Investigating the Influence of Membrane Composition on Protein

Aug 2, 2017 - Patent office boosts fees for filings and challenges. The U.S. Patent and Trademark Office has revamped its fee schedule for the first t...
0 downloads 11 Views 2MB Size
Subscriber access provided by UNIVERSITY OF ADELAIDE LIBRARIES

Article

Investigating the Influence of Membrane Composition on ProteinGlycolipid Binding Using Nanodiscs and Proxy Ligand ESI-MS Ling Han, Luis Morales, Michele R Richards, Elena N Kitova, Simonetta Sipione, and John S. Klassen Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b02094 • Publication Date (Web): 02 Aug 2017 Downloaded from http://pubs.acs.org on August 12, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Investigating the Influence of Membrane Composition on Protein-Glycolipid Binding Using Nanodiscs and Proxy Ligand ESI-MS Ling Han,1 Luis C. Morales,2 Michele R. Richards,1 Elena N. Kitova,1 Simonetta Sipione2 and John S. Klassen1* Alberta Glycomics Centre and 1

Department of Chemistry, University of Alberta, Edmonton, Alberta, Canada T6G 2G2

2

Department of Pharmacology, University of Alberta, Edmonton, AB, Canada T6G 2H7

*Corresponding Author’s address: Department of Chemistry, University of Alberta Edmonton, AB CANADA T6G 2G2 Email: [email protected] Telephone: (780) 492-3501

ACS Paragon Plus Environment

1

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 32

Abstract This work describes a versatile analytical approach, which combines the proxy ligand electrospray ionization mass spectrometry (ESI-MS) assay and model membranes of defined composition, to quantify the influence of lipid bilayer composition on protein–glycolipid binding in vitro. To illustrate the implementation of the assay (experimental design and data analysis), affinities of the monosialoganglioside ligand GM1, incorporated into nanodiscs (NDs), for cholera toxin B subunit homopentamer (CTB5) were measured. A series of NDs containing GM1 and cholesterol were prepared using three different phospholipids (DMPC, DPPC and POPC) and the average GM1 and cholesterol content of each ND were determined. The intrinsic affinities of GM1-containing NDs prepared with the three phospholipids are found to be similar in magnitude, indicating that small differences in the fatty acid chain length and the number of unsaturated bonds do not significantly affect the CTB5–GM1 interaction. Moreover, the measured affinities are similar to the value measured for GM1 pentasaccharide, indicating that neither the ceramide moiety nor the surface of the phospholipid membrane plays a significant role in CTB5 binding. The intrinsic (per binding site) affinity of the CTB5–GM1 interaction was found to decrease with increasing GM1 content of the ND, consistent with the occurrence of GM1 clustering in the membrane, which sterically hinders binding to CTB5. Notably, the addition of cholesterol to GM1-containing NDs did not have a significant effect on the strength of the CTB5–GM1 interaction. This result, which is at odds with the findings of a previous study of CTB5 binding to GM1 in vesicles, suggests that cholesterol does not “mask” GM1, at least not in NDs. These data, in addition to providing new insights into the influence of membrane composition on CTB5–GM1 binding, demonstrate of the potential of the proxy ligand ESI-MS

ACS Paragon Plus Environment

2

Page 3 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

approach for comprehensive and quantitative studies of lectin interactions with glycolipids in native-like, membrane environments. Introduction Cell-surface glycolipids, such as glycosphingolipids, participate in many important biological processes, such as cellular recognition and adhesion, pathogen infection, signal transduction, trafficking and immune response, through interactions with lectins (carbohydrate binding proteins).1,2 Glycolipids are amphipathic molecules, consisting of a hydrophobic lipid moiety (e.g. ceramide),3 which is embedded in the cellular membrane, and a hydrophilic carbohydrate head group, which is exposed to aqueous solution. Although the carbohydrate moiety is primarily responsible for lectin recognition, it is generally accepted that the nature of the membrane environment (i.e., the composition of lipid bilayer) can influence the thermodynamics and kinetics of protein-glycolipid interactions.4,5 For example, glycosphingolipids tend to cluster in plasma membranes resulting in the formation of lipid rafts that are enriched in sphingomyelin and cholesterol; similar effects have also been observed in some model membranes.6-11 This increase of the local glycolipid density has been reported to reduce protein-glycolipid binding due to steric effects resulting from interactions between the oligosaccharide moieties of the glycolipid molecules (an effect referred to as clustering).8 Protein-glycolipid binding has also been shown to depend on the manner in which the glycolipid is displayed in the membrane, i.e., the extent of surface carbohydrate exposure and the accessibility of the aglycone.12,13 Other membrane components may also influence protein-glycolipid interactions. For example, there is evidence from experiment and computational modeling that cholesterol can induce changes in the relative orientation of the carbohydrate moiety of glycosphingolipids at the surface of membranes, thereby modulating (strengthening or weakening) protein-glycolipid binding.13-18

ACS Paragon Plus Environment

3

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 32

Protein-glycolipid interactions can be studied under native-like conditions using model membranes, such as supported lipid bilayers, liposomes, micelles, bicelles and nanodiscs (NDs).8,10,19-30 A variety of detection strategies, including fluorescence8,13 and surface plasmon resonance (SPR) spectroscopy,10,20-22 quartz crystal microbalance (QCM),23,28-30 nuclear magnetic resonance (NMR) spectroscopy,13,24 and flow cytometry,27 have been combined with model membranes to study protein binding to glycolipids. Despite the availability of such assays, there are relatively few quantitative binding data reported for soluble protein interactions with glycolipids in a membrane environment. The dearth of such data is due, at least in part, to the absence of a robust and general assay that is capable of monitoring the stoichiometry of protein-glycolipid interactions. Many lectins possess multiple glycan binding sites and, in the absence of binding stoichiometry information, the multivalent nature of the interactions complicates the interpretation of the binding data. Electrospray ionization mass spectrometry (ESI-MS) has emerged as a versatile, label- and immobilization-free method for quantifying protein-oligosaccharide interactions in vitro. The ESI-MS assay also has the unique feature that it can report directly on the binding stoichiometry.31 Recently, ESI-MS has been applied to glycolipid-containing model membranes, such as with NDs, picodiscs and micelles, to detect glycolipid-lectin interactions and establish binding stoichiometry.23,25,32-35 However, due to differences in the ESI-MS detection efficiencies of the free and glycolipid-bound lectin, reliable quantification of glycolipid-lectin binding by ESI-MS is likely not possible.33 Here, we describe an analytical approach, which combines the proxy ligand electrospray ionization mass spectrometry (ESI-MS) assay33 and model membranes of defined composition, to quantify the influence of lipid bilayer composition on lectin– glycolipid binding. The method relies on direct ESI-MS measurements of lectin binding to a

