Cholesterol secosterol adduction inhibits the misfolding of a mutant prion protein fragment that induces neurodegeneration

Article abstract of DOI:10.1002/anie.200904524

Aldehyde is critical: Atheronal-B, a cholesterol secosterol aldehyde (see structure), completely inhibits the misfolding of a prion protein fragment from its α to β form through a mechanism that involves adduction to the protein. This result offers a fresh view of lipid aldehyde-induced protein misfolding and provides a promising molecular scaffold on which to develop potential prion disease therapeutics.

Full text of DOI:10.1002/anie.200904524

Angewandte
Chemie
DOI: 10.1002/anie.200904524
Prion Misfolding
Cholesterol Secosterol Adduction Inhibits the Misfolding of a Mutant
Prion Protein Fragment that Induces Neurodegeneration**
Johanna C. Scheinost, Daniel P. Witter, Grant E. Boldt, John Offer, and Paul Wentworth,Jr.*
Novel infectious particles termed prions, composed solely of
single proteins, are considered the causative agents in a group
of transmissible spongiform encephalopathies that produce a
lethal decline in motor and cognitive functions.[1–3] Prion
diseases are infectious, inherited, or sporadic in nature but
perhaps the most unsettling aspect of these protein-only
diseases is that they arise at their pathogenic state by
misfolding of a protein from its normal state that may exist
with ubiquitous distribution throughout tissue. Thus, the prion
diseases are now generally accepted to occur by the presence
of an exogenous gene, Prnp, encoding a dimorphic protein
that can exist in a normal (PrPC) or disease-causing “scrapie”
form (PrPSc).[4]
At present there is no therapy, either small molecule or
vaccine-based, for the treatment of prion diseases. However,
molecules that inhibit PrPSc formation from PrPC are viewed
as potential therapeutics for these diseases. Small molecules
such as sulphated polyanions,[5] congo red,[6] heparin sulfate,[7]
pentosan polysulfate[8] and cyclic tetrapyrroles[9] inhibit PrPSc
Figure 1. A) Representation of an open reading frame of the mouse
formation. In addition, short peptides comprised of residues
prion protein (MoPrP) and the synthetic core mutant peptide
from the hydrophobic core of PrPC also inhibit prion protein
MoPrP(89-143, P101L) that initiates GSS in transgenic mice that we
have synthesized and studied in this work. The mutant leucine residue
aggregation.[10,11]
Recently we have discovered a process that we are
studying in the context of a number of disease-related
sporadic amyloidoses. We have shown that in vitro, certain
inflammatory-derived lipid aldehydes, when adducted to pro-
amyloidogenic proteins in their native state, can induce
misfolding and aggregation of the native protein sequences.[12]
Thus, we have shown that cholesterol 5,6-secosterols such as
1, that are derived from oxidation of cholesterol by activated
leukocytes, accelerate the in vitro misfolding of apoB100, the
protein component of low-density lipoprotein (LDL) (Fig-
ure 1b).[13]
101 is black and underlined. HA, HB and HC represent helices A, B
and C, respectively. GPI represents the glycosylphosphatidyl inositol
(GPI) anchor. B) Chemical structures of cholesterol 5,6-secosterol 1
(atheronal-B), carboxylic acid 2, primary alcohol 3, cholesterol (4), and
2-naphthaldehyde (5).
In addition, we have shown that 4-hydroxynonenal and 1
accelerate the in vitro amyloidogenesis of b-amyloid peptides
[14]
(Ab1–40 and Ab1–42
)
through a process that involves a site-
specific modification of Ab.[15] More recently, aldehyde 1 has
been shown to accelerate the aggregation of antibody light
chains[16] and a-synuclein.[17]
Cholesterol secosterol 1 arises in vivo as a product of
chronic inflammation, and it has been shown that inflamma-
tion is a risk factor for prion diseases. Organs that are
undergoing an inflammatory response are more vulnerable to
prion infection, and prion-type infections arise earlier in
inflamed tissue than in the central nervous system (CNS).[18]
Therefore, as part of our ongoing investigation into lipid
aldehyde-induced protein misfolding we wondered whether
aldehyde 1 may accelerate the formation of PrPSc from PrPc in
vitro. Remarkably, what we show herein is exactly the
opposite effect: atheronal-B (1) inhibits the misfolding of a
truncated murine prion protein MoPrP(89-143, P101L). It is
the first example where adduction of a lipid-derived aldehyde
to a protein inhibits its misfolding, and as such offers an
insight into a potential new class of molecular structures that
may serve as therapeutics.
