Vivek Kumar1, Nageswari Yarravarapu2, David J Lapinsky2, Danielle Perley3, Bruce Felts3, Michael J Tomlinson3, Roxanne A Vaughan3, L Keith Henry3, John R Lever4,5, Amy Hauck Newman1. 1. †Medicinal Chemistry Section, National Institute on Drug Abuse, Intramural Research Program, National Institutes of Health, 333 Cassell Drive, Baltimore, Maryland 21224, United States. 2. ‡Division of Pharmaceutical Sciences, Duquesne University, 600 Forbes Avenue, Pittsburgh, Pennsylvania 15282, United States. 3. §Department of Basic Sciences, University of North Dakota School of Medicine and Health Sciences, Grand Forks, North Dakota 58202, United States. 4. ∥Department of Radiology, University of Missouri, One Hospital Drive, Columbia, Missouri 65212, United States. 5. ⊥Harry S. Truman Memorial Veterans' Hospital, 800 Hospital Drive, Columbia, Missouri 65201, United States.
Abstract
Three photoaffinity ligands (PALs) for the human serotonin transporter (hSERT) were synthesized based on the selective serotonin reuptake inhibitor (SSRI), (S)-citalopram (1). The classic 4-azido-3-iodo-phenyl group was appended to either the C-1 or C-5 position of the parent molecule, with variable-length linkers, to generate ligands 15, 22, and 26. These ligands retained high to moderate affinity binding (K(i) = 24-227 nM) for hSERT, as assessed by [(3)H]5-HT transport inhibition. When tested against Ser438Thr hSERT, all three PALs showed dramatic rightward shifts in inhibitory potency, with Ki values ranging from 3.8 to 9.9 μM, consistent with the role of Ser438 as a key residue for high-affinity binding of many SSRIs, including (S)-citalopram. Photoactivation studies demonstrated irreversible adduction to hSERT by all ligands, but the reduced (S)-citalopram inhibition of labeling by [(125)I]15 compared to that by [(125)I]22 and [(125)I]26 suggests differences in binding mode(s). These radioligands will be useful for characterizing the drug-protein binding interactions for (S)-citalopram at hSERT.
Three photoaffinity ligands (PALs) for the humanserotonin transporter (hSERT) were synthesized based on the selective serotonin reuptake inhibitor (SSRI), (S)-citalopram (1). The classic 4-azido-3-iodo-phenyl group was appended to either the C-1 or C-5 position of the parent molecule, with variable-length linkers, to generate ligands 15, 22, and 26. These ligands retained high to moderate affinity binding (K(i) = 24-227 nM) for hSERT, as assessed by [(3)H]5-HT transport inhibition. When tested against Ser438ThrhSERT, all three PALs showed dramatic rightward shifts in inhibitory potency, with Ki values ranging from 3.8 to 9.9 μM, consistent with the role of Ser438 as a key residue for high-affinity binding of many SSRIs, including (S)-citalopram. Photoactivation studies demonstrated irreversible adduction to hSERT by all ligands, but the reduced (S)-citalopram inhibition of labeling by [(125)I]15 compared to that by [(125)I]22 and [(125)I]26 suggests differences in binding mode(s). These radioligands will be useful for characterizing the drug-protein binding interactions for (S)-citalopram at hSERT.
The serotonin transporter
(SERT) is a member of the solute carrier
6 (SLC6) family of transporters that functions to regulate serotonin
neurotransmission and homeostasis.[1−3] In particular, selective
serotonin reuptake inhibitors (SSRIs) and tricyclic antidepressants
(TCAs), widely prescribed medications for treatment of anxiety and
major depressive disorders, principally work by binding to SERT and
inhibiting serotonin reuptake into presynaptic neurons.[4,5] Notably, (S)-citalopram (1, Escitalopram, Figure ) binds with high
affinity and selectivity to hSERT. However, despite its well-documented
clinical success, the molecular interactions between (S)-citalopram and hSERT that determine its reuptake inhibition potency
and selectivity over the norepinephrine transporter (NET) and dopamine
transporter (DAT) remain poorly understood.
Figure 1
Chemical structures of
(S)-citalopram (1) and known DAT and/or
SERT inhibitor PALs (2–7).
Chemical structures of
(S)-citalopram (1) and known DAT and/or
SERT inhibitor PALs (2–7).The 3D structure of SERT as well
as its structure, function, and
regulation, including oligomerization and distribution in cells or
brain tissue, has been studied with small molecular probes containing
electrophiles (e.g., N3, NCS) and radiolabels (e.g., 3H, 125I). In particular, the well-known technique
of photoaffinity labeling utilizes molecular probes that contain a
functional group capable of forming a covalent bond to a biological
target upon photoactivation.[6,7] Of the numerous photoreactive
functional groups that can be employed in the design of photoaffinity
ligands (PALs) (e.g., benzophenones, aliphatic and aromaticdiazirines,
etc.), aryl azides are frequently used given their relatively small
size, ease of synthetic incorporation, and chemical stability.[8]Numerous [125I]-radiolabeled-arylazido
analogues of
dopamine transporter (DAT) inhibitors including cocaine,[9−12] GBR12909 (1-[2-[bis(4-fluorophenyl)methoxy]ethyl]-4-(3-phenylpropyl)piperazine),[13−16] benztropine,[10,17,18] pyrovalerone,[19] bupropion,[20] and methylphenidate[21,22] have been developed to elucidate molecular components of the ligand-binding
site and mechanisms underlying uptake inhibition. The 4′-azido,
3′-iodo-substituted phenyl ring is a common structural motif
found in many PALs that bind DAT irreversibly upon photoactivation
(e.g., RTI 82 (2), MFZ 2-24 (3), and JHC
2-48 (4); Figure ).[9−12] Specifically, tropane-based PALs 2 and 3 have been instrumental in defining drug–protein interactions
at the molecular level and thus further defining the structural components
of DAT that are critical for transport inhibition by cocaine.[23−25]In contrast to DATPALs, the development of SERT-selective
PALs
has been more limited. To the best of our knowledge, only a few PALs
have been synthesized and experimentally validated to label SERT (Figure ). Although tropane-based
PALs 2–4 bind to SERT as well as
DAT, to date, only probes 3 and 4 have been
used to photolabel hSERT.[12,26] Other inhibitor-based
ligands that photolabel SERT include an aryl-azido derivative of the
SSRI paroxetine (5)[27] and
a tricyclic antidepressant-based PAL, [3H]-2-nitroimipramine
(6), that showed covalent incorporation into SERT present
in membrane homogenates of rat brain and liver and human platelets.[28,29] Finally, photoinactivation of serotonin uptake has been demonstrated
by an azidobenzamidine derivative of serotonin (7).[30] However, to date, none of these SERTPALs have
been used to define the structural basis of their binding interactions
at hSERT.On the basis of successful application of PALs 2–4 in elucidating drug–binding
site interactions at
DAT,[23−25] our concept for the present study was that appending
the 4′-azido, 3′-iodo aryl moiety onto different positions
of the (S)-citalopram base might reveal different
sites of adduction on hSERT that would enable a clearer understanding
of the (S)-citalopram binding mechanism. Extensive
structure–activity relationships have been described with analogues
of citalopram that suggest positions C-1 and C-5can accommodate significant
steric bulk without appreciable loss in hSERT binding affinity.[31−34] Herein, we describe the design, synthesis, pharmacological evaluation,
and radiosynthesis of three novel C-1- or C-5-substituted 4′-azido,
3′-iodo analogues of (S)-citalopram and demonstrate
their ability to covalently label hSERT upon photoactivation.