ACS Paragon Plus Environment

4

Page 5 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

water-soluble proxy ligand (Lproxy), with a known affinity, in the presence of a competing glycolipid ligand, which is incorporated in the model membrane. The distribution of free and Lproxy-bound lectin provides a quantitative measure of the amount of lectin bound to glycolipid.33 The interaction between cholera toxin B subunit homopentamer (CTB5) and the monosialoganglioside GM1, which has previously served as a model system for studying the influence of glycolipid and cholesterol content in lipid bilayers and vesicles on binding,8,14 was used in the present study. To illustrate the implementation of the assay, the affinities of GM1, incorporated into NDs of varying compositions, for CTB5 were quantified. Measurements were carried out using a series of NDs of defined compositions to gain insights into the influence of phospholipid structure, and GM1 and cholesterol concentrations on the CTB5–GM1 interaction. Experimental Materials and Methods Proteins Cholera toxin B subunit homopentamer (CTB5, MW 58,040 Da) from Vibrio cholerae was purchased from Sigma-Aldrich Canada (Oakville, Canada). CTB5 was concentrated and dialyzed into a 200 mM aqueous ammonium acetate solution (pH 6.8) using an Amicon 0.5 mL microconcentrator (EMD Millipore, Billerica, MA) with a MW cutoff of 30 kDa, and its concentration was determined using a Pierce BCA assay kit (Thermo Scientific, Ottawa, Canada) following the manufacturer’s instruction. The recombinant membrane scaffold protein (MSP) MSP1E1 (MW 27,494 Da) was produced from the plasmid pMSP1E1 (Addgene, Cambridge, MA) and purified using a procedure described elsewhere.36,37 The resulting MSP1E1 protein was dialyzed against the standard buffer (10 mM Tris/HCl, 0.1 M NaCl, 1 mM EDTA, pH 7.4) and

ACS Paragon Plus Environment

5

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 32

the concentration was estimated from UV absorption at 280 nm, using the extinction coefficient of 32,430 cm-1 M-1.37 Both protein stock solutions were stored at −80 °C until used. Phospholipids, glycolipids, oligosaccharide and cholesterol 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine 1,2-dimyristoyl-sn-glycero-3-phosphocholine

(POPC,

(DMPC,

MW

MW 677.93

760.08 Da)

Da), and

1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC, MW 734.04 Da) were purchased from Avanti Polar Lipids (Alabaster, AL). Cholesterol (MW 386.65 Da) was purchased from Sigma-Aldrich Canada (Oakville, Canada). These lipids were added to HPLC grade methanol/chloroform (1:1, v/v, Thermo Fisher, Ottawa, Canada) to produce 25 mM stock solutions. The GM1 monosialoganglioside, β-D-Gal-(1→3)-β-D-GalNAc-(1→4)-[α-D-Neu5Ac(2→3)]-β-D-Gal-(1→4)-β-D-Glc-ceramide, purified from bovine brain, was purchased from Axxora LLC (Farmingdale, NY). Two isoforms of GM1, i.e., d18:1-18:0 (MW 1545.82 Da) and d20:1-18:0 (MW 1573.87 Da), were identified in the sample, vide infra. N-ω-CD3-octadecanoyl monosialoganglioside GM1 (GM1-D3), also a mixture of 18:1-18:0 (MW 1548.83 Da) and d20:1-18:0 (MW 1576.89 Da) isoforms, was purchased form Matreya LLC (State College, PA). Solid samples of the glycolipids were dissolved in methanol to yield 1 mM stock solutions. GM1 pentasaccharide, D-Glc

β-D-Gal-(1→3)-β-D-GalNAc-(1→4)-[α-D-Neu5Ac-(2→3)]-β-D-Gal-(1→4)-

(GM1os, MW 997.88 Da), was purchased from Elicityl SA (Crolles, France) and was

dissolved in Milli-Q water (Millipore, MA) to prepare a 1 mM stock solution. All stock solutions were stored at −20 °C until used. The structures of the POPC, DMPC DPPC, GM1, GM1os and cholesterol are shown in Figure S1 (Supporting Information). Preparation of nanodiscs

ACS Paragon Plus Environment

6

Page 7 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Nanodiscs composed of POPC, DMPC or DPPC and varying amounts of GM1 and cholesterol were prepared using the protocol described by Sligar and coworkers.36,37 Briefly, GM1, phospholipid, and cholesterol were diluted in methanol at the desired molar ratios. The lipids were dried under a gentle stream of nitrogen to form a lipid film and then dissolved in a re-suspension buffer (pH 7.4) containing 20 mM TrisHCl, 0.5 mM EDTA, 100 mM NaCl and 25 mM sodium cholate (Sigma-Aldrich Canada, Oakville, Canada). The MSP1E1 was added to the mixture at a MSP1E1-to-total lipid molar ratio of 1:100. The ND self-assembly process was initiated by adding pre-washed biobeads (Bio-Rad, Mississauga, Canada) and the mixture was incubating for 4 h on an orbital shaker. Incubation was performed at a temperature that closely matches the phase transition temperature of the phospholipid used (4 °C for POPC, 25 °C for DMPC and 37 °C for DPPC). The supernatant was then recovered and the ND purified using a Superdex 200 10/300 size-exclusion column (GE-Healthcare Life Sciences, Piscataway, NJ) equilibrated with a 200 mM aqueous ammonium acetate solution (pH 6.8) at room temperature. Finally, the fraction corresponding to ND was collected, followed by concentration and dialysis into a 200 mM aqueous ammonium acetate solution (pH 6.8) using an Amicon microconcentrator (EMD Millipore, Billerica, MA) with a 30 kDa MW cut-off. All ND stock solutions were stored at –80 °C until used. Details on the methods used to quantify the NDs concentrations and establish their GM1 and cholesterol contents are given as Supporting Information.38 Mass spectrometry All ESI-MS measurements were performed on a Synapt G2S quadrupole-ion mobility separation-time of flight (Q-IMS-TOF) mass spectrometer (Waters, Manchester, UK) equipped with a nanoflow ESI (nanoESI) source. NanoESI tips with ~5 µm outer diameters (o.d.) were produced from borosilicate capillaries (1.0 mm o.d., 0.68 mm inner diameter) using a P-1000

