pubs.acs.org/BiochemistryPublished on Web 07/02/2010
r2010 American Chemical Society
Biochemistry 2010, 49, 6903–6915
Interaction of Myelin Basic Protein with Actin in the Presence of
Vladimir V. Bamm,‡,#Mumdooh A. M. Ahmed,§,),^,#and George Harauz*,‡,)
‡Department of Molecular and Cellular Biology,§Department of Physics, and
University of Guelph, 50 Stone Road East, Guelph, Ontario N1G 2W1, Canada, and^Department of Physics,
Faculty of Science at Suez, Suez-Canal University, Suez, Egypt.#These authors contributed equally to this work.
Biophysics Interdepartmental Group,
Received March 2, 2010; Revised Manuscript Received June 28, 2010
ABSTRACT: The 18.5 kDa myelin basic protein (MBP), the most abundant splice isoform in human adult
myelin, is a multifunctional, intrinsically disordered protein that maintains compact assembly of the myelin
sheath in the central nervous system. Protein deimination and phosphorylation are two key posttranslational
modifications whose balance determines local myelin microdomain stability and function. It has previously
been shown that MBP in solution causes both polymerization of G-actin to F-actin and bundling of the
roles of different possible interacting domains of membrane-associated MBP, have not yet been investigated.
MBP (full-length charge variants), and of two terminal deletion variants (rmΔC and rmΔN), with actin in
the presence of DPC (dodecylphosphocholine) to mimic a membrane environment. Our results show that
although both charge variants polymerized and bundled actin, the maximal polymerization/bundling due to
to G-actin became higher. Moreover, of the two deletion variants studied in the presence of DPC, the one
lacking the C-terminal domain (rmΔC) was more active compared to the variant lacking the N-terminal
domain (rmΔN) but exhibited weaker binding to actin. Thus, whereas the N-terminal domain of MBP can be
more important for the MBP’s actin polymerization activity and membrane-association, the C-terminal
domain can regulate its interaction with actin.
The myelin basic protein (MBP)1family of proteins has been
known as central nervous system (CNS) self-antigens, which can
induce experimental autoimmune encephalomyelitis in mice,
ses such as multiple sclerosis (MS) (1). The most studied candi-
date autoantigen in MS is the “classic” 18.5 kDa MBP isoform
that originates from transcription start site 3 of the gene (2).
This isoform is preponderant in the adult human CNS, where it
maintains the compact multilamellar myelin assembly by adhe-
sion of the apposing cytoplasmic leaflets of the oligodendrocyte
membrane (3, 4).
The predominant 18.5 kDa isoform (from now on referred to
as MBP for simplicity) belongs to the class of intrinsically dis-
ordered proteins,many ofwhich are multifunctionaland partici-
many different partners in the CNS, including with calcium-
activated calmodulin (Ca-CaM), actin, tubulin, SH3-binding
proteins, and divalent metal cations (7-11). Unmodified MBP
is an extremely positively charged protein (þ19 at neutral pH)
wide variety of posttranslational modifications, such as methyl-
ation, phosphorylation, and deimination (12). The “dynamic
molecular barcode” of these combinatorial modifications modu-
lates the interaction of MBP with the membrane and with other
ligands (13, 14).
strated that MBP is a multifunctional protein. It has been shown
Specifically, in solution, it causes polymerization of G-actin to
F-actin and bundling of the microfilaments. This function is
phosphorylation and deimination, and reversed by binding of
†This work was supported by the Canadian Institutes of Health
Research (CIHR, MOP 74468 to G.H.). V.V.B. was the recipient of a
postdoctoral fellowship from the Multiple Sclerosis Society of Canada,
and M.A.M.A. was the recipient of a doctoral studentship from the
Ministry of Higher Education and Scientific Research of Egypt.
*To whom correspondence should be addressed at the Department of
Molecular and Cellular Biology, University of Guelph. E-mail: gharauz@
uoguelph.ca. Fax: 519-837-1802. Telephone: 519-824-4120 ext 52535.
domain and with G105W substitution; rmΔN, rmMBP(D57-R168-Leu-
Glu-His6), lacking the N-terminal domain; CNS, central nervous system;
EPR, electron paramagnetic resonance spectroscopy; DPC, dodecylphos-
phocholine; F-actin, filamentous actin; G-actin, globular actin; IPTG,
MBP, myelin basic protein; MS, multiple sclerosis; NMR, nuclear mag-
netic resonance; rmMBP, 18.5 kDa recombinant murine MBP (A1-R168-
Leu-Glu-His6); rmC1, recombinant murine MBP, unmodified C1 charge
component; rmC8, recombinant murine MBP, pseudodeiminated C8
charge component; SDS, sodium dodecyl sulfate; SDS-PAGE, sodium
dodecyl sulfate-polyacrylamide gel electrophoresis; Tris, tris(hydroxy-
6904Biochemistry, Vol. 49, No. 32, 2010 Bamm et al.
simultaneously with actin and lipid vesicles, thus supporting its
tethering it to the inner leaflet of the oligodendrocyte mem-
brane(16).Inoligodendrocytes cultured fromtheshiverer mouse
(which lacks MBP), actin filaments appeared to be disorganized,
and the cell processes were smaller than normal with a larger cell
body (19-21). Altogether, these diverse observations suggest
that the interaction of MBP with cytoskeletal proteins is crucial
for oligodendrocyte function.
The interaction of MBP with actin is mostly electrostatic and
has been shown to be inhibited by increasing salt concentration
and by two posttranslational modifications resulting in the de-
crease of the net positive charge of MBP (phosphorylation and
deimination) (8, 17, 18). Since MBP-lipid interaction involves
and since different segments of the MBP polypeptide chain are
tively charged residues of the membrane-associated protein will
be masked or less accessible for interactions with other partners.
Therefore, the interaction of MBP with actin is expected to be
different in aqueous solution and in a lipidic environment; parti-
In this paper, we have studied the interaction parameters
(polymerization and bundling) of recombinant murine MBP
(rmMBP, first the unmodified rmC1 variant) and actin in the
presence of dodecylphosphocholine (DPC) micelles (detergent
with critical micelle concentration of 1 mM) as a membrane-
mimetic condition that can be studied using solution spectros-
copic approaches (27). In previous studies, it has been shown
that one molecule of full-length MBP interacts with 200 mole-
cules of DPC (22, 31). Since one DPC micelle contains 40-60
detergentmolecules, this meansthat each MBP bindsabout four
DPC micelles. Moreover, in the same study the investigators
showed that several MBP-derived peptides bound at least 100
DPC molecules or two micelles (31). These early studies were
performed on MBP purified from brain and comprising hetero-
geneous mixtures of many posttranslationally modified compo-
nents. In the present study, we use well-defined recombinant
MBP variants, starting with the unmodified 18.5 kDa rmC1
protein. We have also determined the effects of charge loss in
MBP (using pseudodeiminated rmMBP variant, rmC8, where
five Arg and one Lys were substituted by Gln) (32), and deletion
of either the N-terminal or the C-terminal domain (rmΔN and
rmΔC, respectively) (33), on the actin-binding affinity and sto-
ichiometry in the presence of DPC. Our results show that, in the
its ability to polymerize and bundle actin, and their binding to
G-actin became more specific with higher affinity. Of the two
domain was more active than the rmΔN variant lacking the
N-terminal domain, which, however, exhibited higher affinity to
actin. We conclude that, whereas the N-terminal domain plays a
more significant role in the actin polymerization and bundling
in the regulation of that activity.