Chemistry
The synthesis of C-1-substituted (S)-citalopramPAL 15 is shown in Scheme . Phthalimide derivative 9(35) was synthesized by stirring 2-(4-aminophenyl)ethan-1-ol
(8) with phthalic anhydride in pyridine, at reflux, and
oxidizing to aldehyde 10 using a Dess–Martin periodinane
reagent.[36,37] (S)-N-Desmethylcitalopram
(11)[31,32] was then coupled to aldehyde 10 using standard reductive amination conditions to provide
tertiary amine 12 in 80% yield, followed by phthalimide
deprotection to give aniline 13. Iodination with iodine
monochloride (ICl) gave the intermediate 14, which was
then treated with NaNO2 followed by NaN3 to
give 4′-azido, 3′-iodo product 15.
Scheme 1
Synthesis of PAL 15
Reagents
and conditions: (a)
phthalic anhydride, pyridine, reflux, 16 h; (b) Dess–Martin
periodinane, CH2Cl2, −78 °C to RT,
3 h; (c) NaBH(OAc)3, HOAc, DCE, RT 18 h; (d) hydrazine,
EtOH, reflux, 3 h; (e) ICl (1 M in CH2Cl2),
0–5 °C to RT; (f) NaNO2, NaN3, TFA,
0–5 °C to RT.
Synthesis of PAL 15
Reagents
and conditions: (a)
phthalic anhydride, pyridine, reflux, 16 h; (b) Dess–Martin
periodinane, CH2Cl2, −78 °C to RT,
3 h; (c) NaBH(OAc)3, HOAc, DCE, RT 18 h; (d) hydrazine,
EtOH, reflux, 3 h; (e) ICl (1 M in CH2Cl2),
0–5 °C to RT; (f) NaNO2, NaN3, TFA,
0–5 °C to RT.The synthesis of
C-5-substituted (S)-citalopramPALs 22 and 26 is shown in Scheme . 3-(4-Aminophenyl)propanoic
acid (16) was initially stirred at reflux with phthalic
anhydride to afford the corresponding phthalimide carboxylic acid 17.[38] Subsequent coupling of 17 to primary amine 18(31,32) via the acid chloride provided amide 19 in 97% yield.
Deprotection of 19 using hydrazine provided aniline intermediate 20, which was reacted with ICl to give 3′-iodo-substituted
product 21. Conversion to azidoPAL 22 was
then performed as previously described for 15 via diazotization
and azide displacement of aniline 21. This final step
went in poor yield (18%); thus, a second strategy was also developed
that featured a decrease in the number of synthetic steps and an increase
in the overall yield. This alternative strategy began with electrophilic
iodination of aniline 16ortho to the
amine to provide the corresponding iodo-substituted aniline 23 in 67% yield, which was subsequently treated with NaNO2 followed by NaN3 in TFA to give iodo-azide 24.[39] Primary amine 18 was then coupled to azido-iodo carboxylic acid 24 using
1,1′-carbonyldiimidazole (CDI) in THF to generate amide 22. Although this route using 24 is preferred
for generating nonradioactive 22 in high yield, the initial
route that provided 20 was utilized in the synthesis
of [125I]-labeled 22. Similarly, 18 was coupled to 4-azido-3-iodobenzoic acid[40] (25) using EDC and HOBT to provide PAL 26 in 67% yield.
Scheme 2
Synthesis of PALs 22 and 26
Reagents and conditions: (a)
phthalic anhydride, pyridine, reflux, 16 h; (b) SOCl2,
3 h; (c) hydrazine, EtOH, reflux, 6 h; (d) ICl, CH2Cl2, 0–5 °C to RT; (e) NaNO2, NaN3, HOAc, H2O; (f) ICl, HOAc, RT; (g) NaNO2, NaN3, TFA; (h) 24, CDI, THF, 0 °C
to RT; (i) EDC, HOBT, Et3N, DMF, 0 °C to RT.
Synthesis of PALs 22 and 26
Reagents and conditions: (a)
phthalic anhydride, pyridine, reflux, 16 h; (b) SOCl2,
3 h; (c) hydrazine, EtOH, reflux, 6 h; (d) ICl, CH2Cl2, 0–5 °C to RT; (e) NaNO2, NaN3, HOAc, H2O; (f) ICl, HOAc, RT; (g) NaNO2, NaN3, TFA; (h) 24, CDI, THF, 0 °C
to RT; (i) EDC, HOBT, Et3N, DMF, 0 °C to RT.Radioiodinated PALs [125I]15, [125I]22, and [125I]26 were prepared
under no-carrier-added conditions as shown in Scheme . Electrophilic radioiodination of anilines 13 and 20 was accomplished by treatment with
[125I]NaI and chloramine-T at ambient temperature for 30
min. Acidification with HOAc followed by treatment with NaNO2 at −5 °C generated the diazonium salts in situ. Decomposition upon warming, in the presence of NaN3,
gave [125I]-labeled aryl azides [125I]15 and [125I]22. This sequence was
accomplished in one flask, as previously described for other radioiodinated
PALs.[11,41,42] Electrophilic
radioiododestannylation of tri-n-butylstannyl azide 27 using chloramine-T as the oxidant in methanolic acetic
acid for 3 min was employed to prepare [125I]26. The organostannane precursor was obtained in 52% yield by treating
nonradioactive iodide 26 with bis(tri-n-butyltin) and bis(triphenylphosphine)palladium(II) dibromide. Since
carryover of trace amounts of 26 would dramatically lower
the specific radioactivity of [125I]26 prepared
from 27,[43] this material was
purified by reversed-phase HPLC prior to radiolabeling.
