrichment of those yeast cells that express the lectin fusion proteins with desired specificity and high affinity can be achieved. We are now exploring the utility of this system for true carbohydrate binding-site affinity maturation. Current Molecular Medicine, 2007, Vol. 7, No. 4 [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] CONCLUSION The field of glycoconjugate-based diagnostics has a long tradition and has provided a multitude of very useful disease markers that are now part of the routine testing procedures in pathogen identification and classification, in cancer diagnosis and monitoring, in liver disease assessment, in the tracking of drug abuse,...) With the advent of better analytical technologies for glycoconjugates, the field is undergoing a revival: old markers are being structurally characterized to gain the knowledge required to improve their diagnostic performance; new markers and testing procedures are being discovered; and technologies geared towards the clinical laboratory are being devised. Structural and quantitative aspects of glycoconjugates often reflect the identity and the condition of the cell that produces them, and we have been, are, and will be tapping into this information resource to diagnose, predict the course of and the response to therapy of a broad range of human diseases. [16] [17] [18] [19] [20] [21] Hanisch, F. G. and Baldus, S. E. (1997) Histol. Histopathol., 12 , 263-281. Reddi, A. L., Sankaranarayanan, K., Arulraj, H. S., Devaraj, N. and Devaraj, H. (2000) Cancer Lett., 149 , 207-211. Cao, Y., Karsten, U. R., Liebrich, W., Haensch, W., Springer, G. F. and Schlag, P. M. (1995) Cancer, 76 , 1700-1708. Toma, V., Sata, T., Vogt, P., Komminoth, P., Heitz, P. U. and Roth, J. (1999) Cancer, 85 , 2151-2159. Brooks, S. A. and Carter, T. M. (2001) Acta Histochem., 103 , 37-51. Brooks, S. A., Lymboura, M., Schumacher, U. and Leathem, A. J. (1996) J. Histochem. Cytochem., 44 , 519-524. Brooks, S. A. and Wilkinson, D. (2003) A cta Histochem., 105 , 205212. Osinaga, E., Babino, A., Grosclaude, J., Cairoli, E., Batthyany, C., Bianchi, S., Signorelli, S., Varangot, M., Muse, I. and Roseto, A. (1996) Int. J. Oncol., 8, 401-406. Thies, A., Moll, I., Berger, J. and Schumacher, U. (2001) Br. J. Cancer, 84 , 819-823. Brooks, S. A. (2000) Histol. Histopathol., 15 , 143-158. Mitchell, B. S. and Schumacher, U. (1999) Histol. Histopathol., 14 , 217-226. Schumacher, U. and Adam, E. (1997) Histochem. J., 29 , 677-684. Brooks, S. A. and Hall, D. M. S. (2002) Clin. Exp. Metastasis, 19 , 487-493. Brockhausen, I., Yang, J., Dickinson, N., Ogata, S. and Itzkowitz, S. H. (1998) Glycoconjug. J., 15 , 595-603. Clement, M., Rocher, J., Loirand, G. and Le Pendu, J. (2004) J. Cell Sci., 117 , 5059-5069. Bresalier, R. S., Ho, S. B., Schoeppner, H. L., Kim, Y. S., Sleisenger, M. H., Brodt, P. and Byrd, J. C. (1996) Gastroenterology, 110 , 1354-1367. Vierbuchen, M. J., Fruechtnicht, W., Brackrock, S., Krause, K. T. and Zienkiewicz, T. J. (1995) Cancer, 76 , 727-735. Cho, S. H., Sahin, A., Hortobagyi, G. N., Hittelman, W. N. and Dhingra, K. (1994) Cancer Res., 54 , 6302-6305. Le Pendu, J., Marionneau, S., Cailleau-Thomas, A., Rocher, J., Le Moullac-Vaidye, B. and Clement, M. (2001) A pmis, 109 , 9-31. Thurin, M. and Kieber-Emmons, T. (2002) Hybrid. Hybridomics, 21 , 111-116. Nakamori, S., Kameyama, M., Imaoka, S., Furukawa, H., Ishikawa, O., Sasaki, Y., Kabuto, T., Iwanaga, T., Matsushita, Y. and Irimura, T. (1993) Cancer Res., 53 , 3632-3637. Nakamori, S., Kameyama, M., Imaoka, S., Furukawa, H., Ishikawa, O., Sasaki, Y., Izumi, Y. and Irimura, T. (1997) Dis. Colon Rectum, 40 , 420-431. Ogawa, J., Inoue, H. and Koide, S. (1997) Cancer, 79 , 1678-1685. Tatsumi, M., Watanabe, A., Sawada, H., Yamada, Y., Shino, Y. and Nakano, H. (1998) C lin. Exp. Metastasis, 16 , 743-750. Nakamori, S., Furukawa, H., Hiratsuka, M., Iwanaga, T., Imaoka, S., Ishikawa, O., Kabuto, T., Sasaki, Y., Kameyama, M., Ishiguro, S. and Irimura, T. (1997) J. Clin. Oncol., 15 , 816-825. Amado, M., Carneiro, F., Seixas, M., Clausen, H. and