Displacement Phenomena in Lectin Affinity Chromatography

Article pubs.acs.org/ac

Displacement Phenomena in Lectin Affinity Chromatography Wonryeon Cho* Department of Chemistry, Wonkwang University, 460 Iksandae-ro, Iksan, Jeonbuk 570-749, Republic of Korea S Supporting Information *

ABSTRACT: The work described here examines displacement phenomena that play a role in lectin affinity chromatography and their potential to impact reproducibility. This was achieved using Lycopersicon esculentum lectin (LEL), a lectin widely used in monitoring cancer. Four small identical LEL columns were coupled in series to form a single affinity chromatography system with the last in the series connected to an absorbance detector. The serial affinity column set (SACS) was then loaded with human plasma proteins. At the completion of loading, the column set was disassembled, the four columns were eluted individually, the captured proteins were trypsin digested, the peptides were deglycosylated with PNGase F, and the parent proteins were identified through mass spectral analyses. Significantly different sets of glycoproteins were selected by each column, some proteins appearing to be exclusively bound to the first column while others were bound further along in the series. Clearly, sample displacement chromatography (SDC) occurs. Glycoproteins were bound at different places in the column train, identifying the presence of glycoforms with different affinity on a single glycoprotein. It is not possible to see these phenomena in the single column mode of chromatography. Moreover, low abundance proteins were enriched, which facilitates detection. The great advantage of this method is that it differentiates between glycoproteins on the basis of their binding affinity. Displacement phenomena are concluded to be a significant component of the separation mechanism in heavily loaded lectin affinity chromatography columns. This further suggests that care must be exercised in sample loading of lectin columns to prevent analyte displacement with nonretained proteins.

R

separation methods that select disease-specific differences in carbohydrate structure which enable biomarker purification from complex mixtures. This problem is exacerbated by the fact that glycoproteins exist in multiple glycoforms of which a small set or even a single glycan identifies a disease. Reproducibility in recognizing a disease-specific glycoform of a glycoprotein at low abundance is a major challenge. One of the more widely used means of purifying glycoprotein glycoforms from complex mixtures is using lectin affinity chromatography. Lectins, such as Lycopersicon esculentum lectin (LEL), are widely known to bind structure specifically to oligosaccharides conjugated to proteins.21 LEL, for example, binds specifically to β-(1 → 4)-linked galactosyl residues in poly-N-acetyllactosamine and GlcNAc residues in N-acetylglucosamine (GlcNAc)n, where n ranges from 1 to 4. The structure and number of residues in a glycan can play a role in affinity selection. For example, in the series (GlcNAc β1−4)1, (GlcNAc β1−4)2, (GlcNAc β1−4)3, to (GlcNAc β1−4)4, the binding affinity of LEL from tomato increases with the size and number of the N-acetylglucosamine oligomers bound to a glycan.22 This allows LEL to differentiate between glycoforms.21 In the case of

ecent advances in high resolution tandem mass spectrometry (MS/MS) are rapidly pushing proteomics toward the analysis of post-translational modifications.1 Even so, there is the problem that biological samples are so complex that they overwhelm the analytical capacity of mass spectrometers.2,3 This problem is addressed more frequently with structurespecific affinity methods that select a small family of proteins. The great advantage of this approach is that sample complexity is reduced enormously in a single step. Although affinity selection provides high levels of analyte enrichment and fractionation, there is still the question of whether these methods have the requisite reproducibility to be used in diagnostics. The work described here focuses on affinity selection of glycoproteins and the potential for high selectivity purification. A post-translational modification (PTM), particularly when aberrant, can be a powerful tool in recognizing the states of diseases.4−9 There is much less overlap between the healthy and the disease states with structure-specific biomarkers than those involving changes in protein concentration alone. With glycoproteins, for example, cystic fibrosis,10 respiratory illnesses,11 diabetes,12 certain types of heart disease,13 Alzheimer’s disease,14 stress,15 renal function diseases,16 arthritis,17 cancer,6 some autoimmune diseases,18 and cellular adhesion related diseases19,20 are all associated with structural deviations in glycosylation. The analytical problem in this approach is to find © 2015 American Chemical Society

Received: January 31, 2015 Accepted: September 7, 2015 Published: September 8, 2015 9612

DOI: 10.1021/acs.analchem.5b00790 Anal. Chem. 2015, 87, 9612−9620

Article

Analytical Chemistry

Figure 1. Scheme of serial affinity column set (SACS) experiment with LEL1 → LEL2 → LEL3→ LEL4 serial columns. A pooled healthy human plasma was loaded directly onto serially connected four soft-gel affinity columns with mobile phase A (0.10 M HEPES buffer, pH 7.5 containing 1 mM CaCl2 and 1 mM MgCl2) at a flow rate of 0.3 mL/min. After washing, the SACS columns were disassembled, and captured proteins were eluted from each of the LEL columns. Then all the fractions were trypsin digested and deglycosylated with PNGase F. The parent proteins from the peptide fractions were identified through mass spectral analysis as described in LC-MS/MS-Based Protein Identification.