Scheme 3
Radiosynthesis
of [125I]15, [125I]22, and [125I]26
Reagents
and conditions: (a)
[125I]NaI, chloramine-T, NaOAc (0.2 M, pH 4.0), RT, 30
min; (b) HOAc (3.0 M), NaNO2 (0.5 M), −5 °C,
20 min; (c) NaN3 (0.5 M), RT, 30 min; (d) Na2S2O5 (50 mM), RT; (e) Pd(PPh3)2Br2, ((n-Bu)3Sn)2, toluene, 105 °C, 4 h; (f) [125I]NaI, chloramine-T,
MeOH, 3% HOAc, RT, 3 min.
Radiosynthesis
of [125I]15, [125I]22, and [125I]26
Reagents
and conditions: (a)
[125I]NaI, chloramine-T, NaOAc (0.2 M, pH 4.0), RT, 30
min; (b) HOAc (3.0 M), NaNO2 (0.5 M), −5 °C,
20 min; (c) NaN3 (0.5 M), RT, 30 min; (d) Na2S2O5 (50 mM), RT; (e) Pd(PPh3)2Br2, ((n-Bu)3Sn)2, toluene, 105 °C, 4 h; (f) [125I]NaI, chloramine-T,
MeOH, 3% HOAc, RT, 3 min.All three radioligands
were purified by reversed-phase HPLC and
then formulated as concentrated solutions in MeOH using solid-phase
extraction. Isolated radiochemical yields were 46–55%. Specific
radioactivities were determined as 2045 to 2098 mCi/μmol, near
the theoretical value of 2175 mCi/μmol, using HPLC to determine
the mass associated with the UV absorbance peak in samples of known
radioactivity. Radiochemical purities were ≥98% by HPLC (Figure ), and the materials
were stable for at least 5 weeks when stored at −20 °C
in the dark. The three radioligands proved to be lipophilic, with
[125I]22 (tR =
13.2 min, k′ = 10.4) and [125I]26 (tR 13.4 min, k′ = 10.6) being substantially less hydrophobic than [125I]15 (tR = 22.4
min, k′ = 18.3). These findings are in accord
with their relative cLogP values (15, 7.02; 22, 6.39; 26, 6.35; generated in ChemDraw).
Figure 2
Reversed-phase
HPLC chromatograms of isolated radioiodinated ligands
illustrating purity and relative lipophilicity. Chromatography performed
using 45% MeOH/CH3CN (1:1, v/v) and 55% water containing
Et3N (2.1% v/v) and HOAc (2.8% v/v) at a flow rate of 3.0
mL/min on a C-18 column.
Reversed-phase
HPLCchromatograms of isolated radioiodinated ligands
illustrating purity and relative lipophilicity. Chromatography performed
using 45% MeOH/CH3CN (1:1, v/v) and 55% watercontaining
Et3N (2.1% v/v) and HOAc (2.8% v/v) at a flow rate of 3.0
mL/min on a C-18 column.
Results and Discussion
Pharmacological Properties of PALs 15, 22, and 26
Inhibition
of [3H]5-HT
uptake by WT hSERT and the hSERTSer438Thr mutant was evaluated to
determine inhibitory constants for PALs 15, 22, 26 and parent compound (S)-citalopram
(1). When tested against WT hSERT, C-5-substituted PALs 26 and 22 showed affinities of 24 and 38 nM,
respectively, values that were reduced by 11- and 17-fold compared
to that of (S)-citalopram (Ki = 2.2 nM) (Figure and Table ). In contrast, C-1 substituted PAL 15 showed a greater
reduction in affinity (Ki = 227 nM, 100-fold)
compared to that of (S)-citalopram, which is in accord
with structure–activity relationships previously described.[32] When tested against the hSERTSer438Thr mutant,
all four compounds showed dramatic rightward shifts in inhibitory
potency (Figure ,
dashed lines) with Ki values ranging from
3.8 to 9.9 μM (Table ). These findings are consistent with previous reports that
Ser438 is a key residue for high-affinity binding of many hSERT antagonists
and that its mutation to Thr results in a significant loss in potency
for (S)-citalopram and many other hSERT inhibitors.[44] Interestingly, C-1-substituted PAL 15 demonstrated lower affinity than (S)-citalopram
or the other PALs for WT hSERT but higher affinity for the S438T mutant.
Moreover, the difference in binding affinities between the mutant
and WT hSERT for 15 was only 17-fold, as compared to
a more dramatic shift for parent ligand 1 (∼3600-fold).
Figure 3
[3H]5-HT uptake inhibition analysis for (S)-citalopram, 15, 22, and 26. HEK-293 Griptite
cells expressing hSERT (solid lines) or hSERT
S438T (dashed lines) were assayed for [3H]5-HT uptake in
the presence of the indicated concentrations of (S)-citalopram (1) (●), 15 (□), 22 (○), or 26 (■). Data shown are
the mean ± SEM of transport activity for each form in the absence
of competitor, set to 100%. Assays were conducted in triplicate and
were repeated at least four times.
