Pyrimethamine Derivatives: Insight into Binding Mechanism and Improved Enhancement of Mutant β- N -acetylhexosaminidase Activity

Article abstract of DOI:10.1021/jm5017895

In order to identify structural features of pyrimethamine (5-(4-chlorophenyl)-6-ethylpyrimidine-2,4-diamine) that contribute to its inhibitory activity (IC₅₀ value) and chaperoning efficacy toward β-N-acetylhexosaminidase, derivatives of the compound were synthesized that differ at the positions bearing the amino, ethyl, and chloro groups. Whereas the amino groups proved to be critical to its inhibitory activity, a variety of substitutions at the chloro position only increased its IC₅₀ by 2-3-fold. Replacing the ethyl group at the 6-position with butyl or methyl groups increased IC₅₀ more than 10-fold. Surprisingly, despite its higher IC₅₀, a derivative lacking the chlorine atom in the para-position was found to enhance enzyme activity in live patient cells a further 25% at concentrations >100 μM, while showing less toxicity. These findings demonstrate the importance of the phenyl group in modulating the chaperoning efficacy and toxicity profile of the derivatives.

Full text of DOI:10.1021/jm5017895

Article
pubs.acs.org/jmc
Pyrimethamine Derivatives: Insight into Binding Mechanism and
Improved Enhancement of Mutant β‑N-acetylhexosaminidase
Activity
Michael B. Tropak,† Jianmin Zhang,‡ Sayuri Yonekawa,† Brigitte A. Rigat,† Virender S. Aulakh,‡
Matthew R. Smith,‡ Hee-Jong Hwang,‡ Marco A. Ciufolini,‡ and Don J. Mahuran*,†,§
†Genetics and Genome Biology, SickKids, PGCRL 686 Bay Street, Toronto, Ontario M5G 0A4, Canada
‡Department of Chemistry, University of British Columbia, 2036 Main Mall Vancouver, British Columbia V6T 1Z1, Canada
§Department of Laboratory Medicine and Pathology, University of Toronto, Medical Science Building, 1 King’s College Circle, 6th
Floor, Toronto, Ontario M5S 1A8, Canada
S
* Supporting Information
ABSTRACT: In order to identify structural features of pyrimethamine (5-(4-
chlorophenyl)-6-ethylpyrimidine-2,4-diamine) that contribute to its inhibitory
activity (IC50 value) and chaperoning efficacy toward β-N-acetylhexos-
aminidase, derivatives of the compound were synthesized that differ at the
positions bearing the amino, ethyl, and chloro groups. Whereas the amino
groups proved to be critical to its inhibitory activity, a variety of substitutions at
the chloro position only increased its IC50 by 2−3-fold. Replacing the ethyl
group at the 6-position with butyl or methyl groups increased IC50 more than
10-fold. Surprisingly, despite its higher IC50, a derivative lacking the chlorine
atom in the para-position was found to enhance enzyme activity in live patient cells a further 25% at concentrations >100 μM,
while showing less toxicity. These findings demonstrate the importance of the phenyl group in modulating the chaperoning
efficacy and toxicity profile of the derivatives.
INTRODUCTION
■
Pyrimethamine (PYR, 1, Figure 1A) is a known human anti-
malarial drug that inhibits the parasite’s (e.g., Plasmodium
falciparum) dihydrofolate reductase (DHFR).1 Interestingly,
the compound also acts as an effective pharmacological
chaperone (PC) for lysosomal β-N-acetylhexosaminidase A
(Hex A).2 Indeed, we have shown that, in some late-onset
forms of GM2 gangliosidosis (Tay-Sachs and Sandhoff
diseases), the Hex A activity in patient fibroblasts can be
enhanced by 2−3-fold.3 Recently it has been reported that
Figure 1. (A) Structure of pyrimethamine (PYR). (B) Amino acid
treating Dutch APPE693Q mice (a mouse model of Alzheimer’s
residues (green) in the active site involved in binding PYR (yellow).
The chlorine atom on PYR is highlighted in red. Residues participating
in formation of the active site from the adjacent subunit are shown in
disease) with a PC for Hex dramatically reduced the
accumulation of a β-amyloid peptide:ganglioside (GM2 and
GM3) complex. This reduction was associated with improved
learning behavior and decreased anxiety in the treated mice.4
Thus, PCs of Hex have the potential to treat a much broader
patient base than just those with late-onset Tay-Sachs or
Sandhoff disease.
marine blue. Amino acid residues (D290, D240, E491) directly
forming hydrogen bonds (denoted with dashed blue lines) with the
amino (blue) groups on PYR are highlighted in green. The ethyl group
of PYR is found in a hydrophobic pocket formed by residues W405
and W424 (green).
Three Hex isozymes arise through the combinatorial
dimerization of two subunits encoded by two evolutionarily
related genes, the α (encoded by the HEXA gene) and/or β
(HEXB gene), i.e, the major Hex A (αβ) and Hex B (ββ)
isozymes and the minor, unstable Hex S (αα) isozyme.5 Of the
three Hex isozymes, only heterodimeric Hex A is able to
efficiently associate with the GM2 activator protein/GM2
ganglioside (GM2) to form the ternary complex and hydrolyze
the non-reducing terminal β-1,4-linked N-acetylgalactosamine
from GM2, producing GM3 ganglioside. In the late-onset
variants of GM2 gangliosidosis, point mutations in either the α-
subunit (Tay-Sachs disease) or the β-subunit (Sandhoff
disease)6 reduce the activity of Hex A below a critical threshold
Received: November 27, 2014
Published: May 18, 2015
© 2015 American Chemical Society
4483
DOI: 10.1021/jm5017895
J. Med. Chem. 2015, 58, 4483−4493

Journal of Medicinal Chemistry
Article
of enzyme activity needed to prevent GM2 storage (5−10% of
normal values).7 Since PYR is a micromolar inhibitor of Hex
isozymes (inhibitory activity (IC50) = 15 μM), it binds to and
stabilizes the folded mutant enzyme in the endoplasmic
reticulum (ER), thereby increasing the number of molecules
that are able to retain their active folded conformation, allowing
the proper disulfide bonds to form and the enzyme to be
transported to the lysosome.2 The presence of disulfide bonds,
the acidic pH of the lysosome, and the presence of substrate all
combine to displace PYR while continuing to stabilize Hex A.
Additionally, the Ki of PYR as an inhibitor of Hex increases at
acidic pH.
We have demonstrated experimentally that PYR treatment of
late-onset Tay-Sachs patients’ cells loaded with a fluorescent
GM2 ganglioside analogue produces an enhancement of the
cells’ ability to turnover the substrate.8 However, the
chaperoning efficacy of PYR is mutation dependent.2
RESULTS AND DISCUSSION
■
To gain insight into the molecular details of how PYR binds to
Hex, and to identify PYR derivatives with improved enzyme
enhancement efficacy in terms of dosing and response, three
regions of 1 (PYR) were modified: (1) the 2,4-diamino groups
on the pyrimidine backbone, (2) the 6-ethyl alkyl side chain,
and (3) the phenyl group itself or the substituents on the
phenyl group attached to the pyrimidine ring.
Synthesis of PYR Analogues. Analogues of PYR retaining
the 2,4-diaminopyrimidine motif were prepared according to
published procedures10 as outlined in Scheme 1, which also lists
Scheme 1. General Procedure for Synthesis of PYR
Analogues and Representative Compounds Examined in
This Study
The structure of Hex B in complex with PYR has only
recently been elucidated.9 Amino acids Asp240 and Glu491 are
within hydrogen-bonding distances of the amino groups in PYR
and are predicted to act as hydrogen bond acceptors (Figure
1B). The ethyl group of PYR is positioned within a
hydrophobic pocket formed by Trp405, Trp424, and Trp489.
The chlorine atom in the para-position on the phenyl of PYR
forms a weak hydrogen bond with a conserved water molecule
near the mouth of the substrate-binding pocket and points
toward the bulk solvent.
PYR has been evaluated in both adult-onset Tay-Sachs
disease (ATSD) and Sandhoff disease (ASD) patients in an
open-label Phase I/II clinical trial to evaluate treatment efficacy
and tolerability.3 The observed increase in Hex A activity levels
in patients’ leukocytes (up to a maximum of 4-fold) was closely
correlated with their plasma levels of PYR, which varied
according to the oral dose of the drug (25, 50, 75, or 100 mg/
day) and the metabolism of each patient; i.e., patients taking
the same dose could have very different levels of plasma PYR.
As an anti-malarial, PYR has 1000-fold higher affinity for the
parasite versus human DHFR. Patients in this study were
supplemented with folic acid (leucovorin) as a precaution
against human DHFR inhibition. Nonetheless, side effects were
seen in most patients treated with PYR at or above 75 mg or
with plasma levels >1.5 mg/L. Therefore, it would be desirable
to obtain derivatives of PYR that enhance Hex activity while
concomittantly minimizing undesirable side effects.
In order to identify features of PYR that contribute to its Hex
inhibitory activity and chaperoning efficacy, i.e., PC activity, as
well as possibly contributing to the detrimental side effects seen
at higher doses, derivatives of the compound were synthesized.