ACS Paragon Plus Environment

7

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 32

micropipette puller (Sutter Instruments, Novato, CA). Details of the instrumental parameters used for the measurements are given as Supporting Information. Proxy ligand ESI-MS assay The proxy ligand ESI-MS assay, which is based on competitive ligand binding and direct ESI-MS measurements, was used to quantify the interactions between CTB5 and GM1-containing NDs.33 Briefly, the binding of the Lproxy (GM1os) to the target protein (CTB5) provides a quantitative measure of the extent of binding of CTB5 to GM1, which is incorporated in the NDs. CTB5 possesses five equivalent GM1 binding sites that exhibit positive cooperativity,39,40 to which both GM1 and GM1os, can bind. The sequential binding of GM1os to CTB5 can be described by the intrinsic (per binding site) association constants (Ka,proxy) and a cooperativity factor ai,40 which reflects the number (i) of ligand-occupied, nearest-neighbour subunits. As reported elsewhere, Ka,proxy has a value of (3.2 ± 0.2) × 106 M-1;40 the association constants corresponding to one and two bound nearest-neighbour subunits are a×Ka,proxy and a2×Ka,proxy, respectively, where a = 1.7.40 The stepwise binding of CTB5 to GM1 contained within NDs can be described by a combination of intermolecular and intramolecular binding reactions, with the corresponding intrinsic association constants, Ka,inter and Ka,intra, respectively, and corresponding cooperativity factors. The Ka,inter term describes the association of CTB5 (or CTB5 bound to one or more GM1os) with the ND (through GM1), while Ka,intra, a unitless quantity, describes the subsequent recruitment of additional GM1 ligands.41-43 To simplify data analysis, it was assumed that the same cooperativity factors apply, regardless of whether the neighbouring subunits are occupied by GM1 or GM1os. In total, the competitive binding of GM1os and GM1 (in NDs) to CTB5 is described by thirty-eight association equilibria and forty-one distinct molecular species, including positional isomers.33,40 A summary of the binding interactions and a

ACS Paragon Plus Environment

8

Page 9 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

list of the corresponding equilibrium and mass balance equations are given Scheme 1 and eqs S9 – S49, Supporting Information. To implement the assay, GM1-ND is titrated into a solution containing a fixed concentration of CTB5 and GM1os and the distribution of CTB5 species is measured by ESI-MS. In principle, the relative abundances all detected CTB5 species could be used to determine the magnitude of Ka,inter and Ka,intra. However, as described elsewhere,33 the CTB5 complexes containing GM1, which originate from ND ions in the gas phase, are expected to have ESI-MS response factors that differ from those of free CTB5 and (CTB5 + qGM1os) complexes, which originate directly from solution. Consequently, the measured distribution of CTB5 ions in the gas phase may not accurately reflect their distribution in solution.33 Because of these non-uniform response factors, only the relative abundances of the (CTB5 + qGM1os) complexes were used to infer the extent of GM1 binding in solution. In practice, changes in the relative abundances of (CTB5 + 4GM1os) and (CTB5 + 5GM1os) provide the most reliable measure of the competition between GM1 and GM1os. Consequently, in the present study, the abundance ratio of CTB5 bound to five and four GM1os (i.e., Rproxy,5-4), which is taken to be equal to the corresponding concentration ratio in solution (eq 1):

Rproxy,5-4 ≡

Ab(CTB5 +5GM1os ) [CTB5 +5GM1os ] = Ab(CTB5 +4GM1os ) [CTB5 +4GM1os ]

(1)

was monitored as a function of GM1 concentration. The values of Ka,inter and Ka,intra were then solved for numerically by minimizing the sum of squares of residuals between the experimental to theoretical Rproxy,5-4 values (eq S50, Supporting Information).33,41 In cases where each ND leaflet contains, on average, ≤1 GM1, the Ka,intra term becomes negligibly small, and the general (multivalent) binding model reduces to the simpler monovalent binding model shown in Scheme

ACS Paragon Plus Environment

9

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 32

S1, Supporting Information.33 Maple 2017.1 (Maplesoft, Waterloo, Canada) was used to fit the binding models to the experimental data. Results and Discussion a. GM1 incorporation efficiency It is essential that the average number of GM1 molecules per ND in a given preparation be accurately known to have confidence in the affinities measured for the CTB5 interactions with the GM1-containing NDs. To accomplish this, a stable isotope labelled internal standard/ESI-MS approach, that was previously used to quantify lipids and glycolipids in membranes, was adapted.38 As described in the Experimental section and Supporting Information, the GM1 content of a given ND preparation was determined by disassembling the disc in a methanol solution (with 20% acetonitrile and 1% formic acid), adding deuterated GM1 (i.e., GM1-D3) as an internal standard, and carrying out ESI-MS analysis. Shown in Figure S3 (Supporting Information) are representative mass spectra acquired for solutions of 0.21 µM, 0.52 µM, and 1.04 µM of (disassembled) ND1 (0.5% GM1 and 99.5% POPC) and GM1-D3 (0.34 µM). The GM1 incorporation efficiency (IEGM1) for ND1, determined from a plot of the abundance ratio of GM1-to-GM1-D3 ions versus expected concentration ratios measured at six different ND1 concentrations, was found to be 1.05 ± 0.03 (Figure S3d, Supporting Information). This value translates to an average of 1.05 GM1 per ND. Using this same approach, the GM1 content of thirty other NDs (ND2 – ND31), which varied in terms of the nature of the phospholipid (DMPC, DPPC or POPC) and the molar percentages (total lipid content) of GM1 (0.5% – 15%) and cholesterol (0% – 40%) used in ND preparation, was determined (Table 1). Inspection of the data in Table 1 reveals that IEGM1 is influenced by the nature of the phospholipid present in the ND and the GM1 concentration (percentage) used to prepare the ND.

