Doron Kletter

Global comparisons of lectin-glycan interactions using a database of analyzed glycan array data

Lectin–glycan interactions have critical functions in multiple normal and pathological processes, but the binding partners and functions for many glycans and lectins are not known. An important step in better understanding glycan–lectin biology is enabling systematic quantification and analysis of the interactions. Glycan arrays can provide the experimental information for such analyses, and the thousands of glycan array datasets available through the Consortium for Functional Glycomics provide the opportunity to extend the analyses to a broad scale. We developed software, based on our previously described Motif Segregation algorithm, for the automated analysis of glycan array data, and we analyzed the entire storehouse of 2883 datasets from the Consortium for Functional Glycomics. We mined the resulting database to make comparisons of specificities across multiple lectins and comparisons between glycans in their lectin receptors. Of the lectins in the database, viral lectins were the most different from other organism types, with specificities nearly always restricted to sialic acids, and mammalian lectins had the most diverse range of specificities. Certain mammalian lectins were unique in their specificities for sulfated glycans. Simple modifications to a lactosamine core structure radically altered the types of lectins that were highly specific for the glycan. Unmodified lactosamine was specifically recognized by plant, fungal, viral, and mammalian lectins; sialylation shifted the binding mainly to viral lectins; and sulfation resulted in mainly mammalian lectins with the highest specificities. We anticipate that this analysis program and database will be valuable in fundamental glycobiology studies, detailed analyses of lectin specificities, and practical applications in translational research.

The Detection and Discovery of Glycan Motifs in Biological Samples Using Lectins and Antibodies: New Methods and Opportunities

Recent research has uncovered unexpected ways that glycans contribute to biology, as well as new strategies for combatting disease using approaches involving glycans. To make full use of glycans for clinical applications, we need more detailed information on the location, nature, and dynamics of glycan expression in vivo. Such studies require the use of specimens acquired directly from patients. Effective studies of clinical specimens require low-volume assays, high precision measurements, and the ability to process many samples. Assays using affinity reagents-lectins and glycan-binding antibodies-can meet these requirements, but further developments are needed to make the methods routine and effective. Recent advances in the use of glycan-binding proteins involve improved determination of specificity using glycan arrays; the availability of databases for mining and analyzing glycan array data; lectin engineering methods; and the ability to quantitatively interpret lectin measurements. Here, we describe many of the challenges and opportunities involved in the application of these new approaches to the study of biological samples. The new tools hold promise for developing methods to improve the outcomes of patients afflicted with diseases characterized by aberrant glycan expression.

Determining Lectin Specificity from Glycan Array Data Using Motif Segregation and GlycoSearch Software

The glycan array is a powerful tool for investigating the specificities of glycan‐binding proteins. By incubating a glycan‐binding protein on a glycan array, the relative binding to hundreds of different oligosaccharides can be quantified in parallel. Based on these data, much information can be obtained about the preference of a glycan‐binding protein for specific subcomponents of oligosaccharides, or motifs. In many cases, the analysis and interpretation of glycan array data can be time consuming and imprecise if done manually. Recently, GlycoSearch software was developed to facilitate the analysis and interpretation of glycan array data based on two previously developed methods, Motif Segregation and Outlier Motif Analysis. Here, the principles behind this method and the use of this new tool for mining glycan array data are described. The automated, objective, and precise analysis of glycan array data should enhance the value of these data for a broad range of research applications. Curr. Protoc. Chem. Biol. 5:157‐169

Prediction of glycan motifs using quantitative analysis of multi-lectin binding: Motifs on MUC1 produced by cultured pancreatic cancer cells.

Lectins are valuable tools for detecting specific glycans in biological samples, but the interpretation of the measurements can be ambiguous due to the complexities of lectin specificities. Here, we present an approach to improve the accuracy of interpretation by converting lectin measurements into quantitative predictions of the presence of various glycan motifs.

Glycans Related to the CA19-9 Antigen Are Increased in Distinct Subsets of Pancreatic Cancers and Improve Diagnostic Accuracy Over CA19-9