MATERIALS AND METHODS
Materials. Electrophoresis grade acrylamide, ultrapure Tris
base, and ultrapure Na2EDTA were purchased from ICN Bio-
medicals (Costa Mesa, CA). Other chemicals were of reagent
The Ni2þ-NTA (nitrilotriacetic acid) agarose beads were pur-
labeling of proteins for NMR spectroscopy, the stable isotopic
from Cambridge Isotope Laboratories (Andover, MA).
Expression and Purification of rmMBP. The unmodified
18.5kDa recombinant murine MBP isoform(rmC1), its pseudo-
deiminated form (rmC8, where five Arg and one Lys were sub-
stituted by Gln to mimic citrulline), and two deletion variants
(rmΔC and rmΔN) (Figure 1) were expressed in Escherichia coli
BL21-CodonPlus(DE3)-RP cells (Stratagene, La Jolla, CA)
and purified by nickel-affinity chromatography, followed by
ion-exchange chromatography to remove minor contaminating
material, as previously described (32-34). Protein eluate from the
for full-length proteins and 3500 Da for the deletion variants)
twice against 2 L of buffer (50 mM Tris-HCl, pH 7.4, 250 mM
NaCl), twice against2 L of100mM NaCl, andfinally fourtimes
against 2 L of ddH2O. Protein concentration was determined by
measuring the absorbance at 280 nm, using the extinction coeffi-
cients ε=0.667 L g-1cm-1, ε=0.672 L g-1cm-1, ε=0.674 L
Purity of the protein preparations was routinely assayed by
SDS-polyacrylamide gel electrophoresis (SDS-PAGE) (35).
Purification of Actin from Chicken Muscle. Chicken
muscle acetone powder was prepared, and actin was extracted
from 8 g of it at a time as described in detail elsewhere (36). The
purity of actin was checked using MALDI-TOF mass spectro-
metry, supported by SDS-PAGE, and it was clear that no
further purification steps were required. Protein concentration
was determined by measuring the absorbance at 280 nm, using
the extinction coefficient ε = 0.62 L g-1cm-1. Actin that was
CaCl2, and 0.2 mM 2-mercaptoethanol) was aliquoted into
1.5 mL microfuge tubes, flash-frozen in liquid nitrogen, and
stored at -80 ?C. The pyrene labeling of actin was performed as
previously published (37).
15NH4Cl, and [13C6]glucose were obtained
FIGURE 1: Sequences of rmMBP variants. The recombinant murine
rmMBP-C1 charge variant (rmC1) has an LE linker and His6tag at
the C-terminus of the protein. Residues denoted with white letters
R and one K in rmC1 were point-substituted by Q to create pseudo-
deiminated rmMBP (rmC8) (32). Parts of the sequence highlighted
withgraycolor and underlined indicate the polypeptide sequences of
rmΔC and rmΔN variants, respectively (33). The rmΔN variant
has an N-terminal Met and LE His6tag. In the rmΔC variant, the
N-terminal Met is missing, and another point mutation of G105W
was introduced (marked with an asterisk), and a His6tag was added
at the C-terminus.
ArticleBiochemistry, Vol. 49, No. 32, 2010 6905
labeled actin employing an automated microplate fluorescence
reader (Polarstar Omega; BMG Labtech GmbH, Offenburg,
a 405-10 nm filter for the emission channel. The stock of 50 μM
G-actin was prepared inG bufferby mixing pyrene-labeled actin
with unlabeled actin to a final labeled proportion of 5%, and
10 μL of actin was added to each well followed by an additional
volume of G buffer, calculated to give a final volume of 100 μL
after addition of rmMBP. Stocks of rmMBP variants (rmC1,
rmC8, rmΔC, and rmΔN) were also prepared in G buffer at
a concentration of 25 μM, and different volumes were auto-
matically injected across the half-area 96-well plate to reach the
final volume of 100 μL in each well. Thus, the concentration of
rmMBP ranged from 0.5 to 22.5 μM, or from 1:0.1 to 1:4.5
[actin]:[rmMBP] molar ratio. The plate was shaken for 30 s, and
the emission intensity was measured for 30 min at 27 ?C. Since
maximal fluorescence intensity was always observed after 5 min,
In the experiments where the effect of DPC on MBP-induced
actin polymerization was studied, stocks of rmC1, rmC8, rmΔC,
and rmΔN were prepared at a concentration of 25 μM in the G
buffer, supplemented with 20 mM DPC.
Actin Bundling and Actin-rmMBP Binding. To test the
bundling and binding of rmMBPs with G- or F-actin, varying
concentrations of rmC1, rmC8, rmΔC, or rmΔN (from 0.5 to
22.5 μM) were added to 5 μM G- or F-actin in a total volume
of 100 μL, in the presence or absence of 20 mM DPC. In the
experiments in which F-actin was used, G-actin was first poly-
merized into F-actin by addition of KCl to a final concentration
of50 mM, alongwith 1 mM EGTAand2 mM MgCl2. Thesam-
centrifugation at 18000g for 2 h. Under these conditions, only
rmMBP bound to highly bundled actin precipitated, but free
rmMBP or unbundled actin stayed in the supernatant. The
supernatants were transferred to separate tubes, and 80 μL was
25% glycerol, 2% SDS, 14.4 mM β-mercaptoethanol, and 0.1%
bromophenolblue).The pellets wereresuspendedin25 μL of 5?
sample buffer and topped up with water to a final volume of
125 μL. Samples (30 μL) from both pellets and supernatants
were analyzed for the presence of rmMBP and actin by 14% dis-
were stained for 2 h (10% (v/v) acetic acid, 45% (v/v) methanol,
ponding toactinmonomers or tormMBP,andsections fromthe
the extraction solution (3% (w/v) SDS, 50% (v/v) 2-propanol)
per band, and the absorbance was measured at 595 nm (38). The
amount of rmC1, rmC8, rmΔC, rmΔN, and actin in the bands
was calculated on the basis of a standard curve obtained by
loading known amounts of the above proteins on the same gel.
To present actin bundling, we plotted the percentage of bund-
amount added to the reaction mixture) vs the molar ratio of
rmMBP/actinin the reaction. To analyze the binding of rmMBP
was plotted vs the total rmMBP concentration ([MBP]total), and
the dissociation constants (Kd) were obtained using nonlinear
curve fitting with weighted regression to a quadratic equation
(Origin Pro 8; OriginLab Corp., Northampton, MA):
½MBP?bound=½actin?total¼ ðð½MBP?totalþ n½actin?totalþ KdÞ
where [MBP]totalis the concentration of rmC1, rmC8, rmΔC, or
rmΔN added to the reaction, [actin]totalis the total actin con-
actin molecule (39).