ACS Paragon Plus Environment

10

Page 11 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

These effects are illustrated in Figures 1a and 1b, where IEGM1, as well as the corresponding average number of GM1 per ND, are plotted versus the GM1 percentage used to prepare the NDs. For all three phospholipids, GM1 incorporation was essentially 100% efficient for the 0.5% and 1.0% GM1-NDs, with an IEGM1 of between 0.98±0.03 and 1.22±0.03. Values >1.0 are attributed to the incomplete incorporation of MSP into the NDs, as revealed by the SEC chromatograms (data not shown). Similar efficiencies were observed for the 5% (0.97±0.08 (DMPC) and 0.95±0.04 (DPPC)) and 10% (0.99±0.05 (DMPC) and 0.91±0.17 (DPPC)) DMPC and DPPC GM1-NDs; significantly lower IEGM1 values were found for the 15% GM1-NDs (0.78±0.04 (DMPC) and 0.77±0.03 (DPPC)). For the GM1-NDs composed of POPC, lower IEGM1 values were observed - 5% (0.57±0.02), 10% (0.77±0.03) and 15% (0.67±0.04) GM1-NDs. To our knowledge, a lower incorporation efficiency of glycosphingolipids into POPC (or other unsaturated phospholipids) membranes, compared to saturated phospholipids such as DMPC and DPPC, has not been previously reported. However, the “kinked” mono-unsaturated acyl chain in POPC is expected to loosen the packing of the lipid bilayer;44 this packing defect could reduce the amount of GM1 that can be incorporated into the ND. It is also known that GM1 associates preferentially with gel-phase regions in mixed phospholipid liposomes.45-47 Consequently, the lower IEGM1 observed for POPC NDs may be related to the greater fluidity of POPC bilayers, compared to those of DMPC and DPPC.48,49 The incorporation of GM1 into NDs prepared from lipid mixtures with 5% – 40% nominal cholesterol content (ND16 – ND31) was also investigated (Table 1). Overall, IEGM1 is found to decrease with increasing (initial) percentage of cholesterol used to prepare the ND, independent of the nature of the phospholipid (Figure 1c). For example, for the POPC-containing NDs, IEGM1 is 0.96±0.01 and 0.97±0.03 at 5% and 10% (initial) cholesterol, respectively, but drops to

ACS Paragon Plus Environment

11

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 32

0.35±0.03 at 40%. A similar trend is observed for DMPC and DPPC. To aid in understanding these results, The cholesterol incorporation efficiency (IEch) was also measured, in the presence and absence of GM1 (Tables 1 and S1, and Figure S4, Supporting Information). In the absence of GM1 (ND32 – ND46), IEch decreased with increasing percentage of cholesterol used to prepare the NDs. Moreover, the POPC NDs exhibited consistently higher IEch than the corresponding DMPC and DPPC NDs. For example, ND32 (prepared with 5% cholesterol and POPC) has an IEch of 0.91±0.01, while for ND37 (prepared with 5% cholesterol and DMPC) and ND42 (prepared with 5% cholesterol and DPPC) the values are 0.66±0.03 and 0.62±0.02, respectively. A similar trend is observed for the NDs prepared with 0.5% GM1, 5% – 40% cholesterol (ND16 – ND31), Table 1. Based on the results of size-exclusion chromatography, it was reported previously that NDs decrease in size with increasing percentages of cholesterol used to prepare the discs.50,51 Therefore, it is possible that this reduction in average ND size has the effect of reducing the number of cholesterol and GM1 molecules ultimately incorporated into the discs. Furthermore, it was shown that cholesterol forms a detergent resistant phase with saturated phospholipids.52 This behavior could explain the lower IEch observed for DMPC and DPPC. b. Affinities of GM1-containing NDs for CTB5 Having established the GM1 and cholesterol compositions of the GM1-containing NDs (ND1 – ND31), the proxy ligand ESI-MS assay was used to quantify the influence of phospholipid structure and GM1 and cholesterol content on the interactions between CTB5 and the GM1 in these NDs. As described in the Experimental section, GM1os served as Lproxy for these measurements. The CTB5 and the GM1os concentrations were such that >92% of the CTB5 binding sites were occupied (~69% of CTB5 bound to five GM1os, ~25% bound to four GM1os

ACS Paragon Plus Environment

12

Page 13 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

and ~5% bound to three GM1os). For NDs containing ≤1GM1 per leaflet (ND1, ND2, ND6, ND7, ND11, ND12, and ND16–ND31), Ka,inter was determined using the reduced binding model (Scheme S1, Supporting Information), while for other NDs (which contained multiple GM1 molecules per leaflet) the general binding model (Scheme 1) was used to establish Ka,inter and Ka,intra. A summary of the binding data is given below. Influence of nature of phospholipid. Shown in Figure 2a is a representative ESI mass spectrum acquired in positive ion mode for an aqueous ammonium acetate solution (200 mM, pH 6.8, 25 °C) of CTB5 (4 µM) and GM1os (20 µM). Signal corresponding to (CTB5 + qGM1os)n+ ions, with q = 3 – 5 and n = 14 – 16, was detected. Upon addition of ND1 (4.1 µM, 0.5% GM1, POPC) to the solution, (CTB5 + GM1 + 4GM1os)n+ ions, at n = 14 – 16, were also detected (Figure 2b). The observation of (CTB5 + GM1 + 4GM1os)n+ ions, which are believed to originate from the gas-phase dissociation of the ND complexes associated with GM1-bound CTB5, suggests that CTB5 binds to a single molecule of GM1 in ND1 under these conditions.33,34 Notably, the addition of the ND resulted in a measurable increase in Rproxy,5-4 (from 2.71 to 5.71). The plot of measured Rproxy,5-4 versus GM1 concentration is shown in Figure 2c; also shown is the theoretical curve corresponding to a Ka,inter of (1.5 ± 0.3) × 106 M-1. The corresponding plot of residuals is shown in Figure S5 (Supporting Information). The results of analogous measurements performed on the DMPC (ND6) and DPPC (ND11) NDs containing 0.5% GM1 are shown in Figure 2c. Analysis of the data yielded Ka,inter values of (2.3 ± 0.3) × 106 M-1 and (1.8 ± 0.6) × 106 M-1, respectively. The similarity in Ka,inter values indicates that, despite influencing the incorporation of GM1 into the ND bilayer, the differences in the structures and properties of DMPC, DPPC and POPC (i.e., fatty acid chain length and the number of unsaturated bonds) do not significantly affect the