These differed from PYR at the positions bearing the amino,
ethyl, and chloro groups. Whereas the amino groups and the
ethyl group at the 5-position proved to be critical for its PC
activity, a variety of substitutions at the chloro position only
increased the IC50 by 2−3-fold. Surprisingly, the derivatives
lacking the pendant chloro group were found to enhance
enzyme activity in live patient cells by a further 25% at
concentrations >100 μM, despite their higher IC50. Combined
with the 3D structure of the Hex A, these derivatives have
provided additional insight into the binding mechanism of PYR,
as well as into features associated with improved chaperoning
efficacy and reduced toxicity.
the structures of representative compounds examined in the
course of this study. Deaminated variants of 1 and 5g were
obtained starting with vigorous acid hydrolysis of the parent
compounds,11 followed by reaction of the emerging pyrimidi-
nones with POCl3 and hydrogenolysis of the resultant
chloropyrimidines (Scheme 2). Pyrimidines such as 11 were
prepared in a like fashion from compounds of the type 10,
which in turn were obtained by reaction of an appropriately
substituted compound 4 with urea in EtOH−EtONa.11
Inhibitory Activity of the Deaminated Analogues of
PYR. Enzyme activity of purified human Hex A in the presence
of the different PYR analogues was monitored using the
Scheme 2. Preparation of Des-amino Analogues of 1 and 5g
4484
DOI: 10.1021/jm5017895
J. Med. Chem. 2015, 58, 4483−4493

Journal of Medicinal Chemistry
Article
colorimetric substrate p-nitrophenyl-β-D-N-acetylgluco-
pyranoside (pNPGlcNAc) (see Experimental Section). Re-
moval of one (cf. 8, 9) or both (cf. 11) amino groups from the
pyrimidine ring resulted in a >10-fold increase in IC50 values
(Table 1). Substitution of the 4-amino group with a carbonyl
Inhibitory Activity of PYR Analogues with Modified
Aryl Groups. Bateman et al. in 2011 demonstrated that the
chlorine atom on the phenyl group makes a hydrogen bond
with a water molecule located near the entrance to the active
site of Hex; however, the phenyl ring was found to project
toward the bulk solvent, suggesting that a larger variety of
substitutions could be accommodated on the underlying ring
without greatly affecting its binding affinity toward the
enzyme.9 Several PYR derivatives, differing in terms of the
pendant group (chloro, methyl, methoxy, or trifluoromethyl)
and their relative position (meta, para, or ortho) on the phenyl
ring, were synthesized and evaluated for inhibitory activity and
chaperoning efficacy. Representative examples are shown in
Table 3. Derivatives incorporating a para-substituted phenyl
Table 1. Biological Activities of Some Des-amino PYR
Analogues
compound
1
6
8
9
11
a
IC50
PC activity
17
3
>2100
>400
>400
>940
b
2.1 0.1
−
−
−
−
a
IC50 (μM) determined using purified human Hex A and pNPGlcNAc
b
(0.5 mM). PC activity: Intracellular enhancement (increase in) of
enzyme activity based on the ratio of Hex A activity in treated ATSD
patient fibroblasts versus mock treated (DMSO). Enzyme activity was
based on the hydrolysis of the α-active-site specific MUGS (1.6 mM)
by lysates from treated and untreated ATSD patient fibroblasts.
Table 3. Effect of Phenyl Group Substitution on the
Biological Activity of PYR Analogues 5
a
a
compd
R1
R2
IC50
17
PC activity
1
4-Cl-C6H4
4-MeO-C6H4
4-CF3-C6H4
Ph
Et
Et
Et
Et
Et
Et
Et
Et
3
2.1 0.1
2.0 0.3
2.2 0.1
2.7 0.1
1.7 0.1
2.9 0.1
2.3 0.1
2.5 0.1
(cf. 6) also resulted in a >120-fold increase in IC50. Both amino
groups are therefore essential for activity. This is consistent
with their role as H-bond donors within the active site of the
enzyme deduced from the 3D structure of the HexB:PYR
complex.9
Inhibitory Activity of PYR Analogues with Modified 6-
Alkyl Side Chains. Shortening of the alkyl side chain from an
ethyl to methyl side arm resulted in a more than 25-fold
increase in IC50 values, i.e., from 17 to 560 μM (Table 2).
5f
5g
5h
5i
5j
5k
5l
18
15
32
6
2
5
4-Me-C6H4
3-Me-C6H4
2-Me-C6H4
3,5-di-Me-C6H4
40 16
34
43
47
7
4
6
a
b
IC50 (μM): see Table 1, footnote a. PC activity: see Table 1,
footnote b.
group (5f, 5g, and 5i), irrespective of the nature of the
Table 2. Effect of the Side-Chain Length on the Biological
Activity of PYR Analogues 5
substituent, displayed an average IC50 value (24
13 μM)
similar to that of PYR (17 3 μM). Collectively, this group of
compounds, including PYR, will be referred to as the PYR
analogues bearing a para-substituted phenyl group. In contrast,
analogues containing an ortho- or meta-substituted phenyl
group (5j, 5k, 5l) and/or lacking a substituent in the para-
a
b
compd
R1
R2
IC50
17
>560
18
PC activity
1
4-Cl-C6H4
4-Cl-C6H4
4-Cl-C6H4
4-Cl-C6H4
4-Cl-C6H4
4-Cl-C6H4
Et
3
2.1 0.1
5a
5b
5c
5d
5e
Me
−
n-Pr
3
1.5 0.1
i-Pr
>410
>1400
−
−
−
position (5h) on average showed a 2.4-fold higher IC50 (41
5
n-Bu
μM) compared to PYR. Collectively, this group of compounds
is referred to as the PYR analogues lacking a para-substituted
phenyl group. The increased IC50 values associated with this
group of compounds may reflect the steric clashes of ortho- and
meta-positioned pendant groups with the residues at the
entrance to the active site, due to the of axial rotation of the
phenyl group. Such steric clashes would be minimized in the
para-positioned PYR analogues.
Chaperoning Efficacy of PYR Derivatives. To compare
the enzyme enhancement efficacy of the PYR derivatives in
Tables 2 and 3, fibroblasts from a patient with ATSD were
treated for 5 days with the different PYR derivatives. Only
compounds with an IC50 value of less than 100 μM were
evaluated. Although the IC50 values of PYR and 5b differed by
less than 20%, they diverged significantly in their ability to
enhance mutant (G269S) Hex A activity in ATSD patient
fibroblasts. Whereas PYR treatment resulted in a maximal
increase of 2.1-fold in Hex A activity, the larger propyl
derivative (5b) showed only a 1.5-fold increase, a relative
decrease of 35% in chaperoning efficacy. This underscores the
importance of the size of the 6-alkyl chain, and given its
importance in binding, it is unlikely that any further
modifications of this group would lead to an improved level
of enhancement. On the other hand, the relatively minor effect
on IC50 values (approximately 2−3-fold increase) associated
CH2NH2
180 20
a
b
IC50 (μM): see Table 1, footnote a. PC activity: see Table 1,
footnote b.
Similarly, increasing the length beyond a propyl or introducing
a branched chain, such as isopropyl, also resulted in a more
than 20-fold increase in IC50 values relative to PYR. By
comparison, the IC50 of the propyl derivative was not
significantly different than that of PYR. These results are in
line with the size of a hydrophobic pocket formed by Trp
residues 405, 424, and 489 that can, in the context of a 6-alkyl
side chain on PYR, accommodate an ethyl or propyl but not a
bulkier group, such as butyl or isopropyl.9,12 A similar trend in
increasing IC50 was observed with thiazoline derivatives that
contained analogously positioned alkyl chains of increasing
size.13
The general acid−base residue of Hex B (Glu355) is located
near the hydrophobic pocket, and its proximity suggested that
replacing the alkyl side chain with an amino group would
improve binding via formation of an additional electrostatic
interaction. However, substitution of the 6-ethyl group with
methylamine resulted in a >10-fold increase in the IC50,
whereas replacement with an azidomethyl group did not
significantly affect the IC50 value.
4485
DOI: 10.1021/jm5017895
J. Med. Chem. 2015, 58, 4483−4493

Journal of Medicinal Chemistry
Article
Figure 2. Comparison of the toxicity of PYR and 5h in patient fibroblasts and their respective in vitro inhibitory activity against human DHFR.
Figure 3. Comparison of in cellulo enhancement of mutant Hex A enzyme activity in ATSD fibroblasts by selected PYR derivatives bearing
modifications on the 5-phenyl group at various concentrations, i.e., determination of the level of “maximal increase” for each derivative. Panels (a) 5j,
circles, and 5h, squares; (b) 5l, circles, and 5k, squares; (c) PYR, circles, and 5i, squares; and (d) 5g, circles, and 5f, squares, all giving the graphical
representations of the dose−response curves. Open symbols represent the β-galactosidase (Bgal) activity ratio, which serves as a control for toxicity,
e.g., open circles, for each drug, and the same filled symbols, e.g., filled circles, represent the Hex A activity ratio between treated and mock treated
cells.
with changing the pendant group linked to the 5-phenyl ring
alludes to the potential for increasing the chaperoning potency
of the PYR derivatives by modifying the aryl group.
increase in mutant Hex A activity in ATSD patient fibroblasts:
2.7- and 2.9-fold, respectively. Among the three, treatment with
compound 5h demonstrated the least decline in cell viability at
the highest concentration employed. The decreased toxicity of
5h compared to PYR was confirmed using the Alamar blue
assay as an independent readout of cell viability (Figure 2).