opportunity to bind. This suggests that the glycoproteins isolated from a lectin column could vary qualitatively as a function of column size and the extent of sample loading. The objective of the work described here was to examine the possibility that displacement phenomena might play a role in the binding and release of glycoproteins in lectin affinity chromatography, the reason being that displacement could play a role in the reproducibility of this method.

cancer, unique ligands are biosynthesized and attached to the glycan that is disease-specific.22 Subsequent to selection by an immobilized lectin, captured glycoproteins (or glycopeptides) can be displaced from sorbents with monosaccharides or glycan displacing agents. Displacing agents vary widely in their affinity for sorbent matrixes and the requisite concentration to achieve displacement.23 In displacement chromatography, displacing agents of intermediate binding affinity will cause the elution of weakly bound analytes while those of strong binding affinity remain adsorbed. The fact that glycoproteins can be displaced from a lectin column by monosaccharides at high concentration brings up the question of whether glycoprotein glycoforms can displace each other. A glycoprotein can be viewed as a glycan conjugated to a polypeptide. The ability of a glycan in a glycoprotein to associate with a lectin should not vary widely between proteoforms with minor structural changes in the protein backbone. In addition, glycoproteins bearing these glycans can differ widely in concentration. Some glycoforms occur at much higher concentration than others. At high glycoprotein concentrations in overloaded lectin columns, there is a strong possibility of glycoform displacement. The significance of this in clinical diagnostics would be that at high loading, weakly bound glycoproteins might not have the

■

MATERIALS AND METHODS Materials and Chemicals. Agarose-bound Lycopersicon esculentum lectin (LEL) sorbent was purchased from Vector Laboratories (Burlingame, CA). Normal pooled plasma from 100 human subjects was generously supplied by the National Institute of Standards and Technology (NIST, Gaithersburg, MD). Acetic acid, sodium hydroxide, formic acid, calcium chloride, magnesium chloride, and HPLC grade acetonitrile (ACN) were purchased from Mallinckrodt Chemicals (Phillipsburg, NJ). Ammonium bicarbonate, glycine, manganese chloride, proteomics grade N-p-tosylphenylalanine chloromethyl ketone (TPCK)-treated trypsin, N-α-tosyl-L-lysine chloromethyl ketone hydrochloride (TLCK), 4-(2-hydroxyethyl)-1piperazine ethanesulfonic acid (HEPES), iodoacetic acid (IAA), and L-cysteine were obtained from Sigma-Aldrich (St. Louis, MO). Dithiothreitol (DTT) and urea were provided by Bio9613


Article


Table 1. Glycoprotein Lists Identified from the Combination of SACS with N- and O-Glycosylation: LEL1→ LEL2→ LEL3→ LEL4 SACS no.

LEL 1 LEL 2 LEL 3

LEL 4 O O O O O O O O

group

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

O O O O O O O O O O O O O O O O O O O O



21 22

O O

O O

O O

D(27) D(27)

P01009|A1AT_HUMAN P01023|A2MG_HUMAN P04114|APOB_HUMAN P10909|CLUS_HUMAN P00738|HPT_HUMAN P02790|HEMO_HUMAN P02743|SAMP_HUMAN P27169|PON1_HUMAN P02765|FETUA_HUMAN P02647|APOA1_HUMAN P02656|APOC3_HUMAN P05090|APOD_HUMAN P02649|APOE_HUMAN O14791|APOL1_HUMAN P04003|C4BPA_HUMAN P02747|C1QC_HUMAN P00736|C1R_HUMAN P09871|C1S_HUMAN P01024|CO3_HUMAN P0C0L5|CO4B_HUMAN; P0C0L4| CO4A_HUMAN P08603|CFAH_HUMAN Q9BXR6|FHR5_HUMAN

23 24 25 26 27 28 29 30 31 32 33

O O O O O O O O O O O



D(27) D(27) D(27) D(27) D(27) D(27) D(27) D(27) D(27) D(27) D(27)

P02675|FIBB_HUMAN P02679|FIBG_HUMAN P23142|FBLN1_HUMAN P69905|HBA_HUMAN P68871|HBB_HUMAN P01876|IGHA1_HUMAN P01857|IGHG1_HUMAN P01859|IGHG2_HUMAN P01871|IGHM_HUMAN P01591|IGJ_HUMAN P80108|PHLD_HUMAN

34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56

O O O O O O O O O O O O

O O O O O O

O O

D(27) D(27) D(27) D(27) D(27) D(27) E(2) E(2 A(4) A(4) A(4) A(4) B(12) B(12) B(12) B(12) B(12) B(12) B(12) B(12) B(12) B(12) F(2) in B (12)

P55058|PLTP_HUMAN P02760|AMBP_HUMAN O75882|ATRN_HUMAN Q9UGM5|FETUB_HUMAN Q15485|FCN2_HUMAN Q08380|LG3BP_HUMAN P01877|IGHA2_HUMAN P04220|MUCB_HUMAN Q14624|ITIH4_HUMAN P51884|LUM_HUMAN P27918|PROP_HUMAN P02787|TRFE_HUMAN P01011|AACT_HUMAN P02763|A1AG1_HUMAN P02671|FIBA_HUMAN P02751|FINC_HUMAN O75636|FCN3_HUMAN Q9Y6R7|FCGBP_HUMAN P01042|KNG1_HUMAN P48740|MASP1_HUMAN P08571|CD14_HUMAN P04275|VWF_HUMAN P22792|CPN2_HUMAN