Phosphorus Assay. To test the amount of DPC bound to
each of the rmMBP variants upon interaction with actin, two
different concentrations of rmC1, rmC8, rmΔC, or rmΔN (from
5 and 22.5 μM) were added to 5 μM G-actin in a total volume of
100 μL, in the presence of 20 mM DPC. The samples were
incubated at room temperature for 30 min, followed by centri-
fugation at 18000g for 2 h. Since the only source of phosphorus
was from DPC, the pellets were resuspended in 100 μL of 0.4%
phosphorus assay (40).
electron microscopy was used to examine the morphology of the
rmMBP-actin assemblies. Samples were prepared as previously
described (41). Briefly, each rmMBP variant was allowed to
actin) in the G buffer. The samples were negatively stained with
uranyl acetate, air-dried, and examined using a Philips CM10
transmission electron microscope.
Solid-State NMR Spectroscopy. For all samples to be stu-
died by solid-state NMR spectroscopy, G-actin was first poly-
and 2 mM MgCl2. When needed, rmMBP samples were mixed
for 30 min at 42 ?C. Both protein solutions were then mixed at a
2 to 1 rmMBP to actin molar ratio, incubated at room tempera-
ture for 1 h, and then centrifuged at 18000g for 90 min to collect
the rmMBP-actin bundles. The pellets were then center-packed
in a regular 3.2 mm magic-angle spinning rotor for NMR mea-
surements. Eight different samples were prepared according to
thismethod using four different forms of rmMBP: the rmC1and
rmC8 charge components and the N-terminal deletion (rmΔN)
and C-terminal deletion (rmΔC) variants. In four of the sam-
ples, unbound rmMBP interacted with F-actin, whereas in the
Table 1: Calculated Physicochemical Parameters of rmMBP Charge and
pI/net chargecalcd ε280, L g-1cm-1
aThe parameters were calculated for the MBP variants based on amino
acid sequence using the ProtParam software tool available at the website
6906Biochemistry, Vol. 49, No. 32, 2010 Bamm et al.
All samples were prepared at the same molar ratio but with a
varying final rmMBP content in the NMR sample.
All NMR measurements were performed on a Bruker Avance
MHz proton frequency and equipped with a Bruker triple-
resonance1H-13C-15N 3.2 mm E-Free magic-angle spinning
probe (41). We collected one-dimensional13C cross-polarization
(CP) and two-dimensional13C-13C dipolar-assisted rotational
resonance (DARR) correlation spectra for all eight samples
at two different temperatures, namely, þ2 and -13 ?C, and at
a spinning frequency of 12 kHz.
Actin Polymerization Effected by rmC1, rmC8, rmΔC,
and rmΔN. In addition to the most-studied function of MBP,
which is to maintain the structural stability of the myelin sheath
by adhesion of the cytoplasmic leaflets of the oligodendrocyte
polymerization of G-actin to F-actin and bundling of the micro-
filaments (8, 16-18, 46-48). However, the mechanism of inter-
domains of MBP when membrane-associated, remains unclear.
Inthispaper,wepresent adetailedcomparisonof theinteraction
of actin with two different charge components and with two
segments of MBP and induce local, ordered secondary struc-
lateral pressure (49, 50), the use of a nonaggregating, monolayer
system such as DPC micelles allows the use of solution fluor-
escence techniques to measure actin polymerization. Vesicles
comprising negatively charged lipids will be aggregated by any
lysolipidmicelles which do not aggregate. For this present study,
we have used the full-length, unmodified recombinant murine
18.5 kDa MBP (rmC1), its pseudodeiminated form (rmC8), and
two deletion variants, rmΔC and rmΔN (Figure 1) for C- and
N-terminal deletions of rmC1, respectively (32, 33).
actin was first used as a measure of actin polymerization caused
by rmMBP. In this series of experiments, salt-induced polymeri-
and its achieved maximal fluorescence intensity was set to 100%
for the purpose of normalization. Data for all four rmMBP vari-
ants are presented in Figures 2 (charge variants) and 3 (deletion
variants). Each pair will be discussed in turn.
We first followed the rate of G-actin polymerization induced
by different concentrations of rmC1 and rmC8 (data not shown)
and found that the maximal fluorescence intensity was always
centrations of rmC1 (Figure 2A,B) and rmC8 (Figure 2C,D)
FIGURE 2: DependenceoffluorescenceenhancementcausedbyactinpolymerizationonthemolarratioofrmMBPvarianttoactin.Allreactions
were carried out in G buffer at 27 ?C and contained 5 μM G-actin with various concentrations of rmMBP variant in the presence or absence
of DPC micelles. (A) Actin polymerization caused by rmC1 alone. (B) Actin polymerization caused by rmC1 in the presence of DPC micelles.
deviation of three readings of the same sample. Data points are expressed as a percentage of the maximal fluorescence intensity observed in the
control experiment in which actin polymerization was induced by F buffer (50 mM KCl, 1 mM EGTA, and 2 mM MgCl2) alone. Note: In the
control experiments with DPC alone at concentrations in the range of 1-20 mM, no actin polymerization was observed.
ArticleBiochemistry, Vol. 49, No. 32, 2010 6907
curve having a sigmoidal shape. Both isoforms caused the same
fluorescence intensity level at a saturation point that was re-
corded at a molar ratio of rmMBP to actin of ∼0.5 for the rmC1
all given concentrations of full-length rmMBP, the initial rate
of actin polymerization was faster in the case of rmC1 (data not
shown), and at low molar ratios, rmC1 was more active than
rmC8 (Figure 2).
It has previously been demonstrated that electrostatic attrac-
tion was the major factor affecting the association of MBP with
actin (17, 18) and that MBP could bind actin filaments to the
protein’s potential to anchor actin to the oligodendrocyte mem-
brane in vivo. In this study, we used DPC, a lysolipid with a net
neutral charge and a critical micelle concentration of 1 mM, as a
membrane-mimic. (As previously indicated, MBP interacts with
and aggregates phospholipid vesicles, thus precluding any solu-
tion spectroscopic approach.) In the presence of DPC, the diffe-
rences between rmC1 and rmC8 in their ability to polymerize
fluorescence intensity level required ∼3 times more rmC8 than
rmC1 in the presence of DPC (Figure 2), and the maximal level
was lower in the case of rmC8 (Figure 4). Since no major diffe-
rences were observed for actin polymerization by rmC1 in the
absence or presence of DPC, this result suggests that the actin-
interacting domains in the MBP polypeptide chain are available
even in a lipidic environment and confirms that the interaction
Thus, in our next set of experiments we tried to identify those
by others that peptides derived from preparations of 18.5 kDa
FIGURE 3: DependenceoffluorescenceenhancementcausedbyactinpolymerizationonthemolarratioofrmMBPtruncatedvarianttoactin.All
reactionswerecarriedout inGbufferat27?Candcontained5μM G-actinwithvariousconcentrationsofthermMBPvariant inthepresenceor
absence of DPC micelles. (A) Actin polymerization caused by rmΔC alone. (B) Actin polymerization caused by rmΔC in the presence of DPC
each panel is a representative (out of three independent experiments) result (mean ( standard deviation of three readings of the same sample),
where data points are expressed as a percentage of the maximal fluorescence intensity observed in the control experiment in which actin
polymerization was induced by F buffer alone. Note: In the control experiments with DPC alone at concentrations in the range of 1-20 mM,
no actin polymerization was observed.
