|Home | About | Journals | Submit | Contact Us | Français|
Lectins, plant-derived glycan-binding proteins, have long been used to detect glycans on cell surfaces. However, the techniques used to characterize serum or cells have largely been limited to mass spectrometry, blots, flow cytometry, and immunohistochemistry. While these lectin-based approaches are well established and they can discriminate a limited number of sugar isomers by concurrently using a limited number of lectins, they are not amenable for adaptation to a high-throughput platform. Fortunately, given the commercial availability of lectins with a variety of glycan specificities, lectins can be printed on a glass substrate in a microarray format to profile accessible cell-surface glycans. This method is an inviting alternative for analysis of a broad range of glycans in a high-throughput fashion and has been demonstrated to be a feasible method of identifying binding-accessible cell surface glycosylation on living cells. The current unit presents a lectin-based microarray approach for analyzing cell surface glycosylation in a high-throughput fashion.
Lectins, plant-derived glycan-binding proteins, have long been used to detect glycans on cell surfaces. However, the techniques used to characterize serum or cells have largely been limited to mass spectrometry, blots, flow cytometry (Batisse et al., 2004; Venable et al., 2005), histochemistry (Carter and Brooks, 2006), and immunohistochemistry (Wearne et al., 2006). While these lectin-based approaches are well established and can discriminate a limited number of sugar isomers by concurrently using a limited number of lectins (Hirabayashi, 2004), they are not amenable for adaptation to a high-throughput platform. Fortunately, given the commercial availability of lectins with a variety of glycan specificities, lectins can be printed on a glass substrate in a microarray format to profile accessible cell surface glycans. This method is an inviting alternative for analysis of a broad range of glycans in a high-throughput fashion and has been demonstrated to be a feasible method of identifying binding-accessible cell surface glycosylation on living cells (Tao et al., 2008). The current unit presents a lectin-based microarray approach for analyzing cell surface glycosylation in a high-throughput fashion.
The lectin microarray is simple in concept similar to other array technologies. Whereas other arrays analyze binding interactions on molecules of similar scales, lectin microarrays can be used to ascertain lectin-glycan interactions for live cells. Nevertheless, the overarching concept of the lectin microarray is direct. The fundamental interaction that captures the cells is between the printed lectin and the cell surface glycans that are accessible to those lectins. Although lectin-glycan interactions are of low affinity (10−6 to 10−7 range for most glycoproteins), the multivalency of lectins can achieve sufficiently high avidity to retain cells on the lectin spot. However, having a low affinity is also advantageous since nonspecific interactions can be easily overcome by the application of shear forces.
1. Dissolve lyophilized lectins to 2 mg/ml in storage buffer as recommended by the manufacturer of the lectins.
2. Dilute the lectin stock solution (from step 1) 1:1 with 2× printing buffer to achieve a final concentration of 1 mg/ml in 1× printing buffer.
3. Transfer 7-μl lectin solution into each well of a 384-well plate for printing. When necessary, include positive controls (e.g., Cy3 and Cy5 dyes).
For long-term storage, keep the lectins divided into aliquots up to 3 months at −20°C.
4. Print lectins in triplicate per block with replicate blocks on Schott NexterionR H slides.
Printing parameters, such as spot/pin size, spot distance, etc., should be optimized based on the arraying technology used. For example, Tao et al. (2008) used a BioRad contact printer with 120-μm diameter quill-style pins and 230-μm spot spacing to print lectins in triplicate per block. Consult the operation guide of the array printer being used for the printing procedure.
5. Keep the printed lectin microarrays at 4°C and protected from light overnight prior to use.
For storage, the printed lectin microarrays can be maintained for up to 6 months at 0°C, without losing activity.
6. Prior to washing, mark the boundaries (using a laboratory marker) of the printed areas on the reverse side of the microarray slide. Gently flow slide washing buffer over the printed slide surface to remove excess lectins using Nalgene wash bottles or a 25-ml pipet.
This is important for the application of silicone gaskets later, since washing the lectin microarray will make printed spots invisible to the naked eye.
7. Submerge the slide surface completely in blocking buffer for 1 hr at room temperature. Remove the slide from the blocking buffer from one side slowly.
