Journ31l of in.nder.r.3du.3- Re--search
..Oluin.e 4, is.ue 1 1 - -u.lui. 21i-i-_
The Effect of Lipid Fluidity Modulating Agents on the Activity of Secretases
Approximately 4 million Americans currently suffer from Alzheimer's Disease (AD), a neurodegenerative
disorder characterized by extracellular deposits of neuritic protein plaques. The plaques are composed primarily of
the 39- to 43-residue amyloid-beta (Ap) peptide generated by cleavage of the amyloid precursor protein (APP) via
P- or y-secretase; however a-secretase cleavage results in soluble extracellular APP fragments. The activity of a-
and p-secretase obtained from rat brain homogenates was observed In vitro with the use of secretase-specific
FRET substrates. Since it is understood that the surrounding lipid environment (i.e. lipid rafts) can regulate
the functioning of a- and p-secretase, lipid modulating agents such as 17-alpha-estradiol, cholesterol,
and progesterone, were administered to detect changes in activity. The observed results illustrated that In
vitro effects of the agents were quite different than In vivo effects cited in the literature. The difference may
arise from the lack of lipid to protein dynamics and interactions. The ability to alter secretase functioning via
lipid modulating agents may one day prove to be a valuable therapeutic approach to battling Alzheimer's Disease.
Alzheimer's Disease (AD) is the most frequent cause of senile dementia in the elderly. In developed countries
around the world, the disease has afflicted approximately 10% of individuals over age of 65 years. One of
the hallmarks of AD includes the development of extracellular deposits of neuritic protein plaques, which
are composed primarily of the 39- to 43-residue amyloid-beta (Ap) peptide (Ladror et al., 1994). The
primary deposition of the peptide into AD plaques has triggered the formation of the amyloid hypothesis stating
that amyloid-beta (Ap) peptide initiates a cascade culminating in neurotoxicity and neurodegeneration-
essentially, the amyloid hypothesis highlights amyloid plaques as being a causative factor in AD
pathogenesis (Simons et al., 2001). The generation of amyloid-beta (Ap) peptide via proteolytic processing by the
so-called secretase enzymes is closely associated with lipid rafts composed of sphingolipids and cholesterol.
These particular lipid domains play an integral role in regulating protein trafficking and processing (Simons
and Ehehalt, 2002). Therefore, modulation of lipid rafts by chemical agents, discussed later, may directly affect
the production of AP and consequently AD plaques.
Amyloid-beta (Ap) peptide is actually a fragment derived from the large type I transmembrane protein APP,
the amyloid precursor protein. The precursor protein may be trafficked through either the non-amyloidogenic or
the amyloidogenic pathway (Figure 1). Most APP is normally guided through the non-amyloidogenic route in
which the precursor protein is cleaved at the a-secretase site resulting in two fragments: APPsa (secreted
ectodomain of APP) and a C-terminal fragment (CTFa). The APPsa fragment is secreted and acts as a
neurotrophic protein, whereas the CTFa is internalized and degraded. Cleavage of APP at the a-secretase
site prevents Ap production, however cleavage of APP via two enzymes, termed &beta: and &gamma:
secretases, leads to the production of A&beta: as the precursor protein is routed through the amyloidogenic
pathway (Wolozin, 2001).
APP Cleaving Pathway
' alternative main ,
Figure 1. The diagram portrays the processing of APP (amyloid precursor protein) as the peptide
travels from the intracellular origin to the extracellular matrix. In the plasma membrane, alpha-,
beta-, and gamma-secretases cleave APP at the corresponding sites. Cleavage by alpha-secretase
results in the normal processing of APP and generation of soluble amyloid-beta peptide.
However, cleavage by beta- and gamma-secretases results in the accumulation of insoluble amyloid-
beta peptide and eventually the formation of AD plaques.