FIGURE 4: Polymerization of G-actin by different forms of rmMBP
(full-length charge components rmC1 and rmC8, deletion variants
average value for triplicate samples ( standard deviation.
6908 Biochemistry, Vol. 49, No. 32, 2010Bamm et al.
MBP purified from brain, specifically the 1-43 and 96-168 seg-
ments, exhibited a very strong actin-polymerization activity (48).
Therefore, here, we used two recombinant deletion variants of
rmC1,onecomprising the N-terminal two-thirds of the sequence
(A1-G105W-His6) and anothermissing theN-terminalone-third
of the sequence (M0-D57-R168-Leu-Glu-His6). These proteins
are referred to as rmΔC and rmΔN, respectively. The results
presented in Figure 3 show that both deletion variants were able
polymerization at lower molar ratios of rmMBP to actin. The
1.5 in the case of the rmΔC variant and at higher than 2 for the
rmΔN protein. At a molar ratio of 2 where actin polymerization
was maximal, both deletion variants yielded an overall lower
slightly higher than rmΔN, suggesting that the interaction
domains are located along the whole 18.5 kDa MBP sequence.
This result agrees with our previous study where, using magic-
angle spinning solid-state NMR spectroscopy, we showed that
the potential interacting fragments can be scattered along the
whole rmC1 sequenceand canundergo ordered secondarystruc-
ture changes upon interaction with actin (41). However, the
comparison of the polymerizing activity of the deletion variants
shows a reduction in the activity of the rmΔN variant, whereas
rmΔC exhibited the same activity (Figure 4), signifying the more
importantroleofthe N-terminal part of lipid-associated MBP in
the polymerization of G-actin.
Actin Microfilament Bundling Effected by rmC1, rmC8,
rmΔC, and rmΔN. MBP has also been shown to assemble
In this study, using a sedimentation assay followed by poly-
acrylamide gel electrophoretic analysis, we compared the ability
of rmC1, rmC8, rmΔN, and rmΔC to bundle F-actin in the
presence and absence of DPC. (Unbundled F-actin cannot be
whatsoever.) We used both G-actin and preformed F-actin (salt-
was added to G-actin than to F-actin (compare Figure 5, panels
A and C, with Figure 5, panels B and D, respectively), because
two separate processes were involved, first polymerization and
then bundling. Although rmC1 appeared to be more active than
rmC8, when we used G-actin as the starting material, both vari-
ants exhibited a similar bundling activity in the case of F-actin,
and both were capable of bundling 100% of the actin available.
In the experiment where rmMBP was added to G-actin in the
to actin ratio of ∼0.6 for rmC1 and 1 for rmC8.
When actin was initially polymerized withsalt and then bund-
the ability of rmC1 to bundle actin starting from G-actin was
significantly decreased but appeared to be slightly higher when
FIGURE 5: Effects of DPC micelles on the bundlingof actin by the rmC1 and rmC8 charge variants of rmMBP. All reactions were carried out in
G buffer at27?Candcontained either5 μM G-actin orpreformedF-actin (polymerizedbysalt), withsubsequent additionofvarious concentra-
total amount of actin added to the reactions was set as 100%. (A) G-Actin polymerization and bundling caused by rmC1. (B) F-Actin bundling
caused by rmC1. (C) G-Actin polymerization and bundling caused by rmC8. (D) F-Actin bundling caused by rmC8. Note: In the control
experiments with DPC alone at concentrations in the range of 1-20 mM, no actin bundling was observed.
Article Biochemistry, Vol. 49, No. 32, 20106909
the starting material was F-actin. In the case of rmC8-induced
G- and F-actin. Thus, in contrast to the differences observed in
of the actin, though a lower concentration of the more cationic
rmC1 was required.
In the same way, the bundling ability of the rmMBP deletion
variants was assayed (Figure 6). Interestingly, we did not find
any effectofDPC on the actinbundling induced byeitherrmΔC
or rmΔN. Although the latter form exhibited slightly decreased
activity at low rmMBP/actin ratios, with a more pronounced
by the rmMBP deletion variants, bundling of actin also seems to
depend on a specific domain in the MBP sequence, particularly
the N-terminal domain that hasa slightly higher bundling ability
at low rmMBP/actin molar ratios (Figure 6A,C).
Additionally, using transmission electron microscopy, we
compared the morphology of actin bundles assembled by differ-
ent rmMBP variants (examples are presented in Figure S1 in the
Supporting Information). At all molar ratios, in either the pre-
sence or absence of DPC, bundles were observed in the reaction
mixtures with all rmMBP variants. Only actin bundled by rmC8
appeared to be slightly different: the bundles were thinner and
lesscompact than others. Also, the presence of DPC micelles did
not affect the morphology of actin bundles.
Binding of rmMBP Variants to G- and F-Actin. In the
present study, we have revisited the mechanism of MBP binding
to G- and F-actin (8, 16-18, 41, 48) and have investigated the
effect of DPC (nonaggregating membrane mimic) on this bind-
ing. Using the sedimentation assay described above (an example
is presented in Figure S2 in the Supporting Information), we
compared and characterized the binding parameters of different
(see Materials and Methods). According to our experimental
design, weused actinas a receptor andeach rmMBP variantas a
based analysis of the sedimentation assay, the concentration of
actin (“receptor”) in the binding experiments was much higher
(5 μM) than the anticipated Kd(nanomolar range). Under these
conditions, the dissociation constant cannot be detected pre-
cisely, but the stoichiometry of binding can be measured accu-
rately. Thus, we present the Kdvalues obtained from different
ing of the different rmMBP variants to actin qualitatively rather
Another important question for investigation was to detect if
DPC micelles stay bound to rmMBP variants upon interaction
with actin. Since the only source of phosphorus in our experi-
ments was DPC, we quantified its amount in the actin-rmMBP
phorus assay (see Materials and Methods section). Surprisingly,
FIGURE 6: EffectsofDPC micellesonthebundlingofactin bythermΔN and rmΔC deletionvariantsofrmMBP.Allreactions werepreparedin
in the presence or absence of DPC micelles. Following a 30 min incubation, bundled actin was pelleted by centrifugation at 18000g for 30 min.
Pellet and supernatant were subjected to SDS-PAGE analysis for actin detection (for details see Materials and Methods). The total amount of
(C) G-Actin polymerization and bundlingcaused byrmΔC. (D) F-Actin bundlingcaused byrmΔC. Note: Inthe control experimentswithDPC
alone at concentrations in the range of 1-20 mM, no actin bundling was observed.
6910 Biochemistry, Vol. 49, No. 32, 2010Bamm et al.
gent molecules per 1 protein, or 1 micelle per rmMBP variant.
This ratio stayed constant with no influence from the initial
rmMBP/actin ratios in the reaction mixtures.