Additionally, the nearly identical IC50 values of the two
compounds for human DHFR argue against attributing the
disparity in toxicity profiles to their differences in inhibitory
activity for the human enzyme. As an alternative target to
account for the differential toxicity of PYR and 5h, it is
interesting to note that PYR is known to significantly decrease
phosphorylation of STAT3, which is necessary for its
activation.14 Furthermore, decreased phosphorylation of
STAT3 is observed in cells treated with 100 μM PYR.
Although these effects were observed in different cells, the
concentrations at which these observations were made are very
similar to the concentrations (30−100 μM) at which significant
toxicity was seen in PYR-treated patient fibroblasts. Activation
of STAT3 is required for cell proliferation and cell survival.15,16
Chaperoning Efficacy of PYR Analogues Bearing a
Modified Aryl Group. The group of PYR analogues bearing a
pendant group in the para-position on the phenyl ring (Table
3: 5f, 5g, and 5i) enhanced intracellular mutant Hex A activity
in ATSD patient fibroblasts (2.0 0.3-fold average enhance-
ment) to a similar degree as PYR (2.1 0.1-fold). In contrast,
the PYR derivatives lacking a pendant group in the para-
position (5j, 5k, 5l, and 5h) on average enhanced intracellular
Hex A activity to a greater degree (30% higher values, i.e., 2.6
0.3-fold enhancement) than either PYR (2.1 0.1-fold) or the
derivatives incorporating a para-substituted phenyl group (2.0
0.3-fold). These results demonstrated that derivatives lacking
a substituent at the para-position on the phenyl moiety
generally showed an enhanced ability to increase the intra-
cellular activity of mutant Hex A in ATSD patients, despite
having decreased inhibitory activities (higher IC50) compared
to PYR. Compounds 5h and 5j produced the greatest maximal
4486
DOI: 10.1021/jm5017895
J. Med. Chem. 2015, 58, 4483−4493

Journal of Medicinal Chemistry
Article
Table 4. Inhibitory Activity of PYR and 5h for Different Hex Isozymes at Acidic and Neutral pH
pH 4.5
pH 4.5
pH 7.0
a
a
b
b
a
a
compd
Hex A
Hex B
9.1 1.2
38
Hex A WT
Hex A G269S
Hex A
3.1 1.0
11
Hex B
c
1
8.9 1.1
30
14
27
4
5
8.5 3.0
4.5 1.2
15
5h
3
5
27
6
2
3
a
b
Hex isozymes were purified from human placenta. The Hex A isozyme, wild-type (WT), or mutant (G269S) was isolated from lysates of normal
c
or ATSD fibroblasts, respectively, by ion exchange chromatography. All IC50s (μM) were determined using the universal MUG (1.6 mM) Hex
substrate.
Figure 4. Steady-state protein levels of the Hex α-subunit are increased to a greater degree in 5h- versus PYR-treated patient fibroblasts. Lysates from
ATSD patient fibroblasts were treated for 5 days with PYR (33 or 11 μg/mL) or 5h (33 or 11 μg/mL) and analyzed by Western blotting using a
proprietary rabbit polyclonal IgG against human Hex A, followed by anti-rabbit peroxidase conjugated IgG and chemiluminescence detection (ECL
kit, Amsersham). Solid and open arrowheads denote position of bands corresponding to the mature (lysosomal) α- and β-subunits of Hex,
respectively.
One could speculate that differential effects of PYR and 5h on
the phosphorylation of STAT3 could account for the variations
in the toxicity profiles of the two compounds. Irrespective of
the mechanism of action in terms of toxicity, the fact remains
that 5h appears to be significantly less toxic than PYR, and that
this decrease in toxicity likely accounts for the increased
chaperoning efficacy of 5h at higher concentrations. Therefore,
5h and PYR were compared in terms of their effects on
intracellular trafficking of Hex A and in cellulo hydrolysis of a
GM2 ganglioside derivative at the concentrations that
corresponded to each compound’s highest level of chaperone
efficacy. A closer examination of the enzyme enhancement
capability of the phenyl derivatives over a range of treatment
concentrations (Figure 3) revealed that the peak increase
occurs at a 2−3-fold lower concentration (30−40 μM) for the
para-positioned functional groups as compared to those
analogues functionalized at other positions (100−140 μM).
This treatment range parallels the differences seen in the IC50
values between the two classes of analogues (i.e., those bearing
versus those lacking a para-substitution on the phenyl ring).
The activity level of another lysosomal enzyme (β-
galactosidase), which is not inhibited by PYR or its derivatives,
was used as a surrogate marker of cell viability. At the highest
concentration evaluated (>400 μM), all PYR derivatives,
including the parent compound but excluding 5h, showed a
>60% reduction in β-galactosidase activity, indicative of
decreased cell viability. The para class of derivatives exemplified
by PYR showed significant signs of declining viability at lower
concentrations (<200 μM) compared to the non-para class of
compounds, such as 5h, which showed a decline in viability at
concentrations >200 μM.
values for the wild-type Hex A and B isozymes were similar,
differing by less than 30% (Table 4).
The inhibitory activity of 5h, like that of PYR, also showed an
approximately 2.5-fold decrease in IC50 values at neutral versus
acidic pH. Although it is a relatively small change, both
compounds would be expected to bind to the mutant enzyme
to a greater degree in the neutral pH of the ER, facilitating
protein folding, versus the acidic environment of the lysosome,
where the complex needs to dissociate in order for the enzyme
to turnover its substrate.
Comparison of Lysosomal Targeting of mutant Hex A
by PYR and 5h. On the basis of protein levels observed by
Western blotting (Figure 4), peak levels of mature, lysosomally
derived Hex α-subunit (∼55 kDa)17 were seen in lysates from
ATSD patient cells treated with 33 μg/mL of 5h and 11 μg/mL
of PYR, relative to mock (1% DMSO) treated cells. Maximal
levels of mature Hex α-subunit were higher in 5h-treated than
in PYR-treated cells. The increased steady-state levels of the
Hex subunit protein paralleled the increase in Hex A enzyme
activity levels observed in similarly treated ATSD cells (Figure
3). These results are suggestive of increased transport of
mutant Hex A to the lysosome in ATSD cells treated with 5h
compared to those treated with PYR. Immunofluorescence
staining of 5h/PYR-treated versus mock treated patient cells
was used as an independent verification of increased transport
of mutant Hex A to the lysosomes in the presence of either
compound (Figure 5). On the basis of fluorescence intensity
alone, increased levels of Hex A (green) were observed in both
5h- and PYR-treated cells, compared to mock treated cells.
Increased colocalization of signals from antibodies directed
against Hex A and LAMP1, a lysosomal marker, was further
evidence of increased transport of mutant Hex A to lysosomes.
Improved Hydrolysis of a Fluorescent Ganglioside
Derivative in Live, PYR-Treated, or 5h-Treated Cells. Cell
feeding experiments using a fluorescent GM2 ganglioside
analogue were used to directly measure intralysosomal
hydrolysis of GM2 ganglioside in live ATSD patient cells
(i.e., an in cellulo assay).8 Fibroblasts from an ATSD patient
Comparison of the Inhibitory Activity of PYR and 5h
against Different Hex Isozymes and Wild-Type versus
Mutant Hex A. As an alternative explanation for the
differences in chaperoning efficacy between PYR and 5h, it
was possible that they exhibited different inhibitory activity
toward wild-type versus mutant Hex A. However, their IC50
4487
DOI: 10.1021/jm5017895
J. Med. Chem. 2015, 58, 4483−4493

Journal of Medicinal Chemistry
Article
lysosomal Hex A activity in response to treatment with either
compound. However, 5h treatment resulted in an additional
∼1.5-fold increase in GM2 hydrolysis in comparison to that of
PYR-treated cells (Figure 6B). A parallel increase is observed in
Hex A activity measured in the same lysates using the
fluorogenic substrate MUGS (Figure 6C). These results are
significant for two reasons. First, there is always a concern that
inhibitory molecules acting as PCs could continue inhibiting
the activity of Hex A in the lysosome. These results clearly
demonstrate that this is not the case for PYR or 5h, as a robust
increase in GM2 hydrolysis was observed, while it is very likely
that, at higher concentrations of either compound, one would
eventually see residual inhibition in lysosomes. However, at
such higher concentrations, cellular toxicity would also become
an issue. Thus, these doses represent the high end of those
expected to be used to treat patients. Second, using two
orthogonal Hex A enzyme assays, the results clearly and
convincingly demonstrate that, at higher concentrations, 5h is
superior to PYR in enhancing lysosomal levels of Hex A
activity. These results are also significant from a therapeutic
perspective. Since 5h is less toxic at higher concentrations than
PYR in cells, it likely could be used at the higher doses in
ATSD and ASD patients, which would further enhance Hex A
activity beyond the critical threshold necessary to prevent the
continuous storage of GM2 gangliosides in their lysosomes.
