Skip to main content
NCBI home page
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2006 Mar 9;103(12):4422–4427. doi: 10.1073/pnas.0509158103

Deimination of membrane-bound myelin basic protein in multiple sclerosis exposes an immunodominant epitope

Abdiwahab A Musse *, Joan M Boggs †,‡, George Harauz *,§
PMCID: PMC1450187  PMID: 16537438

Abstract

The degradation of myelin in the CNS is the hallmark of multiple sclerosis. Reduction in the net positive charge of myelin basic protein (MBP), through deimination, correlates strongly with disease severity and may mediate myelin instability and loss of compaction. Using Cys scanning, spin labeling, EPR spectroscopy, and site-specific proteolysis, we show that in the membrane-bound state the primary immunodominant epitope, V83-T92, of the less cationic recombinant murine MBP C8 mimic (rmC8) forms a more highly surface-exposed and shorter amphipathic α-helix than in the unmodified form, recombinant murine MBP C1 mimic (rmC1), analogous to the most cationic and abundant isomer of MBP in normal myelin. Moreover, cathepsin D digested lipid-associated rmC8 3-fold faster than rmC1, and cleavage at F86–F87 occurred more readily in rmC8 than rmC1. These findings suggest a mechanism for initial loss of myelin stability and the autoimmune pathogenesis of multiple sclerosis.

Keywords: electron paramagnetic resonance, site-directed spin labeling, lipid bilayer, α-helix, cathepsin D

The human inflammatory demyelinating disease multiple sclerosis (MS) is characterized by the active degradation of the myelin sheath in the CNS, resulting in significant neurological deficit (1–4). The resultant lesions contain T cells, B cells, and macrophages that are reactive against myelin antigens. Of the candidate autoantigens, myelin basic protein (MBP) (5) has been widely studied, and its importance to the immunopathogenesis of MS has been extensively reviewed (3, 6, 7). MBP and its peptides are highly encephalitogenic and induce experimental autoimmune encephalomyelitis, used widely as a model for MS, in rodents and primates.

The primary role of MBP in the CNS is generally considered to be maintenance of compaction of the myelin sheath, bringing together the apposing faces of the cytoplasmic leaflet of the oligodendrocyte membrane, by virtue of its extreme net positive charge and through synergistic protein–lipid interactions (8–10). Alteration of MBP cationicity may represent a regulatory mechanism for normal myelin assembly or a degradative mechanism in MS. In this regard, MBP shows extensive posttranslational modifications with varying degrees of deimination, phosphorylation, deamidation, methylation, and N-terminal acylation (5, 11). Such modifications give rise to charge variants denoted C1 to C8, of which the C1 component represents the least modified and the most cationic isomer and is the most abundant form of MBP in adult humans (11, 12). The least cationic component is C8, with extensive deimination of arginyl residues (13), and it is found in elevated levels in both adults with MS and infants (12), suggesting a role in the early stages of myelinogenesis and in demyelination in MS.

The severity of MS correlates strongly with the degree of arginine loss caused by deimination in fulminating Marburg’s syndrome: >90% of the MBP is deiminated, compared with 45% in chronic MS and 20% in normal brain (12, 14). Deiminated MBP is structurally less ordered and more susceptible to proteolysis under solution conditions (15–19), has decreased ability to cause adhesion of apposed membranes, and even causes membrane fragmentation (8, 13, 18, 20–22). However, precise understanding of how this modification contributes toward myelin degradation and autoimmune pathogenesis in MS still remains limited because of the absence of detailed information on the structure of MBP in its natural environment (5).

To this end, we have exploited the charge differences of two recombinant isomers, recombinant murine MBP C1 mimic (rmC1) and recombinant murine MBP C8 mimic (rmC8), of murine MBP corresponding to the most cationic, C1, and least cationic, C8, variants of the protein, respectively. Reduction in net positive charge in rmC8 was accomplished with five R → Q and one K → Q substitutions (“quasi-deimination”) at sites of arginine deimination as found in the C8 isomer of MBP (13, 18). Using site-directed spin labeling (SDSL) and EPR spectroscopy (23, 24), we have shown that this deiminated MBP mimic results in the central and C-terminal regions of the protein being significantly less associated with myelin-like membranes (19). A detailed SDSL/EPR analysis of the primary immunodominant epitope of MBP (V83–T92, murine sequence numbering) of the highly cationic rmC1 variant showed it to be an amphipathic α-helix lying underneath the membrane surface, at a 9° tilt (25). This segment of the protein represents the minimal epitope for T cell recognition of human MBP (residues V86–T98, human sequence numbering) with the highest affinity for the MHC class II haplotype, HLA-DR, believed to be associated with increased susceptibility to MS (26–29). The linear polypeptide D82-ENPVVHFFKNIVTPR-T98 (human numbering, corresponding to D79–T95 in the murine sequence) has been used to induce immunologic tolerance in patients with progressive MS (30).