The slide surface will not retain any liquid except at the edge, which can be easily removed by gently tapping the slide on a paper towel.
8. While the microarrays are in the blocking buffer, prepare the cells of interest. Label cells with CFSE tracking dye by resuspending the cells in 1× PBS with 10 μM CFSE at 1×106 cells/ml. Place the cells in a 37°C water bath for 1.5 min.
From this point forward, keep samples away from light as much as possible.
9. Wash the cells two times, each time with 15 ml 1× PBS to remove excess CFSE dye. Wash once more with 15 ml cell binding buffer.
10. Resuspend the washed cells in cell binding buffer at 5 × 106 cells/ml.
11. Apply adhesive silicone gasket (e.g., Grace BioLabs 16-well silicone gasket) to the printed slide surface.
12. Allow the cells to bind at room temperature for 2 hr total.
The number of cells per block can be adjusted for both availability, as well as printed area size. The number of cells needed can be estimated by accounting for the area of the printed block with the approximate size of the cells used to probe the lectin microarray.
13. Submerge the slide in a large volume of 1 × PBS. Carefully remove the incubation gasket while keeping the slide completely submerged. Invert the slide and shake gently while still submerged to dislodge and wash away unbound cells.
Some gasket adhesives may be too adherent, making delicate removal difficult. The authors used FlexWell silicone gaskets from Grace Biolabs with adhesive specifically designed to work with the Schott Nexterion slide surface. While using FlexWell gaskets is not mandatory, the authors recommend performing a mock run with the gasket and slide.
14. Remove the slide from the slide washing buffer edgewise to reduce the amount of slide washing buffer on the slide.
15. Blot away excess slide washing buffer using a paper towel, avoiding the areas where lectins have been printed. Remove as much slide washing buffer as possible.
After the lectin microarray is removed from the slide washing buffer, lectin spots with bound cells should be visible to the naked eye.
16. Allow the slide to air dry; it will take <5 min.
17. Acquire the slide image on a microarray scanner at 5-μm resolution in the FITC fluorescence channel. Apply appropriate laser power and gain settings while avoiding signal saturation.
Typical cell binding images are shown in Figure 12.9.1.Two repeated probings of HEK 293 cells on lectin microarrays. Lectins were printed in triplicate on the substrate slides. The lower panel shows the lectin layout on the microarray and the binary lectin binding profile for HEK293 cells. For clear visualization, ...
18. Quantitate signal intensities for each lectin spot (e.g., Axon GenePix 5.0).
A binary mode of analysis can be performed according to the methods outlined by Tao et al. (2008).
19. Briefly, normalize lectin intensities for each spot to background fluorescence and plot.
The spots with signal intensities ≤6 standard deviations above 0 are considered nonbinding events; spots with signal intensities >6 standard deviations above 0 are considered positive binding events. Based on the number of replicates per lectin-cell pair, a voting system can be used to determine if the total binding event is positive or negative. A hierarchically clustered binary code of lectin-cell interactions can be constructed for each cell type analyzed.
Supplying divalent cations is necessary for certain lectins to bind their respective glycan ligands. BSA is added to reduce the nonspecific binding.
Printing buffer is used to dissolve lectin stock solutions for printing. Tween-20 and BSA are added to improve spot morphology. Glycerol is added to reduce evaporation and enhance stability in storage. Percentages of reagents in printing buffer are optimized for lectin microarrays but may be adjusted.
50 mM ethanolamine in 50 mM sodium borate, adjust to pH 8.0 using 1 M HCl. Store up to 1 year at room temperature.
If using stored buffer, verify the pH prior to usage.
Ethanolamine is used to block any remaining active amine-chemistry sites on the Schott NexterionR H slide, which is important to eliminate non-specific cell capture.
The surfaces of all vertebrate cells are decorated with a broad selection of glycans with considerable structural diversity called the “glycocalyx.” Due to the abundance of glycans on cell surfaces, it comes as no surprise that many cellular interactions with the surrounding environment are affected by glycosylation. Interactions with neighboring cells, as well as interactions with the extracellular matrix, are two facets of cellular communication that are heavily dependent on glycan binding. Changes in glycosylation have been shown to correlate to cellular differentiation, malignant transformations (Lau and Dennis, 2008), tumor progression (Mayoral et al., 2008; Zhao et al., 2008), and metastasis (Yuan et al., 2008), suggesting the importance of characterizing cell surface glycosylation states for research, diagnostic, and potential therapeutic purposes. Unfortunately, the structural diversity of the cell surface glycome, which confers the biological functional diversity, is in fact responsible for impeding its analysis.
