Microarray polymer profiling (MAPP) is a high-throughput technique for compositional analysis of glycans in biological samples.
Microarray polymer profiling (MAPP) is a robust and reproducible approach to systematically determine the composition and relative abundance of glycans and glycoconjugates within a variety of biological samples, including plant and algal tissues, food materials, and human, animal, and microbial samples. Microarray technology underpins the efficacy of this method by providing a miniaturized, high-throughput screening platform, allowing thousands of interactions between glycans and highly specific glycan-directed molecular probes to be characterized concomitantly, using only small amounts of analytes. Constituent glycans are chemically and enzymatically fractionated, before being sequentially extracted from the sample and directly immobilized onto nitrocellulose membranes. The glycan composition is determined by the attachment of specific glycan-recognizing molecular probes to the extorted and printed molecules. MAPP is complementary to conventional glycan analysis techniques, such as monosaccharide and linkage analysis and mass spectrometry. However, glycan-recognizing molecular probes provide insight into the structural configurations of glycans, which can aid in elucidating biological interactions and functional roles.
Glycans are ubiquitous in all domains of life and exhibit unparalleled diversity in structure and function compared to other macromolecules1. However, due to their complexity, variability in biosynthesis and glycosidic linkages, and the paucity of appropriate methods for dissecting glycan structures, our understanding of this diversity in structures and functions is relatively limited2.
Many glycan analysis techniques are destructive and necessitate the breakdown of glycans into their constituent monosaccharides, which can obscure relevant three-dimensional and biological contexts3. Conversely, monoclonal antibodies (mAbs), carbohydrate binding modules (CBMs), lectins, viral agglutinins, and microbial adhesins, known collectively as glycan-recognizing molecular probes (GRMPs)4, recognize and bind to specific epitopes and can be used as tools to detect and discriminate between glycans within complex multi-glycan matrices5,6.
Here, we present microarray polymer profiling (MAPP), a rapid, versatile, and nondestructive method for glycan analysis that is applicable to a broad spectrum of biological samples. The method aims to provide a robust and high-throughput technology for analyzing glycans from diverse biological and industrial/commercial systems. MAPP unites the recognition specificity of glycan-directed molecular probes with reproducible, high-performance microarray screening technology to allow thousands of molecular interactions to be profiled in parallel. The output of this approach is diagnostic insight into the composition and relative abundance of glycans within a sample or tissue of interest.
MAPP can be used as an independent, stand-alone method, or in conjunction with other biochemical techniques, such as immunofluorescence microscopy7,8,9 and monosaccharide or linkage analysis10,11. The technique can also be used to map epitope specificities of novel GRMPs, using arrays printed with pure and structurally well-defined oligosaccharide standards12. A major advantage of MAPP over other methods, such as enzyme-linked immunosorbent assay (ELISA), is its compatibility with small sample volumes13,14. Moreover, MAPP offers significantly higher-throughput analysis15 and provides an effective form of sample preservation, as printed samples are dry and stable when immobilized onto nitrocellulose16.
The binding of GRMPs is generally dependent on the presence of a number of contiguous sugar residues that collectively form a binding site (epitope) that is unique to a particular polysaccharide class (xylan, mannan, xyloglucan, etc.)17. In contrast, the individual sugar residues (xylose, mannose, glucose) that are quantified using most biochemical techniques, for example monosaccharide composition or methylation analysis, can be components of multiple polysaccharide classes and thus difficult to assign18.
MAPP has been developed in response to a technology gap, namely the ability to rapidly analyze multiple glycans from a variety of sources using small amounts of material. MAPP capitalizes on the extensive repertoire of GRMPs that have been developed and characterized over the last three decades12,19,20,21,22,23,24,25,26,27,28,29,30,31,32. The development of MAPP has been an iterative process, with the technique being steadily refined and optimized. There is now a substantial body of literature describing the application of MAPP to various natural and industrial systems where glycans play central roles 5,6,9,10,21,33,34,35,36,37,38,39. Here, we describe the current state of the art for MAPP.
The main experimental stages of the MAPP method are summarized in Figure 1.
1. Preparation of samples
NOTE: Here, the method is applied to plant tissues for illustrative purposes. The plants selected were Coffea arabica, Allium sativum var. ophioscorodon, and several Thai mango varieties (Aokrong, Kam, Rad, Chokanan, Mamkamdang, Talabnak, Mahachanok, and Nga). The plants were selected for their commercial importance. Their processing for human consumption generates currently underutilized agro-industrial waste, which may provide a source of value-added products, including pure glycans. Thus, MAPP was applied to characterize the glycan composition of the waste plant biomass for bioprospecting purposes.