O O


O O O O O O O O O

O

O

C(8) C(8) C(8) C(8) C(8) C(8) C(8) C(8) D(27) D(27) D(27) D(27) D(27) D(27) D(27) D(27) D(27) D(27) D(27) D(27)

Swiss-prot accession no. and entry name

9614

glyco sylation

names alpha-1-antitrypsin alpha-2-macroglobulin apolipoprotein B-100 clusterin haptoglobin hemopexin serum amyloid p-component serum paraoxonase/arylesterase 1 alpha-2-HS-glycoprotein apolipoprotein A-I apolipoprotein C-III apolipoprotein D apolipoprotein E apolipoprotein L1 C4b-binding protein alpha chain complement C1q subcomponent subunit C complement C1r subcomponent complement C 1s subcomponent complement C3 complement C4-B

N N N N N N, O N N N, O N O N N, O N N O N N N N

complement factor H complement factor H-related protein 5 SV = 1 fibrinogen beta chain fibrinogen gamma chain fibulin-1 hemoglobin subunit alpha hemoglobin subunit beta Ig alpha-1 chain C region Ig gamma-1 chain C region Ig gamma-2 chain C region Ig mu chain C region immunoglobulin J chain phosphatidylinositol-glycan-specific phospholipase D phospholipid transfer protein protein AMBP attractin fetuin-B ficolin-2 galectin-3-binding protein Ig alpha-2 chain C region Ig mu heavy chain disease protein interalpha-trypsin inhibitor heavy chain H4 lumican properdin serotransferrin alpha-1-antichymo trypsin alpha-1-acid glycoprotein 1 fibrinogen alpha chain fibronectin ficolin-3 IgGFc-binding protein kininogen-1 mannan-binding lectin serine protease 1 monocyte differentiation antigen CD14 von Willebrand factor carboxypeptidase N subunit 2

N N N N N N N N, O N N N N N N N, N N N N N N N, N N, N, N N N N, N N N, N N, N, N

O

O O, C O

O

O O O


Article

Analytical Chemistry Table 1. continued no.

LEL 1 LEL 2 LEL 3

LEL 4

57

O

O

group F(2) in B (12)

Swiss-prot accession no. and entry name P18428|LBP_HUMAN

glyco sylation

names lipopolysaccharide-binding protein

N

Table 2. Nonglycoprotein Lists Identified from the Combination of SACS: LEL1→ LEL2→ LEL3→ LEL4 SACS no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14

LEL 1 LEL 2 LEL 3 O O O O O O O O O O O O O O

O O O O O O O O

O O O O O O O O O O

LEL 4 O O

O O

Swiss-prot accession no. and entry name P02652|APOA2_HUMAN P02768|ALBU_HUMAN O43866|CD5L_HUMAN P00739|HPTR_HUMAN P01766|HV305_HUMAN P01834|IGKC_HUMAN P0CG05|LAC2_HUMAN P01602|KV110_HUMAN P02746|C1QB_HUMAN P04264|K2C1_HUMAN P13645|K1C10_HUMAN P0CG04|LAC1_HUMAN P01764|HV303_HUMAN P18136|KV313_HUMAN; P18135| KV312_HUMAN; P01620|KV302_HUMAN

names apolipoprotein A-II serum albumin CD5 antigen-like haptoglobin-related protein Ig heavy chain V-III region BRO Ig kappa chain C region Ig lambda-2 chain C regions Ig kappa chain V−I region HK102 (fragment) complement C1q subcomponent subunit B keratin, type II cytoskeletal 1 keratin, type I cytoskeletal 10 Ig lambda-1 chain C regions Ig heavy chain V-III region VH26 Ig kappa chain V−III region HIC Ig kappa chain V-III region HAH Ig kappa chain V-III region SIE

Concentrator after being adjusted to pH 7.5 with a 0.5 M NH4HCO3 buffer. Samples were then reconstituted with 8 M urea in a 50 mM HEPES buffer containing 10 mM CaCl2 and reduced with 10 mM DTT for 2 h of incubation at 50 °C. Following reduction, alkylation was achieved through addition of iodoacetic acid (IAA) to a final concentration of 20 mM and then the solutions were incubated in darkness for an additional 2 h. IAA was quenched by addition of L-cysteine to the reaction mixture to a final concentration of 40 mM, and the mixture was incubated for 30 min at room temperature. After dilution with a 50 mM HEPES buffer to a final urea concentration of 1 M, proteomics grade trypsin (2%, w/w, enzyme to protein) was added and the solution was incubated overnight at 37 °C. The trypsin digestion reaction was stopped by the addition of TLCK (trypsin/TLCK ratio of 1:1 (w/w)), and the resulting peptide mixture was desalted with HLB Oasis SPE cartridges, dried to 50−100 μL with the Centrivap Concentrator, and adjusted to pH 7.5 with the 0.5 M NH4HCO3 buffer. PNGase F Digestion. N-Linked glycopeptides in digested peptide mixtures were deglycosylated by the treatment with PNGase F. A 50 mM ammonium bicarbonate buffer was added to the desalted tryptic peptides, which adjusted the pH to 7.0− 8.0. Five microliters (2500 U) of PNGase F was added to the pH-adjusted tryptic peptides and the mixture incubated overnight at 37 °C. Following deglycosylation, samples were desalted using HLB Oasis SPE cartridges and concentrated using a C18 microspin column. The PNGase F-treated peptide mixtures were reconstituted in 0.1% formic acid solution and stored at −80 °C until further analysis with an LTQ-Orbitrap mass spectrometer. LC-MS/MS-Based Protein Identification. Proteins were identified with an LTQ-Orbitrap hybrid mass spectrometer (Thermo Fisher, Bremen, Germany). The peptide mixtures resulting from PNGase F treatment were separated on an Agilent 1100 HPLC system using a 75 μm × 120 mm C18 RPC column packed with 5 μm C18 Magic beads. RPC separations were achieved at a flow rate of 300 nL/min using a 60 min linear mobile phase gradient from 98% solvent A with 2%

