Key words
tau - microtubule-associated protein - Trichocomaceae -
Aspergillus nidulans
- Alzheimerʼs disease
Introduction
Alzheimerʼs disease and other neurodegenerative tauopathies are characterized by the
abnormal accumulation of aggregated forms of the microtubule associated protein tau.
Tau aggregation correlates with the type and severity of cognitive impairment in
Alzheimerʼs disease [1]; mutations in the tau gene can
lead to neurodegeneration [2], [3], and tau aggregation can lead to cell death and cognitive defects in
cellular and animal models (reviewed in [4]). It is
therefore believed that inhibiting tau aggregation is a viable therapeutic target
for the treatment of Alzheimerʼs disease and related tauopathies [5], [6].
Tau aggregation is caused when natively unfolded sequences in their
microtubule-binding repeats 3 and 4 undergo a conformational change, potentially as
a result of phosphorylation [7], [8], [9], to β-sheet structure [10], [11] and then interact
with one another to form amyloid-type filaments. It is believed that small molecules
that interact with the β-sheet structure inhibit tau aggregation by
preventing the addition of tau molecules to the growing aggregate. Tau aggregation
inhibitors identified to date belong to chemical classes including phenothiazines
[12], [13], cyanine
dyes [14], [15],
polyphenols [12], pthalocyanines [12], anthraquinones [16],
N-phenylamines [17], phenylthiazolyl-hydrazides [18], rhodanines [19],
quinoxalines [20], and thienopyridazines [21]. These compounds represent a wide variety of
chemical structures, but they share the common feature of multiple aromatic rings.
While it is not known whether these inhibitors will be effective in vivo or
whether they have suitable bioavailability or pharmacokinetic properties to serve
this purpose, it is important to have lead compounds with the appropriate biological
activity for further development.
Fungal natural products and secondary metabolites have historically been a rich
source of compounds with useful biological activities such as antibiotics,
antimicrobials, and antioxidants. Recent advances in genetics and genomics have
greatly facilitated the study of fungal metabolic pathways along with the
identification and purification of biologically interesting compounds. Using an
efficient gene targeting system [22], [23], [24], we have
identified several biosynthetic pathways in Aspergillus nidulans
(Trichocomaceae) that lead to a wide variety of chemical structures [25], [26], [27], [28], [29], [30], [31], [32], [33], [34]. Many of these
compounds contain aromatic ring structures common to previously identified tau
aggregation inhibitors. We therefore sought to determine whether A. nidulans
secondary metabolites may also have tau aggregation inhibition activity.
We assessed the biological activity of 17 compounds using a standard arachidonic acid
induction of tau aggregation in vitro
[35] followed by a filter trap assay [16], [36] and electron
microscopy [37], [38]. The
previously identified tau aggregation inhibitor emodin served as a positive control.
Several of the compounds inhibited aggregation, and the inhibition by three of the
compounds was reproducible and dose-dependent. We also assessed the effect of the
compounds on tauʼs normal function of stabilizing microtubules using a
fluorescence-based assay [39]. While the compounds
reduced the activity of tau in a concentration-dependent manner, tau retained its
ability to stimulate the polymerization of microtubules in the presence of the
compounds, making them interesting candidate compounds for further development.
Lastly, while two of the compounds are structurally similar to compounds that have
been shown to inhibit tau aggregation, the third is quite different structurally and
thus is the founding member of a new class of tau aggregation inhibitors.
Interestingly, this compound is the precursor to the azaphilone chemical class of
molecules, a class that includes compounds with lipoxygenase inhibitor activity
[28], another activity of potential value in the
treatment of dementia.
Results
Because many A. nidulans secondary metabolites have chemical structures
similar to previously identified inhibitors of tau aggregation, we sought to
determine whether these compounds would have biological activity in inhibiting tau
aggregation in vitro. We chose 17 compounds based on their preponderance of
ring structures. These compounds include 8 anthraquinones, 5 xanthones, and 4 other
types of metabolites ([Fig. 1]). One compound (emodin)
had been identified in an earlier study as an inhibitor of tau aggregation [16]. Tau polymerization was initiated in vitro
using a standard arachidonic acid induction assay [35].
Each of the compounds, at a concentration of 200 µM, was preincubated with 2 µM tau
for 20 min before the addition of 75 µM arachidonic acid. The amount of tau
polymerization was determined using a filter trap assay [16], [36].
