GLUT1-mediated venlafaxine-thiamine disulfide system-glucose conjugates with “lock-in” function for central nervous system delivery

Article abstract of DOI:10.1111/cbdd.13128

Venlafaxine, a novel third-generation antidepressant drug, has been described as a reference treatment for major depression, owing to its ability of inhibiting both noradrenalin and serotonin neuronal reuptake, and inhibiting dopamine reuptake slightly. However, its clinical application is hampered by a limited brain distribution. Glucosylation is an effective way to enhance the brain targeting ability of drugs, but the bidirectional transport of glucose transporter 1 (GLUT₁) might decrease the concentrations of venlafaxine-glucose (V-G) in brain before the release of parent drug venlafaxine. To conquer this drawback of GLUT₁, “lock-in” thiamine disulfide system (TDS) was introduced to modify the V-G conjugate. Both conjugates could release venlafaxine when incubated with the various buffers, mice plasma, and brain homogenate. The evaluation in vivo demonstrated that venlafaxine-TDS-glucose (V-TDS-G) had an improved targeting ability and significantly increased the level of venlafaxine in brain compared to the naked venlafaxine and V-G. The relative uptake efficiency (RE) and concentration efficiency (CE) were enhanced to 5.69 and 5.70 times higher than that of naked venlafaxine, respectively. The results of this study suggest that the conjugate strategy based on the glucose-TDS (G-TDS) is available to enhance the delivery of central nervous system (CNS) drugs into brain.

Full text of DOI:10.1111/cbdd.13128

PROFESSOR QIANTAO WANG (Orcid ID : 0000-0002-5553-6246)
PROFESSOR YONG WU (Orcid ID : 0000-0003-0719-4963)
Article type
: Research Article
GLUT1-mediated venlafaxine-TDS-glucose conjugates
with “lock-in” function for CNS delivery
Yi Zhao, Li Zhang, Yao Peng, Qiming Yue, Li Hai, Li Guo, Qiantao Wang*, Yong Wu*
Key Laboratory of Drug Targeting and Drug Delivery System of Education Ministry, Department of Medicinal
Chemistry, West China School of Pharmacy, Sichuan University, Chengdu 610041, P. R. of China
Correspondence
Qiantao Wang, Key Laboratory of Drug Targeting and Drug Delivery System of Education Ministry,
Department of Medicinal Chemistry, West China School of Pharmacy, Sichuan University, Chengdu,
China.
Email: qwang@scu.edu.cn.
and
Yong Wu, Key Laboratory of Drug Targeting and Drug Delivery System of Education Ministry,
Department of Medicinal Chemistry, West China School of Pharmacy, Sichuan University, Chengdu,
China.
Email: wyong@scu.edu.cn.
This article has been accepted for publication and undergone full peer review but has not been
through the copyediting, typesetting, pagination and proofreading process, which may lead to
differences between this version and the Version of Record. Please cite this article as doi:
10.1111/cbdd.13128
This article is protected by copyright. All rights reserved.

Highlights
A novel venlafaxine conjugate
was designed and synthesized in this work. The evaluation
V-TDS-G
in vivo demonstrated that
had an improved targeting ability, which indicated that the
V-TDS-G
G-TDS
could act as a vector to enhance the delivery of CNS drugs into brain.
Abstract
Venlafaxine, a novel third generation antidepressant drug, has been described as a reference treatment
for major depression in clinical, owing to its ability of inhibiting both noradrenalin and serotonin
neuronal reuptake, and inhibiting dopamine reuptake slightly. However, its clinical application is
hampered by a limited brain distribution. Glucosylation is an effective way to enhance the brain
targeting ability of drugs, but the bidirectional transport of glucose transporter 1 (GLUT1) might
decrease the concentrations of venlafaxine-glucose (V-G) in brain before the release of parent drug
venlafaxine. To conquer this drawback of GLUT1, “lock-in” thiamine disulfide system (TDS) was
introduced to modify the V-G conjugate. Both conjugates could release venlafaxine when incubated
with the various buffers, mice plasma and brain homogenate. The evaluation in vivo demonstrated that
venlafaxine-TDS-glucose (V-TDS-G) had an improved targeting ability and significantly increased the
level of venlafaxine in brain compared to the naked venlafaxine and V-G. The relative uptake efficiency
(RE) and concentration efficiency (CE) were enhanced to 5.69 and 5.70 times higher than that of naked
venlafaxine, respectively. The results of this study suggest that the conjugate strategy based on the
glucose-TDS (G-TDS) is available to enhance the delivery of central nervous system (CNS) drugs into
brain.
Keywords: brain targeting, blood-brain barrier, GLUT1, thiamine disulfide system, venlafaxine
1. Introduction
Major depressive disorder (MDD) is a common psychiatric mood disorder, with a lifetime prevalence
of around 17% in the general population [1]. It has become a serious social problem which has attracted
the global attention. The World Health Organization (WHO) listed depression as the fourth leading
This article is protected by copyright. All rights reserved.