In this work, we have investigated how the reduced net positive charge of MBP, as seen in MS, affects the membrane disposition of this primary immunodominant epitope. We derive through SDSL/EPR analysis a membrane topological map of this epitope in the less cationic rmC8 MBP isomer, which differs significantly from that in the highly cationic rmC1. Additionally, we demonstrate enhanced proteolysis of the rmC8 protein in the presence of lipids. These results suggest that the reduction in cationicity of deiminated MBP not only impedes the membrane adhesion and assembly activity of this protein, but also exposes an immunodominant epitope in the membrane-bound protein to proteases. This exposure may cause this highly encephalitogenic epitope to be released to prime the innate immune-derived cells of the CNS and sensitize peripheral T cells.

Results

SDSL EPR of rmC8.

To investigate the membrane-bound conformational state of the immunodominant epitope of the less cationic rmC8 MBP isomer, 10 consecutive Cys replacement mutants (V83C, V84C, H85C, F86C, F87C, K88C, N89C, I90C, V91C, and T92C) were generated for Cys-specific spin labeling. The quasi-deiminated rmC8 differs from the highly cationic rmC1 with R/K → Q substitutions at R25, R33, K119, R127, R157 and R168, all of which are highly conserved arginine deimination sites in human 18.5-kDa MBP (11, 18). Given that there is no codon for citrulline, the substitution of glutamine for arginine is a good compromise for generating site-specific deimination in recombinant proteins (15, 18).

Solution EPR spectra of all mutants resulted in sharp hyperfine lines and narrow center-line widths (ΔH = 2.0 G; data not shown), indicating that this region of the protein was largely unstructured in aqueous solution, consistent with our earlier findings (18). Incubation of the spin-labeled proteins with Cyt-large unilamellar vesicles (LUVs), used to mimic the cytoplasmic leaflet of the myelin sheath, resulted in aggregation of the vesicles. To ensure that all labeled proteins were completely membrane-bound, a high lipid/protein ratio (600:1 molar ratio) was used.

The observed EPR spectra (Fig. 1) of the membrane-bound mutants show markedly broader and more asymmetric lines than in aqueous solution and reduction in peak-to-peak amplitude, indicating anisotropic motion of the spin label (R1). To compare quantitatively the R1 rotational dynamics of the different labeling sites, scaled side-chain mobility (Ms) values were determined from the center-line widths (24). This scaled parameter provides an approximate measure of R1 dynamics from Ms = 1 to Ms = 0 for the most mobile and immobile sites, respectively. Additional motional parameters were also obtained from the peak-to-peak amplitude ratio of the center-line (h0) and the high (h−1) field hyperfine line. The estimated R1 mobility parameters of most side chains showed markedly decreased rotational dynamics upon association with the membrane, with scaled R1 mobility (Ms) values ranging from 0.03 to 0.27 and a peak-to-peak amplitude ratio (h0/h−1) range from 7.5 to 9.7, indicating an increased degree of order in this region upon lipid association. Such anisotropic motion of R1 is expected to arise from a combination of both the formation of secondary structure in this region and the restriction of R1 motion by the interaction of the proteins with the surface of the membrane. In addition, the spectrum of N89R1 shows a more immobilized component that is characteristic of side-chain mobility restriction caused by a tertiary contact, or steric hindrance of probe motion imposed by other side chains. Similar spectral properties were previously observed for this site in the membrane-bound state of rmC1 (25). The spectra of spin labels at positions V83R1 and V84R1 have two components, one with sharp lines because of a mobile population (Fig. 1), and one with broad lines and a larger hyperfine splitting because of a population with more restricted anisotropic motion (Fig. 1). In contrast to the rotational dynamics of other side chains, the estimated Ms and h0/h−1 values for these residues were on average 5-fold higher and 3-fold lower, respectively, than at other sites, because of the contribution of the more mobile population. Given that the proteins were spin-labeled on the column, and free spin probes were washed off before elution of labeled protein, the highly isotropic line shapes seen for V83R1 and V84R1 are not the result of free spin labels. Isotropic line shapes could also arise from contributions of membrane-unassociated spin-labeled proteins; however, the absence of these line shapes in other mutants and the use of aggregated complexes at a high lipid–protein ratio discount such contributions. Thus, the observed spectral properties of V83R1 and V841 may reflect conformational variability caused by two different environments for the N terminus of this epitope when the protein is membrane-bound.

Fig. 1.

Fig. 1.

Open in a new tab

First-derivative EPR spectra of the spin-labeled rmC8 mutants associated with Cyt-LUVs. Most lipid-associated R1 spectra show marked broadening and less defined hyperfine peaks with increased hyperfine splitting compared with that in solution, indicating motional restriction. The spectra for V83R1 and V84R1 contain isotropic line shapes (asterisks), as well as an additional broader spectral component (arrowheads) that indicates anisotropic motional dynamics. The arrows indicate a more immobilized component in the spectrum of N89R1 caused by tertiary contact and/or steric hindrance from other side chains. All spectra were normalized to the amplitude of the center line.

Solvent Accessibility and Membrane Topology.

