This work shows the potential of printed glycan array (PGA) technology for the analysis of circulating anti-carbohydrate antibodies in small animals.
The repertoire of circulating anti-carbohydrate antibodies of a given individual is often associated with its immunological status. Not only the individual immune condition determines the success in combating internal and external potential threat signals, but also the existence of a particular pattern of circulating anti-glycan antibodies (and their serological level variation) could be a significant marker of the onset and progression of certain pathological conditions. Here, we describe a Printed Glycan Array (PGA)-based methodology that offers the opportunity to measure hundreds of glycan targets with very high sensitivity; using a minimal amount of sample, which is a common restriction present when small animals (rats, mice, hamster, etc.) are used as models to address aspects of human diseases. As a representative example of this approach, we show the results obtained from the analysis of the repertoire of natural anti-glycan antibodies in BALB/c mice. We demonstrate that each BALB/c mouse involved in the study, despite being genetically identical and maintained under the same conditions, develops a particular pattern of natural anti-carbohydrate antibodies. This work claims to expand the use of PGA technology to investigate repertoire (specificities) and the levels of circulating anti-carbohydrates antibodies, both in health and during any pathological condition.
Antibodies play a central role in our defense against invading pathogens by directly neutralizing viruses1,2 and bacteria2,3, by activating the complement system4,5 and the enhancement of phagocytosis6. Additionally, they are essential elements in cancer targeting and elimination of malignant cells7, and homeostasis maintenance8,9.
Disorders of the immune system can result in autoimmune and inflammatory diseases10 and cancer11. All these pathological conditions ideally demand a prompt diagnosis for an efficient treatment. In the case of autoimmune disorders, the serological presence of autoantibodies in most of the cases is a predictor for diagnostic of autoimmunity10,12. These antibodies react with the cell surface and extracellular autoantigens and, they are often present for many years before the presentation of autoimmune disease10,12. Immune deficiencies and cancer are also diagnosed with blood tests that either measure the level of immune elements such as antibodies, or their functional activity11.
The identification of the repertoire of circulating antibodies and their serological levels are paramount to set a prognosis and evaluate the progression of all of the mentioned pathological conditions. We have previously demonstrated the potential of PGA technique for the analysis of circulating antibodies in different animal species13–16, minimizing the use of large volumes of serological samples, avoiding the problem associated with antibodies cross-reactivity17 and allowing high-throughput profiling of an extensive repertoire of antibodies15.
Glycan-based immunoassays are mainly conditioned, among other factors, by the origin and production of carbohydrates, which determine the affinity and binding of ligands15,18,19,20,21. Glycan-based immunoassays can be developed in suspension (microspheres)15,21,22 or in flat-activated surfaces15,21,22,23,24. The last include ELISA (the most conventional of these methods) and PGA. There is not much data comparing these methodologies in the same experimental setting15,25,26,27. We have previously compared the efficacy and selectivity of these immunoassays to profile anti-glycan antibodies in individual human plasma samples15. For some antibodies such as those targeting anti-A/B blood group, all the immunoassays could detect them with statistical significance and they positively correlated with each other15,18,21. Meanwhile, anti-P1 antibodies were primarily detected by PGA with the highest discriminative power, and there was no correlation in the determinations made by the different glycan-based immunoassays15,18,21. These differences between methods were mainly related to the antibody/antigen ratio and glycan orientation15. ELISA and suspension arrays are more susceptible to unspecific binding than PGA because there is an excess of antigen over antibodies in these methods15. Additionally, the orientation of glycans in the PGA is more restricted than in ELISA and suspension arrays15. ELISA is convenient when the study includes a limited panel of glycans. Along with suspension arrays, ELISA offers broader flexibility regarding assay reconfiguration. PGA is exceptionally convenient for discovery approaches15,18,21,28. Despite these clear advantages and disadvantages, the three mentioned immunoassays could be used to study different aspects of glycan-antibody interactions. The final goal of the study is the one will guide the selection of the more suitable methodology.