Table 1
Inhibitory Constants of (S)-Citalopram
(1), 15, 22,
and 26 for [3H]5-HT Transport by hSERT and
hSERT Ser438Thra
hSERT (Ki, nM)
hSERT S438T (Ki, nM)
(S)-citalopram
2.2 ± 0.17
7879 ± 1300*
15
227 ± 23#
3811 ± 590*
22
38 ± 2.9#
9879 ± 1200*
26
24 ± 2.2#
5697 ± 745*
Data are K values (nM) (mean ±
SEM) for the ability
of the (S)-citalopram-based PALs to inhibit uptake
of [3H]5-HT, as shown in Figure . K values were calculated using the Cheng–Prusoff equation
in GraphPad Prism 5. Data were analyzed by paired t-tests for statistical significance, where * indicates that the Ki value obtained with the hSERT S438T mutant
is significantly different than the Ki of that compound for WT hSERT (p < 0.001) and
# indicates that the Ki value for the
analogue is significantly different than the Ki for (S)-citalopram obtained with WT hSERT
(p < 0.001). The Ki values obtained with hSERT S438T for the analogues and (S)-citalopram were not statistically different from one
another.
[3H]5-HT uptake inhibition analysis for (S)-citalopram, 15, 22, and 26. HEK-293 Griptite
cells expressing hSERT (solid lines) or hSERTS438T (dashed lines) were assayed for [3H]5-HT uptake in
the presence of the indicated concentrations of (S)-citalopram (1) (●), 15 (□), 22 (○), or 26 (■). Data shown are
the mean ± SEM of transport activity for each form in the absence
of competitor, set to 100%. Assays were conducted in triplicate and
were repeated at least four times.Data are K values (nM) (mean ±
SEM) for the ability
of the (S)-citalopram-based PALs to inhibit uptake
of [3H]5-HT, as shown in Figure . K values were calculated using the Cheng–Prusoff equation
in GraphPad Prism 5. Data were analyzed by paired t-tests for statistical significance, where * indicates that the Ki value obtained with the hSERTS438T mutant
is significantly different than the Ki of that compound for WT hSERT (p < 0.001) and
# indicates that the Ki value for the
analogue is significantly different than the Ki for (S)-citalopram obtained with WT hSERT
(p < 0.001). The Ki values obtained with hSERTS438T for the analogues and (S)-citalopram were not statistically different from one
another.
Photoaffinity Labeling
Experiments with hSERT
Next,
we tested [125I]15, [125I]22, and [125I]26 for irreversible
labeling of HA-hSERT (Figure ). For these studies, HA-hSERT LLCPK1 cells were
incubated with the PALs in the presence or absence of (S)-citalopram and irradiated with UV light to cross-link the ligand
to the protein. Lysates were immunoprecipitated with anti-HA or anti-hSERT
antibodies, and samples were analyzed by SDS-PAGE and autoradiography.
These methods (described in detail in the Experimental
Methods) have been used by our laboratories to verify photolabeling
of DAT and SERT by numerous structurally diverse tropane- and non-tropane-based
ligands.[11,12,17,19−26] The results (Figure ) show that when using HA-hSERT expressing cells, [125I]15, [125I]22, and [125I]26 all covalently label a protein of the expected
hSERT molecular mass of ∼100 kDa, whereas photolabeled bands
were not obtained from nontransfected parent cells (upper panels),
strongly supporting the identity of the photolabeled bands as hSERT.
Immunoblotting (lower panels) verified the presence and absence of
hSERT protein in transfected and untransfected cells, respectively.
The photoaffinity labeled proteins were immunoprecipitated using anti-HA
antibody (Figure )
as well as with the anti-hSERT antibody ST51 (not shown), but they
were not precipitated with nonimmune IgG (not shown), also supporting
the identity of the bands as hSERT. Similar irreversible labeling
and immunoprecipitation results were obtained using tropanePAL 3, which was previously demonstrated to photolabel hSERT.[26] Together, these results demonstrate that hSERT
protein is irreversibly labeled with [125I]15, [125I]22, and [125I]26 upon photoactivation, providing the first demonstration of photoaffinity
labeling of hSERT by analogues of (S)-citalopram.
Figure 4
Photoaffinity
labeling of hSERT. HA-hSERT LLCPK1 cells
or nontransfected (parent) LLCPK1 cells were incubated
with the indicated 125I-labeled ligands (26, 22, 15, 3) in the absence
(−) or presence (+) of 10 μM (S)-citalopram
(S-Cit) followed by cross-linking to SERT with UV light. Cell lysates
were immunoprecipitated with anti-HA antibody followed by SDS-PAGE
and autoradiography to detect [125I] radiolabeled proteins
(upper panels) or were analyzed by immunblotting (IB) with anti-HA
antibody to detect total hSERT (lower panels). Results are representative
of three independent experiments.
Photoaffinity
labeling of hSERT. HA-hSERT LLCPK1 cells
or nontransfected (parent) LLCPK1 cells were incubated
with the indicated 125I-labeled ligands (26, 22, 15, 3) in the absence
(−) or presence (+) of 10 μM (S)-citalopram
(S-Cit) followed by cross-linking to SERT with UV light. Cell lysates
were immunoprecipitated with anti-HA antibody followed by SDS-PAGE
and autoradiography to detect [125I] radiolabeled proteins
(upper panels) or were analyzed by immunblotting (IB) with anti-HA
antibody to detect total hSERT (lower panels). Results are representative
of three independent experiments.For pharmacological characterization of irreversible labeling,
HA-hSERT-LLCPK1 cells were incubated with vehicle or 10
μM (S)-citalopram prior to addition of the
radioiodinated PALs. Incorporation of [125I]22 and [125I]26 was blocked by >99%, demonstrating
that (S)-citalopram fully displaces these C-5 substituted
radioligands, similar to displacement of hSERT labeling by [125I]3.[26] In contrast, 10 μM
(S)-citalopram inhibited [125I]15 labeling of hSERT by only ∼30%, suggesting that this C-1-substituted
PAL exhibits a different binding mode compared to that of the two
C-5-substituted ligands. In a previous SAR study, wherein (S)-citalopram was modified by N-substitution at the C-5
position with a second molecule of (S)-citalopram
to create a homobifunctional ligand, binding data suggested simultaneous
interaction of the ligand with the S1 and S2 binding sites.[32] Moreover, the affinity of (S)-citalopram for S2 is much lower than that for S1.[45] Thus, it is conceivable that [125I]15 may be binding to both the S1 and S2 sites and that the latter might
not be fully inhibited by 10 μM (S)-citalopram.[32,45]In summary, three novel PALs, [125I]15,
[125I]22, and [125I]26, were designed by substituting the parent SSRI, (S)-citalopram, with a 4′-azido, 3′-iodophenyl moiety
in either the C-1 or C-5 position, with variable-length linker chains.