Continuous power saturation EPR spectroscopy was used to determine the relative accessibility (Π) of the 10 spin-labeled side chains (R1) to nickel-ethylenediaminediacetic acid (NiEDDA) and O2 (Fig. 2). Collision of the spin labels with paramagnetic reagents changes the relaxation rate of R1 and reduces the resonance saturation as the magnetic field strength is increased (24, 31). Given the differential solubility of NiEDDA (water-soluble) and O2 (lipid-soluble), their collision frequencies with R1 provide a measure of the degree of surface exposure or membrane insertion of the probed side chain. The overall patterns of ΠNiEDDA and ΠO2 depict an amphipathic orientation, surface-exposed (high ΠNiEDDA and low ΠO2) and relatively buried residues (high ΠO2 and low ΠNiEDDA), which for the most part vary in a periodic manner indicative of a regular secondary structure.

Fig. 2.

Fig. 2.

Open in a new tab

Comparison of the accessibility to NiEDDA and O2 for R1 of rmC8 mutants bound to Cyt-LUVs. Collisional parameters for O2 (ΠO2) (A) and NiEDDA (ΠNiEDDA) (B) as functions of the probed side-chain residues. The observed ΠNiEDDA and ΠO2 were fitted to harmonic sine wave functions to predict the secondary structures from the periodicity of the fit. The shaded areas highlight the region predicted to form an α-helix. Error bars indicate the SDs from triplicate measurements.

To determine the overall membrane topology of this segment of the protein, an immersion depth parameter (Φ), calculated as the natural logarithm of the ratio of ΠNiEDDA and ΠO2, was determined and compared with a standard curve of known immersion depths generated by using various spin-labeled lipids in Cyt-LUVs (Fig. 3A). The dependence of this immersion depth parameter (Φ) on the distance of R1 from the membrane surface follows a hyperbolic tangent function that describes the limiting behavior of Φ that can be applied to sites on the aqueous and the membrane interior sides of the bilayer (19, 31). At the bulk aqueous face, the fitting function was constrained by the Φ observed for 1,2-dipalmitoyl-sn-glycero-3-phospho(TEMPO)choline (TEMPO-PC); thus, additional extrapolation was required to determine the distances of the more aqueous-exposed side chains with lower Φ values than that of TEMPO-PC. This second calibration was achieved by linear regression of the data for TEMPO-PC, 5-doxyl-phosphatidylcholine (PC), and 7-doxyl-PC (Fig. 3A, straight line) (19, 25). Although distance measured in this manner is only an estimate, the observed Φ values (ranging from −0.64 to −1.68) of these more exposed sites are well within the bulk aqueous value of Φ previously determined for 3-carboxyproxyl in solution (31).

Fig. 3.

Fig. 3.

Open in a new tab

Membrane immersion distance measurements. (A) Standard immersion distance calibration curve of doxyl-labeled PC lipids (•) of known depth (TEMPO-PC, 5-doxyl-PC, 7-doxyl-PC, 10-doxyl-PC, and 12-doxyl-PC), using Cyt-LUVs in the presence of unlabeled rmC8. Also shown are the data for spin-labeled rmC8 (○). To determine the immersion distance of the rmC8 mutants, the data from the standard lipids were fitted to a hyperbolic tangent function (curved line) (19). The distances of measured bulk phase Φ values outside the standard curve were estimated from a linear regression (straight line) of the data for TEMPO-PC, 5-doxyl-PC, and 7-doxyl-PC (19). (B) The depth immersion of rmC8 mutants into the lipid bilayer of the Cyt-LUVs (○) fitted to a harmonic wave function. Positive and negative values refer to distances below and above the lipid phosphate headgroups of the membrane, respectively. The shaded area highlights the region predicted to form an amphipathic α-helix. (C) Comparison of the spin-label depths of rmC8 (○) and rmC1 (■), as determined (25), to show that the probed epitope is significantly more surface-exposed in the rmC8 isomer of recombinant murine MBP than in rmC1. Error bars indicate the SDs of triplicate measurements.

Fig. 3B shows the calibrated depth (relative to the phosphate headgroups) of this segment of the protein in rmC8 associated with Cyt-LUVs. Negative and positive values indicate distances above and below the surface of the membrane, respectively. The side chains V83R1, V84R1, H85R1, K88R1, and N89R1 reside at distances >5 Å above the surface of the membrane and are thus highly surface exposed. The side chains F86R1 and T92R1 reside at the interfacial layer near the phosphate headgroups of the membrane, whereas F87R1, I90R1, and V91R1 penetrate into the lipid bilayer in the 5- to 9-Å range. In Fig. 3C, we compare the measured distances of this epitope in rmC8 to those reported for rmC1 (25). Whereas the majority of residues in this epitope in rmC1 lie beneath the surface of the membrane, with an estimated depth penetration of up to 10 Å, this segment is highly surface-exposed in the less cationic rmC8, indicating its topological dependence on the net positive charge of the whole protein.

Predicting Secondary Structure Elements from the Observed EPR Parameters.