Fig. 1 Compounds used in this study. The chemical structures are drawn for
the 17 compounds used. The names of the compounds are included with their
structure along with a compound number. Each compound was purified from a single
HPLC peak. The purity of each compound was estimated from its 1H NMR
spectrum (see Supporting Information) and is listed in Table 1S. In F9775
B, NOESY correlations between H-13/H3-7 and H-13/OH-8 suggested that
H3-7, OH-8, and H-13 are on the same face. The fact that the
specific rotation of F9775 A and B is close to zero indicated that both
compounds are racemic mixtures [48]. Stereocenters
on the side chain of 3′-hydroxyversiconol have not been determined due to the
free rotation of the side chain.
Variecoxanthone, 2,ω-dihydroxyemodin, endocrocin, sterigmatocystin,
asperthecin, chrysophanol, shamixanthone, and asperbenzaldehyde reduced the amount
of tau aggregation ([Fig. 2]). Asperbenzaldehyde,
asperthecin, and 2,ω-dihydroxyemodin had the highest levels of inhibition.
Although emodin inhibits tau aggregation when the glycosaminoglycan heparin is used
as an inducer [16], it did not show appreciable
inhibition in our assay. We have previously shown that arachidonic acid is a more
potent inducer of tau aggregation than heparin [35],
and we believe it is likely that arachidonic acid, coupled with the particular tau
isoform we used, overwhelms the ability of emodin to inhibit tau aggregation (see
Discussion).
Fig. 2 Filter trap assay of tau filament formation. Tau polymerization
reactions were performed with 2 µM tau and 75 µM arachidonic acid either with or
without 200 µM compound. The compounds used are listed on the Y-axis. The
resulting amount of tau filament formation was determined by a filter trap
assay. The values for tau filament formation were normalized to the amount of
the no compound control (dashed line). Negative values indicate that there was
less detectable tau on the filter after treatment with a compound than was
observed with monomeric tau in the absence of arachidonic acid. Values are the
average of three trials ± SD. * P ≤ 0.05; ** p ≤ 0.01; *** p ≤ 0.001.
We then sought to confirm the inhibition of tau aggregation using electron
microscopy, another technique used to quantify tau aggregation reactions [40]. Asperthecin, asperbenzaldehyde, sterigmatocystin,
and 2,ω-dihydroxyemodin seemed to cause the largest reduction in tau
aggregation in comparison with control reactions without a compound ([Fig. 3]). We performed quantitative analysis of the
filament lengths in the presence and absence of compounds. Filament lengths were
placed into 50 nm bins. The first bin is 30–50 nm because it is difficult to
distinguish tau particles below 30 nm in length from the background of the electron
microscope grid. All filaments above 200 nm in length were binned together because
this cutoff should represent particles retained by the filter trap assay. Inhibition
reactions containing the positive control emodin, emericellin, variecoxanthone,
2,ω-dihydroxyemodin, endocrocin, sterigmatocystin, F9775A, F9775B,
asperthecin, aloe emodin, demethylsterigmatocystin, ω-hydroxyemodin,
monodictyphenone, 3′-hydroxyversiconol, and asperbenzaldehyde had filament length
distributions that were distinct from reactions without a compound ([Fig. 4]). These differences tended to include more
objects in the 30–50 nm and 50–100 nm bins and fewer filaments in the > 200 nm
bin. Asperbenzaldehyde had the greatest effect on filament formation with no
filaments detected above 100 nm in length.
Fig. 3 EM images of polymerization reactions. Tau polymerization reactions
were performed with 2 µM tau and 75 µM arachidonic acid either with or without
200 µM compound. Aliquots of the reactions were prepared for negative stain
electron microscopy. The individual images are labeled with the compound
used.
Fig. 4 Filament length distributions. Filament lengths from electron
micrographs of tau polymerization reactions were measured, placed into 50 nm
bins (30–50 nm, 50–100 nm, 100–150 nm, 150–200 nm and > 200 nm), and the
lengths were summed to determine the total amount of filament length in each
bin. The first graph shows the control reaction without a compound, and graphs
(1)-(17) with the different compounds, being labeled with the compound name and
number. The filament distribution for the no-compound control is redrawn on each
graph as a light gray line for comparison. Each point is the average
distribution for images of at least 9 different fields ± SD. * P ≤ 0.05; **
p ≤ 0.01; *** p ≤ 0.001.