The present work aims to extend the use of PGA technology for the analysis of the repertoire of circulating anti-glycan antibodies in small animals. As a representative result, we present here a detailed protocol to assess the repertoire of natural anti-carbohydrate antibodies in adult BALB/c mice by PGA.
1. Glycochips Production
2. Glycan Array Technique
3. Analysis of Glycan Array
Here, we present a summary of representative results obtained from the quantification of the repertoire of natural anti-glycan antibodies in a population of 20 BALB/c mice. The glycochips used in this study contained 419 different glycan structures. Most glycans were synthesized as -CH2CH2CH2NH2 spacer-armed O-glycosides, in several cases as -CH2CH2NH2 or -NHCOCH2NH2 glycosides. All glycan structures were characterized by high resolution (700- or 800 MHz) NMR spectroscopy, purified and tested by HPLC, indicating their >95% purity. We have simultaneously determined IgM + IgG anti-glycan antibodies due to a restriction in the amount of mouse serum. In the PGA, we considered values above 4,000 RFU as a positive signal of antibody binding (this value is ~10% of the top glycans RFU). The results presented in this work follow most of the guidelines for reporting glycan microarray-based data39. Only 17% of carbohydrate structures demonstrated ≥4,000 RFU in the PGA (Figure 2, in red). Most of the glycan structures exposed in the glycochips were not recognized by the repertoire of circulating anti-glycan antibodies of BALB/c mice (Figure 2, in blue and white)28. The conserved pattern of natural anti-carbohydrate antibodies of BALB/c included 12 different glycan specificities, with very high median signal intensities of antibodies binding (≥10,000 RFU Table 1)28.
Figure 1: Schematic representation (not at scale) of the glycan array configuration, printing, and analysis. (A) Printed microchips are developed with a library of 419 different glycan structures, followed by the detection with an appropriate fluorescently labeled secondary antibody. Each slide contains 4 different blocks of sub-arrays (in colors), repeated 6 times. Every single sub-array is formed by 112 different glycan spots (8 rows × 14 columns), including controls. (B) A representative example of the images obtained from microchip scanning using a fluorescence scanner (third part of the image). (C) The process of aligning the "grid" to spots in every single sub-array (template adjustment during quantification). (D) The fluorescence is detected for each spot and results are transferred into a common spreadsheet file. Please click here to view a larger version of this figure.
Figure 2: Repertoire of natural circulating anti-carbohydrate antibodies of BALB/c mice. Mouse serum samples (1:20) were incubated with the glycochips and scanned using a ScanArray reader. Data were analyzed with a microarray analysis system and results were expressed in relative fluorescence units (RFU) as the median ± median absolute deviation (MAD). Blue and white colors represent binding signals lower than 4,000 RFU (background); red color represents signals ≥4,000 RFU (positive binding). F: female; M: male (n = 20). This figure has been reproduced from Bello-Gil, D. et al.28. Please click here to view a larger version of this figure.