Our results showed that the nonradioiodinated compounds were able
to inhibit [3H]5-HT uptake with high to moderate potency
and that the radioiodinated PALs displayed requisite characteristics
as irreversible probes for molecular characterization of the hSERT(S)-citalopram binding site(s). Preincubation of
hSERT with (S)-citalopramcompletely eliminated irreversible
labeling by C-5-substituted [125I]22 and [125I]26, but it reduced irreversible labeling
by C-1-substituted [125I]15 by only ∼30%.
The inability of (S)-citalopram to completely block
[125I]15 hSERT photolabeling indicates that
this PAL has complex hSERT binding properties compared to those of
C-5-substituted PALs [125I]22 and [125I]26 that will require further investigation.
Experimental Methods
General
Reaction
conditions and yields were not optimized.
Anhydrous solvents were purchased from Aldrich and were used without
further purification, except for tetrahydrofuran, which was freshly
distilled from sodium benzophenone ketyl. All other chemicals and
reagents were purchased from Sigma-Aldrich Co. LLC, Combi-Blocks,
TCI America, Acros Organics, and Alfa Aesar. Unless otherwise stated,
final amine products were converted into oxalate salts, typically
by treating the free base in CHCl3 with a 1:1 molar ratio
of oxalic acid in acetone. As described, some of the oxalate salts
were recrystallized from hot methanol or a methanol–acetone
solvent mixture. Flash chromatography was performed using silica gel
(EMD Chemicals, Inc., 230–400 mesh, 60 Å). 1H and 13CNMR spectra were acquired using a Varian Mercury
Plus 400 spectrometer at 400 and 100 MHz, respectively. Chemical shifts
are reported in parts per million (ppm) and referenced according to
deuterated solvent for 1H spectra (CDCl3, 7.26
ppm, or CD3OD-d4, 3.31 ppm)
and 13C spectra (CDCl3, 77.2 ppm, or CD3OD-d4, 49.00 ppm). Infrared (IR)
spectra were obtained (neat) on a PerkinElmer Spectrum Two FTIR spectrometer.
Gas chromatography/mass spectrometry (GC/MS) data were acquired (where
obtainable) using an Agilent Technologies (Santa Clara, CA) 6890N
gas chromatograph equipped with an HP-5MScolumn (cross-linked 5%
PH ME siloxane, 30 m × 0.25 mm i.d. × 0.25 μm film
thickness) and a 5973 mass-selective ion detector in electron-impact
mode. Ultrapure grade helium was used as the carrier gas at a flow
rate of 1.2 mL/min. The injection port and transfer line temperatures
were 250 and 280 °C, respectively, and the oven temperature gradient
used was as follows: the initial temperature (100 °C) was held
for 3 min, increased to 295 °C at 15 °C/min over 13 min,
and finally maintained at 295 °C for 10 min. To acquire high-resolution
mass spectrometry (HRMS) data, 1 μL of sample was mixed with
1 μL of matrix (saturated solution of 2,4,6-trihydroxyacetophenone/2,5-dihydroxybenzoic
acid in 50:50 ethanol/water); then, 1 μL was deposited on a
stainless steel plate and analyzed in positive ion mode in a MALDI
LTQ-XL Orbitrap (Thermo-Scientific, San Jose, CA) using a laser energy
fixed at 6 μJ and a mass resolution of 60 000 at an m/z of 400 Th. Combustion analysis was
performed by Atlantic Microlab, Inc. (Norcross, GA), and the results
agree within ±0.4% of the calculated values, unless indicated
otherwise (S.I.). Melting point determination was conducted using
a Thomas-Hoover melting point apparatus; the melting points are uncorrected.
On the basis of NMR, HRMS, HPLC, and combustion data, all final compounds
are >95% pure.
To a solution of aniline 14 (0.095 g, 0.17 mmol) in TFA (3 mL) was added NaNO2 (0.023
g, 0.34 mmol) at 0 °C, in the dark. The reaction mixture was
then stirred for 45 min at 0–5 °C. NaN3 (0.111
g, 1.71 mmol) and anhydrous ether (3 mL) was then added to the reaction
mixture, followed by stirring for 1 h at 0–5 °C, and the
mixture was then diluted with H2O and ether (10 mL each).
The organic layer was separated, dried, concentrated, and purified
by flash chromatography using 3:7 acetone/CHCl3 to provide
0.086 g of azide 15 as an oil (87%). 1HNMR
(400 MHz, CDCl3) δ 7.59 (d, J =
7.2 Hz, 2H), 7.49 (s, 1H), 7.37 (dd, J = 8.4, 5.6
Hz, 2H), 7.33 (d, J = 8.4 Hz, 1H), 7.16 (d, J = 8.4 Hz, 1H), 7.02–6.98 (m, 3H), 5.18 (d, J = 12.8 Hz, 1H), 5.13 (d, J = 13.2 Hz,
1H), 2.67–2.63 (m, 2H), 2.53–2.49 (m, 2H), 2.39–2.35
(m, 2H), 2.20–2.02 (m, 5H), 1.51–1.26 (m, 2H). 13CNMR (100 MHz, CDCl3) δ 162.3 (d, J = 245.7 Hz), 148.7, 141.2, 140.0, 138.6, 132.3, 130.2,
126.7 (d, J = 8.3 Hz), 125.5, 122.8, 118.9, 118.6,
115.8 (d, J = 21.2 Hz), 112.3, 90.7, 88.2, 71.4,
57.0, 56.0, 53.9, 39.8, 37.9, 29.4, 19.4. IR: azide, 2112 cm–1; HRMScalcd for C27H25FIN5O [M
+ H+], 582.1160; found, 582.1159. Anal. (C27H25FIN5O·2.5H2O): C, H, N.