The comparison of binding of rmC1 and rmC8 variants to
G- and F-actin is shown in Figure 7. In general, both rmC8 and
rmC1 bound to actin under all conditions investigated, but the
stoichiometry and affinity of the binding were different. In the
case of rmC1 binding to actin in the presence of DPC, the sto-
whether the starting material was G- or F-actin (Figure 7A,B).
starting material, it can be seen from the slope and curvature of
the curves presented in Figure 7B that the affinity of binding
increased in the presence of DPC. However, in the experiment
using G-actin as the starting material, the binding of rmC8
could not be assessed in the absence of DPC (Figure 7C), due to
continuous adsorption of more and more of the less cationic
rmMBP molecules to the actin. That binding process never rea-
On the other hand, in the presence of DPC, the binding became
per 1 molecule of actin. One possible explanation for this result
could be that, even though rmC8 was shown to polymerize and
bundle 100% of actin (Figure 5C), the morphology of those
filaments and bundles could be different; e.g., actin filaments
could be of different length and packed more loosely, thus
providing more binding sites forrmC8inthe absence ofmicelles.
This suggestion comes in agreement with our previous report
induced by rmC1 scattered more light than those induced by
rmC8 (17). Here, the results of transmission electron microscopy
confirmed that even in the presence of DPC micelles, where
binding of rmC8 could be assessed, actin bundles appeared to be
packed more loosely than ones assembled by rmC1 (compare
panels A and B in Figure S1 of the Supporting Information).
Moreover, the binding of rmC8 appeared to be very different
interacted with rmMBP. As can be seen from the results pre-
sented in Figure 7D, in the presence of DPC, fewer rmC8 mole-
cules could bind actin, but the binding could be assessed even in
the abscnce of DPC.
Earlier studies have suggested that there were two binding
and anotherat thecarboxy terminus(48).Thus,next weassessed
thebindingoftwo deletionvariantsofMBPtoG- andF-actinin
the presence of DPC (Figure 8). Indeed, both deletion variants
bound to actin, and the binding parameters are presented in
Table 2. Again, the effect of DPC mostly was exhibited in the
decreased stoichiometry of binding and, in the case of the rmΔN
variant, also in increased affinity (based on the curvature of
curves presented in Figure 8A,B). Of the two deletion variants
tested in the membrane-mimetic environment, the binding of
rmΔN (lacking the N-terminal region of the MBP sequence)
Solid-State NMR Spectroscopy. We have previously stu-
died the association of full-length rmC1 with F-actin (in the
absenceof lipids) byFouriertransforminfraredandmagic-angle
spinning solid-state NMR spectroscopy, in which we demon-
strated induced, ordered secondary structure in both protein
partners (41). Here, similar solid-state NMR experiments were
Table 2: Parameters of Binding of rmMBP Variants to Actina
Kd, nM 7.6 ( 2.3
14.3 ( 2.7
66.6 ( 6.8
65.3 ( 9.2
22.4 ( 2.6
95.9 ( 3.8
ND 52.6 ( 4.5
101.2 ( 5.88 276.3 ( 6.0 155.2 ( 13.7 91.7 ( 6.5
25.2 ( 2.4
10.5 ( 1.8
255.2 ( 13.5 93.9 ( 5.7
1.32 ( 0.04 0.99 ( 0.02 2.06 ( 0.05 1.56 ( 0.06 1.40 ( 0.04 1.17 ( 0.02 ND 1.46 ( 0.04 1.27 ( 0.04
1.40 ( 0.03 1.92 ( 0.08
1.47 ( 0.05 1.38 ( 0.03 1.32 ( 0.02 1.72 ( 0.08
1.49 ( 0.04
aBinding parameters were determined from the experiments presented in Figures 7 and 8. Here, “G” and “F” refer to G-actin and F-actin, respectively, and “ND” means “not determined”. Results are presented as value (
standard error based on a nonlinear least-squares fitting algorithm using weighted regression to eq 1.
Article Biochemistry, Vol. 49, No. 32, 20106911
carried out on different rmMBP variants in the presence and
absence of DPC. Ingeneral,the temperaturedependence ofboth
one- and two-dimensional spectra of all samples showed similar
behavior as in our previously studied system of rmC1-F-actin
without lipid (data not shown) (41). The signal-to-noise ratio of
most of the peaks increased at the expense of line width as the
at lower temperature, due to the polymorphic nature of the
interaction between rmMBP and actin (41).
Moreover, the resonance dispersion in two-dimensional spec-
tra (Figure 9) showed a similar trend as the lipid-free rmC1-
variantsunderwent similar,induced secondarystructurechanges
13C-13C correlation spectra for two of the samples under
consideration: DPC-bound rmC1 and DPC-bound rmΔC with
F-actin. In Figure 9A, the resolved fingerprints of some amino
tion of most of the correlation spectra is a common observation
in both DPC-rmC1-F-actin and lipid-free rmC1-F-actin
assemblies. This observation indicates that the prior interaction
of DPC with rmC1, even though it changed the binding affinity
andstoichiometry of the interactionwithactin,didnotaffect the
rmΔC-F-actin (Figure 9B) displayed less crowded correlations
of Figure 9A (∼5 mg), and to the smaller protein size in the case
The 18.5 kDa MBP splice isoform has been thought to be pri-
marily a molecular adhesive in CNS myelin, adhering the appo-
sing cytosolic surfaces of the oligodendrocyte membrane to each
other (3, 4, 14). As recently reviewed, in addition to this main
function, MBP has been shown to interact with a wide variety of
biological ligands including cytoskeletal proteins, calmodulin,
SH3-domain containing proteins, and divalent metal cations
(14, 51, 51-53). It has been shown in previous studies that two
charge components of the “classic” 18.5 kDa MBP, the most
cationic (C1) and the least cationic one (C8), were able to poly-
merize, bundle, and link actin to the phospholipid membranes
(8, 15-17, 46-48). The polymerization of actin was reversible in
the presence of calmodulin (8, 17). In the present study, we have
compared the interaction of recombinant murine MBP charge
the N- and C-terminal domains in the MBP sequence.
MBP is able to polymerize G-actin and bundle F-actin. Of two
FIGURE 7: EffectsofDPCmicellesonthebindingofactintormC1andrmC8chargevariants.AllreactionswerepreparedinGbufferat27?Cand
variant in the presence or absence of DPC micelles. Following a 30 min incubation, bundled actin was pelleted by centrifugation at 18000g for
30 min. Pellet and supernatant were subjected to SDS-PAGE analysis for actin and rmMBP detection (for details see Materials and Methods).
fittings to the experimental data according to eq 1.
6912 Biochemistry, Vol. 49, No. 32, 2010Bamm et al.
FIGURE 8: EffectsofDPCmicellesonthebindingofactintormΔCandrmΔNdeletionvariants.AllreactionswerepreparedinGbufferat27?C
were subjected to SDS-PAGE analysis for actin and rmMBP detection (for details see Materials and Methods). (A) Binding of rmΔN upon
interaction with G-actin. (B) Binding of rmΔN upon interaction with preformed F-actin. (C) Binding of rmΔC upon interaction with G-actin.
(D) Binding of rmΔC upon interaction with preformed F-actin. In all panels, scattered points represent experimentally measured data (mean (
standard deviation of three independent experiments), while solid and dashed lines are the fittings to the experimental data according to eq 1.