Previously, we modeled the entire V83–T92 segment of rmC1 as an amphipathic α-helix, using the periodicity pattern of the observed EPR parameters when the protein was mixed with Cyt-LUVs (25). Here, the observed EPR parameters for membrane-bound rmC8 were also fitted to a harmonic sine wave function. The fitted data reveal the formation of a short α-helix (P = 3.6 ± 0.1 residues per turn), with average helical boundaries encompassing only residues H85–T92 (Table 1). The average amplitude of this short helix was estimated to be ≈13 ± 3 Å, consistent with the diameter of a spin-labeled α-helix (32). The EPR parameters for V83R1 and V84R1 could not be fitted as part of an α-helix, and the observed lineshape spectra and estimated motional parameters of these two residues were consistent with a high degree of backbone disorder. The two populations with different motional dynamics seen in the spectra of V83R1 and V84R1 have also been seen for loop localized spin-labeled residues in the TonB box of the vitamin B12 transporter BtuB (33). Therefore, the multicomponent spectra of V83R1 and V84R1 may reflect a dynamic equilibrium of flexible loop and transient helical conformations at the N terminus of the probed segment of rmC8.

Table 1.

Summary of nonlinear least-squares harmonic wave function analysis of the observed EPR parameters

Parameters Fitting results
p, rpt* ω, °† Residues‡
ΔP1/2 NiEDDA 3.9 92.5 85–92
ΔP1/2 O2 3.1 115.8 85–92
Π NiEDDA 3.8 94.4 85–92
Π O2 3.2 112.5 85–91
Φ 3.7 98.4 85–92
Distance 3.6 100.6 85–91
Average 3.6 ± 0.1 102.3 ± 3.9
Open in a new tab

*The residue per turn (rpt) periodicity of the parameters.

†Angular frequency of the periodicity.

‡Residues whose parameters were used in the fitting process.

Proteolysis of Lipid-Associated MBP Charge Isomers with Cathepsin D.

To gain insight into a possible mechanism of MBP antigen release, we digested the membrane-bound rmC8 and rmC1 isomers with cathepsin D (EC 3.4.23.5) (Fig. 4). Cathepsin D cleaves MBP at F42–F43 and F86–F87, the latter cut site being part of the presently probed epitope (15, 16, 18). Cleavage of the intact protein is expected to release peptides with different molecular masses depending on the cut site (F42–F43 and/or F86–F87). The two bands shown in Fig. 4A are the intact protein (denoted α) and the Mr ≈14.4 kDa peptide (denoted β) expected to result from the single cleavage of the protein at F42–F43. Shown in Fig. 4B are the relative band densities of the intact protein after digestion fitted to an exponential decay function (main plot) and the appearance of the β peptide and its subsequent processing at F86–F87 (Inset). Cathepsin D digested membrane-bound rmC8 ≈3-fold faster than rmC1, consistent with its greater degree of overall exposure. The appearance of the β peptide in the rmC1 digest followed a gradual increase to a steady-state level after 3 h of digestion, indicating that the F86–F87 cut site in the unmodified protein is highly protected in the membrane-bound state. Digestion of rmC8 resulted in a rapid increase in the amount of the β peptide to a steady-state level after only 3 h, followed by a rapid decline in band intensities to nearly zero after 18 h caused by further digestion at F86–F87. These observations suggest that the F86–F87 cathepsin D cut site of this isomer is highly surface-exposed and readily proteolysed even when membrane-bound. These observations corroborate our EPR data with respect to the surface topologies of the V83–P92 epitope of rmC8 vs. rmC1.

Fig. 4.

Fig. 4.

Open in a new tab

Cathepsin D digestion of membrane-bound rmC8 and rmC1. (A) Stained Tricine gel showing the time course digestion of rmC8 and rmC1 in the presence of Cyt-LUVs. Cleavage of the intact protein (denoted α, Mr ≈19.5 kDa) is expected to release a combination of five bands depending on the cut site (F42–F43 and/or F86–F87). Single cleavage at F42–F43 releases the β peptide (Mr ≈14.4 kDa), which can be subsequently digested at F86–F87 (the probed epitope). (B) Quantification of the digestion time course of intact rmC8 (○) and rmC1 (•) fitted to an exponential decay function as well as the appearance of the β peptide (Inset) and its loss caused by subsequent processing at F86–F87 (Inset). Error bars indicate the SD of triplicate measurements.

Discussion

In search of structural insight into the role of MBP deimination in autoimmune pathogenesis, we have characterized the membrane-bound conformational state of the primary immunodominant epitope of MBP (V83–T92, murine numbering) using EPR spectroscopy and proteolysis of a recombinant protein that mimics the natural deiminated component. The T and B cells found in the CNS of MS patients show consistent specificity for this minimal epitope, and antibodies against it are found in both the brain and spinal fluid of MS patients, indicating both humoral and cellular immune responses in disease pathogenesis (34, 35). Detailed requirements for the binding of this epitope to HLA class II molecules and autoimmune T cell receptor (TCR) were provided by the x-ray crystallographic analysis of this peptide complexed with the two MHC class II isotypes, HLA-DRa and HLA-DRb, and TCR (36, 37).