NaNO2 (9.6 mg,
0.14 mmol) was added
to a suspension of aniline 21 (60 mg, 0.10 mmol) in a
mixture of glacial HOAc (3 mL) and H2O (1 mL) at 0 °C,
in the dark. The mixture was stirred for 30 min; then, NaN3 (9.7 mg, 0.15 mmol) was added to the reaction mixture. After stirring
at 0–5 °C for 30 min, the reaction was diluted with water
(10 mL). The mixture was then carefully quenched with saturated aqNaHCO3 to a pH of 9–10. The crude product was extracted
with CHCl3, dried over Na2SO4, concentrated,
and purified by preparative TLC using 15:85 MeOH/CHCl3 to
provide 20 mg of azide 22 as a yellow oil (18%). 1HNMR (400 MHz, CDCl3) δ 7.63 (d, J = 2.0 Hz, 1H), 7.46–7.41 (m, 2H), 7.23 (d, J = 6.8 Hz, 2H), 7.09 (d, J = 8.0 Hz, 1H),
7.04–6.95 (m, 4H), 6.02 (bt, J = 5.0 Hz, 1H),
5.13 (d, J = 12.4 Hz, 1H), 5.07 (d, J = 12.4 Hz, 1H), 4.37 (d, J = 5.6 Hz, 2H), 2.91
(t, J = 7.2 Hz, 2H), 2.64–2.57 (m, 2H), 2.48
(t, J = 7.2 Hz, 2H), 2.39 (s, 6H), 2.19 (m, 2H),1.66–1.47
(m, 2H). 13CNMR (100 MHz, CDCl3) δ 171.6,
161.9 (d, J = 243.4 Hz), 143.4, 141.1, 139.9, 139.8,
139.6, 138.2, 130.0, 127.6, 126.8 (d, J = 10.6 Hz),
122.1, 120.9, 118.4, 115.1 (d, J = 21.2 Hz), 115.0,
90.9, 87.9, 71.9, 59.5, 45.1, 43.5, 39.4, 37.9, 30.5, 22.3. IR: azide,
2114 cm–1; HRMScalcd for C29H31FIN5O2 [M + H+], 628.1579; found,
628.1569. Anal. (C29H31FIN5O2·2H2O·0.5C3H6O):
C, H, N.
Method B
CDI (0.064 g, 0.39 mmol)
was added to a solution
of carboxylic acid 24(39) (0.125
g, 0.39 mmol) in THF (15 mL) under an argon atmosphere. The reaction
mixture was stirred for 6 h and then cooled to 0–5 °C
using an ice bath. Primary amine 18(31,32) (0.129 g, 0.39 mmol) was then added dropwise after dilution with
THF (10 mL). The reaction mixture was allowed to warm to RT and stirred
overnight, followed by subsequent concentration, dilution with water
(20 mL), and extraction with CHCl3 (2 × 20 mL). The
organic layer was dried, concentrated, and purified by flash chromatography
using 3:97 MeOH/CHCl3 as eluent to provide 0.054 g of amide 22 as a yellow oil (22%).
3-(4-Amino-3-iodophenyl)propanoic
Acid (23)[39]
A 1 M
solution of ICl in HOAc (0.240
g, 1.82 mmol) was added dropwise slowly to a solution of 3-(4-aminophenyl)propanoic
acid (16, 0.250 g, 1.51 mmol) in glacial HOAc (8 mL)
at RT. The reaction mixture was stirred overnight and then quenched
with 10% aqNa2SO3. The crude product was extracted
with CH2Cl2 (2 × 20 mL), dried, concentrated,
and purified by flash chromatography using 3:97 MeOH/CHCl3 to provide 0.296 g of acid 23 as a pure white solid
(67%).
3-(4-Azido-3-iodophenyl)propanoic Acid (24)[39]
Azide 24 was prepared
using the same method as described for azide 22 (Method
A) in TFA. The crude product was purified by flash
chromatography using 2:98 MeOH/CHCl3 and used directly
in the next step, Method B, to synthesize 22.
To a solution of 4-azido-3-iodobenzoic
acid[40] (25, 87 mg, 0.30 mmol)
in DMF (2 mL) at 0 °C were added EDC (49 mg, 0.31 mmol) and HOBT
(43 mg, 0.31 mmol). The resulting mixture was stirred in the dark
for 1 h at 0 °C. To this mixture were added a solution of 18(31,32) (99 mg, 0.30 mmol) in DMF (1
mL) and Et3N (0.2 mL), followed by stirring in the dark
at RT overnight. The mixture was then diluted with H2O
and extracted with EtOAc. The combined organic extracts were washed
with 1 M aqNaOH and brine, dried (MgSO4), filtered, concentrated,
and chromatographed using 95:5 EtOAc/Et3N to provide 120
mg of amide 26 as a colorless semisolid (67%). 1HNMR (400 MHz, CDCl3) δ 8.18 (d, J = 1.9 Hz, 1H), 7.81 (dd, J = 8.3, 1.9 Hz, 1H),
7.50–7.40 (m, 2H), 7.30–7.20 (m, 2H), 7.16 (s, 1H),
7.10 (d, J = 8.3 Hz, 1H), 6.97 (t, J = 8.7 Hz, 2H), 6.75 (t, J = 5.5 Hz, 1H), 5.11 (dd, J = 17.1, 12.6 Hz, 2H), 4.56 (d, J = 5.7
Hz, 2H), 2.3–2.0 (m, 10H), 1.5–1.2 (m, 2H). 13CNMR (100 MHz, CDCl3): δ 164.9, 162.9, 144.9, 143.8,
140.9, 139.8, 138.7, 137.6, 131.9, 128.5, 127.4, 126.7, 126.6, 122.1,
120.8, 117.9, 115.1, 114.8, 90.8, 87.5, 71.7, 59.6, 45.4, 43.9, 39.3,
22.3. HRMScalcd for C27H26FIN5O2 [M + H+], 600.1266; found, 600.1268. IR: azide,
2117 cm–1. The oxalate salt was precipitated from
acetone; mp 83–84 °C.