Rad Laboratories (Hercules, CA). PNGase F (glycerol free) was purchased from New England BioLabs (Ipswich, MA). The C18 microspin column was obtained from The Nest Group, Inc. (Southborough, MA). HLB Oasis SPE cartridges were provided by Waters (Milford, MA). The DI water system was purchased from Millipore (Boston, MA). The Centrivap Concentrator was purchased from Labconco, Corp. (Kansas City, MO). Serial Affinity Column Set (SACS). Agarose-bound LEL sorbent was individually self-packed in four 4.6 mm × 50 mm columns. LEL was agarose-bound at the level of 2 mg/mL. The column volume used in these studies was 0.83 mL, suggesting that the column should contain 1.66 mg of bound LEL. But agarose is compressed slightly on packing. This means that the amount of LEL in each column was slightly larger than 1.66 mg. Protein concentration in healthy human plasma was estimated using the Bradford assay to prepare samples with equal amounts of total proteins. For the SACS experiment, the four LEL columns were connected in series. Human plasma from healthy subjects was loaded directly onto a serial column set of soft-gel affinity columns. Mobile phase A (0.10 M HEPES buffer, pH 7.5 containing 1 mM CaCl2 and 1 mM MgCl2) was pumped at a flow rate of 0.3 mL/min. Following extensive washing with mobile phase A to remove nonspecifically and weakly bound proteins, the serial LEL affinity column set was disassembled, and affinity-selected proteins were eluted from each column individually with solution B (0.5 M acetic acid− HCl solution, pH 2.5). This experimental scheme is illustrated in Figure 1. Elution curves were obtained with a 20 AD LC absorbance detector from Shimadzu Scientific Instruments, Inc. (Kyoto, Japan) operated at 280 nm. Saturation Test of Columns. LEL column loading and saturation were determined as previously described4−8,24,25 by increasing sample injections ranging from 25 to 250 μg in increments of 25 μg. Column saturation was achieved at approximately 200 μg of human plasma (data not shown). Trypsin Digestion. Captured proteins were desalted using HLB Oasis SPE cartridges and dried with a Centrivap 9615


Article


Analytical Protocol. One of the ways to observe ligand displacement in chromatography is that subsequent to sample loading, a ligand with a very high association constant is added continuously in the mobile phase. This ligand adsorbs at the inlet of the column and displaces more weakly bound ligands in front of it, setting up a displacement train. An alternative approach is the multiple column sample displacement chromatography (SDC) mode where sample components displace each other.28 Although SDC has been observed with reversed phase and ion exchange chromatography,29 it has not been reported with lectin affinity chromatography. In SDC, the sample is loaded onto a long column composed of a series of segments that can be disassembled and eluted separately. A system capable of SDC was chosen to explore the possibility of glycoprotein displacement in lectin affinity chromatography. SACS of four identical LEL columns (4.6 mm × 50 mm) were coupled in series and loaded as a single column in the order of LEL1→ LEL2→ LEL3→ LEL4, respectively. The LEL4 column was connected to an absorbance detector that monitored elution at 280 nm to minimize perturbation of the baseline during protein desorption. Loading capacity was determined by applying healthy pooled plasma from NIST to the LEL column in 25 μg increments followed by washing at a mobile phase velocity of 0.3 mL/min to preclude resin compression. The amount of proteins required to saturate the LEL column was estimated, using absorbance at 280 nm, by assuming that the molar extinction coefficients of all proteins were equal to that of a bovine serum albumin calibration mixture. Less than 0.1% of the plasma proteome was selected by the LEL column based on a 280 nm absorbance. After loading human plasma to this SACS, the serial LEL column set was washed with 0.10 M HEPES buffer, pH 7.5, containing 1 mM CaCl2 and 1 mM MgCl2. After extensive washing to remove nonspecifically and weakly bound proteins, the SACS was disassembled and affinity-selected proteins were eluted from each column with an acidic mobile phase (0.5 M acetic acid−HCl solution, pH 2.5) at a rate of 0.3 mL/min. Subsequent to elution, lectin-selected proteins were identified in a series of steps involving trypsin digestion of reduced and alkylated proteins, PNGase F deglycosylation, further fractionation of tryptic and PNGase F digests by RPC, and electrospray ionization and mass spectral analysis of peptides by tandem mass spectrometry (ESI-MS/MS). The minimum criterion for the identification of captured proteins from each column was the presence of two or more nonglycosylated or deglycosylated peptides in a sample at a confidence level of 99% or above. Protein Identification. A total of 71 proteins were identified by the LEL1→ LEL2→ LEL3 → LEL4 SACS, of which 57 were glycoproteins (Table 1). The total list of proteins is located in Table S-1 of the Supporting Information. Fourteen out of the 71 proteins captured by the LEL columns have not been reported to be glycoproteins (Table 2). Human serum albumin (HSA) is glycated but was not considered in the glycoprotein set because LEL has not been reported to bind glycated proteins.30,31 Capture of nonglycoproteins by a lectin column would seem to be anomalous, but it is relatively common for nonglycosylated proteins to be associated with other proteins through either biospecific or nonspecific protein−protein interaction (PPI) mechanisms. Among the 57 glycoproteins in Table 1, 51 glycoproteins have been