Our present SDSL/EPR and proteolysis data of the membrane-bound quasi-diminated isomer of MBP (rmC8) reveal that the membrane depth penetration and the length of the amphipathic α-helix formed by this epitope in rmC8 differ significantly from the unmodified rmC1 charge isomer. The rmC8 helix is largely surface-exposed with the helix backbone residing in the membrane/aqueous interface. Whereas the entire V83–T92 sequence of the membrane-bound rmC1 isomer was mapped to an α-helix, the observed EPR parameters for the rmC8 mutants indicated only a short α-helix consisting of residues H85–T92. In rmC8, the V83 and V84 side chains are highly surface-exposed, with measured R1 distances of 8.5 and 12.5 Å above the surface of the membrane, respectively. This result is in contrast to the deep membrane penetration (≈8.0 Å below the surface of the membrane) of these two residues and the rest of this domain in rmC1 (25).

This region of the protein is largely unstructured in aqueous solution (25, 38). Its helicity is stabilized by the low dielectric environment afforded by trifluoroethanol or its immersion into the lipid bilayer when the protein associates with the membrane. Given this environmental dependence of the helical conformation of this segment, the EPR data for membrane-bound V83C and V84C mutants were interpreted as indicating their occurrence in a loop structure and exposure to the aqueous milieu. Movement of these two residues of membrane-bound rmC8 out of the bilayer would be expected to destabilize the helical propensity of the N terminus of the V83–T92 epitope (Fig. 5).

Fig. 5.

Fig. 5.

Open in a new tab

Schematic illustrating the dependence on the overall net positive charge of the protein of the membrane disposition of the primary immunodominant epitope of MBP, segment V83–T92 (murine numbering). The interfacial layer is represented by small triangles. Reduction in global net positive charge of MBP by deimination reduces the overall helicity of this epitope and renders it highly surface-exposed and susceptible to proteolysis. Consequently, there is both loss of CNS myelin compaction and the release of highly encephalitogenic self-antigens, which can sensitize T cells and elicit/exacerbate autoimmune responses against myelin components.

Another striking difference between the membrane-bound rmC1 and rmC8 isomers of MBP is the degree of susceptibility to cathepsin D digestion. Under solution conditions, the digestion rate of MBP depends on its citrulline content (15, 16, 18). For instance, C8 components with 6 and 18 deiminated arginines isolated from MS brain were digested 4- and 35-fold faster than the C1 isomer, respectively (15). Here, although the overall digestion time scale was much longer than under solution conditions the 3-fold increase in the rate of cathepsin D proteolysis of the membrane-bound rmC8 relative to rmC1 is in agreement with earlier data. Given the observed topological difference of the V83–T92 epitope in the membrane-bound rmC1 and rmC8 isomers, we had expected that the relatively buried F86–F87 cut site of rmC1 would be largely protected from cathepsin D digestion, whereas in rmC8 this cut site would be readily digestible. The proteolysis data of the two membrane-bound proteins confirm this prediction and corroborate our EPR data.

The mutations (R/K → Q) in rmC8 used to mimic the deimination of MBP in vivo were made at sites outside of the V83–T92 epitope. Thus, the observed conformational differences of this epitope in the membrane-bound rmC1 and rmC8 are affected by the overall net positive charge and global structural differences of the two membrane-bound proteins. The replacement of only six arginine citrullination sites with glutamine in rmC8 represents a conservative net charge reduction of the protein. Recently, MS analysis has shown that up to 10 of the 19 arginine residues in human MBP are at least partially deiminated in chronic MS (11). Moreover, 18 of the 19 Arg residues in MBP are deiminated in the severe fulminating form of MS (11, 14). Such further reduction in the net positive charge of the protein is expected to lead to more pronounced membrane-bound conformational perturbation, and thus to a higher degree of surface exposure of antigenic epitopes. These highly encephalitogenic epitopes can thus be readily released from the deiminated protein as a consequence of its increased susceptibility to proteolysis even when bound to the membrane.

Traditionally, MS is considered to be an autoimmune disease in which autoreactive immune cells, targeting components of the myelin sheath, propagate a destructive process within the CNS. However, the mechanisms by which myelin antigens are released and targeted immune responses are initiated or sustained in MS patients remain unidentified. It has been suggested that immune responses to viral-derived and/or bacterial-derived antigens with significant structural homology (molecular mimicry) to MBP peptides or other myelin antigens may trigger the initiation of autoimmunity in MS patients (39), but direct involvement of myelin-derived autoantigens is still required for activation of autoreactive T cells and disease induction (40). It has recently been proposed that decreased myelin compaction caused by aberrant posttranslational modification of MBP may initiate a neurodegenerative process that makes myelin sheaths susceptible to degradation in MS (41). In support of this hypothesis, we have shown that the reduction in the net positive charge of MBP results in not only hindrance to compact myelin assembly, but also renders the most immunodominant epitope of this protein highly surface-exposed and readily digestible by myelin-associated proteases, as illustrated schematically in Fig. 5. The antigenic peptides released could initiate or sustain the immune response. Our work thus sheds light onto a plausible role for the observed reduced cationicity of MBP in MS in the generation of self-antigens known to sensitize T cells in MS. On a final note, SDSL/EPR spectroscopy (23, 24) and site-specific proteolysis, as illustrated here, represent powerful tools for probing changes in the topologies of membrane proteins caused by disease-associated alterations.