cause of disability worldwide, and predicted it would be the second leading cause by 2030 [2]. The
major symptoms of this ailment are low mood, lack of interest or pleasure in a variety of activities,
psychomotor agitation or retardation, feelings of guilt, insomnia or hypersomnia, fatigue, extreme
worthlessness, concentration difficulties, hopelessness and suicidal thoughts [3,4]. Currently, the main
treatment for MDD depends on antidepressant medication, such as selective serotonin reuptake
inhibitors (SSRIs), serotonin and norepinephrine reuptake inhibitors (SNRIs), norepinephrine and
dopamine reuptake inhibitors (NDRIs), and monamine oxidase inhibitors (MAOIs) [3].
Venlafaxine (Figure 1), a novel third generation antidepressant drug, has been used as a SSRI for
major depression in clinical, owing to its ability of inhibiting both noradrenalin and serotonin neuronal
reuptake, and inhibiting dopamine reuptake slightly [5,6]. The therapy efficacy of venlafaxine relies
upon its continued presence in CNS for a prolonged period at a considerable concentration. Therefore, it
needs a frequent administration to maintain the effective therapeutic concentration in blood, which may
cause several side effects such as tachycardia, fatigue, headache, dizziness, dry mouth and so on. In
addition, the hydrophilic nature of venlafaxine limits its permeability across blood brain barrier (BBB),
which is formed by tightly connected endothelial cells of brain capillaries with the surrounding
astrocytes and pericytes [7-9], resulting in a poor antidepressant effect in the brain [10]. Therefore,
strategies that can effectively deliver venlafaxine into the brain are of great interests for the treatment of
MDD. Because of the high transport affinity between the transporter and substance, carrier-mediated
transporter (CMT) system seems to be one of the most promising methods to facilitate the delivery of
drugs into brain, among which GLUT1 expressed on the surface of brain capillary endothelial cells is
considered as one of the most efficient transport systems that can transport glucose across the BBB into
brain effectively [11]. Nevertheless, GLUT1 acts as a bidirectional transporter that mediates the
blood-to-brain and brain-to-blood glucose transport in both directions, owing to its expression both on
the luminal and abluminal membranes of the endothelial cells [8].
This article is protected by copyright. All rights reserved.

To overcome the bidirectional delivery of GLUT1, the thiamine disulfide system (TDS) [12,13] with
“lock-in” ability was introduced and attached to the glucose at the C-6 position for the novel brain
targeting carrier. After entering the CNS, the lipophilic TDS part could be reduced by disulfide
reductase, subsequently ring-closed to form a hydrophilic thiazolium quaternary salt, which could not
cross the BBB via diffusion. It means that conjugates with this system would be “locked” in the brain
where they can provide sustained release of the active drug via hydrolysis (Figure 2).
Hence, in this study, we report the design, synthesis and biological evaluation of a novel conjugate
V-TDS-G with “lock-in” function. After altering the behavior of conjugate from bidirectional to
unidirectional transport, it will enhance the concentration of the active drug in brain. The in vivo and in
vitro stability, cytotoxicity, distribution and pharmacokinetic studies in plasma and brain of the new
conjugate were tested and compared with the naked venlafaxine and V-G.
2. Materials and methods
2.1.Materials
Venlafaxine hydrochloride was purchased from Carelife Pharmaceutical Co., Ltd. (Chongqing, P.R.
China). Heparin sodium was obtained from National Institute for Food and Drug Control. Verapamil was
purchased from Biocause Pharmaceutical Co., Ltd. (Hubei). All liquid reagents were distilled before use.
The other chemicals and reagents were obtained from commercial sources.
The melting point was measured on an YRT-3 melting point apparatus (Shantou Keyi instrument &
Equipment Co. Ltd, Shantou, China). IR spectra were obtained on a Perkin Elmer983 (Perkin Elmer,
Norwalk, CT, USA). Elemental analyses were performed by Atlantic Microlab (Atlanta, GA, USA). NMR
spectra were taken on a Varian INOVA 400 MHz (Varian, Palo Alto, CA, USA) using CDCl3 as solvent.
Chemical shifts are expressed in δ (ppm), with tetramethylsilane (TMS) functioning as the internal reference,
and coupling constants (J) were expressed in Hz. Mass spectra were recorded on an Agilent 1946B ESI-MS
instrument (Agilent, Palo Alto, CA, USA). Reversed-phase chromatography performed on C18
chromatographic analysis was carried out using the HPLC system (Alltech, IL, USA) consisted of a RF-530
This article is protected by copyright. All rights reserved.

fluorescence detector (Shimadzu, Japan) and Allchorom plus data operator, respectively. A Diamonsil
column (200 × 4.6 mm, 5 μm) was used.
Kunming mice rats (20-25 g) were provided by the Laboratory Animal Center of Sichuan University
(Chengdu, China). All animal study procedures were approved by the Sichuan University Animal Ethical
Experimentation Committee according to the requirements of the National Act on the use of experimental
animals (China).
2.2.Methods
2.2.1 Synthesis of glucose-TDS-venlafaxine conjugate (V-TDS-G)
2.2.1.1 6-(((benzyloxy)carbonyl)amino)hexanoic acid (2)
The compound 2 was synthesized as reported previously [14].
2.2.1.2. 1-(2-(dimethylamino)-1-(4-methoxyphenyl)ethyl)cyclohexyl
6-(((benzyloxy)carbonyl)amino)hexanoate (3)
To a solution of compound 2 (3.18 g, 12.0 mmol) in CH2Cl2 (25 mL) was added DCC (3.09 g, 15.0
mmol) and DMAP (0.11 g, 0.9 mmol). After stirring for 30 min at room temperature, the solution of
venlafaxine (4.99 g, 18.0 mmol) in CH2Cl2 (15 mL) was added. And the reaction mixture was continued
to be stirred for another 3 h at ambient temperature. Then, the reaction was terminated and filtered. The
filtrate was concentrated and the residue was purified by chromatography to yield 3 (4.89 g, 77.8%) as a
yellow oil. 1H NMR (400 MHz, CDCl3) δ: 1.03-1.11 (m, 2H), 1.27-1.40 (m, 14 H), 2.10 (s, 6H), 2.25(t,
2H, J = 7.2 Hz), 2.43 (dd, 1H, J = 3.6 Hz), 2.84 (t, 1H, J = 11.2 Hz), 3.19 (q, 2H, J = 6.4Hz), 3.77 (s,
3H), 3.90 (dd, 1H, J = 3.6Hz, 10.8 Hz), 4.81 (s, 1H), 5.09 (s, 2H), 6.81 (d, 2H, J = 8.8 Hz), 7.13 (d, 2H,
J = 8.8 Hz), 7.31-7.36 (m, 5H) (Figure S3). HRMS: (ESI+) calculated for C31H44N2O5Na [M+Na]+
547.6812, found 547.6815. Elemental Analysis: C, 70.96; H, 8.45; N, 5.34, found C, 70.90; H, 8.49; N,
5.27.
This article is protected by copyright. All rights reserved.