FIGURE 9: Solid-state magic-angle spinning NMR spectroscopy of assemblies of actin with DPC-rmMBP (two variants). Aliphatic regions of
two-dimensional DARR (dipolar-assisted rotational resonance)13C-13C correlation spectra of uniformly13C,15N-labeled rmMBP in the
were 4 and 14 ms, respectively, and were apodized with a π/2-shifted sine-squared window function prior to Fourier transformation. The first
contour is cut at 5 times the noise floor, and each following contour level is multiplied by 1.15. Both spectra were collected at T = -13 ?C, at
samples in panels A and B, respectively.
Article Biochemistry, Vol. 49, No. 32, 20106913
charge components, rmC1 and rmC8, the more cationic protein
Since a natural lipidic environment is more physiologically
relevant toMBP,which,even when membrane-associated, isstill
significantly mobile and able to interact with other proteins (26),
the novel approach here was to use the membrane-mimicking
conditions (presence of DPC micelles) to compare the actin-
the challenges associated with creating an experimentally tract-
ofDPCasa membrane-mimicking environmentiscommon,and
in previous studies it has been reported that the stoichiometry of
the MBP interaction with DPC was 200 molecules of lipid per
1 protein molecule (22, 31). Since the average number of DPC
molecules per micelle ranges from 40 to 60, each MBP can thus
actin only one DPC micelle stays bound to MBP in the complex
In the presence of DPC, the polymerization ability of the less
highly positively charged rmC1 protein (17). One possible ex-
planation could be that the interaction with DPC micelles masks
some positively chargedresiduesinthe MBP sequence, and since
the charge of rmC8 had already been reduced by þ6, this
In fact, it has previously been shown that several domains in the
polypeptide chain of MBP are capable of penetrating into the
hydrophobic core of the DPC micelles (22, 24, 31). Our previous
rmC1 with actin in an aqueous environment showed that both
MBP and actin had mutually induced ordered secondary struc-
ture(41).The potentialsitesthat might experiencechangesinthe
secondary structure in MBP were distributed over the entire seq-
uence, thussuggesting that the entire moleculemight be involved
in the association with actin.
Here, by studying the ability of two rmMBP deletion variants
to polymerize and bundle actin, we further confirmed that the
entire sequence of 18.5 kDa MBP is involved in the interaction.
Our results showed that though the polymerizing ability of the
deletion variants was reduced compared to the full-length rmC1
isoform, they both were able to bundle 100% of the actin. The
rmΔN variant, the one lacking 56 amino acids at the amino
terminus of the sequence, exhibited lower polymerization and
bundling activity than the rmΔC form, which lacked 63 residues
at the carboxy-terminal end. The difference in actin-assembling
ability appeared to be even more pronounced in the presence
of DPC. Although it seems that the N-terminal part of 18.5 kDa
MBP is more important for its interaction with actin, the inter-
action of the C-terminal part of the protein might be physio-
logically more relevant, because this is the primary binding
shown to cause depolymerization of the MBP-polymerized actin
supported here by the results of binding experiments of the two
presence of DPC micelles. We have shown that both variants
bound actin during polymerization and/or bundling. In the pre-
sence of micelles, the binding was more specific (lower rmMBP/
actin ratio), and the affinity was greater for the rmΔN construct
curvature of curves presented in Figure 8B,D is different, thus
suggestinga greater affinity inthe caseofrmΔN.Additionally,it
is known that the N-terminal part of MBP provides one of the
ing its ability to bind actin. Therefore, in the context of MBP’s
physiological environment, the N-terminal two-thirds of the
polymerization/bundling,and the C-terminalpartcan bind actin
the C-terminal segment of MBP may represent the regulatory
18.5 kDa MBP (28, 54).
Our solid-state NMR spectroscopy data on DPC-bound
rmMBP showed that, although the prior interaction of MBP
with DPC improved its binding specificity, MBP-actin inter-
actions are still polymorphic, even in the membrane-mimicking
environment of DPC micelles (Figure 9) (51).
In the 18.5 kDa splice isoform of MBP, basic residues are
distributed along the sequence, and the charge density is fairly
FIGURE 10: SchematicrepresentationoftheinteractionofMBPwith
membraneand regions potentially involved ininteraction withactin.
The details of basic protein-membrane binding are not known, and
the arrangement shown inthisfigureisone out ofmanypossibilities.
There is scant evidence for MBP dimerization or oligomerization in
sent the amphipathic R-helices that were proposed based on solu-
tion and solid-state NMR spectroscopic studies (26, 28, 52) and
electron paramagnetic resonance investigations (23, 55), including
one that focused on the immunodominant epitope (Pro82-Pro93,
murine 18.5 kDa sequence numbering) (23, 25). The proline-rich
segment immediately following (and slightly overlapping) this epi-
tope forms a transient polyproline type II helix (10), and the region
comprising Pro120-Arg160 represents a primary binding site for
Ca-activated calmodulin (28, 51). In red, we denote the regions of
MBP that might undergo changes in secondary structure upon
interaction with actin (41).
6914Biochemistry, Vol. 49, No. 32, 2010Bamm et al.
into bundles. Moreover, some of those basic residues can be
involved in the interaction of MBP with the oligodendrocyte
anchor actin to the membrane. In Figure 10, we present a model
of a proposed arrangement of MBP in the myelin sheath. First,
resonance and NMR spectroscopic investigations, we know that
the N-terminal part of the protein, particularly the S38-A49
region, is embedded in the hydrophobic portion of the bilayer
(26, 55). The central region of the protein penetrates the mem-
brane and also associates electrostatically with the phospholipid
headgroups (lysine snorkeling), and the C-terminal part pene-
trates the hydrophobic core of the membrane again (23, 26). In
the case of the pseudodeiminated rmC8 component, the net
charge reduction results in the dissociation of some sites in the
C-terminal part of the protein from the lipid bilayer, making it
more accessible for theinteractionwithdifferent ligands(55, 56).
Second, from our previous solid-state NMR structural study of
the interaction of lipid-free rmC1 with actin, we posit that the
potential regions of MBP involved in the interaction with actin
are distributed through the entire molecule, and within the
context of the lipid bilayer (oligodendrocyte membrane) some
Ca2þ-activated calmodulin (28, 54) and thus could play a crucial
ability to link actin to the membrane.
In conclusion, the results of this study confirm that MBP can
act as a linker between actin and membrane and thus participate
in the rearrangement (polymerization and bundling) of actin
withinmyelin. There aremultiplebinding domains located inthe
MBP sequence which are important for the cross-talk between
actin filaments and the oligodendrocyte membrane, which,
in turn, can control cell morphology, thereby maintaining func-
tional and healthy myelin assembly in the CNS.
We are grateful to Drs. Vladimir Ladizhansky (Guelph), John
Dawson (Guelph), and Joan Boggs (Hospital for Sick Children,
SUPPORTING INFORMATION AVAILABLE
led by different rmMBP variants (Figure S1) and a representative
set of SDS-PAGE gels used for the binding assay (Figure S2).
This material is available free of charge via the Internet at http://
1. Martin, R., McFarland, H. F., and McFarlin, D. E. (1992) Immuno-
logical Aspects of Demyelinating Diseases. Annu. Rev. Immunol. 10,
2. Feng, J. M., Fernandes, A. O., Campagnoni, C. W., Hu, Y. H., and
Regulates Signal Transduction in T Lymphocytes. J. Neuroimmunol.