A mixture
of iodide 26 (170 mg, 0.28 mmol), Pd(PPh3)2Br2 (13.5 mg, 0.02 mmol), and bis(tri-n-butyltin) (0.26 mL, 0.51 mmol) in toluene (10 mL) was heated at
105 °C for 4 h. The mixture was then cooled to room temperature,
diluted with saturated aq K2CO3, and then extracted
with EtOAc. The organic layers were then washed with brine, dried
(MgSO4), filtered, concentrated, and chromatographed (EtOAc/Et3N, 95:5) to give 110 mg of the requisite organostannane 27 as a brown oil (52%). Final purification was achieved by
reversed-phase HPLC of 3–5 mg portions on a Waters C-18 Nova-Pak
column (radial compression module, 8 × 100 mm, 6 μm) using
MeOH (42.5%), CH3CN (42.5%), and an aqueous solution (15%)
of Et3N (2.1% v/v) and HOAc (2.8% v/v) at a flow rate of
3 mL/min with UV detection (254 nm). Material with retention time
(tR) 8.5 min was collected, water was
added to lower the organiccomponent to 9% of the total volume, and
the mixture was passed through an activated (MeOH/water) solid-phase
extraction cartridge (Waters Sep-Pak Plus t-C-18)
that was flushed with water (5 mL) and then with air. Elution with
MeOH (6 mL) and concentration under a stream of argon at 60 °C
provided 27 for direct use in the radioiodination reaction. 1HNMR (400 MHz, CDCl3): δ 7.83 (d, J = 2.1 Hz, 1H), 7.74 (dd, J = 8.3, 2.1
Hz, 1H), 7.50–7.40 (m, 2H), 7.25 (s, 2H), 7.19 (s, 1H), 7.13
(d, J = 8.3 Hz, 1H), 6.97 (t, J =
8.7, 2H), 6.42 (t, J = 5.6 Hz, 1H), 5.12 (dd, J = 17.4, 12.5 Hz, 2H), 4.62 (d, J = 5.7
Hz, 2H), 2.30–2.0 (m, 10H), 1.60–1.40 (m, 7H), 1.4–1.2
(m, 7H), 1.2–1.0 (m, 6H), 0.9–0.8 (m, 9H). 13CNMR (100 MHz, CDCl3) δ 166.9, 162.9, 160.5, 149.2,
143.6, 140.9, 139.7, 137.9, 136.3, 135.1, 130.0, 128.5, 127.2, 126.7,
126.6, 121.9, 120.6, 116.5, 114.9, 114.7, 90.7, 71.6, 59.6, 45.3,
43.7, 39.3, 28.9, 27.2, 22.3,13.6, 10.1. IR: azide, 2109 cm–1; HRMScalcd for C39H54FN5O2SnH+, 764.3356; found, 764.3369.No-carrier-added [125I]NaI was obtained
from PerkinElmer, Inc. (Waltham, MA). Radioactivity was measured with
a dose calibrator (CapintecCRC-15W; Ramsey, NJ) employing similar
counting geometries and attenuation correction factors as necessary.
The radio-HPLC system was equipped with flow-through UV absorbance
(254 nm) and radioactivity detectors. A Waters Corp. (Milford, MA)
C-18 Nova-Pak HPLCcolumn (8 × 100 mm, 6 μm; Radial Compression
Module) was used for preparative separations and for analytical work.
HPLC mobile phases were composed of varying proportions of an organiccomponent (MeOH/CH3CN; 1:1, v/v) and an aqueous component
containing Et3N (2.1% v/v) and HOAc (2.8% v/v). Activated
Sep-Pak Light t-C-18 cartridges (Waters Corp.) were
used for solid-phase extraction. Specific radioactivities were calculated
by determining the carrier mass associated with the absorbance peak
for carrier in purified samples of known radioactivity. The carrier
mass in the radioactive samples was assessed by HPLC in reference
to linear (r2 = 1.0) six-point standard
curves (0–3000 or 0–6000 pmol) generated using the corresponding
nonradioactive materials.
A solution of aniline 13 as the oxalate salt (75 μL, 4.0 mM) in NaOAc buffer
(pH 4.0; 0.2 M) was treated with [125I]NaI (2.42 mCi) in
NaOH (20 μL, 10 μM) followed by N-chloro-4-toluenesulfonamide
(chloramine-T) trihydrate (20 μL, 5.0 mM). After 30 min at room
temperature, the mixture was chilled in an ice/MeOH bath and treated
sequentially with ice-cold HOAc (75 μL, 3.0 M) and NaNO2 (30 μL, 0.5 M). After 20 min in the ice bath, NaN3 (30 μL, 0.5 M) was added, and the mixture was allowed
to warm to room temperature. After 30 min, the reaction was quenched
with Na2S2O5 (10 μL, 50 mM).
Purification was achieved by reverse-phase HPLC (45% organic/55% aqueous;
3.0 mL/min). Radioactive material with retention time (tR) 22.4 min and capacity factor (k′)
18.3 corresponded to the chromatographic profile established for nonradioactive 15 and was collected in a 7.5 mL volume. This sample was diluted
to 32 mL with H2O and then passed through an activated
(MeOH/H2O) solid-phase extraction cartridge that was flushed
with H2O (2.0 mL) and then with air. Elution with MeOH
(1.5 mL) gave [125I]15 (1.20 mCi) in 50% radiochemical
yield. A ≥98% radiochemical purity was determined by HPLC,
the radioactive material coeluted with nonradioactive 15, and the specific radioactivity was calculated as 2045 mCi/μmol.
The major nonradioactive products observed during HPLC purification
are likely to be the aryl azide (tR =
10.2 min, k′ = 7.8) and chlorinated aryl azide
(tR = 16.6 min, k′
= 13.3) analogues.
A solution of free-base 20 (75 μL, 4.0 mM) in NaOAc buffer (pH 4.0; 0.2 M) was
treated with [125I]NaI (2.14 mCi) in NaOH (20 μL,
10 μM) and chloramine-T trihydrate (15 μL, 5.0 mM) for
30 min. Further treatments with HOAc/NaNO2, NaN3, and Na2S2O5 were performed as
described above for [125I]15. Purification
was accomplished by HPLC (42% organic/58% aqueous; 3.0 mL/min). Radioactive
material (tR = 20.0 min, k′ = 16.2) corresponding to 22 was collected in
a 12.5 mL volume, diluted to 50 mL with H2O, and processed
by solid-phase extraction as described above to give pure [125I]22 (1.17 mCi) in 55% radiochemical yield. This material
coeluted with nonradioactive 22 on HPLC, and a specific
radioactivity of 2098 mCi/μmol was calculated. Nonradioactive
products observed during HPLC purification are likely to be the aryl
azide (tR = 9.6 min, k′ = 7.3) and chlorinated aryl azide (tR = 15.1 min, k′ = 12.0) analogues.