ACS Paragon Plus Environment

13

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 32

CTB5–GM1 interaction. It is also important to note that the Ka,inter values (Figure 2d) are similar in magnitude, within a factor of two, to the intrinsic Ka (i.e., Ka,proxy) measured for GM1os binding ((3.2 ± 0.2) × 106 M-1).40 This finding indicates that, unlike for some other lectin-ganglioside interactions,13 neither the ceramide moiety of GM1 nor the surface of the phospholipid membrane plays a significant role in CTB5 binding. Influence of GM1 content. To probe the influence of GM1 content on CTB5, affinity measurements were performed on NDs prepared with 1%, 5%, 10% and 15% GM1, which possess between 1.0 and 11.7 GM1 per leaflet. Shown in Figure 3a is a representative ESI mass spectrum measured for a 200 mM aqueous ammonium acetate solution (pH 6.8, 25 °C) of CTB5 (4.0 µM), GM1os (20 µM) and ND5 (0.55 µM, 15% GM1 and POPC). In addition to the ions corresponding to CTB5 bound to GM1os, mixed GM1os/GM1 species, i.e., (CTB5 + 2GM1 + 3GMos)n+, (CTB5 + 3GM1 + 2GMos)n+, and (CTB5 + 4GM1 + GMos)n+ with n = 14 – 16, were also detected, which suggests that CTB5 binds to between two and four GM1 under these conditions. From measurements performed over a range of ND concentrations, a plot of Rproxy,5-4 versus GM1 concentration was constructed (Figure 3b). Analysis of the concentration dependence of Rproxy,5-4 using the general binding model (Scheme 1) gave Ka,inter and Ka,intra values of (3.5 ± 0.9)×105 M-1 and 44.8 ± 12.4, respectively (Table 1). The theoretical curve calculated from these values is also shown in Figure 3b and provides an excellent description of the experimental data (Figure S5, Supporting Information). Notably, Ka,inter determined for this ND is approximately four times smaller than the value measured for ND1 (0.5% GM1). Shown in Figure 4a is a plot of Ka,inter measured for the POPC NDs containing five different percentages of GM1 (Figures S6 and S7, Supporting Information). It can be seen that there is a systematic decrease in the magnitude of Ka,inter with increasing GM1 content of the NDs. Analysis of the

ACS Paragon Plus Environment

14

Page 15 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

DMPC and DPPC NDs produced with varying GM1 content gave similar results (Table 1, Figures 4b and 4c, and Figure S6 and S7, Supporting Information). The decrease in Ka,inter with increasing GM1 content is qualitatively consistent with results described previously.8,10,53,54 For example, Cremer and coworkers reported a decrease in apparent affinity, by a factor of 5 – 8, for CTB5 binding to GM1 in a POPC supported bilayer upon increasing the (nominal) GM1 content from 0.02% to 10.0%.8 This behavior is believed to be related to “clustering” of GM1 in the membrane resulting from hydrogen bonding between the oligosaccharide moieties, which has the effect of sterically hindering GM1 binding to CTB5.8,55,56 Influence of cholesterol content. It has been reported previously that cholesterol binding to GM1 in lipid bilayers induces a conformational change in GM1, whereby the oligosaccharide moiety becomes oriented towards the bilayer.14 This conformational change is believed to “mask” or effectively reduce the number of GM1 in the membrane that are available for protein binding.14 To probe the influence of cholesterol content on the CTB5 binding to GM1 in NDs, affinity measurements performed on a series of GM1 NDs that contained between 4.0 and 19.4 cholesterol molecules (ND16–ND31). Shown in Figure S8a – S8c (Supporting Information) are plots of Rproxy,5-4 versus GM1 concentration measured for the POPC NDs (ND16 – ND21), DMPC NDs (ND22 – ND26) and DPPC NDs (ND27 – ND31). The corresponding Ka,inter values are plotted in Figure S9a – S9c (Supporting Information) and are summarized in Table 1. Inspection of the binding data reveals that the addition of cholesterol to the NDs has little or no effect on Ka,inter, even at the highest cholesterol contents. This result is at odds with the findings of Lingwood and coworkers, wherein the presence of cholesterol in liposomes (0.1% GM1, 5% cholesterol and POPC) was reported to result in a measurable decrease (~1.3-fold) in the amount of GM1 accessible to CTB5.14 While the exact reason for these divergent findings is

ACS Paragon Plus Environment

15

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 32

not known, one possible explanation lies in the GM1 content of the model membranes used. As noted in the present study, the addition of cholesterol results in a decrease in GM1 incorporation efficiency for the NDs. If not corrected for, the lower GM1 content would lead to smaller “apparent” Ka,inter values. If the GM1 content of liposomes is similarly affected by cholesterol, it follows that the observed “masking” of GM1 could, in fact, be due to a reduction in the number of GM1 in the membrane. It is also possible that the influence of cholesterol on GM1 availability in NDs differs from that in liposomes due to differences in membrane curvature, which is known to affect lipid organization.52 It must be stressed, however, that these explanations are speculative and require further investigation. Conclusions In summary, we describe a versatile analytical approach, based on proxy ligand ESI-MS and model membranes of defined composition, to quantitatively investigate the influence of membrane composition on lectin binding to glycolipids. To illustrate implementation, the method was used to evaluate the influence of phospholipid structure and GM1 and cholesterol content on CTB5 binding to GM1 in NDs. The results of this study produced a number of interesting findings. First, the similarity in affinities measured for GM1-containing NDs produced from the three phospholipids indicates that small differences in the fatty acid chain length and/or the number of unsaturated bonds do not significantly affect the CTB5–GM1 interaction. Moreover, the measured affinities are similar to the value measured for GM1 pentasaccharide, indicating that neither the ceramide moiety nor the surface of the membrane significantly influences CTB5 binding. Secondly, the intrinsic affinity of the CTB5–GM1 interaction was found to decrease with increasing GM1 content of the ND, evidence for the occurrence of GM1 clustering in the membrane. Thirdly, the addition of cholesterol to GM1-containing NDs did not have a significant

ACS Paragon Plus Environment

16

Page 17 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

effect on the strength of the CTB5–GM1 interaction, suggesting that cholesterol does not alter the conformation of GM1 in NDs. These data, in addition to providing new insights into the influence of membrane composition on CTB5–GM1 binding, demonstrate of the potential of the proxy ligand ESI-MS approach for comprehensive and quantitative studies of lectin interactions with glycolipids in native-like membrane environments. Finally, it should be noted that the method is not limited to NDs and can be readily applied to other model membranes, such as vesicles and picodiscs, to quantify protein-glycolipid interactions. Acknowledgements The authors acknowledge the Alberta Glycomics Centre for funding. ASSOCIATED CONTENT Supporting Information Methods, structures, binding data, and mass spectra. This information is available free of charge via the Internet at http://pubs.acs.org/. AUTHOR INFORMATION Corresponding Author *E-mail: [email protected]. Phone: (780) 492-3501. Notes The authors declare no competing financial interests.