solvent B to 60% solvent A with 40% solvent B. Solvent A was composed of DI water to which formic acid had been added to a concentration of 0.1%. Solvent B was prepared with ACN to which formic acid had been added to a concentration of 0.1%. The electrospray ionization emitter tip was generated on the prepacked column with a laser puller (model P-2000, Sutter Instrument Co.). The HPLC system was coupled directly to the LTQ-Orbitrap hybrid mass spectrometer equipped with a nanoelectrospray ion source (Proxeon Biosystems, Odense, Denmark). The MS was operated in the data-dependent mode, in which a survey full scan MS spectrum (from m/z 300 to 1600) was acquired in the Orbitrap with a resolution of 60 000 at m/z 400. This was then followed by MS/MS scans of the three most abundant ions with +2 or +3 charge states. Target ions already selected for MS/MS were dynamically excluded for 180 s. The resulting fragment ions were recorded in the linear ion trap. Automated MS/MS data analysis was performed utilizing Protein Pilot software 4.0 using the Pro Group algorithm (ABI) for protein identification. The minimum acceptance criterion for peptide identification was a 99% confidence level. Most of the proteins were identified based on the presence of at least two peptides from a protein identified by the Pro Group algorithm at the 99% confidence level. An unused score cutoff of 4 was the minimum value for identifying proteins with the Protein Pilot 4.0 Software. Proteins are listed in Table 1, Table 2, and Table S-1(Supporting Information) according to their Swiss-Prot entry names and accession numbers.

■

RESULTS Theory. The term affinity chromatography is used to designate the reversible, structure specific binding of a ligand (L) by an immobilized receptor (∼R) on the surface of a supporting material, which is represented as kon

∼R + L XooY ∼RL koff

where kon and koff are the on- and off-rate constants of binding, respectively. At equilibrium, the rates of ligand binding and dissociating from the receptor will be equal, meaning that kon[∼R][L] = koff [∼RL]

and

Ka

kon [∼RL] = koff [∼R][L]

where Ka is the association constant. This equilibrium may also be expressed in terms of a dissociation constant Kd, where Kd = 1/Ka. Whereas the association constant of biotin with avidin is 1015 and that of most antibodies with antigens ranges from 106 to 109, lectins bind to mono- and oligosaccharides with much lower affinity; association constants generally ranging from 102 to 106.26,27 Whereas the off-rate is very low with biotin and antibodies, it is much higher with lectins. In affinity chromatography, ligands with a high association constant will displace those of lower binding affinity in a competitive type of separation. Moreover, this would be most easily observed in a column that is heavily loaded where there is competition for binding sites. For competition-based separations to be achieved with lectins, the difference in binding constants of various glycoforms must be sufficiently large to set up a displacement train in which a strongly bound glycoprotein pushes more weakly bound species ahead of it. Immunoaffinity chromatography has shown that there must be a 10−100 fold difference in binding affinity to achieve displacement type separations 9616


a

type

only LEL1 not LEL1 but LEL2 LEL1, LEL2, LEL3, and LEL4 LEL1, LEL2, and LEL3 LEL1 and LEL4 LEL2 and LEL4

2

2

27

8

12

4

no. of glycoproteins

56−57

40−41

9−35

1−8

46−57

42−45

protein no. inTable1 glycoprotein list

carboxypeptidase N subunit 2, lipopolysaccharide-binding protein

alpha-2-HS-glycoprotein, apolipoprotein A-I, apolipoprotein C−III, apolipoprotein D, apolipoprotein E, apolipoprotein L1, C4b-binding protein alpha chain, complement C1q subcomponent subunit C, complement C1r subcomponent, complement C 1s subcomponent, complement C3, complement C4-B, complement factor H, complement factor H-related protein 5 SV = 1, fibrinogen beta chain, fibrinogen gamma chain, fibulin-1, hemoglobin subunit alpha, hemoglobin subunit beta, Ig alpha-1 chain C region, Ig gamma-1 chain C region, Ig gamma-2 chain C region, Ig mu chain C region, immunoglobulin J chain, phosphatidylinositol-glycan-specific phospholipase D, phospholipid transfer protein, protein AMBP Ig lapha-2 chain C region, Ig mu heavy chain disease protein, carboxypeptidase N subunit 2, lipopolysaccharide-binding protein