The therapeutic potential of 5h is predicated on its ability to
cross the blood brain barrier (BBB). Pharmacokinetic studies in
small animals and quantification of drugs levels in serum and
cerebrospinal fluid of PYR-treated GM2 gangliosidosis3 and
AIDS18 patients have clearly demonstrated that PYR crosses the
BBB, i.e., enters the nervous system. The physiochemical
properties of PYR in terms of number of hydrogen acceptors
(4) and calculated LogP (lipophilicity) value (2.8) are
congruent with other drugs capable of crossing the BBB.19
Thus, the fact that 5h is similar to PYR with regard to number
of hydrogen acceptors (4) and predicted LogP value (PYR, 2.8,
versus 5h, 2.2) suggests that 5h would also be capable of
crossing the BBBthis will need to be verified experimentally.
Recently it has been reported that treating Dutch APPE693Q
mice (a mouse model of Alzheimer’s disease) with a PC for
Hex dramatically reduced the accumulation of a β-amyloid
peptide:ganglioside (GM2 and GM3) complex, improved
learning behavior, and decreased anxiety.4 Indeed, we have
preliminary data demonstrating that both PCs, PYR and 5h, are
able to increase the activity levels of wild-type Hex A by
approximately 1.3-fold (p < 0.05) relative to Hex A in DMSO-
treated fibroblasts from non-GM2 gangliosidosis patients
(Supporting Information, Figure 1S). Taken together with
the data demonstrating the effectiveness of PYR and 5h in
enhancing mutant Hex A activity, these observations suggest
that these compounds not only could be used to treat some
forms of GM2 gangliosidosis but also could be potential
therapeutics for Alzheimer’s disease.
Figure 5. Increased colocalization (yellow) of Hex A (green) protein
with the lysosomal marker LAMP1 (red), following the treatment of
p.G269S/p.G269S α-Hex patient fibroblasts with either 5h (33 μg/
mL) or PYR (11 μg/mL). Nuclei are shown stained in blue (DAPI).
Treated fibroblasts were permeabilized and probed with a proprietary
rabbit polyclonal IgG anti-Hex A pre-absorbed with purified human
Hex B and a LAMP1 mouse antibody. Primary antibody binding was
visualized using the corresponding secondary Alexa Fluor 488 chicken
anti-rabbit (green) or Alexa Fluor 594 chicken anti-mouse conjugated
antibodies (red). Colocalization of Hex A and LAMP1 staining is
shown on panels labeled MERGE.
were first treated with PYR (11 μg/mL) or 5h (33 μg/mL) for
10 days to build up their levels of active mutant Hex A prior to
feeding the cells with the fluorescent analogue of GM2. The
doses of PYR and 5h used to treat cells were chosen on the
basis of the concentrations previously determined to produce
the maximal increases in both Hex A protein and activity
(MUGS) levels in treated cells, i.e., the best balance between
their chaperoning versus toxicity profiles. Unlike in vitro assays
of cell lysates with artificial colorimetric/fluorogenic substrates,
in cellulo hydrolysis of the fluorescent GM2 occurs in the
lysosomes and requires the mutant enzyme to functionally
interact with the GM2 activator protein:GM2 complex. This
assay also allows a read-out of any residual inhibition of Hex A
caused by the chaperone once the complex enters the lysosome.
Following Folch extraction of treated cells, neutral glycolipids
(i.e., glycolipids produced further downstream in the GM2
breakdown pathway) in the organic phase and acidic ganglio-
sides in the aqueous phase were resolved by high-performance
thin-layer chromatography (Figure 6A). Patient cells treated
with 5h (33 μg/mL) demonstrated a greater intracellular
hydrolysis of the fluorescent derivative of GM2 compared to
cells treated with PYR (11 μg/mL), as indicated by the higher
levels of the band corresponding to GM3. Quantification of the
intensity of all bands corresponding to the downstream
products of GM3 hydrolysis, lactosylceramide (LacCer) and
glucosylceramide (GlcCer) (further hydrolysis was inhibited by
the inclusion of an irreversible inhibitor of glucocerebrosidase
in the cell feeding assay8), in both phases clearly shows
increased levels of the products, indicative of increased
CONCLUSION
■
We have experimentally verified the importance of both amino
groups on the pyrimidine ring of PYR and the 6-ethyl side
chain in the binding of the molecule to the Hex isozymes,
consistent with the predictions derived from the crystal
structure of the Hex B:PYR complex. We have demonstrated
that, while the phenyl group and its associated moieties do not
have a great impact on Hex binding, they significantly increase
the ability of the derivatives to enhance the intracellular/
4488
DOI: 10.1021/jm5017895
J. Med. Chem. 2015, 58, 4483−4493

Journal of Medicinal Chemistry
Article
Figure 6. Increased intracellular hydrolysis of a fluorescent derivative of GM2 ganglioside in ATSD patient fibroblasts treated with 5h or PYR.
Hydrolyzed fluorescent GM2 ganglioside derivatives isolated from patient fibroblasts pre-loaded with fluorescent GM2 were resolved by TLC. (A)
Chromatograms showing the resolution of metabolized GM2 derivatives in the two phases generated by Folch extraction. (B) Quantification of
hydrolyzed levels of ganglioside was derived from the optical density of the bands corresponding to GM3 in the aqueous phase and LacCer and
GlcCer in the organic phase. (C) Relative Hex A activity (MUGS) in lysates prepared from the cells used to measure GM2 hydrolysis following
treatment with DMSO (Ctrl), PYR, or 5h.
“t” (triplet), “q” (quartet), “dd” (doublet of doublets), “ddd” (doublet
of doublets of doublets), and “m” (multiplet), and further qualified as
“br” (broad) or “c” (complex). Infrared (IR) spectra (cm−1) were
recorded from films deposited on NaCl plates. Mass spectra (m/z)
were recorded in the electrospray ionization (ESI) or atmospheric
pressure chemical ionization (APCI) mode, as indicated. Melting
points (mp, uncorrected) were measured on a Mel-Temp apparatus.
Commercial reagents and solvents were used without further
purification, except for THF (freshly distilled from Na/Ph2CO
under Ar) and CH2Cl2 (freshly distilled from CaH2 under Ar).
Commercial BuLi solutions were titrated against Ph2CHCOOH.20
The reactions were monitored by thin-layer chromatography (TLC)
using silica gel 60 F254 pre-coated plates, and spots were visualized with
a UV lamp or by using KMnO4 or vanillin stain. Flash chromatography
was performed on Silicycle 230−400 mesh silica gel. All reactions were
carried out under an Ar atmosphere. Oven-dried flasks fitted with
rubber septa for the introduction of substrates/reagents/solvents via
syringe, and equipped with Teflon stirring bars, were employed for
reactions involving air- and moisture-sensitive reagents.
lysosomal activity of mutant G269S Hex A. We have identified
a PYR derivative, 5h, that further increases mutant Hex A
activity by another 30% over the maximum observed with PYR.
Significantly, we see an increase in the intracellular hydrolysis of
GM2 in the lysosomes of fibroblasts treated with this
compound. The maximal increase with 5h is seen at a higher
concentration than with PYR, which is likely caused by the
more pronounced negative effects on cellular viability observed
with PYR at concentrations >100 μM. As opposed to changes
to the amino groups and ethyl arm on the pyrimidine ring,
modifications to the phenyl group do not dramatically change
the strength of binding, they but appear to decrease the toxicity
and thus increase the apparent chaperoning efficacy of the
compounds at higher concentrations. As a starting point,
modification of the phenyl headgroup represents a site for
making further improvements in chaperoning efficacy while
minimizing toxicity.
General Procedure for the Preparation of Pyrimidine-2,4-
diamines. A mixture of an appropriate 2-aryl-3-methoxy-3-alkyl-
acrylonitrile10 (4 mmol), guanidine hydrochloride (2 equiv), and
NaHCO3 (2.2 equiv) in dry DMSO (10 mL) was heated at 90 °C for
5h. The cooled mixture was poured into EtOAc and washed
sequentially with water, saturated NaHCO3 solution, and brine. The
organic phase was dried (MgSO4) and concentrated under reduced
pressure. The crude product was purified by trituration with diethyl
EXPERIMENTAL SECTION
■
Experimental ProtocolsChemistry. Unless otherwise stated,
1H (300 MHz) and 13C (75.5 MHz) NMR spectra were recorded at
room temperature (rt). Chemical shifts are reported in parts per
million (ppm) on the δ scale, and coupling constants, J, are given in
hertz (Hz). Multiplicities are reported as “s” (singlet), “d” (doublet),
4489
DOI: 10.1021/jm5017895
J. Med. Chem. 2015, 58, 4483−4493

Journal of Medicinal Chemistry
Article
ether/hexanes or by recrystallization from methanol/diethyl ether.