[*] J. C. Scheinost, D. P. Witter, Dr. G. E. Boldt, Dr. J. Offer,
Prof. P. Wentworth,Jr.
The Scripps-Oxford Laboratory, Department of Biochemistry, Uni-
versity of Oxford, South Parks Rd., Oxford OX13QU (UK)
Fax: (+44)1865-285-329
E-mail: paul.wentworth@bioch.ox.ac.uk
Homepage: http://www2.bioch.ox.ac.uk/wentworthlab
Prof. P. Wentworth,Jr.
Department of Chemistry and the Skaggs Institute for Chemical
Biology, The Scripps Research Institute
10550 N. Torrey Pines Road, La Jolla CA 92037 (USA)
[**] This work was supported by the NIH AG028300 (P.W.) and by a
grant from The Scripps Research Institute. The authors would like to
thank R. P. Troseth (TSRI) for synthesis of 1 and Dr. P. R. Antrobus
(OU) for high-resolution ES-MS of the synthetic peptides.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200904524.
Angew. Chem. Int. Ed. 2009, 48, 9469 –9472
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
9469

Communications
The model prion protein selected for study was a 55-
residue peptide, MoPrP(89-143, P101L), from a murine
mutant prion protein that can be refolded into at least two
distinct conformations, the b- and a-forms (Figure 1a).[19]
Critically, injection of the b-form of this peptide intracere-
brally into mice induces neuropathological changes that are
hallmarks of Gerstmann–Strꢀussler–Scheinker (GSS) disease,
whereas injection of the a-form causes no neuropathological
effects.[19]
MoPrP(89-143, P101L) was synthesized in multimilligram
amounts using standard Fmoc/tBu solid phase peptide syn-
thesis (SPPS) using the Rink amide linker resulting in a C-
terminal amide (Figure 1a; for full methods see Supporting
Information) and purified to > 98% purity by reversed-phase
HPLC. Aldehyde 1 was synthesised as described previ-
ously;[23] carboxylate 2 and alcohol 3 were prepared by
oxidation and reduction of 1, respectively (see Supporting
Information).
The aqueous buffer conditions for studying the effect of
cholesterol secosterol 1 on MoPrP(89-143, P101L) misfolding,
phosphate buffered saline (PBS, pH 7.4) and 30% v/v 2,2,2,-
trifluoroethanol (TFE), were selected to ensure that the
conformation of the protein at the start of the studies
comprised a high proportion of a-helix. Preliminary studies
revealed that MoPrP(89-143, P101L) is almost completely a-
helical in structure when dissolved in TFE (Supporting
Information, Figure S2). Upon dilution of the protein solution
(in TFE) into PBS (pH 7.4), the fraction of a-helical
structures decreases as a function of TFE concentration,
with the protein being almost completely random coil at ca.
10% v/v TFE in PBS. At 30% TFE in PBS, the conditions
selected for the misfolding assays, the peptide contains
considerable, ca. 30%, a-helical structure, as determined by
far-UV circular dichroism (CD) (Figure S2B).
Incubation of MoPrP(89-143, P101L) (20 mm) in PBS
(pH 7.4) with TFE (30% v/v) at 378C with shaking (900 rpm)
leads to the time-dependent formation of thioflavin T (ThT)-
binding conformers and aggregates with the kinetic profile of
a nucleation-dependent polymerisation process with a lag
phase of ca. 4 d, the time taken to reach 50% of the plateau
intensity (t50) of 5 d and plateau being reached at ca. 6 d
(Figure 2A). In addition, the MoPrP(89-143, P101L) aggre-
gated in this process is resistant to proteinase K digestion, a
feature of scrapie-like prion aggregates (Figure S3).