SobrinhoSimoes, M. (1998) Gastroenterology, 114 , 462-470. Miyake, M., Taki, T., Hitomi, S. and Hakomori, S. (1992) N. Engl. J. Med., 327 , 14-18. Klinger, M., Farhan, H., Just, H., Drobny, H., Himmler, G., Loibner, H., Mudde, G. C., Freissmuth, M. and Sexl, V. (2004) Cancer Res., 64 , 1087-1093. Dennis, J. W., Granovsky, M. and Warren, C. E. (1999) B iochim. Biophys. Acta, 1473 , 21-34. Fernandes, B., Sagman, U., Auger, M., Demetrio, M. and Dennis, J. W. (1991) Cancer Res., 51 , 718-723. Murata, K., Miyoshi, E., Ihara, S., Noura, S., Kameyama, M., Ishikawa, O., Doki, Y., Yamada, T., Ohigashi, H., Sasaki, Y., Higashiyama, M., Tarui, T., Takada, Y., Kannagi, R., Taniguchi, N. and Imaoka, S. (2004) Oncology, 66 , 492-501. Petretti, T., Kemmner, W., Schulze, B. and Schlag, P. M. (2000) Gut, 46 , 359-366. Cappelli, G., Paladini, S. and D'Agata, A. (1999) Tumori, 85 , S19S21. Sawabu, N., Watanabe, H., Yamaguchi, Y., Ohtsubo, K. and Motoo, Y. (2004) Pancreas, 28 , 263-267. Shimono, R., Mori, M., Akazawa, K., Adachi, Y. and Sgimachi, K. (1994) Am. J. Gastroenterol., 89 , 101-105. Alvarez, J. A., Marin, J., Jover, J. M., Fernandez, R., Fradejas, J. and Moreno, M. (1995) Dis. Colon Rectum, 38 , 535-542. Reiter, W., Stieber, P., Reuter, C., Nagel, D., Lau-Werner, U. and Lamerz, R. (2000) Anticancer Res., 20 , 5195-5198. Haglund, C., Roberts, P. J., Jalanko, H. and Kuusela, P. (1992) Scand. J. Gastroenterol., 27 , 169-174. Sperti, C., Pasquali, C., Catalini, S., Cappellazzo, F., Bonadimani, B., Behboo, R. and Pedrazzoli, S. (1993) J. Surg. Oncol., 52 , 137141. Ramage, J. K., Donaghy, A., Farrant, J. M., Iorns, R. and Williams, R. (1995) Gastroenterology, 108 , 865-869. Siqueira, E., Schoen, R. E., Silverman, W., Martini, J., Rabinovitz, M., Weissfeld, J. L., Abu Elmaagd, K., Madariaga, J. R. and Slivka, A. (2002) Gastrointest. Endosc., 56 , 40-47. Yoshida, E. M., Scudamore, C. H., Erb, S. R., Owen, D. A. and Silver, H. K. (1995) Can. J. Surg., 38 , 83-86. Adachi, Y., Iso, Y., Moriyama, M., Kasai, T. and Hashimoto, H. (1998) Hepato-Gastroenterol., 45 , 77-80. Fabris, C., Falleti, E., Pirisi, M., Soardo, G., Toniutto, P., Vitulli, D., Bortolotti, N., Gonano, F. and Bartoli, E. (1995) Clin. Chim. Acta, 243 , 25-33. Decker, D., Bollmann, R., Hirner, A. and Stratmann, H. (1998) Zentralbl Chir., 123 , 855-857. N [23] [24] [25] [26] [27] [28] [29] [30] [31] ACKNOWLEDGEMENTS We thank Markus Aebi for providing us with the necessary time to write this review. Research in the author's labs is funded by the Swiss Federal Institute of Technology (GlycoINIT project) and a Marie Curie Excellence grant to N.C. N.C. holds an honorary fellowship of the Fund for Scientific Research Flanders. W.L. is a postdoctoral fellow of the IWTFlanders. Karl Rumbold is acknowledged for proofreading the manuscript. A complete manually curated bibliography (including >1500 relevant references) is available upon request from the authors. Readers are encouraged to provide us with crucial references that we might have missed, and we regret that space limitations have forced us to omit a lot of interesting work. [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] REFERENCES [1] Brockhausen, I. (2003) in: Glycobiology and Medicine, Advances in Experimental Medicine and Biology, 535 , 163-188. [46] n tio u rib t is D r o F ot [22] 413 414 Current Molecular Medicine, 2007, Vol. 7, No. 4 [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [67] [68] [69] [70] [71] [72] [73] [74] N [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] Hammarstrom, S., Engvall, E., Johansson, B. G., Svensson, S., Sundblad, G. and Goldstein, I. J. (1975) Proc. Natl. Acad. Sci. USA, 72 , 1528-1532. Kuroki, M., Arakawa, F., Haruno, M., Murakami, M., Wakisaka, M., Higuchi, H., Oikawa, S., Nakazato, H. and Matsuoka, Y. (1992) Hybridoma, 11 , 391-407. Yamashita, K., Totani, K., Kuroki, M., Matsuoka, Y., Ueda, I. and Kobata, A. (1987) Cancer Res., 47 , 3451-3459. Yamashita, K., Totani, K., Iwaki, Y., Kuroki, M., Matsuoka, Y., Endo, T. and Kobata, A. (1989) J. Biol. Chem., 264 , 