2. Preparation of alcohol insoluble residue (AIR)
3. Glycan extraction
NOTE: If possible, perform all extraction steps in a tissue lyser with a ball bearing in each tube to aid resuspension. If a tissue lyser is unavailable, extractions can be performed instead with continuous stirring or shaking. It may be necessary to extend the extraction time if this is not possible.
4. Preparation of standards
5. Microarray printing
6. Microarray probing
7. Analysis and quantification
MAPP was applied to determine the glycan composition of agricultural biomass waste, comprising mango peels from several northern Thai varieties, Coffea arabica cherry pulp and coffee bean processing waste, and root, stem, and leaf tissue from Thai black garlic, Allium sativum var. ophioscorodon. Several plant-derived polysaccharides are used in the food industry as functional ingredients42,43. Thus, the aim of this experiment was to deduce whether these abundant and currently underutilized agro-industrial waste materials may provide a source of value-added pure polysaccharides.
AIR material is routinely employed to prepare samples intended for glycan analysis44. There are several advantages to using AIR; treatment with solvents effectively removes endogenous CAZymes, metabolites, small saccharides, lipids, and pigments, resulting in samples enriched with polysaccharides and structural proteins34. Furthermore, producing AIR is a fast and effective way to increase sample longevity, as it is thermostable and can be stored for several years.
Three mixed fractions of constituent glycans were sequentially extracted from plant AIR material using CDTA, NaOH, and cellulase. CDTA chelates Ca2+ ions, which allow the removal of Ca2+ crosslinked de-esterified pectins from plant cell walls45. Alkaline conditions allow predominantly hemicelluloses, such as mannan, xylan, and β-glucan, to be released due to the disruption of hydrogen bonding and saponification of ester linkages between cellulose microfibrils and hemicellulose, and lignin and hemicellulose, respectively46. A recombinant endo-1,4-β-glucanase from Bacillus spp. was used to degrade amorphous regions of the structural cellulose microfibrils, releasing residual glycans bound to cellulose within the cell walls47. Although this method effectively separates glycans into these three broad groups, it should be noted that the samples are not pure; by the very nature of the extraction method, hemicellulose, if present in the sample, will inevitably be extracted and subsequently detected to varying degrees in the CDTA and cellulase fractions. Likewise, some pectin will be detected in the NaOH extraction if present in the sample.
A non-contact, piezoelectric microarray printing robot was used to immobilize extracted glycan fractions onto nitrocellulose via non-covalent attachment11, forming 300 identical microarrays. Defined glycan standards (Table 1) were also included in the printed microarrays as positive controls (Figure 5). The MAPP binding profile obtained for the selected glycan standards corresponds to previously reported epitope specificities. For example, LM21 exhibited strong binding to multiple mannan polysaccharides (galactomannan and glucomannan), while LM22 exhibited only weak binding to galactomannan25. Similarly, LM19 preferentially bound to de-esterified homogalacturonan48 and LM15 bound to tamarind seed xyloglucan23.
The relative abundance of 16 epitopes, diagnostic of non-cellulosic plant cell wall polysaccharides, were detected by the attachment of glycan-directed monoclonal antibodies (Table 2) to printed extracts (Figure 6). The majority of extracted glycans were detected within the alkaline NaOH fraction. Strong binding signals were recorded for mAbs LM10 and LM11, representing xylan/arabinoxylan, within the peels of all the mango varieties tested. Within the garlic samples, LM10 and LM11 bound preferentially to root tissue extract (Garlic R) and exhibited only weak binding to the leaf tissue extract (Garlic L). LM19, representing partially methyl-esterified or un-esterified homogalacturonan, bound strongly to some mango variety extracts (Aokrong and Talabnak), but bound only weakly, or its binding was undetectable, in other varieties (Chokanan, Mamkamdang, Mahachanok, and Nga). In addition, LM19 bound only to the coffee pulp fractions and did not bind to the coffee bean processing waste material, previously thought to be composed of semi-purified coffee pectin (unpublished data).
Figure 1: Major experimental steps in the MAPP method. (A) Samples are homogenized to form fine powders. (B) The homogenized samples are processed to isolate their AIRs. (C) The constituent glycans are sequentially extracted using a tailored extraction regime. (D) The extracted glycan fractions, ink, and GSB are transferred into 384-well plates, according to the plate layout, for printing onto nitrocellulose. (E) The printed microarrays are probed with selected GRMPs. (F) GRMP binding to the printed glycan fractions is quantified and analyzed before data are presented as a heatmap. Please click here to view a larger version of this figure.