Overall yields were between 20 and 30% and were not further
optimized unless the compound proved to be of interest. The
following compounds were thus obtained:
5-(4-Chlorophenyl)-6-methylpyrimidine-2,4-diamine (5a). Known
compound,21 mp >260 °C (lit.18 264−265 °C). IR: 3471, 3306, 3091,
1624, 1553, 1477 1438, 1263, 1086. 1H NMR (acetone-d6): 7.45 (2H,
d, J = 8.3, ArH), 7.26 (2H, d, J = 8.3, ArH), 5.38 (2H, bs, NH2), 5.26
(2H, bs, NH2), 1.89 (3H, s, Me). 13C NMR (DMSO-d6): 162.4 (C),
162.3 (C), 162.2 (C), 135.6 (C), 132.9 (CH), 132.1 (C), 129.3 (CH),
106.2 (C), 22.4 (Me). APCI-MS: 235.3 [M35Cl+H]+ (100%) and
237.2 [M37Cl+H]+ (35%).
5-(4-Chlorophenyl)-6-propylpyrimidine-2,4-diamine (5b). Known
compound,22 mp 173−175 °C (lit.21 171−174 °C). IR: 3456, 3309,
3162, 2965, 1627, 1549, 1428, 1401, 1273, 1088, 1000. 1H NMR
(CDCl3): 7.44 (2H, d, J = 8.5, ArH), 7.18 (2H, d, J = 8.5, ArH), 5.18
(2H, bs, NH2), 4.61 (2H, bs, NH2), 2.26 (2H, t, J = 7.8, CH2CH2Me),
1.55 (2H, sex, J = 7.8, CH2CH2Me), 0.81 (3H, t, J = 7.4, CH2CH2Me).
13C NMR (DMSO-d6): 165.2 (C), 162.14 (C), 162.11 (C), 135.2 (C),
132.8 (CH), 131.9 (C), 129.0 (CH), 105.9 (C), 36.2 (CH2), 21.6
(CH2), 14.1 (Me). ESI-MS: 263.2 [M35Cl+H] (100%) and 265.1
[M35Cl+H]+ (35%).
5-(4-Chlorophenyl)-6-isopropylpyrimidine-2,4-diamine (5c). Mp
240−242 °C. IR: 3440, 3317, 3126, 2967, 1628, 1549, 1425, 1275,
1248, 1086. 1H NMR (CDCl3): 7.43 (2H, d, J = 8.4, ArH), 7.18 (2H,
d, J = 8.4, ArH), 4.74 (2H, bs, NH2), 4.40 (2H, bs, NH2), 2.63 (1H,
sept, J = 6.7, CH(Me)2, 1.07 (6H, d, J = 6.7, CH(Me)2. 13C NMR
(CDCl3): 172.2 (C), 162.1 (C), 161.9 (C), 134.0 (C), 133.8 (C),
132.0 (CH), 129.5 (CH), 106.6 (C), 31.1 (CH), 21.4 (Me). APCI-
MS: 263.4 [M35Cl+H]+ (100%) and 265.3 [M35Cl+H]+ (35%). ESI-
HRMS: calcd for C13H16N435Cl [M+H]+ 263.1063; found 263.1061.
5-(4-Chlorophenyl)-6-butylpyrimidine-2,4-diamine (5d). Known
compound,21 mp 204−207 °C (lit.21 208−210 °C). 1H NMR
(DMSO-d6): 7.45 (2H, app d, J = 8.3, ArH), 7.16 (2H, app d, J =
8.3, ArH), 5.98 (2H, bs, NH2), 5.70 (2H, bs, NH2), 2.10−2.05 (2H,
cm, Ar-CH2(CH2)2CH3), 1.39 (2H, quintet, J = 7.4, Ar-
CH2CH2CH2CH3), 1.10 (2H, sextet, J = 7.4, Ar(CH2)2CH2CH3),
0.70 (3H, t, J = 7.4, Ar(CH2)3CH3). 13C NMR (DMSO-d6): 165.0
(C), 162.6 (C), 162.0 (C), 135.2 (C), 133.1 (CH), 132.3 (C), 129.3
(CH), 106.3 (C), 33.9 (CH2), 30.8 (CH2), 22.4 (CH2), 14.1 (Me).
6-(Aminomethyl)-5-(4-chlorophenyl)pyrimidine-2,4-diamine (5e).
Treatment of 5-(4-chlorophenyl)-6-benzyloxymethyl-pyrimidine-2,4-
diamine (prepared by the general procedure above from benzyloxy-
acetyl chloride and 2-(4-chlorophenyl)acetonitrile) with 45% HBr in
AcOH23 afforded 5-(4-chlorophenyl)-6-bromomethyl-pyrimidine-2,4-
diamine, which was purified by trituration with ether. A portion of this
material (32 mg, 102 mmol) was reacted with NaN3 (66 mg, 1.02
mmol, 10 equiv) in DMSO (5 mL, rt, 12 h). The customary aqueous
extractive workup afforded a solid, which was triturated with ether to
afford 11 mg (30%) of pure 5e, mp 155−158 °C. IR: 3459, 3320,
3168, 2922, 2101, 1623, 1551, 1435, 1260, 1089, 1009. 1H NMR
(CDCl3): 7.45 (2H, d, J = 8.0, ArH), 7.20 (2H, d, J = 8.0, ArH), 5.0
(2H, bs, NH2), 4.73 (2H, bs, NH2), 3.88 (2H, s, CH2). 13C NMR
(CDCl3): 162.5 (C), 161.9 (C), 160.1 (C), 134.6 (C), 133.9 (C),
131.8 (CH), 129.8 (CH), 107.7 (C), 52.1 (CH2). APCI-MS: 276.4
[M35Cl+H]+ (100%) and 278.3 [M37Cl+H]+ (30%). ESI-HRMS: calcd
for C11H11N735Cl [M+H]+ 276.0764; found 276.0764.
and 252.4 [M37Cl+H]+ (100%). ESI-HRMS: calcd for C11H13N535Cl
[M+H]+ 250.0859; found 250.0858.
6-Ethyl-5-(4-(methoxyphenyl)pyrimidine-2,4-diamine (5f). 1H
NMR (DMSO-d6): 7.49 (2H, d, J = 8.4, ArH), 6.98 (2H, d, J = 8.4,
ArH), 5.81 (2H, bs, NH2), 5.40 (2H, bs, NH2), 3.78 (3H, s, OCH3),
2.10 (2H, q, J = 7.0, CH2), 0.95 (3H, t, J = 7.0, CH3). 13C NMR
(DMSO-d6): 167.1 (C), 162.8 (C), 162.4 (C), 158.7 (C), 132.1 (CH),
128.2 (C), 114.8 (CH), 108.5 (C), 55.4 (CH3), 27.9 (CH), 13.6 (Me).
6-Ethyl-5-(4-(trifluoromethyl)phenyl)pyrimidine-2,4-diamine
(5g). Known compound,24 mp 194−197 °C (lit.22 mp not given). IR:
3499, 3316, 3167, 1625, 1554, 1441, 1314, 1123, 1101, 1013; 1H NMR
(CDCl3): 7.71 (2H, d, J = 8.0, ArH), 7.39 (2H, d, J = 8.0, ArH), 4.99
(2H, bs, NH2), 4.58 (2H, bs, NH2), 2.26 (2H, q, J = 7.6, CH2Me),
1.07 (3H, t, J = 7.6, CH2Me). 13C NMR (CDCl3): 168.6 (C), 162.0
(C), 161.8 (C), 139.5 (C), 131.1 (CH), 130.0 (q, JC−F = 32.7, C),
126.2 (q, JC−F = 3.7, CH), 124.0 (q, JC−F = 272.1, CF3), 107.1 (C),
28.2 (CH2), 13.3 (Me). APCI-MS: 283.4 [M+H]+ (100%).
6-Ethyl-5-phenylpyrimidine-2,4-diamine (5h). Known com-
pound,21 optimized yield 86%, mp 242−244 °C (lit.21 237−240
°C). IR: 3423, 3301, 3152, 2980, 1628, 1556, 1430, 1273, 1230, 1069.
1H NMR (CDCl3): 7.48−7.23 (5H, m, ArH), 4.70 (2H, bs, NH2),
4.45 (2H, bs, NH2), 2.30 (2H, q, J = 7.6, CH2Me), 1.07 (3H, t, J = 7.6,
CH2Me). 13C NMR (DMSO-d6): 166.9 (C), 162.5 (C), 136.4 (C),
131.0 (CH), 129.4 (CH), 127.6 (CH), 107.0 (C), 27.9 (CH2), 13.6
(Me). ESI-MS: 215.4 (M+H, 100%).
6-Ethyl-5-(p-tolyl)pyrimidine-2,4-diamine (5i). Mp 212−215 °C.