3. Omlin, F. X., Webster, H. D., Palkovits, C. G., and Cohen, S. R.
(1982) Immunocytochemical Localization of Basic Protein in Major
and Hood, L. (1990) Role of Myelin Basic Protein in the Forma-
tion of Central Nervous System Myelin. Ann. N.Y. Acad. Sci. 605,
5. Radivojac, P., Iakoucheva, L. M., Oldfield, C. J., Obradovic, Z.,
Uversky, V. N., and Dunker, A. K. (2007) Intrinsic Disorder and
Functional Proteomics. Biophys. J. 92, 1439–1456.
6. Receveur-Br? echot, V., Bourhis, J. M., Uversky, V. N., Canard, B.,
and Longhi, S. (2006) Assessing Protein Disorder and Induced
Folding. Proteins 62, 24–45.
7. Boggs, J. M. (2006) Myelin Basic Protein: A Multifunctional Protein.
Cell. Mol. Life Sci. 63, 1945–1961.
Phosphorylation of Myelin Basic Protein by MAPK on its Inter-
actions with Actin and Actin Binding to a Lipid Membrane in Vitro.
Biochemistry 45, 391–401.
9. Harauz, G., Ishiyama, N., Hill, C. M. D., Bates, I. R., Libich, D. S.,
and Far? es, C. (2004) Myelin Basic Protein;Diverse Conformational
Assembly and Multiple Sclerosis. Micron 35, 503–542.
10. Polverini, E., Rangaraj, G., Libich, D. S., Boggs, J. M., and Harauz,
G. (2008) Binding of the Proline-Rich Segment of Myelin Basic
Studies of Ligand Conformation and Effects of Post-Translational
Modifications. Biochemistry 47, 267–282.
11. Smith, G. S. T., Chen, L., Bamm, V. V., Dutcher, J., and Harauz, G.
(2010) The Interaction of Zinc with Membrane-Associated 18.5 kDa
Myelin Basic Protein: An Attenuated Total Reflectance-Fourier Trans-
form Infrared Spectroscopic Study. Amino Acids [2010 Feb 19, Epub
ahead of print].
12. Kim,J.K.,Mastronardi, F. G., Wood, D. D., Lubman, D. M., Zand,
R., and Moscarello, M. A. (2003) Multiple Sclerosis: An Important
Role for Post-Translational Modifications of Myelin Basic Protein in
Pathogenesis. Mol. Cell. Proteomics 2, 453–462.
13. DeBruin, L. S., and Harauz, G. (2007) White Matter Rafting;
Membrane Microdomains in Myelin. Neurochem. Res. 32, 213–228.
14. Harauz, G., Ladizhansky, V., and Boggs, J. M. (2009) Structural
Polymorphism and Multifunctionality of Myelin Basic Protein. Bio-
chemistry 48, 8094–8104.
15. Boggs, J. M. (2008) Myelin Basic Protein Interactions with Actin and
Tubulin in Vitro: Binding, Assembly, and Regulation, in Myelin
Basic Protein (Boggs,J. M., Ed.) pp 149-167, Nova Science Publishers,
16. Boggs, J. M., and Rangaraj, G. (2000) Interaction of Lipid-Bound
Myelin Basic Protein with Actin Filaments and Calmodulin. Bio-
chemistry 39, 7799–7806.
Harauz, G. (2005) Effect of Arginine Loss in Myelin Basic Protein, as
Occurs in its Deiminated Charge Isoform, on Mediation of Actin
Polymerization and Actin Binding to a Lipid Membrane in Vitro.
Biochemistry 44, 3524–3534.
18. Hill, C.M.D., andHarauz,G.(2005)Charge Effects Modulate Actin
Assembly by Classic Myelin Basic Protein Isoforms. Biochem. Bio-
phys. Res. Commun. 329, 362–369.
19. Dyer, C.A.,Philibotte, T.M.,Wolf,M.K.,andBillings-Gagliardi,S.
(1994) Myelin Basic Protein Mediates Extracellular Signals that
Regulate Microtubule Stability in Oligodendrocyte Membrane
Sheets. J. Neurosci. Res. 39, 97–107.
20. Dyer, C.A.,Philibotte, T.M.,Billings-Gagliardi,S.,andWolf, M.K.
(1995) Cytoskeleton in Myelin-Basic Protein-Deficient Shiverer Oligo-
dendrocytes. Dev. Neurosci. 17, 53–62.
21. Dyer, C. A., Philibotte, T., Wolf, M. K., and Billings-Gagliardi, S.
(1997) Regulation of Cytoskeleton by Myelin Components: Studies
on Shiverer Oligodendrocytes Carrying an Mbp Transgene. Dev.
Neurosci. 19, 395–409.
22. Mendz, G. L., Moore, W. J., Kaplin, I. J., Cornell, B. A., Separovic,
F., Miller, D. J., and Brown, L. R. (1988) Characterization of
Dodecylphosphocholine/Myelin Basic ProteinComplexes.Biochemistry.
23. Bates, I. R., Feix, J. B., Boggs, J. M., and Harauz, G. (2004) An
Immunodominant Epitope of Myelin Basic Protein Is an Amphi-
pathic Alpha-Helix. J. Biol. Chem. 279, 5757–5764.
24. Far? es, C., Libich, D. S., and Harauz, G. (2006) Solution NMR
Structure of an Immunodominant Epitope of Myelin Basic Protein.
Conformational Dependence on Environment of an Intrinsically
Unstructured Protein. FEBS J. 273, 601–614.
25. Musse, A. A., Boggs, J. M., and Harauz, G. (2006) Deimination of
an Immunodominant Epitope. Proc. Natl. Acad. Sci. U.S.A. 103,
26. Zhong, L., Bamm, V. V., Ahmed, M. A., Harauz, G., and Ladizhansky,
V.(2007)Solid-StateNMR Spectroscopyof18.5kDaMyelin Basic Pro-
tein Reconstituted with Lipid Vesicles: Spectroscopic Characterisation
ArticleBiochemistry, Vol. 49, No. 32, 20106915 Download full-text
and Spectral Assignments of Solvent-Exposed Protein Fragments.
Biochim. Biophys. Acta 1768, 3193–3205.
27. Libich, D. S., and Harauz, G. (2008) Solution NMR and CD
Spectroscopy of an Intrinsically Disordered, Peripheral Membrane
Protein: Evaluation of Aqueous and Membrane-Mimetic Solvent
Conditions for Studying the Conformational Adaptability of the
18.5 kDa Isoform of Myelin Basic Protein (MBP). Eur. Biophys. J.
28. Libich, D. S., and Harauz, G. (2008) Backbone Dynamics of the
18.5 kDa Isoform of Myelin Basic Protein Reveals Transient R-Helices
and a Calmodulin-Binding Site. Biophys. J. 94, 4847–4866.
29. Harauz, G., Ladizhansky, V. (2008) Structure and Dynamics of the
Myelin Basic Protein Family by Solution and Solid-State NMR, in
Myelin Basic Protein (Boggs, J. M., Ed.) pp 197-232, Nova Science
Publishers, New York.