17873-17881. Fukushima, K., Ohkura, T., Kanai, M., Kuroki, M., Matsuoka, Y., Kobata, A. and Yamashita, K. (1995) G lycobiology, 5, 105-115. Yamashita, K., Fukushima, K., Sakiyama, T., Murata, F., Kuroki, M. and Matsuoka, Y. (1995) Cancer Res., 55 , 1675-1679. Baldus, S. E., Hanisch, F. G., Monaca, E., Karsten, U. R., Zirbes, T. K., Thiele, J. and Dienes, H. P. (1999) Histol. Histopathol., 14 , 1153-1158. Baldus, S. E., Zirbes, T., Glossmann, J., Fromm, S., Hanisch, F. G., Monig, S. P., Schroder, W., Schneider, P. M., Flucke, U., Karsten, U., Thiele, J., Holscher, A. H. and Dienes, H. P. (2001) Oncology, 61 , 147-155. Baldus, S. E., Zirbes, T. K., Hanisch, F. G., Kunze, D., Shafizadeh, S. T., Nolden, S., Monig, S. P., Schneider, P. M., Karsten, U., Thiele, J., Holscher, A. H. and Dienes, H. P. (2000) Cancer, 88 , 1536-1543. Flucke, U., Zirbes, T. K., Schroder, W., Monig, S. P., Koch, V., Schmitz, K., Thiele, J., Dienes, H. P., Holscher, A. H. and Baldus, S. E. (2001) Anticancer Res., 21 , 2189-2193. Guo, J. M., Zhang, X. Y., Chen, H. L., Wang, G. M. and Zhang, Y. K. (2001) J. Cancer Res. Clin. Oncol., 127 , 512-519. Hayden, R. T., Qian, X., Procop, G. W., Roberts, G. D. and Lloyd, R. V. (2002) Diagn. Mol. Pathol., 11 , 119-126. Shinoda, T., Kaufman, L. and Padhye, A. A. (1981) J. Clin. Microbiol., 13 , 513-518. Mercure, S., Senechal, S., Auger, P., Lemay, G. and Montplaisir, S. (1996) J. Clin. Microbiol., 34 , 2106-2112. Heelan, J. S., Sotomayor, E., Coon, K. and D'Arezzo, J. B. (1998) J. Clin. Microbiol., 36 , 1443-1445. Patterson, T. F., Miniter, P., Patterson, J. E., Rappeport, J. M. and Andriole, V. T. (1995) J. Infect. Dis., 171 , 1553-1558. Ertl, P. and Mikkelsen, S. R. (2001) Anal. Chem., 73 , 4241-4248. Harbeck, R. J., Teague, J., Crossen, G. R., Maul, D. M. and Childers, P. L. (1993) J. Clin. Microbiol., 31 , 839-844. Connaughton, M., Lang, S., Tebbs, S. E., Littler, W. A., Lambert, P. A. and Elliott, T. S. J. (2001) J. Infect., 42 , 140-144. Agis, F., Schlich, P., Cartel, J. L., Guidi, C. and Bach, M. A. (1988) Int. J. Lepr. Other Mycobact. Dis., 56 , 527-536. Del Prete, R., Picca, V., Mosca, A., D'Alagni, M. and Miragliotta, G. (1998) Int. J. Tuberc. Lung Dis., 2, 160-163. Chanteau, S., Glaziou, P., Plichart, C., Luquiaud, P., Plichart, R., Faucher, J. F. and Cartel, J. L. (1993) Int. J. Lepr. Other Mycobact. Dis., 61 , 533-541. Cho, S. N., Cellona, R. V., Villahermosa, L. G., Fajardo, T. T., Balagon, M. V. F., Abalos, R. M., Tan, E. V., Walsh, G. P., Kim, J. D. and Brennan, P. J. (2001) Clin. Diagn. Lab. Immunol., 8, 138142. Hamasur, B., Bruchfeld, J., Haile, M., Pawlowski, A., Bjorvan, B., Kallenius, G. and Svenson, S. B. (2001) J. Microbiol. Methods, 45 , 41-52. Okuda, Y., Maekura, R., Hirotani, A., Kitada, S., Yoshimura, K., Hiraga, T., Yamamoto, Y., Itou, M., Ogura, T. and Ogihara, T. (2004) J. Clin. Microbiol., 42 , 1136-1141. Nyame, A. K., Kawar, Z. S. and Cummings, R. D. (2004) Arch. Biochem. Biophys., 426 , 182-200. Nyame, A. K., Leppanen, A. M., Bogitsh, B. J. and Cummings, R. D. (2000) Exp. Parasitol., 96 , 202-212. van Remoortere, A., Vermeer, H. J., van Roon, A. M., Langermans, J. A., Thomas, A. W., Wilson, R. A., van die, I., van den Eijnden, D. H., Agoston, K., Kerekgyarto, J., Vliegenthart, J. F., Kamerling, J. P., van dam, G. J., Hokke, C. H. and Deelder, A. M. (2003) Exp. Parasitol., 105 , 219-225. Alvesbrito, C. F., Simpson, A. J. G., Bahiaoliveira, L. M. G., Rabello, A. L. T., Rocha, R. S., Lambertucci, J. R., Gazzinelli, G., Katz, N. and Correaoliveira, R. (1992) Trans. Roy. Soc. Trop. Med. Hyg., 86 , 53-56. de Vijver, K. K. V., Hokke, C. H., van Remoortere, A., Jacobs, W., Deelder, A. M. and Van Marck, E. A. (2004) Int. J. Parasitol., 34 , 951-961. Shaker, Z. A., Kaddah, M. A., Hanallah, S. B. and El-Khodary, M. I. (1998) Int. J. Parasitol., 28 , 1893-1901. Bandyopadhyay, S., Chatterjee, M., Sundar, S. and Mandal, C. (2003) Glycoconjug. J., 20 , 531-536. Chatterjee, M., Sharma, V., Mandal, C., Sundar, S. and Sen, S. (1998) Glycoconjug. J., 15 , 1141-1147. Bandyopadhyay, S., Chatterjee, M., Pal, S., Waller, R. F., Sundar, S., McConville, M. J. and