1
IR: 3305, 1627, 1567, 1233. H NMR (DMSO): 7.24 (d, 2H, J = 7.9
Hz), 7.06 (d, 2H, J = 7.9 Hz), 6.25 (s, 2H), 5.90 (b.s, 2H), 2.33 (s,
3H), 2.11 (q, 2H, J = 7.6 Hz), 0.96 (t, 3H, J = 7.6 Hz). 13C NMR
(CDCl3): 168.0, 162.4, 161.3, 137.6, 130.3, 129.9, 108.3, 28.0, 21.2,
13.4. LRMS: 229.4 [M+H+]. HRMS: calcd for C13H17N4 229.1453;
found [M+H+].
6-Ethyl-5-(m-tolyl)pyrimidine-2,4-diamine (5j). Mp 200−202 °C.
IR: 3423, 3302, 3144, 2974, 1624, 1556, 1433, 1280, 1232, 1124, 1002.
1H NMR (DMSO-d6) 7.30 (1H, t, J = 7.6, ArH), 7.13 (1H, d, J = 7.6,
ArH), 6.98 (1H, s, ArH), 6.95 (1H, d, J = 7.6, ArH), 5.82 (2H, bs,
NH2), 5.39 (2H, bs, NH2), 2.31 (3H, s, Me), 2.09 (2H, q, J = 7.5,
CH2Me), 0.94 (3H, t, J = 7.5, CH2Me). 13C NMR (CDCl3): 168.4
(C), 162.2 (C), 161.7 (C), 138.9 (C), 135.2 (C), 131.1 (CH), 129.0
(CH), 128.5 (CH), 127.5 (CH), 108.6 (C), 28.2 (CH2), 21.4 (Me),
13.4 (Me). APCI-MS: 229.3 [M+H]+ (100%). ESI-HRMS: calcd for
C13H17N4 [M+H]+ 229.1453; found 229.1451.
6-Ethyl-5-(o-tolyl)pyrimidine-2,4-diamine (5k). Mp 174−176 °C.
1
IR: 3439, 3311, 3175, 2977, 1624, 1549, 1428, 1267, 1113, 984. H
NMR (CDCl3) 7.31−7.29 (3H, m, ArH), 7.13 (1H, d, J 8.1, ArH),
4.70 (2H, bs, NH2), 4.34 (2H, bs, NH2), 2.20 (2H, q, J = 7.6,
CH2Me), 2.14 (3H, s, Me), 1.04 (3H, t, J = 7.6, CH2Me). 13C NMR
(CDCl3): 168.4 (C), 161.89 (C), 161.86 (C), 137.9 (C), 134.2 (C),
131.0 (CH), 130.6 (CH), 128.3 (CH), 126.6 (CH), 107.4 (C), 28.1
(CH2), 19.5 (Me), 12.8 (Me). APCI-MS: 229.4 [M+H]+ (100%). ESI-
HRMS: calcd for C13H17N4 [M+H]+ 229.1453; found 229.1455.
5-(3,5-Dimethylphenyl)-6-ethylpyrimidine-2,4-diamine (5l). Mp
223−225 °C. IR: 3405, 3310, 3169, 2971, 2922, 1600, 1553, 1430,
1
1276, 1003. H NMR (DMSO-d6): 6.94 (1H, s, ArH), 6.75 (2H, s,
ArH), 5.79 (2H, bs, NH2), 5.34 (2H, bs, NH2), 2.27 (3H, s, Me), 2.09
(2H, q, J = 7.5, CH2Me), 0.94 (3H, t, J = 7.5, CH2Me). 13C NMR
(CDCl3): 168.3 (C), 162.3 (C), 161.6 (C), 138.7 (C), 135.1 (C),
129.3 (CH), 128.1 (CH), 108.7 (C), 28.1 (CH2), 21.3 (Me), 13.4
(Me). APCI-MS: 243.4 [M+H]+ (100%). ESI-HRMS: calcd for
C14H19N4 [M+H]+ 243.1610; found 243.1606.
Zinc dust (78 mg, 1.2 mmol) was added to a solution of the above
azido compound (11 mg, 40 mmol) in THF/NH4Cl(aq) (2 mL, 1:1)
and stirred at rt for 1 h. The solution was filtered through Celite,
diluted with water, and extracted with ethyl acetate. The combined
organic extracts were dried (Na2SO4) and concentrated under reduced
pressure. The crude residue was triturated with chloroform to the give
the product as a colorless solid (10 mg, 100%), mp 171−174 °C. IR:
3331, 3206, 2920, 1612, 1546, 1455, 1348, 1257, 1088, 1007.1H NMR
(acetone-d6): 7.55 (2H, d, J = 8.2, ArH), 7.43 (2H, d, J = 8.2, ArH),
6.14 (2H, bs, NH2), 4.35 (2H, s, CH2). 13C NMR (acetone-d6): 163.8
(C), 160.5 (C), 157.3 (C), 134.0 (C), 132.4 (CH), 131.2 (C), 129.8
(CH), 105.8 (C), 51.3 (CH2). APCI-MS 250.5 [M35Cl+H]+ (100%)
2-Amino-6-ethyl-5-(4-chlorophenyl)pyrimidin-4(3H)-one (6).
Known compound prepared as described by Trattner et al. in
1964.25 This material was not purified and was advanced to 8 in crude
1
form (<85% purity). H NMR (DMSO-d6): 7.77 (d, 2H, J = 8.2 Hz,
ArH), 7.44 (d, 2H, J = 8.2 Hz, ArH), 7.0 (br s, 1H, NH), 5.8 (br s, 1H,
NH), 3.35 (br s, 1H, NH), 2.04 (q, 2H, J = 7.5 Hz, CH2), 1.02 (t, 3H,
J = 7.5 Hz, CH3). 13C NMR (DMSO-d6): 165.4 (C), 157.1 (C), 156.1
(C), 138.3 (C), 132.3 (CH), 126.3 (CH), 122.9 (C), 103.2 (C), 23.7
(CH2), 13.4 (CH3).
4490
DOI: 10.1021/jm5017895
J. Med. Chem. 2015, 58, 4483−4493

Journal of Medicinal Chemistry
Article
2-Amino-6-ethyl-5-(4-(trifluoromethyl)phenyl)pyrimidin-4(3H)-
one (7). 1H NMR (DMSO-d6): 7.68 (d, 2H, J = 8.0 Hz), 7.39 (d, 2H, J
= 8.0 Hz), 6.67 (bs, 2H), 2.17 (q, 2H, J = 7.4 Hz), 1.02 (t, 3H, J = 7.4
Hz). 13C NMR (DMSO-d6): 162.7 (C), 155.0 (C), 140.6 (C), 132.0
(CH), 127.5 (q, JC−F = 32 Hz, C), 126.7 (CH), 125.0 (q, JC−F = 3.8
Hz, CH), 124.8 (q, JC−F = 273 Hz, CF3), 112.3 (C), 28.2 (CH2), 13.3
(CH3). IR: 3326, 3125, 1657, 1323. LRMS: [M+H+] 284.4. HRMS:
calcd for 284.1011 C13H13F3N3O; found 284.1021 [M+H+].
2-Amino-4-ethyl-5-(4-chlorophenyl)pyrimidine (8). Known com-
pound prepared as described by Russell et al. in 1954.26 1H NMR
(CDCl3): 8.52 (s, 1H, Pyr-H), 7.49 (d, 2H, J = 8.1 Hz, ArH), 7.22 (d,
2H, J = 8.1 Hz, ArH), 4.79 (bs, 2H, NH2), 2.41 (q, 2H, J = 7.5,
CH2Me), 1.20 (t, 3H, J = 7.6, CH2Me). 13C NMR (CDCl3): 167.3
(C), 161.0 (C), 157.4 (C), 134.5 (C), 132.6 (C), 131.0 (CH), 129.9
(CH), 115.6 (C), 28.1 (CH2), 13.3 (CH3).
performed using bacterially expressed and purified human DHFR,
supplied as a kit from Sigma-Aldrich (USA). Inhibition studies of
purified human O-GlcNAcase by PYR were performed in the
laboratory of Dr. David Vocadlo at Simon Fraser University (Burnaby,
British Columbia, Canada) as described previously.13
Evaluating Intracellular Hex A Activity in Treated Patient
Fibroblasts. The chaperoning efficacy of each PYR derivative, i.e., the
maximum level of increase in mutant Hex A activity in treated versus
non-treated fibroblasts from an ATSD patient, homozygous for the
G269S mutation in the Hex α-subunit, was determined for each test
compound. G269S cells were plated at 70−90% confluence (10 000−
20 000 cells) in 96-well plates. Patient cells were treated for 5 days in
the growth media (DMEM media supplemented with 10% fetal bovine
serum) containing each of the PYR derivatives, serially diluted. DMSO
levels were maintained at 1%. Subsequently, the media containing
compounds were completely aspirated, and cells were washed twice
with PBS prior to lysis of cells in McIlvaine buffer (pH 4.5), containing
0.4% Triton X-100 and 0.025% human serum albumin at 4 °C for 30
min. Aliquots of the crude lysates were used to measure Hex A and β-
galactosidase activity utilizing the fluorogenic substrates, i.e., MU-2-
acetamido-2-deoxy-β-D-glucopyranoside-6-sulfate (MUGS, 1.6 mM),
which is hydrolyzed primarily by the α-active site of Hex A, and MU-β-
galactopyranoside (0.45 mM), respectively. Enzyme reactions were
performed at 37 °C for 1−2 h and subsequently terminated using 0.1
M MAP (pH 10.5) followed by fluorescence measurement (Ex = 365
nm and Em = 450 nm) on a Molecular Devices M2 spectrofluor-
ometer. Intracellular enzyme activity levels were expressed relative to
mock (1% DMSO) treated cells, where 1 = no change. All treatments
were performed in triplicate.