ACS Paragon Plus Environment

17

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 32

References (1)

Sharon, N.; Lis, H. Sci. Am. 1993, 268, 82-89.

(2)

Quiocho, F. A. Annu. Rev. Biochem. 1986, 55, 287-315.

(3)

Varki, A.; Cummings, R.D.; Esko, J.D.; Freeze, H.H.; Stanley, P.; Bertozzi, C.R.; Hart, G.W.; Etzler, M.E. Essentials of Glycobiology 2nd Ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, USA 2009.

(4)

Evans, S. V.; MacKenzie, C. R. J. Mol. Recognit. 1999, 12, 155-168.

(5)

Lingwood, C. A.; Manis, A.; Mahfoud, R.; Khan, F.; Binnington, B.; Mylvaganam, M. Chem. Phys. Lipids 2010, 163, 27-35.

(6)

Simons, K.; Van Meer, G. Biochemistry 1988, 27, 6197-6202.

(7)

Hooper, N. M. Curr. Biol. 1998, 8, R114-116.

(8)

Shi, J.; Yang, T.; Kataoka, S.; Zhang, Y.; Diaz, A. J.; Cremer, P. S. J. Am. Chem. Soc. 2007, 129, 5954-5961.

(9)

Lingwood, D.; Kaiser, H. J.; Levental, I.; Simons, K. Biochem. Soc. Trans. 2009, 37, 955-960.

(10)

Borch, J.; Torta, F.; Sligar, S. G.; Roepstorff, P. Anal. Chem. 2008, 80, 6245-6252.

(11)

Bricarello, D. A.; Mills, E. J.; Petrlova, J.; Voss, J. C.; Parikh, A. N. J. Lipid Res. 2010, 51, 2731-2738.

(12)

Lingwood, C. A. Glycoconjugate J. 1996, 13, 495-503.

(13)

Sanghera, N.; Correia, B. E.; Correia, J. R.; Ludwig, C.; Agarwal, S.; Nakamura, H. K.; Kuwata, K.; Samain, E.; Gill, A. C.; Bonev, B. B.; Pinheiro, T. J. Chem. Biol. 2011, 18, 1422-1431.

ACS Paragon Plus Environment

18

Page 19 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

(14)

Lingwood, D.; Binnington, B.; Rog, T.; Vattulainen, I.; Grzybek, M.; Coskun, U.; Lingwood, C. A.; Simons, K. Nat. Chem. Biol. 2011, 7, 260-262.

(15)

Mahfoud, R.; Manis, A.; Binnington, B.; Ackerley, C.; Lingwood, C. A. J. Biol. Chem. 2010, 285, 36049-36059.

(16)

Šachl, R.; Amaro, M.; Aydogan, G.; Koukalová, A.; Mikhalyov, I. I.; Boldyrev, I. A.; Humpolíčková, J.; Hof, M. BBA-Mol. Cell. Res. 2015, 1853, 850-857.

(17)

Yahi, N.; Aulas, A.; Fantini, J. PLoS ONE 2010, 5, e9079.

(18)

Gallegos, K. M.; Conrady, D. G.; Karve, S. S.; Gunasekera, T. S.; Herr, A. B.; Weiss, A. A. PLoS ONE 2012, 7, e30368.

(19)

Czogalla, A.; Grzybek, M.; Jones, W.; Coskun, U. Biochim. Biophys. Acta. 2014, 1841, 1049-1059.

(20)

Kuziemko, G. M.; Stroh, M.; Stevens, R. C. Biochemistry 1996, 35, 6375-6384.

(21)

MacKenzie, C. R.; Hirama, T.; Lee, K. K.; Altman, E.; Young, N. M. J. Biol. Chem. 1997, 272, 5533-5538.

(22)

Murthy, B. N.; Voelcker, N. H.; Jayaraman, N. Glycobiology 2006, 16, 822-832.

(23)

Janshoff, A.; Steinem, C.; Sieber, M.; el Baya, A.; Schmidt, M. A.; Galla, H. J. Eur. Biophys. J. 1997, 26, 261-270.

(24)

Han, L.; Kitova, E. N.; Klassen, J. S. J. Am. Soc. Mass Spectrom. 2016, 27, 1878-1886.

(25)

Yamaguchi, T.; Uno, T.; Uekusa, Y.; Yagi-Utsumi, M.; Kato, K. Chem. Commun. 2013, 49, 1235-1237.

(26)

Zhang, Y. X.; Liu, L.; Daneshfar, R.; Kitova, E. N.; Li, C. S.; Jia, F.; Cairo, C. W.; Klassen, J. S. Anal. Chem. 2012, 84, 7618-7621.

(27)

Lauer, S.; Goldstein, B.; Nolan, R. L.; Nolan, J. P. Biochemistry 2002, 41, 1742-1751.

ACS Paragon Plus Environment

19

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 32

(28)

Ebara, Y.; Okahata, Y. J. Am. Chem. Soc. 1994, 116, 11209-11212.

(29)

Rydell, G. E.; Dahlin, A. B.; Hook, F.; Larson, G. Glycobiology 2009, 19, 1176-1184.

(30)

Bally, M.; Rydell, G. E.; Zahn, R.; Nasir, W.; Eggeling, C.; Breimer, M. E.; Svensson, L.; Hook, F.; Larson, G. Angew. Chem. Int. Ed. 2012, 51, 12020-12024.

(31)

Kitova, E. N.; El-Hawiet, A.; Schnier, P. D.; Klassen, J. S. J. Am. Soc. Mass Spectrom. 2012, 23, 431-441.