The promotion of Ap production via the amyloidogenic pathway occurs when APP escapes processing at the a-site
and undergoes two sequential cuts by p- and y- secretase. p-secretase first cleaves the precursor protein in
the luminal domain resulting in a C-terminal fragment of 10 kDa (CTFp) and APPsp. Next, the resultant p-
stub becomes the substrate for y secretase cleavage culminating in extracellular Ap secretion (Simons et al.,
2001). The accumulation of Ap and its lack of clearance initiate the formation of amyloid fibrils eventually leading
to production of devastating amyloid plaques.
The regulation of APP processing has been strongly linked to the structure and functioning of lipid domains.
The amyloidogenic pathway that includes p- and y-secretase is closely associated with lipid rafts, while the
non-amyloidogenic a-secretase cleavages occur mainly in phospholipid domains. It has been shown that
modulation of the cholesterol content of these lipid domains can direct APP processing toward either the a-
secretase or p/y-secretase pathway (Wolozin, 2001). There seems to exist a dynamic equilibrium between
raft formation and disassembling, and hence a system either favoring Ap generation or opposing it.
Lipid rafts are enriched not only in cholesterol as previously stated, but also
sphingolipids, glycosylphosphatidylinositol (GPI) proteins. The rafts are located predominantly in the exofacial
leaflet of plasma membrane, connected to a phospholipid domain in the inner cytofacial leaflet of the lipid
bilayer. The difference in fluidity arises as a major distinction between a lipid raft and a phospholipid domain.
Rafts are fluid, but more ordered and tightly packed than the surrounding phospholipid bilayer. The difference
in packing and fluidity is due to the saturation of the hydrocarbon chains in raft sphingolipids and phospholipids
as compared with the unsaturated fatty acids in the liquid-disordered phase i.e. phospholipid domains (Simons
and Ehehalt, 2002).
Experimental evidence prescribed by Riddell et al. and Wahrle et al. substantiates the presence of p- and y-
secretase, respectively, in lipid rafts. Using both detergent and non-detergent methods, Riddell et al. found BACE1
(a type of p-secretase) protein and activity in a light membrane raft fraction containing other components of
the amyloidogenic pathway. Furthermore, depletion of raft membrane cholesterol discontinued association
between BACE1 and the light membrane fraction, supplanting crucial evidence of the steroid's role in stabilizing
lipid raft structure (Riddell et al., 2001).
The non-amyloidogenic a-secretase has residence in non-raft phospholipid domains. Studies performed by Kojro
et al. exhibit a strong inverse correlation between cholesterol and a-secretase activity. Simply, the data
demonstrate that the effect of cholesterol on a-secretase activity is reversible and dependent on membrane
fluidity. Fractionation studies also examined the prime location of ADAM1O (a type of a-secretase) using
detergent methods via Triton X-100. The results concluded the presence of the non-amyloidogenic enzyme in
high-density fractions opposite that of p- or y-secretase found in low-density fractions associated with lipid
raft domains (Kojro et al., 2001).
In this paper, the intense association between lipid domains and secretase activities will be examined. The
activities of a-secretase and p-secretase under various lipid domain conditions will be investigated via
fluorescent studies involving FRET substrates. The specific goal of this paper is to elucidate and demonstrate
the relationship between lipid domains, more specifically lipid rafts, and secretase enzymes associated with AD.
Rat Brain Homogenate Preparation
Commercially available brains were obtained from normal strain rat donors with no evidence of neurological
disorder. After thawing the brain to room temperature, the sample was homogenized in a glass homogenizer for
1 minute in a tissue buffer containing 25mM Tris pH 7.5, 0.3 M sucrose, and 2.5mM EDTA. The tissue/buffer ratio
was 1:5 g/ml. The homogenized samples were centrifuged twice for 15 minutes at 5000 x g at 40C. The
supernatant containing the lipid-intact secretases were preserved and the pellets were discarded.
FRET Substrate Cleaving Mechanism
2)Figure 2. TheL schematic* represents the basis upon which the secretase activity assay methodology
were added to dtct changes in enzyme activity.