2.2.1.3. 1-(2-(dimethylamino)-1-(4-methoxyphenyl)ethyl)cyclohexyl 6-aminohexanoate (4)
To a solution of compound 3 (1.10 g, 2.10 mmol) in methanol, Pd/C (0.13 g, 10%) was added. Then, the
mixture was stirred in hydrogen atmosphere at room temperature overnight. Pd/C was filtered, and the
filtrate was concentrated to give 4 (780 mg, 95.3%) as a colorless oil, which was used without purification.
1H NMR (400 MHz, CDCl3) δ: 1.02-1.06 (m, 2H), 1.24-1.62 (m, 14H), 2.11 (s, 6H), 2.16 (s, 2H), 2.25 (t, 2H,
J = 7.6 Hz), 2.44 (dd, 1H, J = 3.2 Hz), 2.71 (t, 2H, J = 7.2 Hz), 2.85 (t, 1H, J = 11.2 Hz), 3.77 (s, 3H), 3.90 (dd,
1H, J = 3.2 Hz, 10.8 Hz), 6.81 (d, 2H, J = 8.4 Hz), 7.13 (d, 2H, J = 8.4 Hz) (Figure S4). HRMS: (ESI+)
calculated for C23H38N2O3Na [M+Na]+ 413.5492, found 413.5490. Elemental Analysis: C, 70.73; H, 9.81; N,
7.17, found C, 70.79; H, 9.85; N, 7.10.
2.2.1.4. 1-(2-(dimethylamino)-1-(4-methoxyphenyl)ethyl)cyclohexyl
6-(2-(N-(3-(benzyldisulfanyl)-5-hydroxypent-2-en-2-yl)formamido)acetamido)hexanoate (5)
To a solution of compound 13 (0.40 g, 1.17 mmol) in THF (15 mL) was added NMM (128 µl, 1.17
mmol) and IBCF (148 µl, 1.17 mmol) at -15 oC. After stirring for 3 min, compound 4 (0.46 g, 1.17
mmol) in THF (15 mL) was added to the above mixture at -15 oC. The reaction mixture was stirred
overnight at room temperature. THF was removed in vacuo and CH2Cl2 was added to redissolve the
residue. Then, the CH2Cl2 layer was washed with saturated aqueous NaHCO3, HCl (aq., pH 3), brine,
and dried by anhydrous Na2SO4. The organic layer was concentrated and purified by flash
chromatography to give compound 5 (0.40 g, 47.6%) as a white solid. m.p.: 70-72 oC. 1H NMR (400
MHz, CDCl3) δ: 1.00-1.10 (m, 2H), 1.20-1.65 (m, 14 H), 2.02 (s, 3H), 2.33 (t, 2H, J = 6.4 Hz), 2.37 (s,
6H), 2.54 (d, 2H, J = 12.8 Hz), 2.71 (t, 2H, J = 6.0 Hz), 3.20-3.28 (dq, 2H, J = 2.4 Hz, 6.0 Hz), 3.73 (t,
2H, J = 6.0 Hz), 3.80 (s, 3H), 3.91 (s, 2H), 4.04 (s, 2H), 4.13 (t, 1H, J = 8.4 Hz), 4.25 (brs, 1H), 6.88 (d,
2H, J =8.4 Hz), 6.93 (s, 1H), 7.20 (d, 2H, J = 8.4 Hz), 7.26-7.35 (m, 5H), 8.02 (s, 1H) (Figure S5).
HRMS: (ESI+) calculated for C38H55N3O6S2Na [M+Na]+ 736.9796, found 736.9794. Elemental
Analysis: C, 63.92; H, 7.76; N, 5.89; S, 8.98, found C, 63.97; H, 7.71; N, 5.83; S, 8.90.
2.2.1.5. 3-(benzyldisulfanyl)-4-(N-(2-((6-((1-(2-(dimethylamino)-1-(4-methoxyphenyl)ethyl)cyclohexyl
)oxy)-6-oxohexyl)amino)-2-oxoethyl)formamido)pent-3-en-1-yl
(((2R,3R,4S,5R)-3,4,5,6-tetrakis((trimethylsilyl)oxy)tetrahydro-2H-pyran-2-yl)methyl)
succinate (6)
This article is protected by copyright. All rights reserved.