Mandal, C. (2004) Diagn. Microbiol. Infect. Dis., 50 , 15-24. Sarkari, B., Chance, M. and Hommel, M. (2002) Acta Trop., 82 , 339348. Bruschi, F., Moretti, A., Wassom, D. and Fioretti, D. P. (2001) Parasite-J. Soc. Fr. Parasitol., 8, S141-S143. Restrepo, B. I., Obregon-Henao, A., Mesa, M., Gil, D. L., Ortiz, B. L., Mejia, J. S., Villota, G. E., Sanzon, F. and Teale, J. M. (2000) Int. J. Parasit., 30 , 689-696. Prabhakaran, V., Rajshekhar, V., Murrell, K. D. and Oommen, A. (2004) Trans. Roy. Soc. Trop. Med. Hyg., 98 , 478-484. Sato, C. and Furuya, K. (1994) Jpn. J. Med. Sci. Biol., 47 , 65-71. Sato, C., Kawase, S. and Yano, S. (1999) Jpn. J. Infect. Dis., 52 , 156-159. [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] n tio u rib t is D r o F ot [66] [75] Galizia, G., Lieto, E., Ferraraccio, F., Castellano, P., De Vita, F., Orditura, M., Romano, C. and Pignatelli, C. (2003) Dig. Surg., 20 , 71-74. Ishibashi, R., Sakai, T., Yamashita, Y., Maekawa, T., Hideshima, T. and Shirakusa, T. (1999) Int. Surg., 84 , 151-154. Harada, T., Kubota, T. and Aso, T. (2002) Fertil. Steril., 78 , 733-739. Hirakata, Y., Kobayashi, J., Sugama, Y. and Kitamura, S. (1995) Eur. Respir. J., 8, 689-696. Holtzman, R. N. N., Heymann, A. D., Bordone, F., Marinoni, G., Barillari, P. and Wahl, S. J. (2001) Arch. Pathol. Lab. Med., 125 , 944-947. Angel, C. A., Pratt, C. B., Rao, B. N., Schell, M. J., Parham, D. M., Lobe, T. E. and Fleming, I. D. (1992) Cancer, 69 , 1487-1491. Ichihara, T., Sakamoto, J., Nakao, A., Furukawa, K., Watanabe, T., Suzuki, N., Horisawa, M., Nagura, H., Lloyd, K. O. and Takagi, H. (1993) Cancer, 71 , 71-81. Nishihara, S., Narimatsu, H., Iwasaki, H., Yazawa, S., Akamatsu, S. ando, T., Seno, T. and Narimatsu, I. (1994) J. Biol. Chem., 269 , 29271-29278. Mare, L. and Trinchera, M. (2004) Eur. J. Biochem., 271 , 186-194. Yin, B. W., Dnistrian, A. and Lloyd, K. O. (2002) Int. J. Cancer, 98 , 737-740. Yin, B. W. and Lloyd, K. O. (2001) J. Biol. Chem., 276 , 2737127375. Maggino, T. and Gadducci, A. (2000) Eur. J. Gynaecol. Oncol., 21 , 64-69. Bast, R. C., Feeney, M., Lazarus, H., Nadler, L. M., Colvin, R. B. and Knapp, R. C. (1981) J. Clin. Invest., 68 , 1331-1337. Buamah, P. (2000) J. Surg. Oncol., 75 , 264-265. D'Aloia, A., Faggiano, P., Aurigemma, G., Bontempi, L., Ruggeri, G., Metra, M., Nodari, S. and Dei Cas, L. (2003) J. Am. Coll. Cardiol., 41 , 1805-1811. Kui Wong, N., Easton, R. L., Panico, M., Sutton-Smith, M., Morrison, J. C., Lattanzio, F. A., Morris, H. R., Clark, G. F., Dell, A. and Patankar, M. S. (2003) J. Biol. Chem., 278 , 28619-28634. Sung, C. C., Pearl, D. K., Coons, S. W., Scheithauer, B. W., Johnson, P. C. and Yates, A. J. (1994) Cancer, 74 , 3010-3022. Ragupathi, G. (1996) Cancer Immunol. Immunother., 43 , 152-157. Kobayashi, Y., Tsukazaki, K., Kubushiro, K., Sakayori, M. and Nozawa, S. (1996) Clin. Cancer Res., 2, 749-754. Mondal, S. and Saha, S. (2000) J. Exp. Clin. Cancer Res., 19 , 317327. Taga, H., Hirai, H., Ishizuka, H. and Kaneda, H. (1988) Tumor Biol., 9, 110-115 Taketa, K. and Hirai, H. (1989) Electrophoresis, 10 , 562-567. Du, M. Q., Hutchinson, W. L., Johnson, P. J. and Williams, R. (1991) Cancer, 67 , 476-480. Vanstaden, L., Bukofzer, S., Kew, M. C. and Savage, N. (1992) J. Gastroenterol. Hepatol., 7, 260-265. Hirai, H. and Taketa, K. (1992) J. Chromatogr., 604 , 91-94. Taketa, K., Endo, Y., Sekiya, C., Tanikawa, K., Koji, T., Taga, H., Satomura, S., Matsuura, S., Kawai, T. and Hirai, H. (1993) Cancer Res., 53 , 5419-5423. Sato, Y., Nakata, K., Kato, Y., Shima, M., Ishii, N., Koji, T., Taketa, K., Endo, Y. and Nagataki, S. (1993) N. Engl. J. Med., 328 , 18021806. Shiraki, K., Takase, K., Tameda, Y., Hamada, M., Kosaka, Y. and Nakano, T. (1995) Hepatology, 22 , 802-807. Magne, D., Seta, N., Lebrun, D., Durand, G. and Durand, D. (1992) Clin. Chem., 38 , 1418-1424. Shimizu, K., Taniichi, T., Satomura, S., Matsuura, S., Taga, H. and Taketa, K. (1993) Clin. Chim. Acta, 214 , 3-12. Albanese, E. A., Bachl, B. L. and Mulcahy, G. M. (1995) Ann. Clin. Lab. Sci., 25 , 