A solution of organostannane 27 (0.11 mg, 0.15 μmol)
in MeOH (25 μL) was treated with [125I]NaI (2.07
mCi) in dilute NaOH (20 μL, 10 μM) followed by a mixture
of aqueous chloramine-T trihydrate (15 μL, 5.0 mM) and 3% glacial
HOAc in MeOH (85 μL). After 3 min at room temperature, the reaction
was quenched with Na2S2O5 (10 μL,
50 mM). Purification was accomplished by HPLC (45% organic/55% aqueous;
3.0 mL/min). The major radioactive product (tR 13.4 min, k′ = 10.6) corresponded
to 26 and was resolved from minor nonradioactive and
radioactive side products. This material was collected in 7.5 mL of
mobile phase, diluted to 35 mL with water, and processed by solid-phase
extraction as described above for [125I]22 to give 0.96 mCi (46%) of pure [125I]26 that
coeluted with nonradioactive 26 on HPLC and had a specific
radioactivity of 2096 mCi/μmol.
Whole-Cell Competition
Uptake Assay
HEK-293 GripTite
cells (Invitrogen) were plated in 24-well CulturPlates (PerkinElmer)
at 50 000 cells/well and transiently transfected with wild-type
hSERT or hSERTS438T-containing pcDNA3-based plasmid constructs using
the Trans-IT LTI transfection system (Mirus Bio). Assays were conducted
36–48 h after transfection, where cells were washed with MKRHG
buffer (5 mM Tris, 7.5 mM HEPES, 120 mM NaCl, 5.4 mM KCl, 1.2 mM CaCl2, 1.2 mM MgSO4, 10 mM dextrose, pH 7.4) and preincubated
with the competitor compounds for 5 min followed by addition of [3H]5-HT at a final concentration of 50 nM. Uptake was allowed
to proceed for 10 min and then terminated by washing twice with ice-cold
MKRHG. [3H]5-HT uptake was quantified by dissolving cells
in Microscint-20 (PerkinElmer) and using the Packard TopCount NXT
(PerkinElmer) to measure radioactivity (CPM). Data were normalized
to percent activity in the absence of a drug competitor for each form.
Assays were carried out in triplicate and were repeated at least four
times. Ki values were determined using
Prism 5 (GraphPad).
hSERT Photoaffinity Labeling
Photoaffinity
labeling
of hSERT was performed using methods that have been verified previously
for assessing irreversible adduction of a variety of ligands for both
DAT and SERT.[11,12,17,19−26] These procedures were performed using Lewis lung carcinoma porcine
kidney (LLCPK1) cells stably expressing hemagluttinin (HA)-hSERT
(generous gift of Dr. James Foster) and using untransfected parent
LLCPK1 cells for negative controls. Cells were grown to
90% confluence in 12-well plates and washed with KRH buffer. [125I]15, [125I]22, or
[125I]26 was added to a final concentration
of 10 nM and incubated for 1.5 h at 4 °C. In some experiments,
hSERTs were irreversibly labeled in parallel with 10 nM [125I]3 as a positive control.[26] For pharmacological displacement, 10 μM (S)-citalopram was added 30 min prior to addition of the radioligands.
Cells were irradiated with shortwave ultraviolet light (254 nm, Fotodyne
UV Lamp model 3-6000) for 5 min at a distance of 15–20 mm to
photoactivate the radioligand, washed twice with 1 mL of ice-cold
KRH buffer, and lysed by addition of 0.5 mL of RIPA buffer (50 mM
NaF, 2 mM EDTA, 125 mM Na3PO4, 1.25% Triton
X-100, and 1.25% sodium deoxycholate) containing protease inhibitors
for 30 min on ice. Lysates were centrifuged at 20 000g for 15 min at 4 °C to remove insoluble material and
subjected to immunoprecipitation and immunoblotting using anti-HA
monoclonal antibody (Covance) or anti-SERT antibody ST51 (mAB Technologies)
as previously described.[26] Immunoprecipitated
samples were separated on 4–20% SDS-polyacrylamide gels followed
by autoradiography using Hyperfilm MP film (GE Healthcare) for 1–3
days at −80 °C. For immunoblotting, cell lysates were
separated on 4–20% SDS-polyacrylamide gels transferred to 0.45
μm poly(vinylidene difluoride) membranes and probed for hSERT
with anti-HA antibody.
Authors: Peng Zhang; George Cyriac; Theresa Kopajtic; Yongfang Zhao; Jonathan A Javitch; Jonathan L Katz; Amy Hauck Newman Journal: J Med Chem Date: 2010-08-26 Impact factor: 7.446
Authors: David J Lapinsky; Ranganadh Velagaleti; Nageswari Yarravarapu; Yi Liu; Yurong Huang; Christopher K Surratt; John R Lever; James D Foster; Rejwi Acharya; Roxanne A Vaughan; Howard M Deutsch Journal: Bioorg Med Chem Date: 2010-11-04 Impact factor: 3.641
Authors: Rejwi Acharya Dahal; Akula Bala Pramod; Babita Sharma; Danielle Krout; James D Foster; Joo Hwan Cha; Jianjing Cao; Amy Hauck Newman; John R Lever; Roxanne A Vaughan; L Keith Henry Journal: J Biol Chem Date: 2014-08-31 Impact factor: 5.157
Authors: Nageswari Yarravarapu; Laura Geffert; Christopher K Surratt; Michael Cascio; David J Lapinsky Journal: Bioorg Med Chem Lett Date: 2018-09-22 Impact factor: 2.823
Authors: Hafsteinn Rannversson; Jacob Andersen; Lena Sørensen; Benny Bang-Andersen; Minyoung Park; Thomas Huber; Thomas P Sakmar; Kristian Strømgaard Journal: Nat Commun Date: 2016-04-19 Impact factor: 14.919