Figure 2: Example of a 384-well plate layout for sample, ink, and GSB loading with four dilutions per extracted glycan sample/standard. Different colors denote samples arising from different extraction reagents, while different shades represent serial dilutions. The first number in the code represents the sample number, while the end number represents the dilution number (D1 denotes dilution one, D2 denotes dilution two, and so on). For example, a well labeled '12D3' represents glycan sample 12, dilution three. Well plates should be divided into eight identical sections comprising six columns and eight rows. The first section of the first plate should contain only ink and buffer and resemble the example plate layout. Extracted glycan samples can then be loaded into subsequent plate sections according to the plate layout. Different extraction reagents should not be loaded into the same plate section. If there are insufficient samples to fill an entire section, fill all remaining wells in that section with buffer; do not leave any wells empty. If multiple plates are required, the next section after all samples have been loaded should contain three alternating columns of ink and GSB-this may not be section eight, depending on the number of samples being printed. Please click here to view a larger version of this figure.
Figure 3: Schematic representation of printed microarray design. Please click here to view a larger version of this figure.
Figure 4: Representative microarrays. (A) No binding. (B) The binding signal is obscured by high background signal. (C) Generalized blue/purple staining due to oversaturation with NBT/BCIP. (D) Defective probing due to high substrate concentration. (E) Defective printing due to the unclean print head. (F) Strong binding to few samples. (G) Strong binding to many samples. Please click here to view a larger version of this figure.
Figure 5: Monoclonal antibody binding to defined glycan standards, included to validate the printing and probing process. Please click here to view a larger version of this figure.
Figure 6: MAPP of glycans extracted from agricultural biomass waste. The samples include coffee pulp waste (coffee pulp and coffee pectin), mango peels from several Thai varieties (AO, Aokrong; KO, Kam; RD, Rad; CH, Chokanan; MA, Mamkamdang; TL, Talabnak; MH, Mahachanok; NG, Nga) and black garlic leaves (Garlic L), stem (Garlic S), bulb (Garlic BG), and roots (Garlic R), using CDTA, NaOH, and cellulase (Bacillus spp. cellulase 5A). Please click here to view a larger version of this figure.
Table 1: Defined commercial polysaccharide standards used in MAPP analysis as positive controls. Please click here to download this Table.
Table 2: Glycan-directed monoclonal antibodies selected for the interrogation of extracted plant glycan microarrays. Please click here to download this Table.
The MAPP technique described here is now a well-established method for glycan analysis. The basic principles were first described in 200711, but the technique has undergone continuous development in order to capitalize on the latest innovations in microarray technology, molecular probe development, and advances in our understanding of glycan biochemistry. In general, glycans, especially polysaccharides, are more challenging to analyze than proteins and nucleotides due to their structural complexity and heterogeneity45, as well as the fact that they cannot readily be sequenced or synthesized1. In many cases, no single technique can decipher glycan complexity conclusively; thus, MAPP is often used with other methods. This is one of the reasons why AIR preparation is usually chosen as a starting point for MAPP, since AIR is compatible with most other glycan analysis methods34, facilitating the subsequent comparison of datasets.
Due to homogenization of the sample prior to AIR preparation, some spatial information is invariably lost. However, as polysaccharides are sequentially released from samples, the presence of epitopes in the fractions obtained provides information about the molecular architecture and composition of that sample17. Thus, selecting an appropriate extraction regime is critical to the success of the method. Multiple parameters determine the suitability of the extraction method: cellular structure, time, temperature, pH, pressure, ionic strength of the solvent, and fineness of the solid particulate sample49. It is recommended that a range of increasingly more aggressive solvents is used to maximize the likelihood of successfully extracting constituent glycans and building a representative compositional picture of the sample. For most samples, CDTA, NaOH, and cellulase are sufficient to remove plant-derived storage and cell wall polysaccharides33,50,51,52. For some tissue samples, a hybrid extraction regime that also includes CaCl2, HCl, and Na2CO3 has been shown to be successful53, while marine microalgal samples may require the addition of ethylenediaminetetraacetic acid (EDTA)10.
Microarrays should include a range of pure, defined glycan standards to be used as positive controls5. Included standards should be modified according to the nature of the sample. Once printed, appropriate GRMPs need to be selected. The generation of hybridoma mAbs to polysaccharide structures is challenging54; glycan-binding antibodies are difficult to raise and can have low affinity55. Fortunately, gene sequence information for CBMs can be obtained with relative ease for recombinant expression4 and engineering their binding specificities56,57. Whilst an impressive catalog of GRMPs has been developed, with most now available from commercial sources, relative to the diversity of glycan structures existing in nature, only a small proportion have been produced and successfully characterised58. This can limit the ability to detect and discriminate between certain structures. It is advisable to perform an initial probing experiment using one or two probes representative of each major glycan structure anticipated to be present, for which the binding specificity is well characterized. In subsequent probing experiments, the probe list can be expanded to cover a broader range of glycans and delve deeper into fine structures.