Glycan ID (#) |
Structure | Common name | Median and MAD as RFU | Number of mice showing RFU ≥4000 (%) | |||
60 | 6-O-Su-Galβ-spb | 61113 | 1156 | 100 | |||
271 | Galβ1-6Galβ1-4Glcβ-sp | 53622 | 1934 | 100 | |||
802 | Galβ1-3GalNAc(furc)β-sp | 51348 | 2324 | 100 | |||
176 | 3-O-Su-Galβ1-4(6-O-Su)Glcβ-sp | 43008 | 9342 | 100 | |||
166 | GlcAβ1-6Galβ-sp | 39105 | 2993 | 85 | |||
150 | 3-O-Su-Galβ1-3GalNAcα-sp | 37943 | 3232 | 100 | |||
437 | GalNAcα1-3(Fucα1-2)Galβ1-3GalNAcβ-sp | A(type 4) | 33886 | 3193 | 90 | ||
125 | 6-Bn-Galβ1-4GlcNAcβ-sp | 32674 | 5389 | 95 | |||
154 | 3-O-Su-Galβ1-3GlcNAcβ-sp | 32651 | 3954 | 100 | |||
177 | 3-O-Su-Galβ1-4(6-O-Su)GlcNAcβ-sp | 32496 | 7215 | 100 | |||
287 | 3-O-Su-Galβ1-3(Fucα1-4)GlcNAcβ-sp | SuLea | 20063 | 4962 | 95 | ||
234 | Galβ1-4(Fucα1-3)GlcNAcβ-sp | Lex | 13573 | 2635 | 80 |
Table 1: Top rank glycan structures recognized by natural antibodies of BALB/c mice. Glycans with binding signals above 4,000 RFU in at least 80% of examined mice (n = 20). bsp means aminoethyl, aminopropyl or glycyl spacer. cfuranose; all other monosaccharides are in a pyranose form; Fuc residue has L-configuration, all other monosaccharides – D-configuration. This table has been modified from Bello-Gil, D. et al.28.
Supplementary Table 1: List of glycans, their binding to natural circulating antibodies (IgM + IgG) of BALB/c mice (n = 20), expressed in relative fluorescence units (RFU) as median ± MAD, and the number of animals exceeding cut off (≥4000 RFU). This table has been reproduced from Bello-Gil, D. et al.28. Please click here to download this table
Glycan microarrays have become indispensable tools for studying protein-glycan interactions40. The present work describes a protocol based on PGA technology to study the repertoire of circulating of anti-carbohydrate antibodies in BALB/c mice. Since PGA offers the possibility to screen large numbers of biologically unknown glycans, it is an exceptionally convenient discovery tool13,15,28. The proposed method offers the possibility to measure, in the same experimental setting, hundreds of glycan structures using a reduced amount of serological sample (50 µL). This is especially critical in the case of small animals (little circulating blood volume), or when it's necessary to extract blood several times from the same experimental animal.
We demonstrated, as representative results, that genetically identical mice should not be considered as immunological equivalents; because they develop different patterns of natural anti-carbohydrate antibodies (only 12 glycan specificities were conserved). Serological levels for the rest of the repertoire of natural anti-carbohydrate antibodies varied considerably among the examined animals. Analysis of the gut microbiota of inbred animals41 could explain this heterogeneity42,43,44,45,46. If the production of natural anti-glycan antibodies is mediated by the antigenic stimulation of microbiota, and this is different among inbred mice41, fine specificity of these antibodies will not be identical.
The main drawback for PGA development is the access to well-defined glycan structures40,47. Glycans produced in biological systems are heterogeneous40,47,48, and their biosynthesis relies on the differential expression of carbohydrate enzymes, resulting in heterogeneous mixtures of glycoforms, each with a distinct physiological activity47. The complex composition and configuration of the glycans present in the biological systems make their productions challenging40,47,48. Along with chemo-enzymatic synthesis, glycans isolated from natural sources will continue to be the major source of glycans for arrays development40. Low synthetic yields and the complex purification process from glycoproteins and glycosphingolipids make the efficient production of glycans at large scale difficult40,47,48. Hence, the availability and the prices of glycans continue being a very limiting condition to expand the use of PGA as a discovery tool.
Additionally, within the protocol, critical steps mostly relating to the correct distribution of solutions (serum, secondary antibodies) over the glycochip surface must be executed with caution. The methodology requires, at least, 1 mL of these solutions, to homogenously soak all dry areas of the glycochip surface. This is crucial to obtain minimal differences between glycan replicates and also to avoid excessive background during quantification.
Despite the mentioned limitations, PGA is a very sensitive tool for approaches related to study protein-glycan interactions40, or to study the repertoire of anti-glycan antibodies in a particular experimental setting or condition13,15,28. This study can be extrapolated to different species (including human samples) 13,15,23,28, providing a versatile methodology for identifying the repertoire of circulating anti-carbohydrate antibodies.