158-168. Li, D., Mallory, T. and Satomura, S. (2001) Clin. Chim. Acta, 313 , 15-19. Okuda, K., Tanaka, M., Kanazawa, N., Nagashima, J., Satomura, S., Kinoshita, H., Eriguchi, N., Aoyagi, S. and Kojiro, M. (1999) Int. J. Oncol., 14 , 265-271. Yamagata, Y., Shimizu, K., Nakamura, K., Henmi, F., Satomura, S., Matsuura, S. and Tanaka, M. (2003) Clin. Chim. Acta, 327 , 59-67. Poon, T. C. W., Mok, T. S. K., Chan, A. T. C., Chan, C. M. L., Leong, V., Tsui, S. H. T., Leung, T. W. T., Wong, H. T. M., Ho, S. K. W. and Johnson, P. J. (2002) Clin. Chem., 48 , 1021-1027. Yamamoto, T., Amuro, Y., Matsuda, Y., Nakaoka, H., Shimomura, S., Hada, T. and Higashino, K. (1991) Am. J. Gastroenterol., 86 , 495-499. Watt, K. W., Lee, P. J., M'Timkulu, T., Chan, W. P. and Loor, R. (1986) Proc. Natl. Acad. Sci. USA, 83 , 3166-3170. Troyer, D. A., Mubiru, J., Leach, R. J. and Naylor, S. L. (2004) Dis. Markers, 20 , 117-128. Huber, P. R., Schmid, H. P., Mattarelli, G., Strittmatter, B., Vansteenbrugge, G. J. and Maurer, A. (1995) Prostate, 27 , 212-219. Sumi, S., Arai, K., Kitahara, S. and Yoshida, K. (1999) J. Chromatogr. B Biomed. Sci. Appl., 727 , 9-14. Peracaula, R., Tabares, G., Royle, L., Harvey, D. J., Dwek, R. A., Rudd, P. M. and de Llorens, R. (2003) G lycobiology, 13 , 457-470. Jankovic, M. M. and Kosanovic, M. M. (2005) Clin. Biochem., 38 , 58-65. Basu, P. S., Majhi, R. and Batabyal, S. K. (2003) Clin. Biochem., 36 , 373-376. Ohyama, C., Hosono, M., Nitta, K., Oh-eda, M., Yoshikawa, K., Habuchi, T., Arai, Y. and Fukuda, M. (2004) G lycobiology, 14 , 671679. Gold, P. and Freedman, S. O. (1965) J. Exp. Med., 122 , 467-481. Thomson, D. M., Krupey, J., Freedman, S. O. and Gold, P. (1969) Proc. Natl. Acad. Sci. USA, 64 , 161-167. Schulz et al. Clinical Laboratory Testing in Human Medicine Based [133] [134] [135] [136] [137] [138] [139] [140] [141] [142] [143] [144] [145] [146] [147] [148] [149] Riordan, J. R., Rommens, J. M., Kerem, B., Alon, N., Rozmahel, R., Grzelczak, Z., Zielenski, J., Lok, S., Plavsic, N., Chou, J. L. and al., e. (1989) S cience, 245 , 1066-1073. Verkman, A. S., Song, Y. and Thiagarajah, J. R. (2003) Am. J. Physiol. Cell Physiol ., 284 , C2-15 Wine, J. J. and Joo, N. S. (2004) Proc. Am. Thorac. Soc., 1, 47-53. Lyon, E. and Miller, C. (2003) Arch. Pathol. Lab. Med., 127 , 11331139. Davril, M., Degroote, S., Humbert, P., Galabert, C., Dumur, V., Lafitte, J. J., Lamblin, G. and Rousse, P. (1999) G lycobiology, 9, 311-321. Boat, T. F., Cheng, P. W., Iyer, R. N., Carlson, D. M. and Polony, I. (1976) Arch. Biochem. Biophys., 177 , 95-104. Chace, K. V., Flux, M. and Sachdev, G. P. (1985) B iochemistry, 24 , 7334-7341. Zhang, Y., Doranz, B., Yankaskas, J. R. and Engelhardt, J. F. (1995) J. Clin. Invest., 96 , 2997-3004. Mohapatra, N. K., Cheng, P. W., Parker, J. C., Paradiso, A. M., Yankaskas, J. R., Boucher, R. C. and Boat, T. F. (1995) Pediatr. Res., 38 , 42-48. Mendicino, J. and Sangadala, S. (1999) Mol. Cell. Biochem., 201 , 141-149. Cheng, P. W., Boat, T. F., Cranfill, K., Yankaskas, J. R. and Boucher, R. C. (1989) J. Clin. Invest., 84 , 68-72. Holmen, J. M., Karlsson, N. G., Abdullah, L. H., Randell, S. H., Sheehan, J. K., Hansson, G. C. and Davis, C. W. (2004) Am. J. Physiol. Cell. Mol. Physiol., 287 , L824-834. Schulz, B. L., Sloane, A. J., Robinson, L. J., Sebastian, L. T., Glanville, A. R., Song, Y., Verkman, A. S., Harry, J. L., Packer, N. H. and Karlsson, N. G. (2005) Biochem. J., 387 , 911-919. Tiddens, H. A. (2002) Pediatr. Pulmonol., 34 , 228-231. Helbich, T. H., Heinz-Peer, G., Fleischmann, D., Wojnarowski, C., Wunderbaldinger, P., Huber, S., Eichler, I. and Herold, C. J. (1999) AJR Am. J. Roentgenol., 173 , 81-88. Zalewska, A., Zwierz, K., Zolkowski, K. and Gindzienski, A. (2000) Acta Biochim. Pol., 47 , 1067-1079. Prakobphol, A., Thomsson, K. A., Hansson, G. C., Rosen, S. D., Singer, M. S., Phillips, N. J., Medzihradszky, K. F., Burlingame, A. L., Leffler, H. and Fisher, S. J. (1998) B iochemistry, 37 , 4916-4927. Murray, P. A., Prakobphol, A., Lee, T., Hoover, C. I. and Fisher, S. J. (1992) Infect. Immun., 60 , 31-38. Thomsson, K. A., Prakobphol, A., Leffler, H., Reddy, M. S., Levine, M. J., Fisher, S. J. and Hansson, G. C. (2002) G lycobiology, 12 , 114. Schulz, B. L., Packer, N. H. and Karlsson, N. G. (2002) Anal. Chem., 74 , 6088-6097. Klein, A., Carnoy, C., Wieruszeski, J. M., Strecker, G., Strang, A. M., van Halbeek, H., Roussel, P. and Lamblin, G. (1992) Biochemistry, 31 , 6152-6165. Thomsson, K. A., Schulz, B. L., Packer, N. H. and Karlsson, N. G. (2005) G lycobiology, 15 , 791-804. Seemann, R., Zimmer, S., Bizhang, M. and Kage, A. (2001) Caries Res., 35 , 156-161. Sewell, A. C. (1980) Eur. J. Pediatr. 