Figure 2. Inhibition of moPrP55 misfolding by atheronal-B (1). A) ThT
fluorescence of PrP55 misfolding in the presence of varying concen-
trations of 1. PrP55 (20 mm) was incubated with shaking (900 rpm) at
~
!
*
^
378C with 1; & 0 mm, 10 mm, 50 mm, 100 mm, 150 mm).
B,C) Far-UV CD spectra (0–15 days) of PrP55 (100 mm) incubated in
the absence (B) or presence (C) of 1 (100 mm). D) ThT fluorescence of
PrP55 (100 mm) misfolding in the presence of compounds 1–5
(100 mm). PrP55 alone (c), 1 (c), 2 (c) 3 (c), 4 (c), 5
(c). E) Far-UV CD kinetics of PrP55 aggregation in the presence of
structurally related compounds 1–5 plotted as wavelength at minimal
MRE (mean residue elipticity) in CD spectra vs. time. Color scheme as
in (D). F) ThT fluorescence of the aggregation of PrP55 (25 mm) in the
presence of cholesterol–sphingomyelin liposomes (400 mm of each
lipid) with or without 1 (100 mm). c PrP55 in the presence of
cholesterol–sphingomyelin liposomes; c PrP55 in the presence of
cholesterol–sphingomyelin–1 liposomes; c cholesterol–sphingomye-
lin–1 liposomes; c cholesterol–sphingomyelin liposomes.
When the protein (20 mm) is incubated with increasing
concentrations of atheronal-B (1) (10–150 mm) the ThT
fluorescence profile reveals
a concentration-dependent
decrease in plateau fluorescence intensity and an increase in
the t50 (Figure 2A). With increasing concentration of the lipid
aldehyde, the lag phase also becomes longer (ca. 10 d). At the
highest concentration of 1 (150 mm), PrP55 does not show any
ThT fluorescence within the measured timeframe (19 d),
demonstrating a complete inhibitory effect of 1 on the
misfolding/aggregation of PrP55.
Far-UV CD confirms that in the buffer system selected to
study PrP55, in the absence of 1, PrP is initially rich in a-
helices (t = 0 d), with two characteristic local minima (206 nm
and 222 nm) (Figure 2B and Figure S4). Upon heating and
agitation (378C, 900 rpm), a gradual loss in secondary
structure is observed (t = 3 d) followed by conversion into a
b-sheet-rich conformation (t = 11 d) characterised by a mean
molar elipticity minimum at 217 nm. These changes in the far-
UV CD spectra of PrP55 are slowed in the presence of
atheronal-B (1), with conversion into b-sheet only occurring
at 15 d (Figure 2C). The far-UV CD data supports the ThT
fluorescence analysis, insofar the atheronal-B induced
changes in protein secondary structure observed by far-UV
CD occur in the same time frame as the observed changes in
ThT fluorescence.
9470 www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 9469 –9472

Angewandte
Chemie
To assess the structural aspects of atheronal-B induced
inhibition of prion protein misfolding we next studied how the
misfolding of MoPrP(89-143, P101L) was affected by a panel
of structural analogs of 1 (compounds 2–5, Figure 1B). These
studies showed that only aldehyde 1 causes any change in the
misfolding of PrP55 as shown by both time-dependent ThT
fluorescence and far-UV CD spectra (Figures 2D and E and
Figure S4).
Given that the carboxylate and hydroxy analogs of
atheronal-B (2 and 3, respectively) as well as cholesterol
(4), which all lack an aldehyde group but are otherwise
structurally very similar to 1, do not demonstrate an
inhibitory effect, it becomes apparent that the aldehyde
function present in 1 is a critical moiety required for the
inhibition process to occur. Furthermore, the fact that
possessing close structural simile with 1, as does the hydroxy
compound 3, is not sufficient to induce a measurable change
in the aggregation kinetics of the protein, suggests that any
binding event associated with the secocholesterol core is
secondary and minor to an initial presumed covalent inter-
action between the aldehyde of 1 and the protein. In addition,
the fact that the hydrophobic aldehyde 2-naphthaldehyde (5)
has no discernible effect on the misfolding properties of PrP55
suggests that, while the aldehyde group of 1 is essential for its
effect, not all aldehyde-containing hydrophobic compounds
will induce this effect.