Many of the compounds reduced tau aggregation as measured by either the filter trap
assay or quantitative electron microscopy, supporting our rationale for choosing
this set of compounds. Also, four compounds showed a reduction in tau aggregation
with both the filter trap assay and electron microscopy: 2,ω-dihydroxyemodin,
sterigmatocystin, asperthecin, and asperbenzaldehyde. We therefore sought to
determine the IC50 for each compound. The IC50 for
sterigmatocystin could not be determined because it did not demonstrate a consistent
and reproducible concentration-dependent decrease in tau filament formation (data
not shown). 2,ω-Dihydroxyemodin, asperthecin, and asperbenzaldehyde gave
reasonably consistent and reproducible concentration-dependent decreases in tau
filament formation and had IC50 values of 205 ± 28 µM, 39 ± 2 µM, and
177 ± 103 µM, respectively ([Fig. 5]).
Fig. 5 IC50 determination. Polymerization reactions at 2 µM tau
and 75 µM arachidonic acid were performed at several different concentrations of
the compounds, and the resulting amount of filament formation was determined by
a filter trap assay. The amount of polymerization was normalized to controls
without a compound (100 %). The normalized data was plotted against the log of
the inhibitor concentration and fit to a dose-response curve (solid line) to
determine the IC50 for A 2,ω-dihydroxyemodin,
B asperthecin, and C asperbenzaldehyde. Data points are the
average of three trials ± SD.
We also determined whether these three compounds would interfere with tauʼs normal
function of stabilizing microtubules. Microtubule polymerization was monitored using
DAPI fluorescence in the presence of tau with and without a compound ([Fig. 6]). Although 2,ω-dihydroxyemodin,
asperthecin, and asperbenzaldehyde showed a dose-dependent reduction in tauʼs
stabilization of microtubules, significant tubulin stabilization remained even at
higher concentrations of inhibitors.
Fig. 6 Microtubule assembly. Tubulin was incubated with either tau protein
alone (•) or tau in the presence of A 2,ω-dihydroxyemodin,
B asperthecin, or C asperbenzaldehyde at compound
concentrations of 50 µM (▪) or 100 µM (▴). Microtubule assembly was monitored by
DAPI fluorescence (y-axis) over time (x-axis) and normalized to microtubule
polymerization in the presence of paclitaxel. Every third time point is shown.
Each point is the average of three independent trials ± SD. The data are fit to
a Gompertz growth curve (solid, dashed, and dotted lines for no compound, 50 µM
and 100 µM compound, respectively). * P ≤ 0.05; ** p ≤ 0.01; *** p ≤ 0.001.
Discussion
Tau filament formation is associated with many neurodegenerative diseases and is
likely a significant contributor to the accompanying neuronal dysfunction and death.
There are currently no available treatments to inhibit tau aggregation in
Alzheimerʼs disease or other related dementias. There is great interest in
identifying small molecules that may inhibit or reverse tau aggregation. Previous
studies have identified several classes of compounds that show promise as
aggregation inhibitors including anthraquinones, phenothiazines, and a
benzothiazolidine derivative [13], [14], [16]. Many of the
previously identified aggregation inhibitors share the common feature of containing
aromatic ring structures. It is believed that these ring structures could interact
with the β-structure that characterizes aggregation prone tau conformations
thereby preventing the propagation of filament formation. While these compounds show
promise, there remains a need for additional lead compounds for further development
into potential therapeutics.
Many fungal natural products, including low molecular weight secondary metabolites,
have been shown to have pharmacological properties such as acting as antibiotics,
cholesterol synthesis inhibitors, immunosuppressants, and others [41]. Recent advances in genetic manipulation of silent
secondary metabolite synthesis gene clusters have enhanced our capabilities to
produce, purify and characterize fungal natural products and also allow the testing
of these small molecules for biological or pharmacological activities. We have
manipulated the genome of Aspergillus nidulans to produce a wide variety of
natural products in amounts that allow ready purification [22], [42]. Many of these small compounds
have aromatic ring structures similar to those found in tau aggregation
inhibitors.
2,ω-Dihydroxyemodin and asperthecin belong to the anthraquinone class of
compounds. Other anthraquinones have previously been shown to inhibit tau
aggregation [16]. Notably, however, in this study both
of these compounds proved to be more potent inhibitors of tau polymerization than
emodin, which has been previously shown to be a strong inhibitor with an
IC50 in the range of 1–5 µM [16].