30. Boggs, J. M., Bates, I. R., Musse, A. A., and Harauz, G. (2008)
Studies by Electron Paramagnetic Resonance Spectroscopy and
Implications for Generation of Autoimmunity in Multiple Sclerosis,
Publishers, New York.
31. Mendz, G. L., Moore, W. J., Brown, L. R., and Martenson, R. E.
(1984) Interaction of Myelin Basic Protein with Micelles of Dodecyl-
phosphocholine. Biochemistry 23, 6041–6046.
32. Bates, I. R., Libich, D. S., Wood, D. D., Moscarello, M. A., and
Myelin Basic Protein as a Mimic of the Deiminated Form Implicated
in Multiple Sclerosis. Protein Expression Purif. 25, 330–341.
33. Hill, C. M. D., Haines, J. D., Antler, C. E., Bates, I. R., Libich, D. S.,
and Harauz, G. (2003) Terminal Deletion Mutants of Myelin Basic
Protein: New Insights into Self-Association and Phospholipid Inter-
actions. Micron 34, 25–37.
34. Bates, I. R., Matharu, P., Ishiyama, N., Rochon, D., Wood, D. D.,
Polverini, E., Moscarello, M. A., Viner, N. J., and Harauz, G. (2000)
Characterization of a Recombinant Murine 18.5-kDa Myelin Basic
Protein. Protein Expression Purif. 20, 285–299.
35. Laemmli, U. K. (1970) Cleavage of Structural Proteins during the
Assembly of Bacteriophage T4. Nature 227, 680–685.
36. Pardee, J. D., and Spudich, J. A. (1982) Purification of Muscle Actin.
Methods Cell Biol. 24, 271–289.
37. Kouyama, T., and Mihashi, K. (1981) Fluorimetry Study of N-(1-
Pyrenyl) Iodoacetamide-Labelled F-Actin. Eur. J. Biochem. 114,
38. Ball, E. H. (1986) Quantitation of Proteins by Elution of Coomassie
Brilliant Blue R from Stained Bands after Sodium Dodecyl Sulfate-
Polyacrylamide Gel Electrophoresis. Anal. Biochem. 155, 23–27.
39. Bligh, S. W., Haley, T., and Lowe, P. N. (2003) Measurement of
Dissociation Constants of Inhibitors Binding to Src SH2 Domain
Protein by Non-Covalent Electrospray Ionization Mass Spectro-
metry. J. Mol. Recognit. 16, 139–148.
40. Bartlett, G. (1959) Phosphorus Assay in Column Chromatography.
J. Biol. Chem. 234, 466–468.
41. Ahmed, M. A., Bamm, V. V., Shi, L., Steiner-Mosonyi, M., Dawson,
J. F., Brown, L., Harauz, G., and Ladizhansky, V. (2009) Induced
Secondary Structure and Polymorphism in an Intrinsically Dis-
ordered Structural Linker of the CNS: Solid-State NMR and FTIR
Spectroscopy of Myelin Basic Protein Bound to Actin. Biophys. J. 96,
42. Napolitano, L., Lebaron, F., and Scaletti, J. (1967) Preservation of
Myelin Lamellar Structure in the Absence of Lipid. A Correlated
Chemical and Morphological Study. J. Cell Biol. 34, 817–826.
43. Moscarello, M. A., Brady, G. W., Fein, D. B., Wood, D. D., and Cruz,
T. F. (1986) The Role of Charge Microheterogeneity of Basic Protein in
A Possible Role in Multiple Sclerosis. J. Neurosci. Res. 15, 87–99.
44. Smith, R. (1992) The Basic Protein of CNS Myelin: Its Structure and
Ligand Binding. J. Neurochem. 59, 1589–1608.
45. Moscarello, M. A. (1997) Myelin Basic Protein, the “Executive”
Molecule of the Myelin Membrane, in Cell Biology and Pathology
of Myelin: Evolving Biological Concepts and Therapeutic App-
roaches (Juurlink, B. H. J., Devon, R. M., Doucette, J. R., Nazarali,
A. J., Schreyer, D. J., and Verge, V. M. K., Eds.) pp 13-25, Plenum
Press, New York.
46. Barylko, B., and Dobrowolski, Z. (1984) Ca2þ-Calmodulin-Dependent
Regulation of F-Actin-Myelin Basic Protein Interaction. Eur. J. Cell
Biol. 35, 327–335.
47. Dobrowolski, Z., Osinska, H., Mossakowska, M., and Barylko, B.
(1986) Ca2þ-Calmodulin-Dependent Polymerization of Actin by
Myelin Basic Protein. Eur. J. Cell Biol. 42, 17–26.
48. Roth, G. A., Gonzalez, M. D., Monferran, C. G., De Santis, M. L.,
and Cumar, F. A. (1993) Myelin Basic Protein Domains Involved in
the Interaction with Actin. Neurochem. Int. 23, 459–465.
49. Rosetti, C. M., Maggio, B., and Oliveira, R. G. (2008) The Self-
Organization of Lipids and Proteins of Myelin at the Membrane
Interface. Molecular Factors Underlying the Microheterogeneity of
Domain Segregation. Biochim. Biophys. Acta 1778, 1665–1675.
50. Rosetti, C. M., Maggio, B., and Wilke, N. (2010) Micron-Scale Phase
Segregation in Lipid Monolayers Induced by Myelin Basic Protein in
the Presence of a Cholesterol Analog. Biochim. Biophys. Acta 1798,
V., and Harauz, G. (2010) Fuzzy Complexes of Myelin Basic Protein;
NMR Spectroscopic Investigations of a Polymorphic Organizational
Linker of the Central Nervous System. Biochem. Cell Biol. 88, 143–155
(Special Issue on Protein Folding: Principles and Diseases).
52. Harauz, G., and Libich, D. S. (2009) The Classic Basic Protein of
Myelin;Conserved Structural Motifs and the Dynamic Molecular
Barcode Involved in Membrane Adhesion and Protein-Protein Inter-
actions. Curr. Protein Pept. Sci. 10, 196–215.
53. Boggs, J. M., Rangaraj, G., Heng, Y.-M., and Harauz, G. (2009)
Myelin Basic Protein Binds Microtubules to a Membrane Surface in
Vitro:EffectofPhosphorylation andDeimination. J.Neurochem.108
(Suppl. 1), 64.
54. Majava, V., Petoukhov, M. V., Hayashi, N., Pirila, P., Svergun, D. I.,
and Kursula, P.(2008)Interaction betweentheC-TerminalRegionof
Human Myelin Basic Protein and Calmodulin: Analysis of Complex
Formation and Solution Structure. BMC Struct. Biol. 8, 10.
55. Bates, I. R., Boggs, J. M., Feix, J. B., and Harauz, G. (2003)
Membrane-Anchoring and Charge Effects in the Interaction of
Myelin Basic Protein with Lipid Bilayers Studied by Site-Directed
Spin Labeling. J. Biol. Chem. 278, 29041–29047.
A. A., Smith, G. S. T., Harauz, G., and Boggs, J. M. (2010) SDSL-
EPR Study of a C-Terminal Segment of Myelin Basic Protein in a
Myelin Mimetic Environment, 54th Annual Biophysical Society
Meeting, San Francisco, CA.