To a solution of glucose-TMS derivative 18 (79 mg, 0.14 mml) in CH2Cl2 (5 mL) was added DCC
(35 mg, 0.17 mmol) and DMAP (1.7 mg, 0.014 mmol) at -5 oC, and the mixture was stirred for 30 min.
Then compound 5 (50 mg, 0.07 mmol) in CH2Cl2 (5 mL) was added dropwise. After stirring 5 h at 30
oC, the reaction mixture was filtered and concentrated in vacuo. Column chromatography yielded
product 6 (80 mg, 90.4%) as a yellow oil. 1H NMR (400 MHz, CDCl3) δ: 0.15 (s, 36H), 1.06 (m, 2H),
1.25-1.62 (m, 14H), 1.98 (s, 3H), 2.13 (d, 6H, J = 12.8 Hz), 2.25 (t, 2H, J = 7.6 Hz,), 2.46 (d, 2H, J = 14
Hz), 2.57 (t, 2H, J = 6 Hz), 2.67 (m, 2H), 2.75 (t, 2H, J = 6.4 Hz), 3.22 (q, 2H, J = 6.8 Hz), 3.36 (dd, 1H,
J = 3.2 Hz, 9.2 Hz), 3.42 (t, 1H, J = 8.8Hz), 3.78 (s, 3H), 3.78 (t, 1H, J = 8.8 Hz), 3.90-3.94 (m, 2H),
3.93 (s, 2H), 3.98 (s, 2H), 4.05 (dd, 1H, J = 5.6 Hz, 12 Hz), 4.17 (t, 2H, J = 6Hz), 4.34 (dd, 1H, J = 2 Hz,
12 Hz), 5.00 (d, 1H, J = 3.2 Hz), 6.79 (t, 1H, J = 6.4 Hz), 6.82 (d, 2H, J = 8.8 Hz), 7.14 (d, 2H, J = 8.4
Hz), 7.26-7.34 (m, 5H), 7.90 (s, 1H) (Figure S6). HRMS: (ESI+) calculated for C60H101N3O14S2Si4Na
[M+Na]+ 1287.9174, found 1287.9170. Elemental Analysis: C, 56.97; H, 8.05; N, 3.32; S, 5.07; Si,
8.88, found C, 56.90; H, 8.09; N, 3.37; S, 5.02; Si, 8.95.
2.2.1.6. 3-(benzyldisulfanyl)-4-(N-(2-((6-((1-(2-(dimethylamino)-1-(4-methoxyphenyl)ethyl)cyclohexyl
)oxy)-6-oxohexyl)amino)-2-oxoethyl)formamido)pent-3-en-1-yl
(((2R,3S,4S,5R)-3,4,5,6-tetrahydroxytetrahydro-2H-pyran-2-yl)methyl) succinate (V-TDS-G)
To a solution of compound 6 (54 mg, 0.043 mmol) in CH2Cl2/CH3NO2 (1 mL/1 mL) was added
trifluoracetic acid (64 μL, 0.86 mmol) at 0 °C and then the reaction was stirred for 1 h at room
temperature. The mixture was concentrated and the residue was purified by chromatography to give
conjugate V-TDS-G (34 mg, 80.0%) as a white solid. m.p.: 76-78 °C. 1H NMR (400 MHz, CDCl3) δ:
0.80-1.03 (m, 2H), 1.19-1.62 (m, 14 H), 1.99 (s, 3H), 2.36 (s, 4H), 2.54 (s, 6H), 2.61 (s, 2H), 2.67 (s,
2H), 2.89 (s, 2H), 3.23 (d, 2H, J = 6.0 Hz), 3.29-3.41 (m, 1H), 3.48 (s, 1H), 3.50 (s, 1H), 3.70 (t, 1H, J =
12 Hz), 3.79 (s, 3H), 3.89 (s, 2H), 3.99 (s, 2H), 4.16 (s, 2H), 4.20 (d, 2H, J = 11.2 Hz), 4.36 (t, 1H, J =
13.6 Hz), 4.56(d, 1H, J = 7.2 Hz), 5.19 (d, 1H, J = 3.2 Hz), 6.89 (d, 2H, J = 8 Hz), 7.04 (s, 1H, J = 6.4
Hz), 7.22 (d, 2H, J = 8.4 Hz), 7.25-7.33 (m, 5H), 7.85 (s, 1H) (Figure S15). IR (KBr, cm-1) ν: 3448,
3426, 2862, 2062, 1727, 1648, 1452, 1400, 1105, 701, 631, 558 (Figure S16). HRMS: (ESI+) calculated
for C48H69N3O14S2Na [M+Na]+ 999.1929, found 999.1926. Elemental Analysis: C, 59.06; H, 7.12; N,
4.30; S, 6.57, found C, 59.00; H, 7.18; N, 4.36; S, 6.52.
This article is protected by copyright. All rights reserved.

2.2.2 HPLC analysis of venlafaxine and conjugates
Venlafaxine and conjugates (V-TDS-G and V-G) were determined by LC-10A liquid
chromatographic system (Shimadzu Japan). A C18 analytical column (4.6 mm × 200 mm, 5µm) was
maintained at 25 oC and the sample volume injected was 20 µl. In brief, chromatographic separation
was carried out using isocratic elution with a mobile phase composed of methanol/water (55:45),
containing potassium dihydrogen phosphate (340 mg) and triethylamine (0.3 mL) per 100 mL. The
detection wavelength was 263 nm and samples were delivered at a flow rate of 1.0 mL/min.
2.2.3 In vitro stability in various buffers
The in vitro chemical stability of conjugates V-G and V-TDS-G was investigated by incubating the
conjugates in pH 2.0, 4.0, 6.0, 7.4 and 8.0 phosphate buffers, respectively. Briefly, 1 mL of the
conjugates methanol solution (250 μM) was added into 4 mL different buffer separately. The entire
system was kept at 37 ± 0.5 °C with continuous shaking at 100 rpm/min. 200 μL medium samples were
taken at the following time points: 0, 0.5, 1, 3, 6, 12, and 24 hours, and then replaced by fresh medium.
The samples were analyzed by HPLC method described before.
2.2.4 Metabolic stability in plasma and brain homogenate
The preliminary metabolic stability of conjugates V-G and V-TDS-G was studied by incubating in
plasma and brain homogenate, respectively. 1 mL of the conjugates methanol solution (250 μM) was
added into 4 mL plasma or brain homogenate, with 10 μL verapamil (30 μg/mL) as the internal
standard. Then the entire system was kept at 37 ± 0.5 °C with continuous shaking at 100 rpm/min. 200
μL samples were taken at the following time points: 0, 0.5, 1, 3, 6, 12, and 24 h, and then replaced by
plasma or brain homogenate. 200 μL acetonitrile was added to each sample. After centrifuging for 15
min, the supernatants of the samples were analyzed by HPLC method described before.
2.2.5 Evaluation of cytotoxicity
Human embryonic kidney 293T cells (HEK293T, Shanghai Institutes for Biological Sciences, CAS)
were cultured in Dulbecco’s Modified Eagles Medium (DMEM) supplemented with 10% fetal bovine
serum (FBS), 100 U/mL streptomycin, and 100 U/mL penicillin at 37 °C in a humidified 5% CO2
atmosphere.
This article is protected by copyright. All rights reserved.