Although our compounds were much more effective at inhibiting tau aggregation than
emodin, it was surprising that their IC50 values, in the range of
40–200 µM, were much higher than those published for emodin. One likely explanation
is that our assay conditions strongly drive aggregation and provide a particularly
rigorous test for inhibition. The previous study was done using the
glycosaminoglycan heparin sulfate as an inducer [16],
and previous studies have shown that arachidonic acid induces approximately three
times the amount of aggregation as heparin sulfate under similar conditions [35]. The previous emodin study also used 0N3R and 0N4R
isoforms of tau [16], while our current study uses 2N4R
tau. Our previous studies show that in the presence of arachidonic acid, 2N4R tau
is
much more prone to aggregation than the 0N3R and 0N4R isoforms [43]. These factors together could account for the
differences in IC50 values and provide an explanation for why the
IC50 values for our compounds are much higher than those previously
reported even though they are more potent than emodin and, by inference, other
previously reported tau aggregation inhibitors.
In our analysis, 2 of the 8 anthraquinones were active in blocking tau aggregation
while none of the xanthone compounds were as effective. This could indicate the
importance of the chemistry of the central ring. Anthraquinones contain 2 keto
groups on their central ring, while xanthones contain a single keto group and an
ether linkage. It is possible that the keto groups in the central ring are better
aligned to interact with the beta-strand forming regions of tau than the xanthone
ring. It is interesting that the compounds contain ring structures when considering
that heparin sulfate also contains ring structures and arachidonic acid can adopt
conformations in solution that could also be considered to be ring-like in shape.
It
is possible that the inhibitory compounds bind to tau in a manner similar to that
of
the inducers but block the addition of more molecules of tau.
Asperbenzaldehyde has a single aromatic ring and is a stable intermediate in the
asperfuranone biosynthetic pathway. It is particularly interesting because, as an
azaphilone intermediate, it is the founding member of a new class of compounds that
inhibit tau aggregation. Azaphilones are a particularly interesting group of
compounds because they exhibit a great variety of biologically important activities
including antioxidant and anti-inflammatory activities (reviewed in [44]). Another interesting property of asperbenzaldehyde
is that, while it has weak lipoxygenase-1 (LOX-1) inhibitory activity, simple
modifications convert it into a series of strong LOX-1 inhibitors [28]. Inhibition of LOX-1 may help to reduce fatty acid
metabolites of arachidonic acid and docosahexanoic acid that are elevated in
Alzheimerʼs disease [45], and it is possible that
derivatives of asperbenzaldehyde may have more than one positive therapeutic
activity in tau dementias.
One goal of tau therapeutics is that they should prevent undesirable tau aggregation
while still retaining normal tau function of stabilizing the assembly of tubulin
into microtubules. These compounds do reduce tau function but do not completely
inhibit it. It is perhaps not surprising that some diminution occurs in the presence
of compounds because the sequences likely responsible for tau aggregation are
275VQIINK280 and 306VQIVYK311 which
reside in the microtubule binding repeats [10], [11]. If the compounds are interacting with these
sequences as a mechanism for inhibiting tau aggregation, they would also likely have
some impact on microtubule stabilization. It is encouraging that these compounds do
allow some retention of tau function and thus are promising as future seed compounds
to obtain derivatives with even greater ability to inhibit tau aggregation with
fewer effects on tauʼs normal function of stabilizing microtubules.
In conclusion, this report shows that selected secondary metabolites from fungi
include seed compounds that may be modified for further therapeutic benefit. We have
also provided another example of the wide application of fungal natural products.
Lastly, we have identified an azaphilone precursor, asperbenzaldehyde, that inhibits
both tau aggregation and LOX-1. The fact that asperbenzaldehyde has a quite
different chemical structure from previously identified small molecule inhibitors
of
tau aggregation opens a new direction for tau inhibitor discovery.
Materials and Methods
Chemicals and reagents
Full length 2N4R tau (441 amino acids) was expressed in E. coli and
purified as described previously [46]. Arachidonic
acid (ARA) was purchased from Cayman Chemicals. Compounds were purified from
Aspergillus nidulans as described previously [28], [29], [31], [32], [47], [48] from a single peak of HPLC
chromatography, and purity was estimated by NMR (see Table 1S and
Fig. 1S, Supporting Information).
ARA-induced tau polymerization reactions
2 µM recombinant tau protein was incubated in polymerization buffer which
contained 10 mM HEPES (pH 7.64), 5 mM DTT, 100 mM NaCl, 0.1 mM EDTA, and 3.75 %
ethanol. Compounds dissolved in DMSO were added to the tau solution at final
concentrations of 200 µM, 100 µM, 50 µM, or 25 µM as described in Results.
Compounds were allowed to incubate with tau for 20 min before the addition of
75 µM arachidonic acid to initiate tau filament formation [35]. The reactions were allowed to proceed at room
temperature for 16 h before analysis.