Previous studies with amyloid-b (Ab1–40) have shown that
the aldehyde of secosterol 1 binds covalently to lysine
residues Lys16 and Lys28 and the N-terminal amine of
Asp1, but it is only when Lys16 is adducted that amyloido-
genesis occurs.[25] Given that PrP55 contains a cluster of four
lysine residues (100, 103, 105 and 109) (Figure 1A), we
investigated whether this same adduction reaction is affecting
the aggregation of PrP55. Thus, PrP55 was quiescently
incubated in the presence of 1 for 3 h in PBS (pH 7.4)
followed by the addition of an excess of NaBH4. Sodium
borohydride was selected for Schiff base reduction because it
would simultaneously reduce unreacted aldehyde 1 in buffer
and thus prevent any adduction occurring during the reduc-
tion and centrifugation process. After ultracentrifugation, the
pellet was separated from supernatant. Both the pellet
(dissolved in 30% CH3CN/H2O) and the supernatant were
then subjected to analytical HPLC, the peaks collected and
analyzed by MALDI-TOF mass spectrometry (Figure 3 and
Figure S5). The supernatant contained only one peptide
species, i.e. unreacted peptide (77% by area), whereas the
pellet contained unmodified PrP55 (9%) as well as several
atheronal-B adducted peptides (14% combined). The max-
imum number of molecules detected in the aggregates was
two molecules of 1 attached to PrP55. While it is likely that
atheronal-B (1) can attach to PrP55 at more than two loci,
such multiply adducted forms may be below the detection
limit of the analytical HPLC and mass spectrometry. How-
ever, an alternative explanation could be that specific mono-
and bis-adducted forms of PrP55 get trapped into the
misfolded protein during inhibition of aggregation and
hence appear in the pellet. As described above, this phenom-
enon of only certain adducted forms of a protein being
Figure 3. HPLC traces of the supernatant (c) and the pellet (c)
after sodium borohydride reduction of MoPrP(89-143, P101L) (150 mm)
incubated with 1 (150 mm). After reduction, the sample was subjected
to ultracentrifugation and the pellet and the supernatant evaluated by
analytical HPLC and MALDI-TOF mass spectrometry. In the pellet,
adductions of up to two molecules of 1 per molecule of PrP55 could
be detected.
We next investigated whether secosterol 1 could inhibit
misfolding of MoPrP(89-143, P101L) in lipid rafts. Lipid rafts
are cholesterol and sphingolipid-rich membrane domains,
where both GPI-anchored and phosphatidylinositol-specific
phospholipase C (PI-PLC) processed prion locates on cell
membranes, and where the conformational conversion of
PrPC to PrPSc is thought to occur.[26] We prepared unilamellar
liposomes[27] incorporating cholesterol, sphingomyelin and 1
(molar lipid ratio 4:4:1) and studied the misfolding of PrP55
(25 mm) in the presence of these liposomes in PBS (pH 7.4) by
ThT fluorescence. Liposomes that contain atheronal-B (1)
inhibit aggregation of PrP55 and lead to a reduced ThT
fluorescence plateau compared to PrP55 in the presence of
liposomes lacking 1 (Figure 2F).
The data obtained from ThT assays and far-UV CD
demonstrates that atheronal-B is able to retard the misfolding
of the PrP55 truncated prion protein. The CD spectra
revealed that the a-helical form of PrP55 is first converted
into a random coil-rich structure before formation of b-sheets.
Given that not only the switch from random coil to b-sheets is
retarded by aldehyde 1 but also the initial loss in secondary
structure, it seems plausible that the inhibitory effect due to
adduction of 1, results in stabilisation of the initial a-helical
secondary structure. However, both thermal and urea dena-
turation studies of PrP55 have shown no measurable differ-
ence in the overall thermodynamic stability of the PrP55 in
the presence of atheronal-B (1) and so the search for this
inhibitory effect is still ongoing (Figure S6). One can imagine
a scenario where adduction of 1 at specific sites on a protein
will lead to a local increase in hydrophobicity due to the
cholesterol secosterol structure. If this added hydrophobicity
stabilizes a proximate protein domain, a reduction in the free
energy of the a-form of the protein (DGa, the ground state of
the protein) will occur and the rate of transformation into the
b-form will be reduced, because the DDG between the DGa
and the free energy of the transition state for unfolding/
misfolding (DG°) will be increased. Alternatively, for a folded
protein to undergo partial unfolding before refolding into a
more thermodynamically stable form, as occurs in the a to b
transformation of the MoPrP(89-143, P101L) protein, the
protein has to pass through a number of kinetic free energy
amyloidogenic was something we observed with Ab1–40
.