Background & Aims The cancer antigen 19-9 (CA19-9) is the current best biomarker for pancreatic cancer, but it is not increased in approximately 25% of pancreatic cancer patients at a cut-off value that provides a 25% false-positive rate. We hypothesized that antigens related to the CA19-9 antigen, which is a glycan called sialyl-Lewis A (sLeA), are increased in distinct subsets of pancreatic cancers. Methods We profiled the levels of multiple glycans and mucin glycoforms in plasma from 200 subjects with either pancreatic cancer or benign pancreatic disease, and we validated selected findings in additional cohorts of 116 and 100 subjects, the latter run with the investigators blinded to diagnoses and including cancers that exclusively were early stage. Results We found significant increases in 2 glycans: an isomer of sLeA called sialyl-Lewis X, present both in sulfated and nonsulfated forms, and the sialylated form of a marker for pluripotent stem cells, type 1 N-acetyl-lactosamine. The glycans performed as well as sLeA as individual markers and were increased in distinct groups of patients, resulting in a 3-marker panel that significantly improved upon any individual biomarker. The panel showed 85% sensitivity and 90% specificity in the combined discovery and validation cohorts, relative to 54% sensitivity and 86% specificity for sLeA; and it showed 80% sensitivity and 84% specificity in the independent test cohort, as opposed to 66% sensitivity and 72% specificity for sLeA. Conclusions Glycans related to sLeA are increased in distinct subsets of pancreatic cancers and yield improved diagnostic accuracy compared with CA19-9.

Glycan Motif Profiling Reveals Plasma Sialyl-Lewis X Elevations in Pancreatic Cancers That Are Negative for Sialyl-Lewis A

The sialyl-Lewis A (sLeA) glycan forms the basis of the CA19–9 assay and is the current best biomarker for pancreatic cancer, but because it is not elevated in ∼25% of pancreatic cancers, it is not useful for early diagnosis. We hypothesized that sLeA-low tumors secrete glycans that are related to sLeA but not detectable by CA19–9 antibodies. We used a method called motif profiling to predict that a structural isomer of sLeA called sialyl-Lewis X (sLeX) is elevated in the plasma of some sLeA-low cancers. We corroborated this prediction in a set of 48 plasma samples and in a blinded set of 200 samples. An antibody sandwich assay formed by the capture and detection of sLeX was elevated in 13 of 69 cancers that were not elevated in sLeA, and a novel hybrid assay of sLeA capture and sLeX detected 24 of 69 sLeA-low cancers. A two-marker panel based on combined sLeA and sLeX detection differentiated 109 pancreatic cancers from 91 benign pancreatic diseases with 79% accuracy (74% sensitivity and 78% specificity), significantly better than sLeA alone, which yielded 68% accuracy (65% sensitivity and 71% specificity). Furthermore, sLeX staining was evident in tumors that do not elevate plasma sLeA, including those with poorly differentiated ductal adenocarcinoma. Thus, glycan-based biomarkers could characterize distinct subgroups of patients. In addition, the combined use of sLeA and sLeX, or related glycans, could lead to a biomarker panel that is useful in the clinical diagnosis of pancreatic cancer. Précis: This paper shows that a structural isomer of the current best biomarker for pancreatic cancer, CA19–9, is elevated in the plasma of patients who are low in CA19–9, potentially enabling more comprehensive detection and classification of pancreatic cancers.

Parsimonious Charge Deconvolution for Native Mass Spectrometry

Charge deconvolution infers the mass from mass over charge (m/z) measurements in electrospray ionization mass spectra. When applied over a wide input m/z or broad target mass range, charge-deconvolution algorithms can produce artifacts, such as false masses at one-half or one-third of the correct mass. Indeed, a maximum entropy term in the objective function of MaxEnt, the most commonly used charge deconvolution algorithm, favors a deconvolved spectrum with many peaks over one with fewer peaks. Here we describe a new “parsimonious” charge deconvolution algorithm that produces fewer artifacts. The algorithm is especially well-suited to high-resolution native mass spectrometry of intact glycoproteins and protein complexes. Deconvolution of native mass spectra poses special challenges due to salt and small molecule adducts, multimers, wide mass ranges, and fewer and lower charge states. We demonstrate the performance of the new deconvolution algorithm on a range of samples. On the heavily glycosylated plasma properdin glycoprotein, the new algorithm could deconvolve monomer and dimer simultaneously and, when focused on the m/z range of the monomer, gave accurate and interpretable masses for glycoforms that had previously been analyzed manually using m/z peaks rather than deconvolved masses. On therapeutic antibodies, the new algorithm facilitated the analysis of extensions, truncations, and Fab glycosylation. The algorithm facilitates the use of native mass spectrometry for the qualitative and quantitative analysis of protein and protein assemblies.

Exploring the Specificities of Glycan-Binding Proteins Using Glycan Array Data and the GlycoSearch Software

The glycan array is a powerful tool for investigating the specificities of glycan-binding proteins. By incubating a glycan-binding protein on a glycan array, the relative binding to hundreds of different oligosaccharides can be quantified in parallel. Based on these data, much information can be obtained about the preference of a glycan-binding protein for specific subcomponents of oligosaccharides or motifs. In many cases, the analysis and interpretation of glycan array data can be time consuming and imprecise if done manually. Recently we developed software, called GlycoSearch, to facilitate the analysis and interpretation of glycan array data based on the previously developed methods called Motif Segregation and Outlier Motif Analysis. Here we describe the principles behind the software, the use of the software, and an example application. The automated, objective, and precise analysis of glycan array data should enhance the value of the data for a broad range of research applications.