Filter trap assay
Tau polymerization reactions, as described above, were diluted to 20 ng/300 µL in
TBS and passed through a pre-wetted nitrocellulose membrane (Bio-Rad
Laboratories) using vacuum force in a dot-blot apparatus (Bio-Rad Laboratories).
The membranes were washed thrice with TBS-0.05 % Tween20 (TBST) and then blocked
in 5 % nonfat dry milk in TBST for 1 h. The membranes were then incubated with
primary antibody mixture [Tau 5 [49] at 1 : 50 000
dilution, Tau 12 [50] at 1 : 250 000 dilution, and
Tau 7 [51] at 1 : 250 000 dilution; antibodies were
a kind gift from Dr. Lester I. Binder] overnight at 4 °C. Membranes were washed
thrice in TBST and incubated with secondary antibody, HRP-linked goat anti-mouse
IgG (Thermo Scientific), for 1 h at room temperature. The membranes were washed
twice in TBS-Tween buffer, and a final wash was made with TBS. The blot was
developed using an ECL (enhanced chemiluminescence) Western blotting analysis
system (GE Healthcare). Images were captured with a Kodak Image Station 4000R
and were quantified using the histogram function of Adobe Photoshop 7.0.
Statistical analyses were performed using 1-way ANOVA with Dunnettʼs multiple
comparison test to compare the triplicate values to control values.
Transmission electron microscopy
Polymerization reaction samples were diluted 1 : 10 in polymerization buffer and
fixed with 2 % glutaraldehyde for 5 min. 10 µL of each sample was added to a
Formvar carbon-coated grid for 1 min. The grid was blotted on filter paper,
washed with water, blotted, washed with 2 % uranyl acetate and blotted before
staining with 2 % uranyl acetate for 1 min followed by a final blotting on
filter paper. The grids were examined with a Technai F20 XT field emission
transmission electron microscope (FEI Co.). Images were taken with the Gatan
Digital Micrograph imaging system. The images were collected at a magnification
of 3600×. The filaments were quantified using Image-Pro Plus 6.0 software [35]. The perimeter of the filaments was determined,
and the length was obtained by dividing the perimeter in half. For quantitative
analysis, filament lengths were placed into bins as described in Results.
Statistical analyses were performed using two-way ANOVA with Bonferroni
post-tests to compare replicate means for each bin size with the no compound
data serving as reference values.
Tubulin polymerization assay
Polymerization of tubulin was measured using a tubulin polymerization assay kit
(Cytoskeleton, Inc.). The reactions were measured in 96-well Costar black
polystyrene flat-bottom plates (Corning, Inc.). Each well contained porcine
tubulin at 2 mg/mL and GTP at 1 mM in 80 mM PIPES buffer (pH 6.9) with 2 mM
MgCl2 and 0.5 mM EGTA. Tau was added to the wells at a
concentration of 1 µM, and the test compounds diluted in DMSO were added at a
final concentration of 200 µM. An equal amount of DMSO was added to each of the
wells. Another well containing 1 µM paclitaxel (99.1 % pure; Cytoskeleton, Inc.)
and tubulin served as a positive control for polymerization and as a standard
for normalizing results for each of the experiments (not shown). Control
reactions containing only buffer without either tau or a compound, and buffer
with the individual test compounds but without tau were also performed (not
shown). Reaction plates were placed at 37 °C and shaken for 5 s in a FlexStation
II fluorometer (Molecular Devices Corp.). The fluorescence was measured at a
constant interval of 1 minute for an hour with an excitation wavelength
(λ
ex) of 355 nm and an emission wavelength
(λ
em) of 455 nm. This gave 60 readings for every well
sample. The resulting data were normalized to the amount of microtubule
polymerization observed in the presence of tau without a compound and fit to the
Gompertz equation [43]. Experiments were performed
in triplicate and averaged. Statistical analyses and individual growth
parameters can be found in Table 2S, Supporting Information.
Supporting information
Data on the estimated purity of the tested compounds based on 1H NMR
spectra and the spectrum of emericellin used to determine compound purity are
available as Supporting Information.
Acknowledgements
The Tau5, Tau12, and Tau7 antibodies were kindly provided by Lester I. Binder
(Northwestern University). This work was supported in part by grant PO1-GM084077
from the National Institute of General Medical Sciences (C. C. C. W. and B. R. O.)
and by the H. L. Snyder Medical Foundation (B. R. O. and T. C. G.).