The cytotoxicity of conjugates and venlafaxine was assayed in HEK293T cells using the
(3,4,5-dimethylthiazol-yl)-2,5-diphenyl-tetrazolium (MTT) assay. Briefly, Human 293T embryonic
kidney cells were seeded in 96-well plates at a density of 8 000 cells/well and cultured for 24 h at 37 oC.
The cells were then treated with V-G, V-TDS-G and venlafaxine (400 μM, 200 μM, 100 μM, 50 μM, 25
μM, 12.5 μM, 6.25 μM, and 3.125 μM) respectively and incubated for 24 h. After incubation, 20 μL
MTT solution (5.0 mg/mL) was added to the medium and incubated for another 4 h at 37 °C. After
removal of the culture medium, the reduced MTT dye was solubi1ized by DMSO (100 μL) and the
absorbance was read at 570 nm wavelength on an automatic microplate spectrophotometer. The cell
viability = A570 nm for the treated cells/A570 nm for the control cells × 100%.
2.2.6 Distribution and pharmacokinetic studies in mice
Kunming mice weighing 20-22 g were randomly divided into groups. All animal experiments were
performed by following institutional guidelines and approved by the University Committee on Use and
Care of Animals, Sichuan University. The aqueous solutions of venlafaxine and conjugates (V-TDS-G
and V-G) were prepared with 0.1% Tween-80 to give a final concentration of 1 mg/mL. The mice were
injected the venlafaxine, V-TDS-G and V-G solutions (10 mg/kg, calculated as venlafaxine) through
the tail vein.
At appropriate time interval (5 min, 15 min, 30 min, 45 min, 1 h, 3 h, 6 h and 24 h), the animal was
sacrificed and blood sample withdrawn from the orbital sinus was collected in heparinized tube. Plasma
was immediately separated by centrifugation and diluted with saline to 1:4. Meanwhile, the brain tissue
was removed and flushed with water for three times to remove the blood remained and then the brain
was rolled over on the filter paper carefully to remove the main vessel. Each brain tissue was
homogenized and diluted with methanol to 1:4 (g/mL). An aliquot of 10 μL verapamil (30 μg/mL) was
added into 0.5 mL plasma or 0.5 mL brain homogenate and the mixture was vortexed with 1.5 mL
methanol for 5 min. Before analyzing, hydrolysis was performed. The plasma samples and brain
homogenates were added with an aliquot of 40 μL NaOH (6 M) aqueous solution. After 5 min of
hydrolysis at room temperature, venlafaxine was released from the conjugates, then another aliquot of 6
M HCl (40 μL) aqueous solution was added to neutralize NaOH. Subsequently, the mixture was
vortexed with 1.5 mL methanol for another 5 min. After that, the mixture was centrifuged at 15000 rpm
This article is protected by copyright. All rights reserved.

for 15 min. The separated supernatant was evaporated to dryness at 40 °C under nitrogen flow. The
residues were redissolved in 200 μL methanol and centrifuged at 15000 rpm for 15 min, and then 20 μL
of the supernatant was injected into the HPLC system as described earlier for analysis. What’s more, the
concentrations of released venlafaxine and conjugates in blood and brain were also analyzed before
hydrolysis.
2.2.7 Date analysis
The area under the curve (AUC0–t) and the maximal concentration (Cmax) were calculated by Data and
Statistics (DAS, Shanghai, China). The statistical analysis of the samples was performed by using a
Student’s t-test with p-values < 0.05 as the minimal level of significance. The relative uptake efficiency
(REbrain) and concentration efficiency (CEbrain) were calculated to evaluate the brain targeting property
of V-TDS-G and V-G. The values of REbrain and CEbrain are defined as follows:
REbrain = (AUC0 t, brain Con/(AUC0 t, brain)Ven,
)
−
−
CEbrain = (Cmax, brain Con/(Cmax, brain)Ven.
)
The relative uptake efficiency (Replasma) and concentration efficiency (CEplasma) were calculated to
evaluate the plasma targeting property of conjugates. The values of REplasma and CEplasma are defined as
follows:
REplasma = (AUC0 t, plasma Con/(AUC0 t, plasma)Ven,
)
−
−
CEplasma = (Cmax, plasma Con
)
/(Cmax, plasma)Ven.
where Con and Ven represented conjugates samples and naked venlafaxine, respectively.
This article is protected by copyright. All rights reserved.