Abstract B31: Development of fucose based pancreatic cancer biomarkers using modified lectins

Pancreatic cancer has poor prognosis because of ineffective and delayed diagnosis as well as treatment. Accurately identifying molecular changes could help with early detection and evaluation of treatment efficacy. Glycoprotein fucosylation is associated with various cancers. Interestingly, information at structural level such as diversity in fucose linkages can provide extensive details of disease state. But, lack of diagnostic reagents has impeded the research in identifying and interpreting these changes. This study is aimed to develop fucose based cancer biomarker reagents to identify important fucose linkages associated with cancer progression. We hypothesized that a panel of lectins, each specific for a particular type of fucose presentation, could identify differences between glycoproteins that are not discernable using any individual lectin or measurements of total fucose. Using a database of glycan array data, we identified lectins with high specificity for diverse linkages of fucose. We selected three for further testing: CGL2 for detecting fucose in a 2’ linkage, CCL2 for detecting fucose in a 3’ linkage, and RSL for detecting fucose in all linkages. To improve quality, production consistency, and test various modes of detection the lectins were recombinantly expressed with avitag (biotin) and poly-histidine fusion tags. The choice of terminus as well as the type of fusion tag could affect the molecular architecture and thereby lectin-glycan interaction. Therefore, two variants for each lectin (N terminus biotin and C terminus histidine and vice-versa) were expressed. The initial purification by size-exclusion chromatography showed similar results between the two tagging schemes for each protein. To test the accessibility of fusion tags and lectin multimerization with secondary detections (streptavidin or anti-poly-histindine), the two were pre-incubated prior to separation on an analytical scale size exclusion column resulting in protein elution that was entirely shifted to a higher molecular weight. Specificity for recombinant lectins to their respective linkages was performed using thermostability assay that resulted in a stability shift of lectins with respective glycan target but not with off-target glycan. To examine if the use of particular tag or secondary reagent was optimal for detection we performed competitive assay. Lectin binding to spotted glycoproteins in microarray assay was reduced upon pre-incubation with a competing sugar but not upon pre-incubation with a non-specific sugar. This suggests that recombinant lectins with fusion tags were functional. Lectins produced by recombinant expression showed specific glycan-binding using either a biotin or poly-histidine fusion tag at either terminus, but detection using the C-terminus tag was more reliable in each case. We probed the glycans on various glycoproteins with each lectin and predicted that the pattern of lectin binding would perform better than individual lectin measurements for identifying differences in fucose presentation. We tested this prediction using monoclonal antibodies against selected fucose-bearing glycans. Based on the known specificities of the antibodies, we could confirm differences in glycans predicted by the pattern of lectin binding but not by individual lectin measurements, thus confirming the ability to distinguish fucose presentations using a panel of lectins. This technique promises to help uncover details about the presentation of fucose on specific glycoproteins. The ability of these lectins to distinguish subtle changes in glycan presentation with enough precision at early stage and enable comparisons between tumor grades could prove valuable diagnostic reagents for pancreatic cancer. Citation Format: Sudhir Singh, Kuntal Pal, Elliot Ensink, Jessica Yadav, Doron Kletter, Marshall Bern, Anand Mehta, Karsten Melcher, Brian B. Haab. Development of fucose based pancreatic cancer biomarkers using modified lectins. [abstract]. In: Proceedings of the AACR Special Conference on Pancreatic Cancer: Innovations in Research and Treatment; May 18-21, 2014; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2015;75(13 Suppl):Abstract nr B31.

SlideDeckFinder: identifying related slide decks based on visual appearance and composition patterns

This paper introduces SlideDeckFinder, a tool integrated into a users email client enabling the search for similarities between slide decks. The similarity calculations are based on visual correspondence (both from text and images/graphics) as well as slide (re-)composition patterns. The individual slides of different slide decks are first compared by matching their respective visual features extracted from any content such as text and images. The resulting similarity scores between pairs of slides are then the input for calculating the similarity between whole slide decks. Hidden Markov models (HMMs) are used to represent the transformation (in terms of re-arrangements or insertions of new slides) from one slide deck to another, where the state emissions probabilities of the HMM correspond to slide similarity and the transition probabilities represent the likely slide sequence within slide decks. The Viterbi algorithm is finally used to calculate the most likely state sequence (i.e. recomposition pattern) between the slide decks and thus the similarity score. SlideDeckFinder has been evaluated both on its accuracy to compare visual appearance of slides with respect to human perception and its performance to retrieve related slide deck variants.