alpha-1-antitrypsin, alpha-2-macroglobulin, apolipoprotein B-100, clusterin, haptoglobin, hemopexin, serum amyloid P-component, serum paraoxonase/arylesterase 1

alpha-1-antichymotrypsin, alpha-1-acid glycoprotein 1, fibrinogen alpha chain, fibronectin, ficolin-3, IgGFc-binding protein, kininogen-1, mannan-binding lectin serine protease 1, monocyte differentiation antigen CD14, von Willebrand factor, carboxypeptidase N subunit 2, lipopolysaccharide-binding protein

interalpha-trypsin inhibitor heavy chain H4, lumican, properdin, serotransferrin

The Venn diagram of the captured glycoproteins is represented in Figure 2 according to grouping.

group F

group E

group D

group C

group B

group A

capture columns

Table 3. Grouping of the Glycoproteins Captured from LEL1→ LEL2→ LEL3→ LEL4 SACSa

Analytical Chemistry Article

9617


Article


only the first three LEL columns. The fact that the group D proteins were bound to all four columns in the LEL tandem column set means that the LEL2 and LEL3 columns were saturated as well with those proteins. Also, it was concluded above that there is evidence of competitive binding on LEL1 and that proteins bound to LEL1 had affinity for LEL higher than that of group B proteins. This suggests that any protein bound to LEL1 exclusively should bind before group B proteins on downstream LEL columns. However, this was not the case. Multiple proteins were bound to LEL1 and also to LEL4, behind the group B proteins that bound to LEL2. This is very important. “Von Willebrand factor” protein was the only protein that bound exclusively to LEL2. Failure to bind sequentially suggests that proteins bound to LEL1 and LEL4 (that is group E in Table 1, Table 3, and Figure 2) or LEL2 and LEL4 (designated as group F in Table 1, Table 3, and Figure 2) probably carry glycans that differ in structure and LEL binding affinity. This explains why the groups E and F glycoproteins (Table 1, Table 3, and Figure 2) bind nonsequentially in the LEL train.

reported to be involved in protein−protein interactions.32 Analysis of the 14 nonglycoproteins in Table 2 using software such as Metacore, Ingenuity Pathway Analysis (IPA), Intact, and String that identify protein−protein association (data not shown) lead to the conclusion that these 14 nonglycoproteins were retained by the LEL columns because they form protein− protein complexes with glycoproteins. Among the 57 glycoproteins selected, 45 glycoproteins (79% of the total 57) were captured by the LEL1 column, while 4 (7% of the total 57) defined as group A (proteins 42−45 in Table 1, Table 3, and Figure 2) bound exclusively to the LEL1

■

DISCUSSION Glycosylation variants are generally separated independently using multiple lectin chromatography columns that differ in selectivity, each column being eluted with a denaturing mobile phase that produces a single chromatographic peak. This study has shown that at least in the case of LEL, it is possible to separate glycoproteins from human plasma into multiple components with a single lectin by placing several, short columns in series and overloading the first few columns in the set. Like ion exchange and reversed phase chromatography, overloading columns causes a displacement train to be established along the length of the column in which proteins of lower binding affinity are displaced by those of higher binding affinity. This displacement phenomenon and the use of displacing agents to push ligands from a column were first described by Horvath many years ago.33,34 When the chromatography column is broken into a series of small segments that are eluted separately, the need for a displacing agent is circumvented as described by Hodges,28 albeit with some loss of resolution relative to the Horvath method. The great advantage of Hodges’s sample displacement chromatography (SDC) method is that there is no need for specialized displacers. Also, SDC is quicker. With LEL, glycoproteins of higher off-rates are displacing those of lower binding affinity during column loading. Although not demonstrated in this work, it is highly likely that this phenomenon is shared by all lectins. The fact that glycoproteins can undergo SDC on lectin columns has multiple ramifications. One is in analytical lectin chromatography with small columns. Overloading columns will cause the displacement of weakly bound species from columns. This will make the concentration ratio of ligands captured by the column a function of sample loading, compromising reproducibility. The general rule would be that lectin columns should never be overloaded when being used in the analytical mode. A second implication is that subsequent to loading, the elution position of a glycoprotein in sample displacement chromatography will be a function of the amount of glycoproteins in the sample having a higher binding affinity for the lectin. As the concentration of higher affinity glycoproteins increases in a sample, those of lower affinity

Figure 2. The Venn diagram of captured glycoproteins according to grouping. Bold boundary lines represent the four columns with red, blue, green, and black being the LEL1, LEL2, LEL3, and LEL4 columns, respectively. Captured glycoproteins are shown with colorcoded areas in this diagram while protein groups A−F are represented with light blue, yellow, orange, gray, and pink areas subsequently. The grouping is also listed in Table 3. The numbers in parentheses represent the numbers of identified glycoproteins from lectin columns and groups on LEL1→ LEL2→ LEL3→ LEL4 SACS experiment.