Methods

Expression, Purification, and SDSL of rmC8 Mutants.

The single cysteine mutant proteins of rmC8 were purified from BL21-CodonPlus (DE3)-RP Escherichia coli (18, 42). Spin labeling of mutants was performed on the column matrix (19, 25), using a 10-fold molar excess of [1-oxyl-2,2,5,5-tetramethyl-d-pyrroline-3-methyl]methanethiosulfonate (Toronto Research Chemicals, Toronto) in 20 mM Hepes-NaOH (pH 7.4), 6 M urea, and 10 mM NaCl, at 4°C overnight. The column was then repacked and washed to remove the excess spin probe, and labeled proteins were eluted and dialyzed against 20 mM Hepes-NaOH (pH 7.4) and 10 mM NaCl.

Preparation of LUVs.

LUVs with lipid composition similar to that of the cytoplasmic monolayer of myelin (Cyt-LUVs; cholesterol/phosphatidylethanolamine/phosphatidylserine/PC/sphingomyelin/phosphatidylinositol in 44:27:13:11:3:2% molar percentages) were prepared by extrusion (8, 19, 25). Lipids were purchased from Avanti Polar Lipids and prepared in 20 mM Hepes-NaOH, 10 mM NaCl, pH 7.4, for EPR experiments or in 50 mM sodium acetate, 10 mM NaCl, pH 4.0, for proteolysis with cathepsin D. For distance calibration, the Cyt-LUVs contained 0.2% spin-labeled lipid, also purchased from Avanti Polar Lipids.

Proteolysis of Lipid-Associated MBP Charge Isomers with Cathepsin D.

Lipid-associated rmC1 and rmC8 were digested with bovine kidney cathepsin D (EC 3.4.23.5) (Calbiochem) under similar conditions as described (18). Briefly, 100 ng of cathepsin D was mixed with an aliquot of Cyt-LUVs in 50 mM sodium acetate buffer, pH 4.0, to which subsequently were added 3 μg of rmC1 or rmC8 at a final molar MBP/lipid ratio of 1:600, to give a 20-μl reaction volume in separate tubes for each time period. The samples were mixed, gently flushed with N2 to avoid lipid oxidation during the reaction period, and placed in a water bath at 37°C. The reaction was allowed to proceed for up to 18 h, during which time sample tubes were removed at various time intervals, and reaction was stopped by the addition of 5 μl of 5× Tricine-PAGE loading buffer, followed by 5 min of boiling. The extent of digestion of each protein was monitored by Tricine-PAGE (16.5%). Stained gels were digitized, and densitometry was performed with scion imaging software, release Beta 4.02 (Scion, Frederick, MD) for quantitative comparison.

EPR Spectroscopy.

All EPR measurements were made with a Bruker ECS 106 spectrometer equipped with a loop-gap resonator, and spectra were recorded at room temperature with a modulation amplitude of 1.0 G. MBP was added to Cyt-LUVs at a molar protein/lipid ratio of 1:600, and the sample pellets were taken up in TPX methylpentene capillary tubes. Details of sample preparation, data collection, and analysis have been described (19, 25, 31, 43).

Predicting Secondary Structure from EPR Parameters.

The secondary structure elements within the probed epitope of MBP were predicted from the observed EPR parameters by using a method adopted from Cornette et al. (44). Briefly, the periodicity and the angular frequency of the observed EPR parameters were obtained through a least-squares fit of the data by using a harmonic wave function:

graphic file with name zpq01206-1464-m01.jpg

where a is the amplitude, b is the phase, p is the period, and c is an offset value. All calculations were made by using a Microsoft excel 2002 template spreadsheet developed in-house by Uwe Oehler (University of Guelph).

Acknowledgments

This work was supported by the Natural Sciences and Engineering Research Council of Canada and the Canadian Institutes of Health Research (G.H. and J.M.B.). A.A.M. was a recipient of a Multiple Sclerosis Society of Canada Ph.D. Studentship.

Abbreviations

LUV

large unilamellar vesicle

MBP

myelin basic protein

MS

multiple sclerosis

NiEDDA

nickel-ethylenediaminediacetic acid

rmC1

recombinant murine MBP C1 mimic

rmC8

recombinant murine MBP C8 mimic

SDSL

site-directed spin labeling

PC

phosphatidylcholine

TEMPO-PC

1,2-dipalmitoyl-sn-glycero-3-phospho(TEMPO)choline.

Conflict of interest statement: No conflicts declared.

This paper was submitted directly (Track II) to the PNAS office.

See Commentary on page 4339.