We also anticipate the potentiality that this approach may bring in the early diagnosis and derived treatment in some of the pathological conditions where antibodies directed to glycan structures seem to play an important role.
The authors have nothing to disclose.
This work was supported by "Fondo de Investigaciones Sanitarias" (FIS) grant PI13/01098 from Carlos III Health Institute, Spanish Ministry of Health. DB-G was benefited from a post-doctoral research position funded by the European Union Seventh Framework Programme (FP7/2007-2013) under the Grant Agreement 603049 (TRANSLINK). Work of NK, NS, and NB was supported by grant #14-50-00131 of Russian Science Foundation. DB-G wants to express his gratitude to Marta Broto, J. Pablo Salvador and Ana Sanchis for excellent technical assistance, and Alexander Rakitko for assistance in statistical analysis. With the support of the "Pla de Doctorats Industrials de la Secretaria d'Universitats i Recerca del Departament d'Empresa i Coneixement de la Generalitat de Catalunya (grant number 2018 DI 021). We thank CERCA Programme / Generalitat de Catalunya for institutional support.
Antibodies | |||
biotinylated goat anti-human Igs | Thermo Fisher Scientific, Waltham, MA, USA | Ref. #: 31782 | |
biotinylated goat anti-mouse IgM + IgG | Thermo Fisher Scientific | Ref. #: 31807 | |
Equipment | |||
Robotic Arrayer sciFLEXARRAYER S5 | Scienion AG, Berlin, Germany | http://www.scienion.com/products/sciflexarrayer/ | |
Stain Tray (slide incubation chamber) | Simport, Beloeil, QC, Canada | Ref. #: M920-2 | |
Centrifuge | Eppendorf, Hamburg, Germany | Ref. #: 5810 R | |
Pipettes | Gilson, Middleton, WI, USA | http://www.gilson.com/en/Pipette/ | |
Slide Scanner | PerkinElmer, Waltham, MA, USA | ScanArray GX Plus | |
Shaking incubator | Cole-Parmer, Staffordshire, UK | Ref. #: SI50 | |
Biological samples | |||
BALB/c mice sera | This paper | N/ A | |
Complex Immunoglobulin Preparation (CIP) | Immuno-Gem, Moscow, Russia | http://www.biomedservice.ru/price/goods/1/17531 | |
Chemicals, Reagents and Glycans | |||
Glycan library | Institute of Bioorganic Chemistry (IBCh), Moscow, Russia | N/ A | |
Bovine serum albumin (BSA) | Sigma-Aldrich, St. Louis, MO, | Ref. #: A9418 | |
Ethanolamine | Sigma-Aldrich | Ref. #: 411000 | |
Tween-20 | Merck Chemicals & Life Science S.A., Madrid, Spain | Ref. #: 655204 | |
Phospahte buffered saline (PBS) | VWR International Eurolab S.L, Barcelona, Spain | Ref. #: E404 | |
Sodium azide | Sigma-Aldrich | Ref. #: S2002 | |
Streptavidin Alexa Fluor 555 conjugate | Thermo Fisher Scientific | Ref. #: S21381 | |
Streptavidin Cy5 conjugate | GE Healthcare, Little Chalfont, Buckinghamshire, UK | Ref. #: PA45001 | |
Materials | |||
N-hydroxysuccinimide-derivatized glass slides H | Schott-Nexterion, Jena, Germany | Ref. #: 1070936 | |
Whatman filter paper | Sigma-Aldrich | Ref. #: WHA10347509 | |
1.5 mL tubes | Eppendorf | Ref. #: 0030120086 | |
Software and algorithms | |||
ScanArray Express Microarray Analysis System | PerkinElmer | http://www.per | |
kinelmer.com/microarray | |||
Hierarchical Clustering Explorer application | University of Maryland, MD, USA | http://www.cs.umd.edu/hcil/hce/ |