(32)

Leney, A. C.; Darestani, R. R.; Li, J.; Nikjah, S.; Kitova, E. N.; Zou, C. X.; Cairo, C. W.; Xiong, Z. J.; Prive, G. G.; Klassen, J. S. Anal. Chem. 2015, 87, 4402-4408.

(33)

Han, L.; Kitova, E. N.; Li, J.; Nikjah, S.; Lin, H.; Pluvinage, B.; Boraston, A. B.; Klassen, J. S. Anal. Chem. 2015, 87, 4888-4896.

(34)

Leney, A. C.; Fan, X. X.; Kitova, E. N.; Klassen, J. S. Anal. Chem. 2014, 86, 5271-5277.

(35)

Li, J.; Fan, X. X.; Kitova, E. N.; Zou, C. X.; Cairo, C. W.; Eugenio, L.; Ng, K. K. S.; Xiong, Z. J.; Prive, G. G.; Klassen, J. S. Anal. Chem. 2016, 88, 4742-4750.

(36)

Denisov, I. G.; Grinkova, Y. V.; Lazarides, A. A.; Sligar, S. G. J. Am. Chem. Soc. 2004, 126, 3477-3487.

(37)

Ritchie, T. K.; Grinkova, Y. V.; Bayburt, T. H.; Denisov, I. G.; Zolnerciks, J. K.; Atkins, W. M.; Sligar, S. G. Methods Enzymol. 2009, 464, 211-231.

(38)

Gu, J. H.; Tifft, C. J.; Soldin, S. J. Clin. Biochem. 2008, 41, 413-417.

(39)

Merritt, E. A.; Kuhn, P.; Sarfaty, S.; Erbe, J. L.; Holmes, R. K.; Ho, W. G. J. J. Mol. Biol. 1998, 282, 1043-1059.

(40)

Lin, H.; Kitova, E. N.; Klassen, J. S. J. Am. Soc. Mass Spectrom. 2014, 25, 104-110.

(41)

Kitov, P. I.; Bundle, D. R. J. Am. Chem. Soc. 2003, 125, 16271-16284.

ACS Paragon Plus Environment

20

Page 21 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

(42)

Reynolds, M.; Perez, S. C. R. Chim. 2011, 14, 74-95.

(43)

Ercolani, G.; Mandolini, L.; Mencarelli, P.; Roelens, S. J. Am. Chem. Soc. 1993, 115, 3901-3908.

(44)

Vanni, S.; Hirose, H.; Barelli, H.; Antonny, B.; Gautier, R. Nat. Commun. 2014, 5, 4916.

(45)

Fricke, N.; Dimova, R. Biophys. J. 2016, 111, 1935-1945.

(46)

Sonnino, S.; Prinetti, A. Front Physiol. 2010, 1, 153.

(47)

Aureli, M.; Mauri, L.; Ciampa, M. G.; Prinetti, A.; Toffano, G.; Secchieri, C.; Sonnino, S. Mol. Neurobiol. 2016, 53, 1824-1842.

(48)

Shaw, A. W.; McLean, M. A.; Sligar, S. G. FEBS Lett. 2004, 556, 260-264.

(49)

Denisov, I. G.; McLean, M. A.; Shaw, A. W.; Grinkova, Y. V.; Sligar, S. G. J. Phys. Chem. B 2005, 109, 15580-15588.

(50)

Roy, J.; Pondenis, H.; Fan, T. M.; Das, A. Biochemistry 2015, 54, 6299-6302.

(51)

Midtgaard, S. R.; Pedersen, M. C.; Arleth, L. Biophys. J. 2015, 109, 308-318.

(52)

Stepien, P.; Polit, A.; Wisniewska-Becker, A. Biochim. Biophys. Acta 2015, 1848, 60-66.

(53)

Worstell, N. C.; Krishnan, P.; Weatherston, J. D.; Wu, H.-J. PLoS One 2016, 11, e0153265.

(54)

Bricarello, D. A.; Mills, E. J.; Petrlova, J.; Voss, J. C.; Parikh, A. N. J. Lipid Res. 2010, 51, 2731-2738.

(55)

Yuan, C.; Johnston, L. J. Biophys. J. 2001, 81, 1059-1069.

(56)

Sharom, F. J.; Grant, C. W. M. Biochim. Biophys. Acta. 1978, 507, 280-293.

ACS Paragon Plus Environment

21

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 32

Table 1. Incorporation efficiency of GM1 (IEGM1) and cholesterol (IEch) in GM1-containing nanodiscs (NDs), prepared with varying mole percentages of GM1 and cholesterol in POPC, DMPC and DPPC, and values of Ka,inter (M-1) and Ka,intra (unitless) determined using the proxy ligand ESI-MS assay for binding to CTB5 to the NDs in 200 mM aqueous ammonium acetate solutions (pH 6.8 and 25 °C). ND

% GM1

%

% a

cholesterol

a

phospholipid

a

IEGM1 b

IEch b

Ka,inter (M-1) b

Ka,intra b,c

ND1

0.5

0

99.5 POPC

1.05 ± 0.03

0

(1.5 ± 0.3)×106

n. d.

ND2

1

0

99 POPC

0.98 ± 0.03

0

(1.1 ± 0.4)×106

n. d.

ND3

5

0

95 POPC

0.57 ± 0.02

0

(0.56 ± 0.18)×106

16.2 ± 5.8

ND4

10

0

90 POPC

0.77 ± 0.03

0

(0.47 ± 0.07)×106

31.9 ± 4.9

ND5

15

0

85 POPC

0.67 ± 0.04

0

(0.35 ± 0.09)×106

44.8 ± 12.4

ND6

0.5

0

99.5 DMPC

1.22 ± 0.03

0

(2.3 ± 0.3)×106

n. d.

ND7

1

0

99 DMPC

1.10 ± 0.05

0

(1.9 ± 0.4)×106

n. d.

ND8

5

0

95 DMPC

0.97 ± 0.08

0

(0.76 ± 0.11)×106

21.0 ± 6.2

ND9

10

0

90 DMPC

0.99 ± 0.05

0

(0.56 ± 0.08)×106

41.8 ± 7.2

ND10

15

0

85 DMPC

0.78 ± 0.04

0

(0.40 ± 0.11)×106

52.1 ± 14.1

ND11

0.5

0

99.5 DPPC

1.17 ± 0.07

0

(1.8 ± 0.6)×106

n. d.