Angew. Chem. Int. Ed. 2009, 48, 9469 –9472
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
9471

Communications
[2] P. Brown, D. C. Gajdusek, Curr. Top. Microbiol. Immunol. 1991,
172, 1.
[3] M. E. Bruce, R. G. Will, J. W. Ironside, I. McConnell, D.
Drummond, A. Suttie, L. McCardle, A. Chree, J. Hope, C.
Birkett, S. Cousens, H. Fraser, C. J. Bostock, Nature 1997, 389,
498.
barriers. If adduction of 1 leads to an elevation in the highest
DG° (free energy of the transition state) or raises another
energy barrier above the level of the transition state in the
absence of 1, then the rate at which the a to b transformation
occurs will also be lowered. Our denaturing studies are only
able to offer partial information into these two possible
mechanisms, because while we are starting from the a-form,
denaturation leads to a completely unfolded form, not the b-
form. However, given that the denaturation profile from the
a-form of MoPrP(89-132, P101L) either heat or urea-medi-
ated is essentially unchanged in the presence of 1, this
suggests atheronal-B is not lowering the DGa. This drives us
to our present conclusion that atheronal-B is elevating a
kinetic barrier, DG°, on the pathway to the b-form.
[4] S. B. Prusiner, Science 1991, 252, 1515.
[5] B. Caughey, G. J. Raymond, J. Virol. 1993, 67, 643.
[6] B. Caughey, R. E. Race, J. Neurochem. 1992, 59, 768.
[7] R. Gabizon, Z. Meiner, M. Halimi, S. A. Ben-Sasson, J. Cell.
Physiol. 1993, 157, 319.
[8] H. Diringer, B. Ehlers, J. Gen. Virol. 1991, 72, 457.
[9] S. A. Priola, A. Raines, W. S. Caughey, Science 2000, 287, 1503.
[10] J. Chabry, B. Caughey, B. Chesebro, J. Biol. Chem. 1998, 273,
13203.
[11] J. Chabry, S. A. Priola, K. Wehrly, J. Nishio, J. Hope, B. Chesebro,
J. Virol. 1999, 73, 6245.
In conclusion, we have shown that the misfolding of a
truncated murine prion protein is retarded in the presence of
the lipid aldehyde atheronal-B (1), but not by structurally and
functionally related compounds (2–5). This effect was seen
both in preparations where atheronal-B was added into
solution and where it was present in liposomes, mimicking
membrane lipid raft domains. This work builds upon our work
studying the effect of lipid-derived aldehydes, that arise from
lipid peroxidation, and their effects on amyloidogenic pep-
tides and proteins. In all previous studies lipid aldehydes,
when adducted to proamyloidogenic peptides accelerate or
trigger the misfolding event. This acceleration occurs with
proteins that are both natively unfolded, such as Ab1–40 and
Ab1–42, and natively folded forms such as apoB100, a-synuclein
and antibody light chains. The fact that aldehyde 1 is able to
inhibit the misfolding of the MoPrP(89-143, P101L) is an
exciting and unexpected result but expands the potential
impact and scope of lipid aldehydes and their impact on
protein misfolding in vivo. One wonders for example, whether
these aldehydes, generated as a by-product of the immune
response in addition to being iatrogenic, could in fact play a
role in protection against certain unwanted infections, such as
prion disease. If such a case were to be true then the right
levels and loci for production would differentiate the
iatrogenic from beneficial effects.
[12] J. Bieschke, Q. Zhang, D. A. Bosco, R. A. Lerner, E. T. Powers,
P. Wentworth, Jr., J. W. Kelly, Acc. Chem. Res. 2006, 39, 611.
[13] P. Wentworth, Jr., J. Nieva, C. Takeuchi, R. Galve, A. D.