3. Results and discussion
3.1.Synthesis and characterization of conjugates
3.1.1. General procedure for synthesis of conjugate V-TDS-G
The synthetic route of conjugate V-TDS-G was illustrated in Scheme 1. The starting material were
commercially available aminocaproic acid and venlafaxine hydrochloride. In the presence of DCC and
DMAP, condensation of venlafaxine and 6-Cbz-adipic acid 2 gave 3. After deprotecting the Cbz group,
amine 4 was generated. Subsequently, condensation of compound 4 and the TDS part 13 was carried
out in the condition of IBCF/NMM to give 5. Intermediate 5 coupled with glucose-TMS derivative 18
to generate 6, which was then deprotected the trimethylsilyl (TMS) group under TFA, resulting in the
desired conjugate V-TDS-G.
3.1.2. General procedure for synthesis of conjugate V-G
In the presence of IBCF/NMM, intermediate 4 coupled with glucose-TMS derivative 18 to generate
7, which was then removed the trimethylsilyl (TMS) group, giving venlafaxine-glucose conjugate V-G
(Scheme S1). The detailed synthetic process and 1H NMR (Figure S7 and Figure S13) and IR (Figure
S14) spectrums were seen in the Supporting Information.
3.1.3. General procedure for synthesis of the TDS intermediate 13
The synthesis of key intermediate 13, the TDS part, was described in Scheme S2. The starting
material 5-(2-hydroxyethyl)-4-methylthiazole 10 is commercially available, which was carried out with
t-butyl bromoacetate in acetonitrile to generate quaternary thiazolium salt 11 in satisfied yield. With the
protection of argon gas, compound 11 was hydrolyzed in aqueous solution of sodium hydroxide,
following substituted with thiosulfate salt 9 prepared from benzyl bromide 8 and sodium thiosulfate,
resulting in the compound 12. Then the t-butyl protecting group was removed to give the TDS
intermediate 13. The 1H NMR spectrums of compound 11-13 (Figure S8, S9 and S10) were shown in
the Supporting Information.
This article is protected by copyright. All rights reserved.

3.1.4. General procedure for synthesis of glucose-TMS intermediates 18
Glucose-TMS derivative 18 was prepared by the synthetic route shown in Scheme S3. D-glucose 14
was totally etherified with chlorotrimethylsilane (TMSCl) and hexamethyl disilazane in pyridine to
give penta-O-trimethylsilyl-D-glucopyranose 15, which was treated by the mixture of acetone, acetic
acid and methanol to deprotect the C-6 TMS group selectively to obtain compound 16. Then, the linker
4-(benzyloxy)-4-oxobutanoic acid was connected to intermediate 16 to generate compound 17, which
was deprotected the benzyl group to form glucose-TMS derivative 18. The detailed synthesis of glucose
derivative intermediates (15-18) and 1H NMR (Figure S11 and S12) were outlined in the Supporting
Information.
3.2.In vitro stability in various buffers
In order to investigate the chemical stability of the two conjugates, V-G and V-TDS-G were
incubated in pH 2.0, 4.0, 6.0, 7.4 and 8.0 phosphate buffers, respectively. These solutions were
maintained at 37 °C, and the aliquots were withdrawn in predetermined time intervals, then the
concentrations of the conjugates were determined by HPLC method.
It could be seen from the degradation curves (Figure 3) that both conjugates appeared to be highly
stable in pH 4.0 and 6.0 buffer solution, moderately stable in 2.0 and instable in pH 7.4 and 8.0 buffer
solution. The pseudo first order rate constants (Kdisapp) and half-lives (t1/2) of conjugates in aqueous
solutions were calculated by linear regression of the peak area against time, and were illustrated in
Table S1. According to this result, the slow hydrolysis of the conjugates ensured that the parent drug
venlafaxine could be slowly released and sustained in the physiological environment at pH 7.4.
3.3.Metabolic stability in plasma and brain homogenate
The metabolic stability of the V-G and V-TDS-G was studied in the mice plasma and brain
homogenate at 37 °C. The degradation curves were shown in Figure 4. The Kdisapp and t1/2 of the
conjugates in aqueous solutions were calculated by linear regression of the peak area against time
(Table S2). As expected, the two conjugates showed considerable stability in plasma, where t1/2 were
found to be 73.74 and 57.76 hours for conjugates V-G and V-TDS-G respectively. The long half-life
time guaranteed that they had sufficient time to reach the brain before they were metabolized.
This article is protected by copyright. All rights reserved.

As shown in Figure 4b, conjugate V-G presents a moderate stability in brain homogenate, illustrating
that it was metabolized with a low rate. Therefore, it may take some time to release the active drug
venlafaxine after V-G was transported into the brain. What’s more, the accumulation of V-G in brain
may cause that a portion of the conjugates were excreted again to the outside of the brain, which
decreased the concentration of parent drug in brain. Quite the opposite, conjugate V-TDS-G converted
swiftly after incubating in brain homogenate and showed a dramatic degradation. There was only
26.09% of V-TDS-G left after 3 h. It may be attributed to the disulfide reductase that catalyzed the
reduction and cyclization reaction of V-TDS-G to become the thiazolium. It turned out that the
conjugate could be “locked” in the brain upon completion of this transformation, instead of being
transported outside the brain. As the above studies in vitro demonstrated that the conjugate V-TDS-G
possessed of favorable physicochemical properties, we proceeded with the distribution studies in vivo.
3.4.Cytotoxicity
Cytotoxicity is one of the most critical factors to be considered in selecting conjugates for biomedical
applications. The toxicity of V-G and V-TDS-G against HEK293T cells was evaluated by MTT assays
and the results were shown in Figure 5. At the dose level of 0 ~ 200 μM, the conjugates had almost no
toxicity on HEK293T cells. Even when the dose level was up to 400 μM, the cell viability remained to
be more than 50%. The results suggested that both conjugates had limited toxicity.
3.5.Distribution and pharmacokinetic studies in mice
The curves in Figure 6 showed that the venlafaxine concentrations in blood from the two conjugates
were higher than that of the direct venlafaxine administration. The concentrations of the conjugates also
declined much slower. Because the conjugates were more stable in plasma and had lower degradation
rates, they had higher chances to be rapidly transported across the BBB. Meanwhile, the AUC0–t of
venlafaxine from the two conjugates were higher than that of the naked venlafaxine within 24 h after i.v.
administration, given the values of 1.41 and 1.14 times for conjugates V-G and V-TDS-G respectively
(Table 1). The Cmax for free venlafaxine from the two conjugates were 1.64 and 1.76 times higher than
that of the naked venlafaxine. These data further indicated that the conjugates were stable in plasma.
The mean residence times (MRT) and half-life (t1/2) of venlafaxine from V-G and V-TDS-G were lower
than that of venlafaxine, which might due to the rapid transportation of the conjugates from blood to
This article is protected by copyright. All rights reserved.