Figure 2. The schematic represents the basis upon which the secretase activity assay methodology
is based. The mechanism of FRET substrate cleavage occurs is shown. Each FRET substrate has
two fluorophores, Cl and C2; moreover, the emission spectrum of the first fluorophore overlaps with
the excitation spectrum of the second fluorophore. Therefore, the decrease in fluorescence from
the intact peptide is proportional to the increase in enzyme activity.
FRET Substrate Cleaving Assay
After In order to experimentally observe the activity of the secretase enzymes under various conditions,
FRET (fluorescence resonance energy transfer) substrates specific for a-secretase and P-secretase were
obtained from Enzyme Systems (Livermore, CA). The mechanism of the FRET substrates is illustrated in Figure
2. The activity assays were performed in standard 96-well opaque plates at 370C. For standard experiments
each well contained 250uL buffer of 40mM Tris at pH 7.5, lOuL of lOuM FRET Substrate (either a or P), and
lOuL brain homogenate. The experiment was run in a fluorescent spectrophotometer at excitation and
emission wavelengths of 320nm and 405nm, respectively, for 60 minutes cycling at 1-minute intervals. lOuL
of various modulating agents such as P-methyl-cyclodextrin, cholesterol, 17-alpha-estradiol, or progesterone,
were added to detect changes in enzyme activity.
-- -- ----:: . -.-.-.--- - - - -- - - ----.- . . - ----------. ---------.
0 10 20 0 40 s 6 70
Figure 3. The detection of alpha- and beta-secretase activity in rat brain homogenates observed in
a fluorescent spectrophotometer using FRET substrates specific to each enzyme. The substrates have
an excitation wavelength of 320nm and an emission wavelength of 405nm. The experiment was run
for 60 minutes, cycling at 1-minute intervals. The graphic display above is used as a standard of
control for further experiments.
Standard results for alpha- and beta-secretase activity are shown in Figure 3. The curves represent the
innate activity for each enzyme under In vitro conditions. The graph ideally shows a hyperbolic beta-secretase
curve alongside linear alpha-secretase activity, and a flat control line indicating zero activity. After administering
lOuL of specific modulating agents, the results obtained were compared to the basis above.
a " *A%"
0 to 2o 30 40 10 4 0 70
Figure 4. The effect of 10uL of 10mM 17-alpha-estradiol in ethanol on the activities of alpha- and
beta-secretase obtained from rat brain homogenates. Both alpha- and beta-secretase activities
decreased upon estradiol addition, however the decline in activity for alpha-secretase was dramatic.
Upon addition of 17-alpha-estradiol, Figure 4 illustrates that alpha- and beta-secretase activity decreased, while
the control remained constant. Moreover, alpha-secretase activity decreased significantly when compared to
the decline in beta-secretase activity. Further, each activity curve is markedly linear, and the hyperbolicity of
beta-secretase in the standard curve diminishes. The effects of progesterone are very similar to the results
obtained upon 17-alpha-estradiol administration.
m. * tm..k
o 10 20 30 40 s0 s0 70
Figure 5. The illustration above portrays the effects of progesterone on rat alpha- and beta-
secretase activities. An application of homogenateuL of 1mM progesterone in ethanol was administered. It
was observed that progesterone attenuates the activity of both alpha- and beta-secretase. However,
was observed that progesterone attenuates the activity of both alpha- and beta-secretase. However,
the alpha-secretase activity decline more prominently than beta-secretase activity. The results are
very similar to the effects of 17-alpha-estradiol.
Progesterone also caused the activities of both alpha- and beta-secretase to decrease. Figure 5 portrays
the dramatic decrease in alpha-secretase activity, while a much less significant decline in beta-secretase activity as
a result of progesterone. The hyperbolicity is again reduced and linearized.