ND12

1

0

99 DPPC

1.07 ± 0.05

0

(1.3 ± 0.3)×106

n. d.

ND13

5

0

95 DPPC

0.95 ± 0.04

0

(0.68 ± 0.13)×106

23.0 ± 4.8

ND14

10

0

90 DPPC

0.91 ± 0.17

0

(0.48 ± 0.14)×106

31.5 ± 9.5

ND15

15

0

85 DPPC

0.77 ± 0.03

0

(0.40 ± 0.06)×106

46.5 ± 7.8

ND16

0.5

5

94.5 POPC

0.96 ± 0.01

0.73 ± 0.02

(1.2 ± 0.4)×106

n. d.

ND17

0.5

10

89.5 POPC

0.97 ± 0.03

0.73 ± 0.02

(1.1 ± 0.3)×106

n. d.

ND18

0.5

15

84.5 POPC

0.91 ± 0.03

0.50 ± 0.02

(2.0 ± 0.3)×106

n. d.

ND19

0.5

20

79.5 POPC

0.82 ± 0.06

0.49 ± 0.02

(1.9 ± 0.6)×106

n. d.

ND20

0.5

25

74.5 POPC

0.36 ± 0.01

0.34 ± 0.01

(1.5 ± 0.4)×106

n. d.

ND21

0.5

40

59.5 POPC

0.35 ± 0.03

0.16 ± 0.01

(1.5 ± 0.6)×106

n. d.

ND22

0.5

5

94.5 DMPC

1.09 ± 0.06

0.58 ± 0.02

(1.8 ± 0.6)×106

n. d.

ND23

0.5

10

89.5 DMPC

1.05 ± 0.09

0.44 ± 0.01

(1.6 ± 0.2)×106

n. d.

ACS Paragon Plus Environment

22

Page 23 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

ND24

0.5

15

84.5 DMPC

0.54 ± 0.02

0.25 ± 0.01

(1.5 ± 0.3)×106

n. d.

ND25

0.5

20

79.5 DMPC

0.41 ± 0.02

0.21 ± 0.01

(1.1 ± 0.2)×106

n. d.

ND26

0.5

25

74.5 DMPC

0.38 ± 0.02

0.17 ± 0.01

(1.3 ± 0.4)×106

n. d.

ND27

0.5

5

94.5 DPPC

1.05 ± 0.05

0.41 ± 0.02

(1.9 ± 0.1)×106

n. d.

ND28

0.5

10

89.5 DPPC

0.96 ± 0.08

0.31 ± 0.01

(2.4 ± 0.6)×106

n. d.

ND29

0.5

15

84.5 DPPC

0.75 ± 0.04

0.17 ± 0.01

(1.5 ± 0.5)×106

n. d.

ND30

0.5

20

79.5 DPPC

0.45 ± 0.01

0.16 ± 0.01

(1.4 ± 0.3)×106

n. d.

ND31

0.5

25

74.5 DPPC

0.53 ± 0.02

0.14 ± 0.01

(1.6 ± 0.5)×106

n. d.

a. The percentages of GM1, cholesterol and phospholipid are based on the initial molar ratios used for ND preparation. b. Uncertainties correspond to one standard deviation. c. n. d. ≡ not determined.

ACS Paragon Plus Environment

23

Analytical Chemistry

Page 24 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment

24

Page 25 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Scheme 1.

General interaction map for CTB5 (≡P) binding to GM1os (≡Lproxy) and GM1(≡L)-ND binding. The intermolecular and intramolecular association constant (Ka,inter and Ka,intra, respectively), along with the cooperativity factor and corresponding statistical coefficients, are given for each interaction. The symbols α, β, γ and δ are used to distinguish the positional isomers.

ACS Paragon Plus Environment

25

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 32

Figure Captions Figure 1.

Plots of (a) GM1 incorporation efficiency (IEGM1) and (b) number of GM1 per ND, versus mole percentages (total lipid content) of GM1 (nominal %GM1) for NDs (ND1 – ND15) prepared using different initial amounts of GM1 and phospholipid (DMPC, DPPC and POPC). (c) Plots of IEGM1 and number of GM1 per ND versus nominal %cholesterol for NDs prepared with 0.5% GM1, 0% – 40% cholesterol and DMPC, DPPC or POPC (ND1, ND6, ND11 and ND16 – ND31). The dashed lines represent the expected values based on the initial molar ratios used for ND preparation. The errors correspond to one standard deviation.

Figure 2.

(a) and (b) ESI mass spectra acquired in positive ion mode for aqueous ammonium acetate solutions (200 mM, 25 °C and pH 6.8) containing CTB5 (4 µM), GM1 pentasaccharide (GM1os, Lproxy, 20 µM) (a) without and (b) with ND1 (containing 4.1 µM GM1). (c) Plots of Rproxy,5-4 versus GM1 concentration for NDs (ND1, ND6 and ND11) prepared with 0.5% GM1 and a phospholipid (POPC, DMPC or DPPC). (d) Bar graph of Ka,inter values measured for CTB5 binding to ND1, ND6, ND11, and GM1os. The errors correspond to one standard deviation.

Figure 3.

(a) ESI mass spectrum acquired in positive ion mode for aqueous ammonium acetate solutions (200 mM, 25 °C and pH 6.8) containing CTB5 (4 µM), GM1 pentasaccharide (GM1os, Lproxy, 20 µM) with 0.55 µM ND5 (POPC ND, 11.1 µM GM1). (b) Plot of Rproxy,5-4 versus GM1 concentration measured for ND5. The error bars correspond to one standard deviation.

ACS Paragon Plus Environment

26

Page 27 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 4.

Bar graph of Ka,inter values measured for CTB5 binding to GM1-containing NDs: (a) POPC NDs, ND1 – ND5, (b) DMPC NDs, ND6 – ND10, and (c) DPPC NDs, ND11 – ND15. The errors correspond to one standard deviation.

ACS Paragon Plus Environment

27

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 32

Figure 1

ACS Paragon Plus Environment

28

Page 29 of 32

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 2

ACS Paragon Plus Environment

29

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 32

Figure 3

ACS Paragon Plus Environment

30

Page 31 of 32

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 4

ACS Paragon Plus Environment

31

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 32

TOC graphic

ACS Paragon Plus Environment

32