Western Blotting and Immunofluorescence Imaging of
Treated Patient Fibroblasts. Both methodologies were as
previously described.29 For immunofluorescence, fibroblasts were
first fixed with 2.5% paraformaldehyde in PBS, permeabilized with
0.5% saponin, and blocked with 10% normal goat serum. The cells
were then probed with a proprietary Hex A rabbit antibody, which was
pre-absorbed with purified human Hex B to produce an α-specific
antibody, and a LAMP1 mouse antibody (from the Developmental
Studies Hybridoma Bank developed under the auspices of the NICHD
and maintained by the University of Iowa, Department of Biological
Sciences, Iowa City, IA 52242). Primary antibody binding was
visualized using the corresponding secondary antibodies Alexa Fluor
488 anti-rabbit IgG (green) and Alexa Fluor 594 anti-mouse IgG (red)
(Invitrogen) in the blocking solution.
Intracellular Hydrolysis of Fluorescent Analogues of GM2
Ganglioside. Intracellular hydrolysis of GM2 ganglioside was
monitored using a fluorescent analogue of GM2 ganglioside (kindly
provided by W. Wakarchuk, U. Ryerson, Toronto, Canada) as
previously described.8 ATSD patient fibroblasts (three confluent 10
cm plates) were treated with either PYR or one of its derivatives for 10
days. Prior to evaluation of the GM2 hydrolysis of the cell’s
endogenous mutant lysosomal Hex A, the cells were treated with
conduritol β-epoxide (CBE, an irreversible glucocerebrosidase
inhibitor) to limit hydrolysis beyond GlcCer and loaded with a
fluorescent derivative of GM2 ganglioside for 8 h. Neutral glycolipids
and gangliosides were separated by Folch extraction30 of the treated
cells and divided into two pools consisting of gangliosides in the
aqueous layer and neutral glycolipids in the organic layer. Gangliosides
and glycolipids were resolved by high-performance thin-layer
chromatography and visualized/quantified by fluorescence imaging
(Storm Imager, Molecular Devices).
2-Amino-4-ethyl-5-(4-(trifluoromethyl)phenyl)pyrimidine (9). 1H
NMR (CDCl3): 8.20 (s, 1H, Pyr-H), 7.74 (d, 2H, J = 8.0, ArH), 7.40
(d, 2H, J = 8.0, ArH), 6.35 (bs, 2H, NH2) 2.72 (q, 2H, J = 7.5,
CH2Me), 1.27 (t, 3H, J = 7.6, CH2Me). 13C NMR (DMSO-d6): 170.0
(C), 159.0 (C), 156.7 (C), 138.2 (C), 130.6 (q, J = 32 Hz, C), 129.7
(CH), 126.0 (q, J = 3.7 Hz, CH), 123.8 (q, J = 272 Hz, CF3), 123.6
(C), 27.6 (CH2), 12.7 (CH3). IR: 3328, 3184, 1322. LRMS: 268.4 [M
+H+]. HRMS: calcd for C13H13F3N3 [M+H+] 268.1056; found
268.1062 [M+H+].
4-Amino-6-ethyl-5-(4-(trifluoromethyl)phenyl)pyrimidin-2(1H)-
one (10). 1H NMR (DMSO-d6): 7.77 (d, 2H, J = 8.0 Hz), 7.43 (d, 2H,
J = 8.0 Hz), 7.06 (bs, 1H), 5.85 (bs, 1H), 2.03 (q, 2H, J = 7.6 Hz),
0.94 (t, 3H, J = 7.4 Hz). 13C NMR (DMSO-d6): 165.2, 156.8, 156.2,
138.1, 132.3, 128.6 (q, JC−F = 32 Hz, C), 126.3 (q, JC−F = 3.8 Hz, CH),
124.8 (JC−F = 273.1 Hz, CF3), 103.2, 23.8 (CH2), 13.4 (CH3). IR:
3393, 1619, 1324. LRMS: 284.4 [M+H+]. HRMS: calcd for
C13H13F3N3O [M+H+] 284.1011; found 284.1013.
4-Ethyl-5-(4-(trifluoromethyl)phenyl)pyrimidine (11). 1H NMR
(DMSO-d6): 9.16 (1H, s, Pyr-H), 8.63 (1H, s, Pyr-H), 7.88 (2H, app
d, J = 8.0, ArH), 7.69 (2H, app d, J = 8.0, ArH), 2.71 (2H, q, J = 7.5,
CH2Me), 1.15 (3H, t, J = 7.5, CH2Me). 13C NMR (DMSO-d6): 167.9
(C), 157.7 (C), 156.3 (C), 139.8 (C), 132.8 (C), 130.2 (CH), 128.8
(q, JC−F = 32.0, C), 125.6 (q, JC−F = 3.7, CH), 124.3 (q, JC−F = 272.9,
CF3), 27.6 (CH2), 12.3 (CH3). IR: 3060, 2929, 2856, 1324 (v. strong).
ESI-MS: 253 [M+H]+ (100%). ESI-HRMS: calcd for C13H12N2F3 [M
+H]+ 253.0953; found 253.0954.
Monitoring Inhibitory Activity of PYR Derivatives against
Hex A and Other Enzymes. All test compounds were dissolved in
DMSO. For enzyme reactions, the compounds were diluted such that
the concentration of the diluent was 1% DMSO. Inhibitory activity of
the compounds was monitored using purified human Hex A isozyme
isolated from human placenta,27 together with the artificial substrate
pNPGlcNAc. Enzyme reactions were performed at 37 °C in McIlvaine
citrate phosphate buffer pH 4.5, containing 0.025% human serum
albumin, using 10 ng/mL of human Hex A, one of the diluted
compounds, and 1.6 mM colorimetric substrate pNPGlcNAc.
Reactions were stopped with 0.1 M 2-amino-2-methyl-1 propanol
(MAP) pH 10.5, and the absorbance was read in a Molecular Devices
M2 spectrofluorometer with absorbance set to 405 nm. IC50 values
were extracted from the dose−response curve following non-linear
curve fitting according to a built-in function within Prism Graphpad
5.0. Analysis of the inhibitory activity of compounds against mutant or
wild-type human Hex A or wild-type human Hex B was performed
using the Hex isozymes isolated and enriched from the lysates of either
unaffected or ATSD patient fibroblasts by DEAE ion exchange
chromatography. Alternatively, affinity-purified Hex B or Hex A from
human placenta was used for the assay.28 IC50 values comparing the
inhibitory activity of PYR versus its derivatives for the different mutant
or wild-type Hex isozymes were determined and analyzed as described
above, except that the fluorescent substrate 4-methylumbelliferyl-2-
acetamido-2-deoxy-β-D-glucopyranoside (MUG, 1.6 mM), which is
recognized by both the α- and β-active sites of the Hex isozymes, was
used. Evaluation of the inhibitory activity of the compounds at neutral
pH was performed as described above, except that pH 7 McIlvaine
buffer was utilized. Inhibition studies requiring human DHFR were
ASSOCIATED CONTENT
■
S
* Supporting Information
Figure S1, comparison of in cellulo enhancement of wild-type
Hex A enzyme activity in fibroblasts treated with either DMSO,
PYR, or 5h (n = 4); spectral data and NMR spectra for
representative compounds. The Supporting Information is
available free of charge on the ACS Publications website at
DOI: 10.1021/jm5017895.
4491
DOI: 10.1021/jm5017895
J. Med. Chem. 2015, 58, 4483−4493

Journal of Medicinal Chemistry
Article
with pyrimethamine, a potential pharmacological chaperone. J. Med.
Chem. 2011, 54, 1421−1429.
AUTHOR INFORMATION
■
Corresponding Author
*Phone: (416) 813-6161. E-mail: hex@sickkids.ca.
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
(10) Tarnchompoo, B.; Sirichaiwat, C.; Phupong, W.; Intaraudom,
C.; Sirawaraporn, W.; Kamchonwongpaisan, S.; Vanichtanankul, J.;
Thebtaranonth, Y.; Yuthavong, Y. Development of 2,4-diaminopyr-
imidines as antimalarials based on inhibition of the S108N and C59R
+S108N mutants of dihydrofolate reductase from pyrimethamine-
resistant Plasmodium falciparum. J. Med. Chem. 2002, 45, 1244−1252.
(11) Griffin, R. J.; Schwalbe, C. H.; Stevens, M. F. G.; Wong, K. P.
Structural studies on bio-active compounds.III. Re-examination of the
hydrolysis of the antimalarial drug pyrimethamine and related
derivatives and crystal-structure of a hydrolysis product. J. Chem.