134 , 183-194. Paschke, E. and Stockler, S. (1992) Wien. Klin. Wochen., 104 , 658-664. Meikle, P. J., Ranieri, E., Simonsen, H., Rozaklis, T., Ramsay, S. L., Whitfield, P. D., Fuller, M., Christensen, E., Skovby, F. and Hopwood, J. J. (2004) P ediatrics, 114 , 909-916. Giudici, T. A., Sunico, H. and Blaskovics, M. (1996) J. Inherit. Metab. Dis., 19 , 263-266. Ramsay, S. L., Meikle, P. J. and Hopwood, J. J. (2003) Mol. Genet. Metab., 78 , 193-204. Jaeken, J., Vanderschueren-Lodeweyckx, M., Casaer, P., Snoeck, L. and Corbeel, L. (1980) Pediatr. Res., 14 , 179. Jaeken, J., van Eijk, H. G., van der Heul, C., Corbeel, L., Eeckels, R. and Eggermont, E. (1984) Clin. Chim. Acta, 144 , 245-247. Stibler, H. and Kristiansson, B. (1991) Acta Paediatr. Scand., 3238. Wada, Y., Gu, J. G., Okamoto, N. and Inui, K. (1994) Biol. Mass Spectrom., 23 , 108-109. Yamashita, K., Ohkura, T., Ideo, H., Ohno, K. and Kanai, M. (1993) J. Biochem. (Tokyo), 114 , 766-769. Lacey, J. M., Bergen, H. R., Magera, M. J., Naylor, S. and O'Brien, J. F. (2001) Clin. Chem., 47 , 513-518. Muntoni, F. (2004) A cta Myol., 23 , 79-84. Zhang, W. L., Vajsar, J., Cao, P. J., Breningstall, G., Diesen, C., Dobyns, W., Herrmann, R., Lehesjoki, A. E., Steinbrecher, A., Talim, B., Toda, T., Topaloglu, H., Voit, T. and Schachter, H. Y. (2003) Clin. Biochem., 36 , 339-344. Quentin, E., Gladen, A., Roden, L. and Kresse, H. (1990) Proc. Natl. Acad. Sci. USA, 87 , 1342-1346. Lind, T., Tufaro, F., McCormick, C., Lindahl, U. and Lidholt, K. (1998) J. Biol. Chem., 273 , 26265-26268. McCormick, C., Duncan, G., Goutsos, K. T. and Tufaro, F. (2000) Proc. Natl. Acad. Sci. USA, 97 , 668-673. Bessler, M., Schaefer, A. and Keller, P. (2001) in Transf. Med. Rev., 15 , 255-267. Krauss, J. S. (2003) Ann. Clin. Lab. Sci., 33 , 401-406. Shin, D. J., Lee, J. J., Choy, H. E. and Hong, Y. J. (2004) B iochem. Biophys. Res. Commun., 324 , 753-760. Stibler, H., Allgulander, C., Borg, S. and Kjellin, K. G. (1978) Acta Med. Scand., 204 , 49-56. Stibler, H. (1991) Clin. Chem., 37 , 2029-2037. Arndt, T. (2001) Clin. Chem., 47 , 13-27. Bergen, H. R., Lacey, J. M., O'Brien, J. F. and Naylor, S. (2001) Anal. Biochem., 296 , 122-129. Alte, D., Luedemann, J., Rose, H. J. and John, U. (2004) Alcoholism (NY), 28 , 931-940. Current Molecular Medicine, 2007, Vol. 7, No. 4 [180] [181] [182] [183] [184] [185] [186] [187] [188] [189] [190] [191] [192] [193] [194] [195] [196] [197] [198] [199] Anttila, P., Jarvi, K., Latvala, J. and Niemela, O. (2004) A lcohol Alcohol, 39 , 59-63. Koch, H., Meerkerk, G. J., Zaat, J. O. M., Ham, M. F., Scholten, R. and Assendelft, W. J. J. (2004) A lcohol Alcohol, 39 , 75-85. Arndt, T., Kropf, J., Brandt, R., Gressner, A. M., Hackler, R., Herold, M., Van Pelt, J., Martensson, O., Salzmann, K. and Velmans, M. H. (1998) A lcohol Alcohol, 33 , 639-645. Wuyts, B., Delanghe, J. R., Kasvosve, I., Wauters, A., Neels, H. and Janssens, J. (2001) Clin. Chem., 47 , 247-255. Lanz, C., Kuhn, M., Deiss, V. and Thormann, W. (2004) Electrophoresis, 25 , 2309-2318. Martello, S., Trettene, M., Cittadini, F., Bortolotti, F. B., De Giorgio, F., Chiarotti, M. and Tagliaro, F. (2004) Forensic Sci. Int., 141 , 153157. Iourin, O., Mattu, T. S., Mian, N., Keir, G., Winchester, B., Dwek, R. A. and Rudd, P. M. (1996) Glycoconjug. J., 13 , 1031-1042. Fermo, I., Germagnoli, L., Soldarini, A., Dorigatti, F. and Paroni, R. (2004) Electrophoresis, 25 , 469-475. Renner, F. and Kanitz, R. D. (1997) Clin. Chem., 43 , 