affinity column. Another 12 (21% of the total) defined as group B (proteins 46−57 in Table 1, Table 3, and Figure 2) failed to bind to the LEL1 column but bound to the LEL2 column. Also, in group B, one protein bound exclusively to LEL2 while the remaining 11 proteins bound to the other downstream columns as well (yellow areas in Figure 2). These results are interpreted as follows. First, it is concluded that binding to the serial affinity column set must occur sequentially in the order LEL1 → LEL2 → LEL3 → LEL4. Second, binding sites on LEL1 were concluded to be saturated based on the fact that proteins were bound to the rest of the columns in the lectin column train. Third, it is further concluded that there was competitive binding with displacement on the LEL1 column based on the failure of group B glycoproteins to bind to the first column (LEL1). Clearly the glycoproteins in group A have a higher affinity for LEL1 than those in group B. Fourth, the fact that no group A proteins were found on LEL2 means that they occurred in low abundance and were not displaced by any of the other 41 proteins bound to LEL1. This further suggests that group A proteins are among those bearing the highest affinity ligands (glycans) in the mixture. Fifth, the group A proteins did not saturate the binding sites on LEL1. Lastly, the sixth conclusion is that the rest of the 41 proteins bound to LEL1 had a LEL binding affinity higher than that of the group B proteins. This implies that all the proteins bound to LEL1 have a LEL binding affinity higher than that of the group B proteins. It was also noted that eight proteins defined as group C (proteins 1−8 in Table 1, Table 3, and Figure 2) bound to all four columns while another 27 proteins defined as group D (proteins 9−35 in Table 1, Table 3, and Figure 2) bound to 9618



■

will be displaced further along the affinity column train. Affinity-selected glycoproteins will not always be found at the same elution position in sample displacement chromatography. A third ramification seen in these studies is that a single glycoprotein can be found at multiple positions in the SDC train when it bears glycans that differ in binding affinity. With LEL, this would most likely be in the order (GlcNAc β1−4)4, (GlcNAc β1−4)3, (GlcNAc β1−4)2, and (GlcNAc β1−4)1 for Nacetylglucosamine oligomers. High affinity glycans on a protein will displace glycoforms of lower binding affinity and higher offrates. This suggests that the study of protein glycoforms would be achieved best by using a combination of an antibody targeting the polypeptide portion of the molecule to isolate all the members of a glycoform family and a lectin to differentiate between glycan family members. Another ramification is that SDC can fail to differentiate between glycoforms of proteins with multiple glycosylation sites when a high affinity glycan is conjugated to a protein at one site and a low affinity glycan at another site in the same protein. The high affinity glycan will dominate binding of the protein to the lectin column and the low affinity glycan will be unrecognized. This problem can be circumvented by trypsin digestion. SDC lectin affinity chromatography of a trypsin digest will select peptides bearing glycans based on their affinity for the lectin. But it is impossible to determine whether these glycans were on the same glycoform or different glycoforms of the protein. The only way to address this problem would be by carrying out a lectin-based SDC separation at the glycoprotein level, trypsin digesting the individual fractions, and then executing a peptide level SDC separation, which would reveal the presence of both high and low affinity ligands in the same glycoprotein fraction. Lastly, it should be possible to isolate high affinity glycoforms exclusively by adding an intermediate affinity glycan to displace all glycoproteins of lower binding affinity subsequent to initial sample loading. This exploits both SDC and conventional displacement chromatography.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (grant number 2013R1A1A3012431).