brain during the testing time. The concentration curves of the conjugates and released venlafaxine in
plasma versus time after administration were shown in Figure S1, and the pharmacokinetic parameters
were listed in Table S3. The concentration of venlafaxine was higher than that of the conjugates groups
(Figure S1), which was ascribed to the stability of both conjugates in plasma. Hence, the conjugates had
enough time to reach the brain before they were metabolized. What’s more, the venlafaxine released
from V-TDS-G was less than that of V-G, which suggested that more V-TDS-G might be delivered
into brain and locked in it.
Table 1: Pharmacokinetic parameters in plasma. (n=3)
Parameters
AUC(0-t) (µg/mL·h)
MRT (h)
t1/2 (h)
Venlafaxine
340.56 ± 3.21
10.41± 0.06
6.46± 0.99
0.08
Venlafaxine from V-G
479.98 ± 49.49*
9.85 ± 0.84
3.78 ± 0.81
0.25
Venlafaxine from V-TDS-G
388.09 ± 23.66**
4.41 ± 0.06*
3.44 ± 1.04
Tmax (h)
0.25
Cmax (µg/mL)
RE
62.98 ± 12.41
-
103.09 ± 4.40*
1.41
110.81 ± 8.89***
1.14
CE
-
1.64
1.76
*P<0.05 versus venlafaxine, **P<0.01 versus venlafaxine, ***P<0.001 versus venlafaxine.
The in vivo pharmacokinetic evaluation of conjugates gives new insight into the possibility of
utilizing glucose as a means for brain targeting. It is obvious that the conjugates could be delivered to
the brain after i.v. administration (Figure 7). The concentration of venlafaxine was much higher when
conjugates were administrated than when naked venlafaxine was used at any interval during 24 h. The
AUC0-t and Cmax of venlafaxine for both conjugates were fairly higher than that of naked venlafaxine
(Table 2). The relative uptake efficiencies (REs) were enhanced for venlafaxine from conjugates
V-TDS-G and V-G, for which the values were 2.58 and 5.69 times higher, respectively, than that of the
naked venlafaxine. The concentration efficiencies (CEs) were also increased to 1.92 and 5.70 times
This article is protected by copyright. All rights reserved.

higher than that of venlafaxine. More importantly, conjugates V-TDS-G was able to penetrate the
blood-brain barrier and release venlafaxine in the brain in vivo (Figure S2). Although both AUC0-t and
Cmax of the conjugates in plasma were about half that of the parent drug, the opposite trend was seen in
the brain (Table S4). After 1 h, the brain concentration of venlafaxine released from either conjugate
was approximately twice that found in brain after administration of the venlafaxine, which can have a
slow release, long-term effect. What’s more, conjugates V-TDS-G was degraded to venlafaxine when it
reached the brain, which was mainly in the reduction and cyclization reaction of disulfide reductase and
locking it in the brain. Although, the concentration of V-TDS-G in brain was lower than that of V-G,
the venlafaxine released from V-TDS-G was more than the later. These data further indicated our
design was feasible, that both conjugates could maintain sufficient levels and deliver venlafaxine into
the brain.
Table 2: Pharmacokinetic parameters of venlafaxine in brain homogenate. (n=3)
Parameters
AUC(0-t) (µg/mL·h)
MRT (h)
t1/2 (h)
Venlafaxine
81.74 ± 2.93
5.82 ± 0.35
7.60 ± 1.11
0.50
Venlafaxine from V-G
211.18 ± 17.47***
8.76 ± 0.49*
9.24 ± 1.09
0.83
Venlafaxine from V-TDS-G
465.29 ± 32.01*
6.25 ± 0.29**
2.89 ± 0.34
0.92
Tmax (h)
Cmax (µg/g)
RE
10.64 ± 0.92
-
20.44 ± 2.28*
2.58
60.70 ± 4.34*
5.69
CE
-
1.92
5.70
*P<0.05 versus venlafaxine, **P<0.01 versus venlafaxine, ***P<0.001 versus venlafaxine.
What’s more, these data also indicated that conjugate V-TDS-G with “lock-in” function could
deliver venlafaxine across the BBB and keep it in the brain. And the pharmacokinetic parameters
suggested that the brain targeted ability of conjugate V-TDS-G is better than that of V-G without
“lock-in” ability. Moreover, the higher RE and CE of venlafaxine from conjugate V-TDS-G
This article is protected by copyright. All rights reserved.