485-490. Helander, A., Eriksson, G., Stibler, H. and Jeppsson, J. O. (2001) Clin. Chem., 47 , 1225-1233. Winearls, C. G., Oliver, D. O., Pippard, M. J., Reid, C., Downing, M. R. and Cotes, P. M. (1986) Lancet, 2, 1175-1178. Pascual, J. A., Belalcazar, V., de Bolos, C., Gutierrez, R., Llop, E. and Segura, J. (2004) Ther. Drug Monit., 26 , 175-179. Lasne, F., Martin, L., Crepin, N. and de Ceaurriz, J. (2002) Anal. Biochem., 311 , 119-126. Takeuchi, M., Takasaki, S., Miyazaki, H., Kato, T., Hoshi, S., Kochibe, N. and Kobata, A. (1988) J. Biol. Chem., 263 , 3657-3663. Takeuchi, M. and Kobata, A. (1991) G lycobiology, 1, 337-346. Koury, M. J. (2003) Trends Biotechnol., 21 , 462-464. Nimtz, M., Martin, W., Wray, V., Kloppel, K. D., Augustin, J. and Conradt, H. S. (1993) Eur. J. Biochem., 213 , 39-56. Souillard, A., Audran, M., Bressolle, F., Gareau, R., Duvallet, A. and Chanal, J. L. (1996) Br. J. Clin. Pharmacol., 42 , 355-364. Breidbach, A., Catlin, D. H., Green, G. A., Tregub, I., Truong, H. and Gorzek, J. (2003) Clin. Chem., 49 , 901-907. Nowicki, M., Kokot, F., Kokot, M., Bar, A. and Dulawa, J. (1994) Int. Urol. Nephrol., 26 , 691-699. Neumayr, G., Pfister, R., Hoertnagl, H., Mitterbauer, G., Getzner, W., Ulmer, H., Gaenzer, H. and Joannidis, M. (2003) Int. J. Sports Med., 24 , 131-137. Gore, C. J., Parisotto, R., Ashenden, M. J., Stray-Gundersen, J., Sharpe, K., Hopkins, W., Emslie, K. R., Howe, C., Trout, G. J., Kazlauskas, R. and Hahn, A. G. (2003) Haematologica, 88 , 333344. Sendid, B., Colombel, J. F., Jacquinot, P. M., Faille, C., Fruit, J., Cortot, A., Lucidarme, D., Camus, D. and Poulain, D. (1996) Clin. Diagn. Lab. Immunol., 3, 219-226. Annese, V., Piepoli, A., Perri, F., Lombardi, G., Latiano, A., Napolitano, G., Corritore, G., Vandewalle, P., Poulain, D., Colombel, J. F. and Andriulli, A. (2004) Aliment Pharmacol. Ther., 20 , 1143-1152. Kim, J. E., Kim, K. S. and Seo, J. K. (2003) Korean J. Gastroenterol., 42 , 297-302. Konrad, A., Rutten, C., Flogerzi, B., Styner, M., Goke, B. and Seibold, F. (2004) Inflamm. Bowel. Dis., 10 , 97-105. Galassi, G., Susuki, K., Quaglino, D. and Yuki, N. (2004) Eur. J. Neurol., 11 , 790-791. Ang, C. W., Laman, J. D., Willison, H. J., Wagner, E. R., Endtz, H. P., De Klerk, M. A., Tio-Gillen, A. P., Van den Braak, N., Jacobs, B. C. and Van Doorn, P. A. (2002) Infect. Immun., 70 , 1202-1208. Hirano, M., Kusunoki, S., Asai, H., Tonomura, Y., Morita, D. and Ueno, S. (2003) Neurology, 60 , 1719-1720. Nishimoto, Y., Odaka, M., Hirata, K. and Yuki, N. (2004) J. Neuroimmunol., 148 , 200-205. Nagashima, T., Koga, M., Odaka, M., Hirata, K. and Yuki, N. (2004) J. Neurol. Sci., 219 , 139-145. Pincus, T. (1995) Br. J. Rheumatol., 34 (Suppl 2), 59-73. Young, A., Dixey, J., Kulinskaya, E., Cox, N., Davies, P., Devlin, J., Emery, P., Gough, A., James, D., Prouse, P., Williams, P. and Winfield, J. (2002) Ann. Rheum. Dis., 61 , 335-340. Munro, R., Hampson, R., McEntegart, A., Thomson, E. A., Madhok, R. and Capell, H. (1998) Ann. Rheum. Dis., 57 , 88-93. Zeidler, H. K., Kvien, T. K., Hannonen, P., Wollheim, F. A., Forre, O., Geidel, H., Hafstrom, I., Kaltwasser, J. P., Leirisalo-Repo, M., Manger, B., Laasonen, L., Markert, E. R., Prestele, H. and Kurki, P. (1998) Br. J. Rheumatol., 37 , 874-882. Tsakonas, E., Fitzgerald, A. A., Fitzcharles, M. A., Cividino, A., Thorne, J. C., M'Seffar, A., Joseph, L., Bombardier, C. and Esdaile, J. M. (2000) J. Rheumatol., 27 , 623-629. Smolen, J. S., Aletaha, D. and Machold, K. P. (2005) Best Pract. Res. Clin. Rheumatol., 19 , 163-177. Keane, J., Gershon, S., Wise, R. P., Mirabile-Levens, E., Kasznica, J., Schwieterman, W. D., Siegel, J. N. and Braun, M. M. (2001) N. Engl. J. Med., 345 , 1098-1104. Gardam, M. and Iverson, K. (2003) J. Rheumatol., 30 , 1397-1399. Parekh, R. B., Dwek, R. A., Sutton, B. J., Fernandes, D. L., Leung, A., Stanworth, D., Rademacher, T. W., Mizuochi, T., Taniguchi, T., Matsuta, K., Takeuchi, F.; Nagano, Y.; Miyamotom T.; and Kobata, A. (1985) Nature, 316 , 452-457. Hansler, M., Kotz, K. and Hantzschel, H. (1995) Electrophoresis, 16 , 