Soc., Perkin Trans. 1 1985, 2267−2276.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
(12) Mark, B. L.; Mahuran, D. J.; Cherney, M. M.; Zhao, D.; Knapp,
S.; James, M. N. Crystal structure of human beta-hexosaminidase B:
understanding the molecular basis of Sandhoff and Tay-Sachs disease.
J. Mol. Biol. 2003, 327, 1093−1109.
(13) Macauley, M. S.; Whitworth, G. E.; Debowski, A. W.; Chin, D.;
Vocadlo, D. J. O-GlcNAcase uses substrate-assisted catalysis: kinetic
analysis and development of highly selective mechanism-inspired
inhibitors. J. Biol. Chem. 2005, 280, 25313−25322.
(14) Takakura, A.; Nelson, E. A.; Haque, N.; Humphreys, B. D.;
Zandi-Nejad, K.; Frank, D. A.; Zhou, J. Pyrimethamine inhibits adult
polycystic kidney disease by modulating STAT signaling pathways.
Hum. Mol. Genet. 2011, 20, 4143−4154.
(15) Reich, N. C.; Liu, L. Tracking STAT nuclear traffic. Nat. Rev.
Immunol. 2006, 6, 602−612.
(16) Horvath, C. M. STAT proteins and transcriptional responses to
extracellular signals. Trends Biochem. Sci. 2000, 25, 496−502.
(17) Hasilik, A.; Neufeld, E. F. Biosynthesis of lysosomal enzymes in
fibroblasts: synthesis as precursors of higher molecular weight. J. Biol.
Chem. 1980, 255, 4937−4945.
(18) Weiss, L. M.; Harris, C.; Berger, M.; Tanowitz, H. B.; Wittner,
M. Pyrimethamine concentrations in serum and cerebrospinal fluid
during treatment of acute Toxoplasma encephalitis in patients with
AIDS. J. Infect. Dis. 1988, 157, 580−583.
(19) Pajouhesh, H.; Lenz, G. R. Medicinal chemical properties of
successful central nervous system drugs. NeuroRx 2005, 2, 541−553.
(20) Kofron, W. G.; Baclawski, L. M. Convenient method for
estimation of alkyllithium concentrations. J. Org. Chem. 1976, 41,
1879−1880.
(21) Russell, P. B.; Hitchings, G. H. 2,4-diaminopyrimidines as
antimalarials.III. 5-aryl derivatives. J. Am. Chem. Soc. 1951, 73, 3763−
3770.
Funds from CIHR “Team” grant (CTP-82944; M.C., M.B.T
and D.M.) were used to support the research presented in the
manuscript along with a kind donation from the Uger estate
(D.M.). We are grateful for the fluorescent GM2 ganglioside
derivative kindly synthesized by Warren Wakarchuk (Ryerson
University, Ontario, CAN). We thank Kam S. Ho for helping
with the preparations of compounds 5h−5l. We also thank Dr.
Brian Mark (University of Manitoba, Manitoba, CAN) for his
helpful discussions and suggestions, particularly with regard to
Figure 1B.
ABBREVIATIONS USED
■
PC, pharmacological chaperone; MU, 4-methylumbelliferyl;
PYR, pyrimethamine; ATSD, adult-onset Tay-Sachs disease;
ASD, adult-onset Sandhoff disease
REFERENCES
■
(1) Kongsaeree, P.; Khongsuk, P.; Leartsakulpanich, U.; Chitnumsub,
P.; Tarnchompoo, B.; Walkinshaw, M. D.; Yuthavong, Y. Crystal
structure of dihydrofolate reductase from Plasmodium vivax: pyrimeth-
amine displacement linked with mutation-induced resistance. Proc.
Natl. Acad. Sci. U.S.A. 2005, 102, 13046−13051.
(2) Maegawa, G. H.; Tropak, M.; Buttner, J.; Stockley, T.; Kok, F.;
Clarke, J. T.; Mahuran, D. J. Pyrimethamine as a potential
pharmacological chaperone for late-onset forms of GM2 gangliosi-
dosis. J. Biol. Chem. 2007, 282, 9150−9161.
(3) Clarke, J. T.; Mahuran, D. J.; Sathe, S.; Kolodny, E. H.; Rigat, B.
A.; Raiman, J. A.; Tropak, M. B. An open-label Phase I/II clinical trial
of pyrimethamine for the treatment of patients affected with chronic
GM2 gangliosidosis (Tay-Sachs or Sandhoff variants). Mol. Genet.
Metab. 2011, 102, 6−12.
(22) Rogers, E. F.; Leanza, W. J.; Sarett, L. H. Antiparasitic agents
0.1. Some new 2,4-diaminopyrimidines. J. Org. Chem. 1957, 22, 1492−
1494.
(23) Liu, B.; Liu, M.; Xin, Z. L.; Zhao, H. Y.; Serby, M. D.; Kosogof,
C.; Nelson, L. T. J.; Szczepankiewicz, B. G.; Kaszubska, W.; Schaefer,
V. G.; Falls, H. D.; Lin, C. W.; Collins, C. A.; Sham, H. L.; Liu, G.
Optimization of 2,4-diaminopyrimidines as GHS-R antagonists: Side
chain exploration. Bioorg. Med. Chem. Lett. 2006, 16, 1864−1868.
(24) Barlin, G. B.; Kotecka, B.; Rieckmann, K. H. Potential
antimalarials.XXII. Some 2,4-diamino-5-(3- and 4-trifluoromethyl-
phenyl and 3,4-methylenedioxyphenyl)pyrimidines. Aus. J. Chem.
1996, 49, 647−650.
(4) Knight, E. M.; Williams, H. N.; Stevens, A. C.; Kim, S. H.;
Kottwitz, J. C.; Morant, A. D.; Steele, J. W.; Klein, W. L.; Yanagisawa,
K.; Boyd, R. E.; Lockhart, D. J.; Sjoberg, E. R.; Ehrlich, M. E.;
Wustman, B. A.; Gandy, S. Evidence that small molecule enhancement
of beta-hexosaminidase activity corrects the behavioral phenotype in
Dutch APP mice through reduction of ganglioside-bound Abeta. Mol.
Psych. 2014, 20, 109−117.
(5) Sandhoff, K.; Harzer, K. Gangliosides and gangliosidoses:
principles of molecular and metabolic pathogenesis. J. Neurosci.
2013, 33, 10195−10208.
(25) Trattner, R. B.; Hitchings, G. H.; Sharefkin, D. M.; Elion, G. B.
Deamination studies on pyrimidine and condensed pyrimidine
systems. J. Org. Chem. 1964, 29, 2674−2677.
(6) Mahuran, D. J. Biochemical consequences of mutations causing
the GM2 gangliosidoses. Biochim. Biophys. Acta 1999, 1455, 105−138.
(7) Leinekugel, P.; Michel, S.; Conzelmann, E.; Sandhoff, K.
Quantitative correlation between the residual activity of beta-
hexosaminidase A and arylsulfatase A and the severity of the resulting
lysosomal storage disease. Hum. Genet. 1992, 88, 513−23.
(8) Tropak, M. B.; Bukovac, S. W.; Rigat, B. A.; Yonekawa, S.;
Wakarchuk, W.; Mahuran, D. J. A sensitive fluorescence-based assay
for monitoring GM2 ganglioside hydrolysis in live patient cells and
their lysates. Glycobiology 2010, 20, 356−65.
(26) Russell, P. B. The effects of substitution on the ultra-violet
absorption spectra of phenylpyrimidines. J. Chem. Soc. 1954, 2951−
2955.
(27) Mahuran, D.; Lowden, J. A. The subunit and polypeptide
structure of hexosaminidases from human placenta. Can. J. Biochem.
1980, 58, 287−294.
(28) Tropak, M. B.; Reid, S. P.; Guiral, M.; Withers, S. G.; Mahuran,
D. Pharmacological enhancement of beta-hexosaminidase activity in
fibroblasts from adult Tay-Sachs and Sandhoff Patients. J. Biol. Chem.
2004, 279, 13478−13487.
(9) Bateman, K. S.; Cherney, M. M.; Mahuran, D. J.; Tropak, M.;
James, M. N. Crystal structure of beta-hexosaminidase B in complex
4492
DOI: 10.1021/jm5017895
J. Med. Chem. 2015, 58, 4483−4493

Journal of Medicinal Chemistry
Article
(29) Tropak, M. B.; Kornhaber, G. J.; Rigat, B. A.; Maegawa, G. H.;
Buttner, J. D.; Blanchard, J. E.; Murphy, C.; Tuske, S. J.; Coales, S. J.;
Hamuro, Y.; Brown, E. D.; Mahuran, D. J. Identification of
pharmacological chaperones for Gaucher disease and characterization
of their effects on beta-glucocerebrosidase by hydrogen/deuterium
exchange mass spectrometry. ChemBioChem 2008, 9, 2650−2662.
(30) Lees, M.; Folch, J.; Stanley, G. H.; Carr, S. A simple procedure
for the preparation of brain sulphatides. J. Neurochem. 1959, 4, 9−18.
4493
DOI: 10.1021/jm5017895
J. Med. Chem. 2015, 58, 4483−4493