Although mundane, ensuring that microarrays are washed thoroughly after each incubation step is fundamental to the success of the probing procedure. The ineffective removal of non-specifically bound probes is likely to obscure the result by causing a high background signal following color development. In this case, it is necessary to repeat the probing procedure, starting with a new microarray. Furthermore, arrays should be touched sparingly and only by holding the edges with forceps; the nitrocellulose membrane is brittle and easily damaged. The color development solution collects in cracks and creases, causing oversaturation, which impedes array analysis.
MAPP is quick, adaptable, and convenient. This method is compatible with animal, microbial, or plant glycans derived from any biological or industrial system, so long as they can be extracted and immobilized onto nitrocellulose, and for which one has appropriate molecular probes. The data generated provide detailed, semi-quantitative, compositional insight, which cannot readily be obtained via other glycan analysis methods.
The authors have nothing to disclose.
The authors would like to acknowledge ArrayJet for their expert advice regarding microarray robotics. SS and JS would like to acknowledge the support from Fundamental Fund 2022 (FF65/004), Chiang Mai University.
1,3:1,4-β-D-Glucan, Lichenan (icelandic moss) | Megazyme | P-LICHN | |
1,4-β-D-Mannan | Megazyme | P-MANCB | |
384-well microtiter plate | Greiner Bio-One | M1686 | |
5-bromo-4-chloro-3-indolyl-phosphate (BCIP) | Melford | B74100-1.0 | |
Acetone | Sigma | 270725 | |
Alkaline Phosphatase AffiniPure Goat Anti-Mouse IgG (H+L) | Jackson ImmunoResearch | 115-055-003 | |
Alkaline Phosphatase AffiniPure Goat Anti-Rat IgG (H+L) | Jackson ImmunoResearch | 112-055-003 | |
Alkaline Phosphatase AffiniPure Rabbit Anti-His Tag | Jackson ImmunoResearch | 300-055-240 | |
Arabinoxylan (wheat) | Megazyme | P-WAXYL | |
Array-Pro Analyzer Software | Media Cybernetics | Version 6.3 | |
Bacillus sp. Cellulase 5A (BCel5A) | NZYTech | CZ0564 | |
BAM antibodies | SeaProbes | Various | |
Black drawing ink (indian ink) | Winsor & Newton | GWD030 | |
Carbohydrate binding modules | NZYTech | Various | |
CCRC antibodies | CarboSource | Various | |
CDTA | Sigma | 319945 | |
Chloroform | Sigma | PHR1552 | |
Ethanol | Sigma | 1.11727 | |
Galactan (potato) | Megazyme | P-GALPOT | |
Galactomannan (carob) | Megazyme | P-GALML | |
Glycerol solution | Sigma | 49781-5L | |
Gum tragacanth (legumes) | Sigma-Aldrich | G1128 | |
INCh antibodies | INRA | Various | |
LM and JIM antibodies | PlantProbes | Various | |
Marathon Argus Microarray Printer | ArrayJet | ||
Methanol | Sigma | 34860 | |
Monoclonal antibodies | Biosupplies Australia | Various | |
NaBH4 | Sigma | 452882 | |
NaOH | Sigma | S5881 | |
Nitro-blue tetrazolium (NBT) | Melford | N66000-1.0 | |
Nitrocellulose membrane | Thermo Fisher Scientific | 88018 | |
Pectin (degree of methyl esterification 46%) | Danisco | NA | |
ProClin 200 | Sigma | 48171-U | |
Rhamnogalacturonan (soybean pectic fibre) | Megazyme | P-RHAGN | |
Rotating mixer | Fisher Scientific | 88-861-050 | |
Rotating/rocking Shaker | Cole-Parmer | ||
Skimmed milk powder | Marvel | ||
Spin filter | Costar Spin-X | 8160 | |
Stainless steel beads | Qiagen | 69989 | |
TissueLyser II | Qiagen | 85300 | |
Tris | Sigma | 93362 | |
Triton X-100 | Sigma | T8787-250ML | |
Tween 20 | Sigma | P9416-100ML | |
Xylan (beechwood) | Megazyme | P-XYLNBE | |
Xyloglucan (tamarind) | Megazyme | P-XYGLN | |
β-Glucan (oat) | Megazyme | P-BGOM |