Wentworth, R. B. Dilley, G. A. DeLaria, A. Saven, B. M.
Babior, K. D. Janda, A. Eschenmoser, R. A. Lerner, Science
2003, 302, 1053.
[14] Q. Zhang, E. T. Powers, J. Nieva, M. E. Huff, M. A. Dendle, J.
Bieschke, C. G. Glabe, A. Eschenmoser, P. Wentworth, Jr., R. A.
Lerner, J. W. Kelly, Proc. Natl. Acad. Sci. USA 2004, 101, 4752.
[15] J. C. Scheinost, H. Wang, G. E. Boldt, J. Offer, P. J. Wentworth,
Angew. Chem. 2008, 120, 3983; Angew. Chem. Int. Ed. 2008, 47,
3919.
[16] J. Nieva, A. Shafton, L. J. Altobell III, S. Tripuraneni, J. K.
Rogel, A. D. Wentworth, R. A. Lerner, P. Wentworth, Jr.,
Biochemistry 2008, 47, 7695.
[17] D. A. Bosco, D. M. Fowler, Q. Zhang, J. Nieva, E. T. Powers, P.
Wentworth, Jr., R. A. Lerner, J. W. Kelly, Nat. Chem. Biol. 2006,
2, 249.
[18] M. Heikenwalder, N. Zeller, H. Seeger, M. Prinz, P.-C. Klohn, P.
Schwarz, N. H. Ruddle, C. Weissmann, A. Aguzzi, Science 2005,
307, 1107.
[19] K. Kaneko, H. L. Ball, H. Wille, H. Zhang, D. Groth, M. Torchia,
P. Tremblay, J. Safar, S. B. Prusiner, S. J. DeArmond, J. Mol. Biol.
2000, 295, 997.
[20] A. Taraboulos, M. Rogers, D. R. Borchelt, M. P. McKinley, M.
Scott, D. Serban, S. B. Prusiner, Proc. Natl. Acad. Sci. USA 1990,
87, 8262.
[21] M. Rogers, F. Yehiely, M. Scott, S. B. Prusiner, Proc. Natl. Acad.
Sci. USA 1993, 90, 3182.
Arguably, the most important aspect of this study is that if
the data for the murine prion fragment can be transposed to
proteins that induce prion disease in humans, atheronal-B will
offer a new molecular scaffold on which to start structure–
activity studies to develop a new class of compounds that may
prove to be useful to treat prion disease in vivo.
[22] D. D. Laws, H.-M. L. Bitter, K. Liu, H. L. Ball, K. Kaneko, H.
Wille, F. E. Cohen, S. B. Prusiner, A. Pines, D. E. Wemmer, Proc.
Natl. Acad. Sci. USA 2001, 98, 11686.
[23] P. Wentworth, Jr., J. Nieva, C. Takeuchi, R. Galve, A. D.
Wentworth, R. B. Dilley, G. A. DeLaria, A. Saven, B. M.
Babior, K. D. Janda, A. Eschenmoser, R. A. Lerner, Science
2003, 302, 1053.
Received: August 13, 2009
Published online: November 6, 2009
[24] R. Roland, H. Simone, W. Gerhard, G. Rudi, W. T. Kurt, FEBS
Lett. 1997, 413, 282.
[25] Q. Zhang, E. T. Powers, J. Nieva, M. E. Huff, M. A. Dendle, J.
Bieschke, C. G. Glabe, A. Eschenmoser, P. Wentworth, Jr., R. A.
Lerner, J. W. Kelly, Proc. Natl. Acad. Sci. USA 2004, 101, 4752.
[26] T. J. T. Pinheiro, Chem. Phys. Lipids 2006, 141, 66.
[27] E. Wachtel, D. Bach, R. F. Epand, A. Tishbee, R. M. Epand,
Biochemistry 2006, 45, 1345.
Keywords: atheronal · cholesterol · prion proteins ·
protein misfolding
.
[1] A. L. Horwich, J. S. Weissman, Cell 1997, 89, 499.
9472 www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 9469 –9472