811-812. Martin, K., Talukder, R., Hay, F. C. and Axford, J. S. (2001) J. Rheumatol., 28 , 1531-1536. Kanoh, Y., Mashiko, T., Danbara, M., Takayama, Y., Ohtani, S., Imasaki, T., Abe, T. and Akahoshi, T. (2004) Oncology, 66 , 365370. [151] [152] [153] [154] [155] [156] [157] [158] N [159] [160] [161] [162] [163] [164] [165] [166] [167] [168] [169] [170] [171] [172] [173] [174] [175] [176] [177] [178] [179] [200] [201] [202] [203] [204] [205] [206] [207] [208] [209] [210] [211] [212] [213] [214] [215] [216] [217] [218] [219] [220] [221] [222] n tio u rib t is D r o F ot [150] 415 416 Current Molecular Medicine, 2007, Vol. 7, No. 4 [223] [224] [225] [226] [227] [228] [229] [230] [231] [232] [233] [234] [235] [236] [237] [238] [239] Alavi, A., Arden, N., Spector, T. D. and Axford, J. S. (2000) J. Rheumatol., 27 , 1379-1385. Alavi, A., Axford, J. S. and Pool, A. J. (2004) J. Rheumatol., 31 , 1513-1520. Saez-Valero, J., Barquero, M. S., Marcos, A., McLean, C. A. and Small, D. H. (2000) J. Neurol. Neurosurg. Psychiatry, 69 , 664-667. Fodero, L. R., Saez-Valero, J., McLean, C. A., Martins, R. N., Beyreuther, K., Masters, C. L., Robertson, T. A. and Small, D. H. (2002) J. Neurochem., 81 , 441-448. Sherman, M. and Takayama, Y. (2004) Gastroenterol. Clin. North Am., 33 , 671-691. Mann, A. C., Record, C. O., Self, C. H. and Turner, G. A. (1994) Clin. Chim. Acta, 227 , 69-78. Callewaert, N., Van Vlierberghe, H., Van Hecke, A., Laroy, W., Delanghe, J. and Contreras, R. (2004) Nat. Med., 10 , 429-434. Hada, T., Kondo, M., Yasukawa, K., Amuro, Y. and Higashino, K. (1999) Clin. Chim. Acta, 281 , 37-46. Ryden, I., Pahlsson, P. and Lindgren, S. (2002) Clin. Chem., 48 , 2195-2201. Seta, N., Gayno, S., JezequelCuer, M., Toueg, M. L., Erlinger, S. and Durand, G. (1997) J. Hepatol., 26 , 265-271. Kudo, M., Todo, A., Ikekubo, K. and Hino, M. (1992) Am. J. Gastroenterol., 87 , 865-870. Dell, A. and Morris, H. R. (2001) S cience, 291 , 2351-2356. Zaia, J. (2004) Mass Spectrom. Rev., 23 , 161-227. Rashed, M. S. (2001) J. Chromatogr. B Biomed. Sci. Appli., 758 , 27-48. Rinaldo, P., Tortorelli, S. and Matern, D. (2004) Curr. Opin. Pediatr., 16 , 427-433. Julka, S. and Regier, F. (2004) J. Proteome Res., 3, 350-363. Joshi, H., Harrison, M. J., Schulz, B. L., Cooper, C. A., Packer, N. H. and Karlsson, N. G. (2004) Proteomics, 4, 1650-1664. Schulz et al. [240] [241] [242] [243] [244] [245] [246] [247] [248] [249] [250] [251] [252] [253] [254] Lohmann, K. K. and von der Leith, C. W. (2004) Nucleic Acids Res., 32 , W261-266. Cottrell, J. S. (1994) Pept. Res., 7, 115-124. Cooper, C. A., Joshi, H. J., Harrison, M. J., Wilkins, M. R. and Packer, N. H. (2003) N ucleic Acids Res., 31 , 511-513. Loss, A., Bunsmann, P., Bohne, A., Loss, A., Schwarzer, E., Lang, E. and von der Leith, C. W. (2002) Nucleic Acids Res., 30 , 405408. Righetti, P. G., Castagna, A. and Herbert, B. (2001) Anal. Chem., 73 , 320A-326A. Lee, W. C. and Lee, K. H. (2004) A nal. Biochem ., 324 , 1-10. Hortin, G. L. and Trimpe, B. L. (1990) Anal. Biochem., 188 , 271-277. Kaji, H., Saito, H., Yamauchi, Y., Shinkawa, T., Taoka, M., Hirabayashi, J., Kasai, K., Takahashi, N. and Isobe, T. (2003) Nat. Biotechnol., 21 , 667-672. Hagglund, P., Bunkenborg, J., Elortza, F., Jensen, O. N. and Roepstorff, P. (2004) J. Proteome Res., 3, 556-566. Zhang, H., Li, X. J., Martin, D. B. and Aebersold, R. (2003) Nat. Biotechnol., 21 , 660-666. Liu, M. S. and Amirkhanian, V. D. (2003) Electrophoresis, 24 , 9395. Callewaert, N., Contreras, R., Mitnik-Gankin, L., Carey, L., Matsudaira, P. and Ehrlich, D. (2004) Electrophoresis, 25 , 31283131. Zamfir, A. and Peter-Katalinic, J. (2004) Electrophoresis, 25 , 19491963. Lee, J., Park, S. H. and Stanley, P. (2002) Glycoconj. J., 19 , 211219. Ravn, P., Danielczyk, A., Jensen, K. B., Kristensen, P., Christensen, P. A., Larsen, M., Karsten, U. and Goletz, S. (2004) J. Mol. Biol., 343 , 985-996. t is D r o F ot N n tio u rib Received: November 15, 2006 Revised: December 21, 2006 Accepted: February 20, 2007 ...
View Full Document

{[ snackBarMessage ]}

Ask a homework question - tutors are online