demonstrated that the exposed glucose residues could recognize GLUT1, which could deliver the
conjugates into the brain. Furthermore, the TDS system could be reduced and ring-closed to be a
thiazolium quaternary salt in the brain. Thus, the conjugates V-TDS-G would be “locked” in the brain
and prevent the bidirectional delivery of the drug. Consequently, it can serve sustained release of
venlafaxine, improving the concentration of parent drug venlafaxine in the brain.
4. Conclusion
Generally, to overcome the bidirectional delivery of GLUT1, the thiamine disulfide system (TDS),
with the ability of “lock-in”, was introduced and attached to the glucose moiety at the C-6 position by
ester bond. And a novel venlafaxine conjugate
was designed and successfully synthesized in
V-TDS-G
this work. After i.v. administration of the conjugates, the results showed that the concentrations of
venlafaxine in brain were remarkably higher than administration of the naked venlafaxine and with
V-G
the same dose. The conjugates exhibited excellent transport ability across the BBB and were able to
deliver venlafaxine into CNS. Meanwhile, according to the pharmacokinetic parameters and
biodistribution data, the TDS modified conjugates was locked in the brain through reduction and
cyclization processes, contributing to the brain bioavailability of venlafaxine. Moreover, the results
indicated that the glucose-thiamine disulfide system (
) could act as a vector to enhance the
G-TDS
delivery of CNS drugs into brain. Given it efficiency and ease of synthesis, this approach may be
applied in design of other brain targeting drugs.
Acknowledgement
This work was supported by the National Natural Science Foundation of China (No.81573286,
No.81773577 & No.21472130).
Conflict of Interest
The authors declare no competing financial interest.
This article is protected by copyright. All rights reserved.

Supporting Information
Additional Supporting Information may be found online in the supporting information tab for this
article.
Figure Legends
Figure 1: Structure of venlafaxine.
Figure 2: Structure of V-TDS-G and the brain “lock-in” pathways.
Scheme 1: Synthesis of conjugate V-TDS-G. Reagents and conditions: (a) CbzCl, NaOH (aq.), H2O, 30
min; (b) venlafaxine, DCC, DMAP, CH2Cl2, 3 h; (c) Pd/C, H2, CH3OH, overnight; (d) 13, IBCF, NMM,
THF, -15 °C - r.t., overnight; (e) 18, DCC, DMAP, CH2Cl2, 5 h; (f) TFA, CH2Cl2/CH3NO2, 0 °C - r.t., 1
h.
Figure 3: Degradation curves of conjugates in various buffers at 37 °C. (a) V-G; (b) V-TDS-G. Data
represent the mean ± SD (n = 3).
Figure 4: Degradation curves of conjugates in plasma and brain homogenate at 37 °C. (a) in plasma; (b)
in brain homogenate. (n=3). Data represent the mean ± SD (n = 3).
Figure 5: Cell viability upon addition of conjugates in HEK293T cells. Cytotoxicity of conjugates
against HEK293T cells over a 24 h incubation period determined by MTT assay. Cell viability was
expressed as a percentage of the control cell culture. Values are represented as mean ± SD (n=3).
Figure 6: Concentration curves in plasma versus time after administration of venlafaxine, V-G and
V-TDS-G. (n=3).
Figure 7: Concentration curves in brain homogenate versus time after administration of venlafaxine,
conjugates V-G and V-TDS-G. (n=3).
This article is protected by copyright. All rights reserved.

References
[1] Z. J. Weng, Y. Y. Zheng, J. Q. Li, Chem. Biol. Drug Des. 2015, 85, 454.
[2] C. D. Mathers, D. Loncar, PLoS Med. 2006, 3, e442.
[3] A. K. Dhar, R. Mahesh, A. Jindal, T. Devadoss, S. Bhatt, Chem. Biol. Drug Des. 2014, 84, 721.
[4] J. Y. Ma, Y. C. Quan, H. G. Jin, X. H. Zhen, X. W. Zhang, L. P. Guan, Chem. Biol. Drug Des.
2016, 87, 342.
[5] H. A. Panahi, H. S. Alaei, Int. J. Pharm. 2014, 476, 178.
[6] D. M. Gaber, N. Nafee, O. Y. Abdallah, Int. J. Pharm. 2015, 488, 86.
[7] J. Stockwell, N. Abdi, X. F. Lu, O. Maheshwari, C. Taghibiglou, Chem. Biol. Drug Des. 2014, 83,
507.
[8] Y. Zhao, B. Y. Qu, X. Y. Wu, X. C. Li, Q. Q. Liu, X. X. Jin, L. Guo, L. Hai, Y. Wu, Eur. J. Med.
Chem. 2014, 82, 314.
[9] C. Saraiva, C. Praca, R. Ferreira, T. Santos, L. Ferreira, L. Bernardino, J. Control. Release 2016,
235, 34.
[10] S. Haque, S. Md, J. K. Sahni, J. Ali, S. Baboota, J. Psychiatr. Res. 2014, 48, 1.
[11] B. Y. Qu, X. C. Li, M. Guan, X. Li, L. Hai, Y. Wu, Eur. J. Med. Chem. 2014, 72, 110.
[12] T. Ishikura, T. Senou, H. Ishihara, T. Kato, T. Ito, Int. J. Pharm. 1995, 116, 51.
[13] X. C. Li, B. Y. Qu, X. X. Jin, L. Hai, Y. Wu, J. Drug Target. 2014, 22, 251.
[14] A. Tarozzi, C. Marchetti, B. Nicolini, M. D'Amico, N. Ticchi, L. Pruccoli, V. Tumiatti, E. Simoni,
A. Lodola, M. Mor, A. Milelli, A. Minarini, Eur. J. Med. Chem. 2016, 117, 283.
This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.