References

  • 1.Trapp B. D., Peterson J., Ransohoff R. M., Rudick R., Mork S., Bo L. N. Engl. J. Med. 1998;338:278–285. doi: 10.1056/NEJM199801293380502. [DOI] [PubMed] [Google Scholar]
  • 2.Kornek B., Lassmann H. Brain Res. Bull. 2003;61:321–326. doi: 10.1016/s0361-9230(03)00095-9. [DOI] [PubMed] [Google Scholar]
  • 3.Lutton J. D., Winston R., Rodman T. C. Exp. Biol. Med. 2004;229:12–20. doi: 10.1177/153537020422900102. [DOI] [PubMed] [Google Scholar]
  • 4.Hafler D. A., Slavik J. M., Anderson D. E., O’Connor K. C., De Jager P., Baecher-Allan C. Immunol. Rev. 2005;204:208–231. doi: 10.1111/j.0105-2896.2005.00240.x. [DOI] [PubMed] [Google Scholar]
  • 5.Harauz G., Ishiyama N., Hill C. M., Bates I. R., Libich D. S., Farès C. Micron. 2004;35:503–542. doi: 10.1016/j.micron.2004.04.005. [DOI] [PubMed] [Google Scholar]
  • 6.Steinman L., Waisman A., Altmann D. Mol. Med. Today. 1995;1:79–83. doi: 10.1016/s1357-4310(95)92366-7. [DOI] [PubMed] [Google Scholar]
  • 7.Sospedra M., Martin R. Annu. Rev. Immunol. 2005;23:683–747. doi: 10.1146/annurev.immunol.23.021704.115707. [DOI] [PubMed] [Google Scholar]
  • 8.Boggs J. M., Yip P. M., Rangaraj G., Jo E. Biochemistry. 1997;36:5065–5071. doi: 10.1021/bi962649f. [DOI] [PubMed] [Google Scholar]
  • 9.Riccio P., Fasano A., Borenshtein N., Bleve-Zacheo T., Kirschner D. A. J. Neurosci. Res. 2000;59:513–521. doi: 10.1002/(SICI)1097-4547(20000215)59:4<513::AID-JNR6>3.0.CO;2-M. [DOI] [PubMed] [Google Scholar]
  • 10.Hu Y., Doudevski I., Wood D., Moscarello M., Husted C., Genain C., Zasadzinski J. A., Israelachvili J. Proc. Natl. Acad. Sci. USA. 2004;101:13466–13471. doi: 10.1073/pnas.0405665101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Kim J. K., Mastronardi F. G., Wood D. D., Lubman D. M., Zand R., Moscarello M. A. Mol. Cell. Proteomics. 2003;2:453–462. doi: 10.1074/mcp.M200050-MCP200. [DOI] [PubMed] [Google Scholar]
  • 12.Moscarello M. A., Wood D. D., Ackerley C., Boulias C. J. Clin. Invest. 1994;94:146–154. doi: 10.1172/JCI117300. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Wood D. D., Moscarello M. A. J. Biol. Chem. 1989;264:5121–5127. [PubMed] [Google Scholar]
  • 14.Wood D. D., Bilbao J. M., O’Connors P., Moscarello M. A. Ann. Neurol. 1996;40:18–24. doi: 10.1002/ana.410400106. [DOI] [PubMed] [Google Scholar]
  • 15.Cao L., Goodin R., Wood D., Moscarello M. A., Whitaker J. N. Biochemistry. 1999;38:6157–6163. doi: 10.1021/bi982960s. [DOI] [PubMed] [Google Scholar]
  • 16.Pritzker L. B., Joshi S., Gowan J. J., Harauz G., Moscarello M. A. Biochemistry. 2000;39:5374–5381. doi: 10.1021/bi9925569. [DOI] [PubMed] [Google Scholar]
  • 17.Pritzker L. B., Joshi S., Harauz G., Moscarello M. A. Biochemistry. 2000;39:5382–5388. doi: 10.1021/bi9925571. [DOI] [PubMed] [Google Scholar]
  • 18.Bates I. R., Libich D. S., Wood D. D., Moscarello M. A., Harauz G. Protein Expression Purif. 2002;25:330–341. doi: 10.1016/s1046-5928(02)00017-7. [DOI] [PubMed] [Google Scholar]
  • 19.Bates I. R., Boggs J. M., Feix J. B., Harauz G. J. Biol. Chem. 2003;278:29041–29047. doi: 10.1074/jbc.M302766200. [DOI] [PubMed] [Google Scholar]
  • 20.Beniac D. R., Wood D. D., Palaniyar N., Ottensmeyer F. P., Moscarello M. A., Harauz G. J. Struct. Biol. 2000;129:80–95. doi: 10.1006/jsbi.1999.4200. [DOI] [PubMed] [Google Scholar]
  • 21.Boggs J. M., Rangaraj G., Koshy K. M., Ackerley C., Wood D. D., Moscarello M. A. J. Neurosci. Res. 1999;57:529–535. [PubMed] [Google Scholar]
  • 22.Shanshiashvili L. V., Suknidze N. C., Machaidze G. G., Mikeladze D. G., Ramsden J. J. Arch. Biochem. Biophys. 2003;419:170–177. doi: 10.1016/j.abb.2003.09.020. [DOI] [PubMed] [Google Scholar]