■

REFERENCES

(1) Gajadhar, A. S.; White, F. M. Curr. Opin. Biotechnol. 2014, 28, 83−87. (2) Cech, N. B.; Enke, C. G. Mass Spectrom. Rev. 2001, 20, 362−387. (3) Smith, R. Int. J. Mass Spectrom. 2000, 200, 509−544. (4) Cho, W.; Jung, K.; Regnier, F. E. Anal. Chem. 2008, 80, 5286− 5292. (5) Jung, K.; Cho, W.; Regnier, F. E. J. Proteome Res. 2009, 8, 643− 650. (6) Cho, W.; Jung, K.; Regnier, F. E. J. Proteome Res. 2010, 9, 5960− 5968. (7) Cho, W.; Jung, K.; Regnier, F. E. J. Sep. Sci. 2010, 33, 1438−1447. (8) Drake, P. M.; Schilling, B.; Niles, R. K.; Prakobphol, A.; Li, B.; Jung, K.; Cho, W.; Braten, M.; Inerowicz, H. D.; Williams, K.; Albertolle, M.; Held, J. M.; Iacovides, D.; Sorensen, D. J.; Griffith, O. L.; Johansen, E.; Zawadzka, A. M.; Cusack, M. P.; Allen, S.; Gormley, M.; Hall, S. C.; Witkowska, H. E.; Gray, J. W.; Regnier, F.; Gibson, B. W.; Fisher, S. J. J. Proteome Res. 2012, 11, 2508−2520. (9) Dwek, M. V.; Ross, H. A.; Leathem, A. J. Proteomics 2001, 1, 756−762. (10) Kirkham, S.; Sheehan, J.; Knight, D.; Richardson, P.; Thornton, D. Biochem. J. 2002, 361, 537−546. (11) Hilgendorff, A.; Doerner, M.; Rawer, D.; Leick, J.; Trotter, A.; Ebsen, M.; Ruppert, C.; Günther, A.; Gortner, L.; Reiss, I. Crit. Care Med. 2006, 34, 203−210. (12) Kim, M.-R.; Yu, S.-A.; Kim, M.-Y.; Choi, K. M.; Kim, C.-W. Biotechnol. Bioprocess Eng. 2014, 19, 83−92. (13) Trobec, K.; Kerec Kos, M.; von Haehling, S.; Springer, J.; Anker, S. D.; Lainscak, M. PLoS One 2013, 8, e79603. (14) Gauthier-Kemper, A.; Weissmann, C.; Golovyashkina, N.; SeböLemke, Z.; Drewes, G.; Gerke, V.; Heinisch, J. J.; Brandt, R. J. Cell Biol. 2011, 192, 647−661. (15) Söndergaard, H. P.; Kushnir, M. M.; Aronsson, B.; Sandstedt, P.; Bergquist, J. BMC Res. Notes 2012, 5, 186. (16) Šebeková, K.; Blažíček, P.; Syrova, D.; Krivošíková, Z.; Spustova, V.; Heidland, A.; Schinzel, R. Kidney Int. 2001, 59, S58−S62. (17) Cretu, D.; Diamandis, E. P.; Chandran, V. Crit. Rev. Clin. Lab. Sci. 2013, 50, 51−63. (18) Sato, M.; Toyozaki, T.; Odaka, K.; Uehara, T.; Arano, Y.; Hasegawa, H.; Yoshida, K.; Imanaka-Yoshida, K.; Yoshida, T.; Hiroe, M.; Tadokoro, H.; Irie, T.; Tanada, S.; Komuro, I. Circulation 2002, 106, 1397−1402. (19) Schwartz, K. J.; Peck, R. F.; Bangs, J. D. Eukaryotic Cell 2013, 12, 132−141. (20) Nakamura, K.; Sakagami, H.; Asanuma-Date, K.; Nagasawa, N.; Nakahara, Y.; Akiyama, H.; Ogawa, H. Carbohydr. Res. 2013, 382, 77− 85. (21) Fakla, I.; Hever, A.; Molnár, J.; Fischer, J. Anticancer Res. 1998, 18, 3107−3111. (22) Yamamoto, K.; Qi, W.-M.; Yokoo, Y.; Miyata, H.; Udayanga, K. G. S.; Kawano, J.; Yokoyama, T.; Hoshi, N.; Kitagawa, H. J. Vet. Med. Sci. 2010, 72, 1119−1127. (23) Hage, D. S.; Anguizola, J. A.; Bi, C.; Li, R.; Matsuda, R.; Papastavros, E.; Pfaunmiller, E.; Vargas, J.; Zheng, X. J. Pharm. Biomed. Anal. 2012, 69, 93−105.

■

CONCLUSIONS It is concluded that in addition to ion exchange and reversed phase chromatography, proteins can be fractionated by sample displacement chromatography using LEL serial affinity chromatography set (SACS) to select glycans according to their binding affinity. The great advantage of this approach is that it allows the fractionation of glycoforms using single lectin SACS. This will be particularly useful in determining biomarkers in which the disease-specific feature is a unique glycan, or a group of glycans. But biomarker-based diagnostics is generally associated with quantification. This leads to the second conclusion associated with this work. Because reversible ligand binding is common to all lectins, displacement will occur in all types of lectin affinity chromatography, and reproducibility will be compromised by overloading. This is a particular concern with small columns and lectin arrays that have low surface area and limited amounts of immobilized lectin.

■

Article

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.5b00790. Additional information as noted in the text (PDF) 9619


Article

Analytical Chemistry (24) Pandey, A.; Mann, M. Nature (London, U. K.) 2000, 405, 837− 846. (25) Jung, K.; Cho, W. Anal. Chem. (Washington, DC, U. S.) 2013, 85, 7125−7132. (26) Lemieux, R. U. Acc. Chem. Res. 1996, 29, 373−380. (27) Mandal, D. K.; Kishore, N.; Brewer, C. F. Biochemistry 1994, 33, 1149−1156. (28) Hodges, R. S.; Burke, T. W. L.; Mant, C. T. J. Chromatogr. 1991, 548, 267−280. (29) Srajer Gajdosik, M.; Clifton, J.; Josic, D. J. Chromatogr. A 2012, 1239, 1−9. (30) Wang, L.; Li, F.; Sun, W.; Wu, S.; Wang, X.; Zhang, L.; Zheng, D.; Wang, J.; Gao, Y. Mol. Cell. Proteomics 2006, 5, 560−562. (31) Yang, Z.; Hancock, W. S. J. Chromatogr. A 2004, 1053, 79−88. (32) Chanphai, P.; Bekale, L.; Tajmir-Riahi, H. A. Eur. Polym. J. 2015, 67, 224−231. (33) Buczek, P.; Horvath, M. P. J. Biol. Chem. 2006, 281, 40124− 40134. (34) Buczek, P.; Orr, R. S.; Pyper, S. R.; Shum, M.; Kimmel, E.; Ota, I.; Gerum, S. E.; Horvath, M. P. J. Mol. Biol. 2005, 350, 938−952.

9620


Displacement Phenomena in Lectin Affinity Chromatography

Recommend Documents