Traditional means of investigating glycosylation using flow cytometry, blotting, histochemistry, and mass spectrometry either address singular glycosylation specificities or fail to scale to high-throughput applications. Lectin microarrays help overcome both limitations by using the wide range of available lectins with a correspondingly wide range of glycan specificities to profile cell surfaces (Tao et al., 2008). Lectin microarrays showed promise in early studies analyzing glycosylation states of cell lysates, bacteria, as well as proof of principle experiments using Lec mutant Chinese Hamster Ovary (CHO) cells (Hsu et al., 2006; Tateno et al., 2007). Although lectin microarrays have been used to profile glycoproteins and cell lysates (Kuno et al., 2005; Ebe et al., 2006; Lee et al., 2006; Pilobello et al., 2007), their use for profiling live cells have been limited to proof-of-principle studies (Zheng et al., 2005; Hsu et al., 2006; Lee et al., 2006; Tateno et al., 2007). For example, lectin microarrays accurately and predictably distinguish between glycosylation-defective CHO mutants (Tateno et al., 2007), and Hsu and colleagues used a 21-lectin microarray to distinguish O-antigens on lipopolysaccharides (LPS) of bacterial cell surfaces (Hsu et al., 2006). However, extensive profiling of mammalian cell lines, as well as primary cell populations, was still largely unexplored territory until recently. Tao and colleagues showed that in addition to the profiling potential of lectin microarrays, the lectin binding profile can also used for multiple biologically relevant applications including development into potentially clinically relevant diagnostic tools with predictive power (Tao et al., 2008).
The fabrication process of the lectin microarray is similar to that of existing protein microarrays (Zhu et al., 2001; Chen et al., 2008; Hu et al., 2009). Printing buffers described in this protocol have been optimized for the lectins used in Tao et al. (2008). The composition of the printing solution can be adjusted if spot morphology is unacceptable. If there is no cell binding and a printing error is suspected, lectin spots can be validated by probing the microarray with an amine-reactive dye, such as 549 nm NHS-Ester (Invitrogen). Moreover, individual lectins can also be tested via flow cytometry, using fluorescently conjugated lectins (e.g., using FITC), and cells in solution analyzed by FACS.
The authors have found that one of the most important considerations in probing the microarray is the number of cells used. Each lectin block should be seeded with cells at the recommended density when possible to ensure proper coverage of the printed spots with cells. As mentioned in the protocol, the addition of periodic agitation can reduce the minimum number of cells necessary, but a static incubation is preferred when possible. Fluorescent labeling of cells can be verified by analyzing labeled cell samples on a flow cytometer.
Cell binding will vary based on numerous factors including lectin stability, glycan availability, lectin binding avidity, and cell density. See Tao et al. (2008) for examples of data obtained from lectin microarrays for both spot binding, as well as binary data analysis.
Duration of the assay can be separated into three segments: fabrication, experimental, and data acquisition. The time necessary for the fabrication process will depend on the equipment used, the size of the printing batch, and the number of lectins used per slide. Minimally, the overnight incubation requires that the preparation of the lectin chip will take one day. The experimental stages of the lectin microarray will vary in time based on the number of samples used. For two to six samples, expect the slide blocking, cell labeling, and microarray probing to consume ~2.5 hr. Lastly, the length of time for data acquisition and analysis will vary based on the equipment used and the number of slides to be analyzed. Scanning at 5-μm resolution, each slide should take ~30 min. Time duration for data analysis is entirely dependent on the user.
This work was supported by grants from the NIH AI 44129, CA 108835, & P01 AI072677 to J.P. Schneck, U54RR020839-01 to H. Zhu, and the State Key Development Program for Basic Research of China (grant 2010 B 529205) & SRF for ROCS, SEM to S.C. Tao. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health. Yu Li and Sheng-ce Tao contributed equally to this article.
Current Protocols in Protein Science 12.9.1-12.9.7, February 2011