Figure 6. Rat brain alpha- and beta-secretase activities upon a 10uL application of a beta-
methyl-cyclodextrin (50mM) and cholesterol (10mM) solution in ethanol are shown above. The
plasma membrane. It is observed that alpha-secretase activity declines significantly upon
00 . - CM4*
methyl-cyclodextrin. Cholesterol causes a significant reduction in alpha-secretase activity, while beta-
secretase activity is relatively the same. When compared to Figure 3, the activity of beta-secretase becomes
linear and shows no characteristic standard hyperbolic curve. Alpha-secretase becomes more predominant than
Figure 6. Rat brain alpha- and beta-secretase activities upon a leul application of a beta-
methyl-cyclodetCtrin (50mM) and cholesterol (uomM) solution in ethanol are shown above. The
beta-secretase in the standard controls in Figures 4, 5, and 6. The control groups in the experiments
with modulating agents did not seem to correlate too well with the standard experiment shown in Figure 3.
Although the activity of secretase enzymes is innately connected to various lipid domains and rafts, the In
vitro effects of lipid fluidity modulating agents are much different than In vivo effects. Ideally, the activity of
alpha-secretase should be either increased or stabilized, while the beta-secretase activity should have
decreased upon addition of both 17-alpha-estradiol and progesterone, and extraction of cholesterol from
the membrane. Hence, modulating agents destabilizing the structure of lipid rafts should cause alpha-
secretase stability and beta-secretase instability. The difference between In vitro effects in this paper and In
vivo effects cited in literature is attributed to the lack of lipid dynamics and lipid-protein interactions existing In vivo.
Furthermore, the large decrease in alpha-secretase activity compared to small reductions in beta-secretase activity
is accredited to the In vitro stability of different lipid domains. The lipid environment surrounding beta-secretase
In vitro is hypothesized to be less susceptible to modulating agent perturbation, than the lipid
environment surrounding alpha-secretase. Furthermore, the neuroprotective effects of 17-alpha-estradiol
and progesterone were seemingly reversed. The compounds must have acted very similarly in both systems since
the results seen in Figures 4 and 5 are also very similar. It is thought that 17-alpha-estradiol and
progesterone somehow stabilized lipid raft formation in the membrane destabilizing alpha-secretase activity.
The addition of cholesterol via beta-methyl-cyclodextrin caused drastic decreases in alpha-secretase activity,
while the beta-secretase activity was approximately unchanged. The dramatic decrease in alpha-secretase
activity can be the result of raft degradation initiated by [Cholesterol] Inside << [Cholesterol] Outside. Since it
is assumed that alpha-secretase activity is located in a cholesterol-depleted environment, addition of
cholesterol would shift the lipid membrane equilibrium towards raft formation and stabilization, thus
destabilizing alpha secretase functionality. Additionally, the small decrease in beta-secretase activity upon addition
of cholesterol via beta-methyl-cyclodextrin can be associated with stable lipid raft constructions, and a condition
in which [Cholesterol] Inside -[Cholesterol] Outside.
1. Dahlqvist A. Method for assay of intestinal disaccharidases. Anal Biochem 7:18-25,1964
2. Doumas BT, Bayse DD, Carter RJ, et al: A candidate reference method for determination of total protein in serum
I. Development and validation. Clin Chem 27: 1642-1650, 1981.
3. Eichholz A: Structural and functional organization of the brush border of intestinal epithelial cells. 3.
Enzymatic activities and chemical composition of various fractions of tri-disrupted brush borders. Biochem
Biophys Acta 135:475-482, 1967.
4. Lin, Cheng-mao. Effect of Dietary Nucleotide Supplementation on In Vivo and In Vitro Immune Function in
Protein-Malnourished Mice. University of Florida. Ph.D. dissertation. December 1995.
5. Lin RI, Schjeide OA. Micro estimation of RNA by the cupric ion catalyzed orcinol reaction. Anal Biochem 27:473, 1969.
Back to the Journal of Undergraduate Research
College of Liberal Arts and Sciences I University Scholars Program I University of Florida |
ï¿½ University of Florida, Gainesville, FL 32611; (352) 846-2032.
LW1 UNIVERSITY of
The Fordhonfifï¿½sr fV? I G4lkor Nahoff