  • 23.Malmberg N. J., Falke J. J. Annu. Rev. Biophys. Biomol. Struct. 2005;34:71–90. doi: 10.1146/annurev.biophys.34.040204.144534. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Hubbell W. L., Cafiso D. S., Altenbach C. Nat. Struct. Biol. 2000;7:735–739. doi: 10.1038/78956. [DOI] [PubMed] [Google Scholar]
  • 25.Bates I. R., Feix J. B., Boggs J. M., Harauz G. J. Biol. Chem. 2004;279:5757–5764. doi: 10.1074/jbc.M311504200. [DOI] [PubMed] [Google Scholar]
  • 26.Valli A., Sette A., Kappos L., Oseroff C., Sidney J., Miescher G., Hochberger M., Albert E. D., Adorini L. J. Clin. Invest. 1993;91:616–628. doi: 10.1172/JCI116242. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Wucherpfennig K. W., Sette A., Southwood S., Oseroff C., Matsui M., Strominger J. L., Hafler D. A. J. Exp. Med. 1994;179:279–290. doi: 10.1084/jem.179.1.279. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Ota K., Matsui M., Milford E. L., Mackin G. A., Weiner H. L., Hafler D. A. Nature. 1990;346:183–187. doi: 10.1038/346183a0. [DOI] [PubMed] [Google Scholar]
  • 29.Lang H. L., Jacobsen H., Ikemizu S., Andersson C., Harlos K., Madsen L., Hjorth P., Sondergaard L., Svejgaard A., Wucherpfennig K., et al. Nat. Immunol. 2002;3:940–943. doi: 10.1038/ni835. [DOI] [PubMed] [Google Scholar]
  • 30.Warren K. G., Catz I. Multiple Sclerosis. 2000;6:300–311. doi: 10.1177/135245850000600502. [DOI] [PubMed] [Google Scholar]
  • 31.Frazier A. A., Wisner M. A., Malmberg N. J., Victor K. G., Fanucci G. E., Nalefski E. A., Falke J. J., Cafiso D. S. Biochemistry. 2002;41:6282–6292. doi: 10.1021/bi0160821. [DOI] [PubMed] [Google Scholar]
  • 32.Rabenstein M. D., Shin Y. K. Proc. Natl. Acad. Sci. USA. 1995;92:8239–8243. doi: 10.1073/pnas.92.18.8239. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Merianos H. J., Cadieux N., Lin C. H., Kadner R. J., Cafiso D. S. Nat. Struct. Biol. 2000;7:205–209. doi: 10.1038/73309. [DOI] [PubMed] [Google Scholar]
  • 34.Martino G., Olsson T., Fredrikson S., Hojeberg B., Kostulas V., Grimaldi L. M., Link H. Eur. J. Immunol. 1991;21:2971–2976. doi: 10.1002/eji.1830211211. [DOI] [PubMed] [Google Scholar]
  • 35.Wucherpfennig K. W., Catz I., Hausmann S., Strominger J. L., Steinman L., Warren K. G. J. Clin. Invest. 1997;100:1114–1122. doi: 10.1172/JCI119622. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Hahn M., Nicholson M. J., Pyrdol J., Wucherpfennig K. W. Nat. Immunol. 2005;6:490–496. doi: 10.1038/ni1187. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Li Y., Huang Y., Lue J., Quandt J. A., Martin R., Mariuzza R. A. EMBO J. 2005;24:2968–2979. doi: 10.1038/sj.emboj.7600771. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Farès C., Libich D. S., Harauz G. FEBS J. 2006;273:601–614. doi: 10.1111/j.1742-4658.2005.05093.x. [DOI] [PubMed] [Google Scholar]
  • 39.Wucherpfennig K. W., Strominger J. L. Cell. 1995;80:695–705. doi: 10.1016/0092-8674(95)90348-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Carrizosa A. M., Nicholson L. B., Farzan M., Southwood S., Sette A., Sobel R. A., Kuchroo V. K. J. Immunol. 1998;161:3307–3314. [PubMed] [Google Scholar]
  • 41.Mastronardi F. G., Moscarello M. A. J. Neurosci. Res. 2005;80:301–308. doi: 10.1002/jnr.20420. [DOI] [PubMed] [Google Scholar]
  • 42.Bates I. R., Matharu P., Ishiyama N., Rochon D., Wood D. D., Polverini E., Moscarello M. A., Viner N. J., Harauz G. Protein Expression Purif. 2000;20:285–299. doi: 10.1006/prep.2000.1307. [DOI] [PubMed] [Google Scholar]
  • 43.Oh K. J., Altenbach C., Collier R. J., Hubbell W. L. Methods Mol. Biol. 2000;145:147–169. doi: 10.1385/1-59259-052-7:147. [DOI] [PubMed] [Google Scholar]
  • 44.Cornette J. L., Cease K. B., Margalit H., Spouge J. L., Berzofsky J. A., DeLisi C. J. Mol. Biol. 1987;195:659–685. doi: 10.1016/0022-2836(87)90189-6. [DOI] [PubMed] [Google Scholar]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES