Enhanced anti-tumor efficiency of gemcitabine prodrug by FAPα-mediated activation

Article abstract of DOI:10.1016/j.ijpharm.2019.01.032

Gemcitabine (Gem) as an anti-cancer agent has been limited by its short circulation time and rapid metabolism that reflects in low tumor uptake and low therapeutic efficiency. To improve its anti-tumor activity, a novel FAPα enzyme-activated prodrug of Z-

Full text of DOI:10.1016/j.ijpharm.2019.01.032

Accepted Manuscript
Enhanced anti-tumor efficiency of gemcitabine prodrug by FAPα-mediated ac-
tivation
Jing Sun, Dan Yang, Shi-He Cui, Hai-Tao Zhang, Yu Fu, Jian-Cheng Wang,
Qiang Zhang
PII:
S0378-5173(19)30067-5
https://doi.org/10.1016/j.ijpharm.2019.01.032
IJP 18094
DOI:
Reference:
International Journal of Pharmaceutics
To appear in:
Received Date:
Revised Date:
Accepted Date:
21 September 2018
28 December 2018
10 January 2019
Please cite this article as: J. Sun, D. Yang, S-H. Cui, H-T. Zhang, Y. Fu, J-C. Wang, Q. Zhang, Enhanced anti-tumor
efficiency of gemcitabine prodrug by FAPα-mediated activation, International Journal of Pharmaceutics (2019),
doi: https://doi.org/10.1016/j.ijpharm.2019.01.032
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Enhanced anti-tumor efficiency of gemcitabine prodrug by
FAPα-mediated activation
Jing Sun#1, Dan Yang#1, Shi-He Cui1, Hai-Tao Zhang1,2, Yu Fu1, Jian-Cheng Wang1*, Qiang
Zhang1.
1. Beijing Key Laboratory of Molecular Pharmaceutics and New Drug Delivery Systems, School
of Pharmaceutical Sciences, Peking University Health Science Center, Beijing 100191, China
2. School of Pharmaceutical Sciences, Central South University, Changsha, Hunan 410013, China
*Corresponding Author:
Jian-Cheng Wang, Ph.D, Tel/Fax: 86-10-82805932, Email: wang-jc@bjmu.edu.cn;
#These authors equally contributed to this work.
Abstract
Gemcitabine (Gem) as an anti-cancer agent has been limited by its short circulation
time and rapid metabolism that reflects in low tumor uptake and low therapeutic
efficiency. To improve its anti-tumor activity, a novel FAPα enzyme-activated prodrug
of Z-GP-Gem modified at 4-amino group of Gem was developed, which could
effectively release parent Gem based on the specific cleavage via FAPα
enzyme-activation in tumor microenvironment. Compared to Gem, the Z-GP-Gem
prodrug exhibited significantly enhanced inhibition of both tumor growth and
pulmonary metastasis in BALB/c mice bearing orthotopic breast 4T1 tumors. The
Z-GP-Gem prodrug has a prolonged circulation time and a high tumor uptake based
on the modification of Z-GP dipeptide at 4-amino group of Gem. These eventually
caused a marked improvement in the systemic toxicity and the tumor growth
inhibition in 4T1 cells. More interestingly, the unexpected depletion of
tumor-associated fibroblast (TAF) was observed during the treatment of Z-GP-Gem
prodrug in animal model. Therefore, these findings demonstrated that the
FAPα-activated prodrug Z-GP-Gem would be a desirable approach for tumor therapy
by intravenous administration.
1

Keywords: FAPα, prodrug, gemcitabine, antitumor activity, targeting
1. Introduction
Gemcitabine (Gem), one of nucleoside analog drugs, has been approved by FDA
for the first-line treatment of advanced or metastatic pancreatic, non-small cell lung,
ovarian and breast cancer (Abdayem et al., 2014). It exerted the anti-proliferative
capability by disturbing DNA synthesis (Muggia et al., 2012). However, the short
plasma half-life and drug resistance still impeded the clinical application of Gem
(Maiti et al., 2013; Zhao et al., 2013; Hoang et al., 2016). It has been unclosed that the
specific recognition of deoxycytidine deaminase to naked 4’-amino group of Gem
always led to rapid metabolism and resulted in short half-life in plasma. Also, the
invalid uptake of Gem in hENT-deficient patients was discovered, that associated with
a special hENT-dependent endocytosis mechanism. Thus, “prodrugs” has been
extensively considered as a promising strategy to resolve these problems.
Up to now, some prodrugs of Gem have achieved satisfactory clinical application
based on their improved physicochemical, pharmacokinetic or biological effects (Pratt
et al., 2013; Hou et al., 2017). Among the prodrug strategies of Gem, both 4-amino
and 5’-hydroxyl groups were often used as the key sites to be conjugated with
designed fragments (Fan et al., 2015; Karampelas et al., 2017). Considering the
phosphorylation of 5’-hydroxyl group is important step for intracellular activation, the
4-amino group was usually selected to be modified prior to other sites to protect from
deamination and increase its stability because the 4-amino group could be
identified/attacked by cytidine deaminase during circulation (Moysan et al., 2013). In
previous studies, several successful cases of Gem prodrug with 4-amino modification
have been developed to promote the stability and uptake by introducing different
hydrophobic fragments. It was reported that the lipophilic amide prodrug of Gem (Tao
et al., 2012; Dubey et al., 2016), by conjugating of saturated/unsaturated fatty acids to
4-amino group of Gem, have displayed the enzymatic stability during circulation and
the promoted anti-tumor activities in MCF-7 cells.
Recently, tumor microenvironment (TME) has been gradually known as a key
2

basis for cancer occurrence and progression. It is usually comprised of surrounding
blood vessels, the extracellular matrix (ECM), signaling molecules and normal cells.
As one of the important ingredients in TME, fibroblast activation protein α enzyme
(FAPα) (Van Rite et al., 2011; Hui and Chen 2015; Lee et al., 2015), overexpressed on
the surface of tumor-associated fibroblasts (TAF) (Lai et al., 2012), in over 90% of
human epithelial carcinomas, has been increasingly studied as a remarkable target site
of tumor therapeutics (Teichgraber et al., 2015; Tansi et al., 2016). It has been verified
that FAPα could exclusively hydrolyze the N-terminal benzyloxycarbonyl (Z)-blocked
peptides, such as N-terminal carbobenzoxy-glycyl-L-proline-4-(Z)-blocked dipeptides
(Z-GP) (Collins et al., 2004; Aertgeerts et al., 2005), which could be used for the
design as the enzyme-activate prodrugs. Cai’s lab (Huang et al., 2011) designed a
FAPα-activated prodrug of Z-GP-DOX for tumor therapy without obvious cardio
toxic effect. It was reported that the Z-GP-DOX could selectively release DOX by the
recognition of FAPα enzymes in TME and enhance anti-cancer efficacy in 4T1
tumor-bearing mice.
Considering the advantages of 4-amino group modification and the characteristics
of Z-GP dipeptide specially cleaved by FAPα enzymes in TME, we intended to
synthesis a novel prodrug (Z-GP-Gem) with the conjugation of Z-GP peptide and
Gem for tumor therapy in this study. It was hypothesized that the prepared Z-GP-Gem
prodrug could not only improve the physicochemical properties and plasma half-life
in vivo by the introduction of Z-GP peptide at the 4-amino group of Gem, but also
enhance the chemotherapeutic efficiency in breast cancer treatment based on the
FAPα enzyme-activated release of parent Gem.
2. Materials and Methods
2.1 Materials
Gemcitabine hydrochloride was purchased from Saan Chemical Technology
(Shanghai) Co., Ltd. Acetonitrile (HPLC, gradient grade) was purchased from
Sigma-Aldrich. Methanol (99.9%, ACS/HPLC grade) was obtained from J&K
scientific LTD, China. Hoechst 33342 was purchased from Molecular Probes Inc.
3

(Oregon, USA). Recombinant human FAPα (rhFAPα) was purchased from R&D
Systems (Minneapolis, MN). Cells were cultured in RPMI-1640 medium (Macgene,
Beijing, China) supplemented with 10% fetal bovine serum (FBS), 100 units per mL
penicillin and 100 mg/mL streptomycin at 37 °C in a humidified atmosphere
containing 5% CO2. The cells for all experiments were in the logarithmic phase of
growth. BALB/c mice (6-8 weeks old) were purchased from Vital River Laboratory
Animal Center (Beijing, China) and were raised in a specific pathogen free (SPF)
environment. All care and handling of animals was in compliance with the approval of
Institutional Authority for Laboratory Animal Care of Peking University.
2.2 Synthesis of Z-GP-Gem
The Z-GP-Gem prodrug was designed and synthesized by conjugating the Z-GP
peptide to the 4’-amino group of gemcitabine according to the similar method
reported in several literatures (Dreiocker et al., 2010; May et al., 2016). The detailed
data of both operation conditions and structural confirms of all the intermediates were
depicted in support information (seen in section 1 in Support information). Briefly,
gemcitabine was firstly selectively protected by the TBDMS-Cl reagent under the
condition of imidazole in dry DMF solution to afford to the TBDMS-gem; meanwhile,
the Z-GP peptide was obtained by the amidation between Cbz-L-Glycin and L-Proline
under the condition of the NHS/DCC activation reagent; after purified by
chromatographic column, fixed-point amidation between the protected gem
intermediate and Z-GP peptide occurred by the condition of DCC/DMAP in 1,
4-dioxane solution. Finally, the Tetrabutylammonium fluoride reagent was induced to
de-protection to obtain the objective product Z-GP-Gem prodrug.
2.3 Immunofluorescence detection of FAPα expression
U87 cells and PC-3 cells were respectively inoculated subcutaneously of the
BALB/c mice while 4T1 cells were inoculated under the fourth breast pad. Each
mouse was inoculated with 3×106 corresponding cells. When the tumor volume
reached 500 mm3, tumor tissues were taken out after the mice were sacrificed.
4

Sequentially, frozen sections were prepared and stored at -80 °C. Sections were fixed
with 4% paraformaldehyde for 15 min at room temperature and blocked with 5% BSA
for 30 min. Then sections were incubated with FAPα rabbit monoclonal antibody
(1:100) at 4 ℃ overnight. After overnight incubation, slides were washed in PBS and
incubated in the dark for 2 h at 37 °C, with secondary antibody FITC-labeled goat
anti-rabbit IgG (H+L) (1:100). After washing in PBS, adding 0.1% Triton-PBS for 50
s, then tissues were stained with Hoechst 33342 (5 μg/mL) and Rhodamine phalloidin
(1:40) for 15 min at room temperature. After washing with PBS, the sections were
mounted in glycerol: PBS (V/V = 9:1) solutions, observe and quantify FAPα by using
STED laser confocal microscope.
2.4 Western blot detection of FAPα expression
Tumor tissues were obtained as 2.3. The tissues were homogenized in RIPA lysis
buffer with 1% proteinase inhibitor PMSF on ice, and then incubated on ice for 30
min and vortexed every 10 min. The total protein was gathered by centrifugation at
12,000 g for 30 min and their concentration was determined with BCA protein assay
kit. Total protein (30 μg) was loaded on 10% sodium dodecyl sulfate polyacrylamide
gel and electrophoresed at 80 mV for 30 min and 120 mV for 2 h. Then the proteins
were transferred to Polyvinylidene Fluoride (PVDF) membranes at 200 mA for 2 h,
and blocked with 5% skimmed milk on a shaker for 1 h. The membranes were
incubated with the FAPα monoclonal antibodies (1:1000) at 4 °C overnight, followed
by incubation with HRP-conjugated goat anti-rabbit antibodies (1:2000) at room
temperature for 2 h. Finally, the membranes were exposed using a GE Healthcare
Bio-Sciences AB System. β-Actin was used as endogenous control.
2.5 Degradation of Z-GP-Gem by tumor tissue homogenate or FAPα enzyme.
20 μL Z-GP-Gem solution (dissolved in mobile phase solution, 1 mg/mL) was
added to 1980 μL tumor homogenate, rat plasma and PBS, respectively. Briefly, all
samples were incubated at 37 °C for 48 h, and 100 μL homogenate, plasma and PBS
sample was removed at scheduled time (0.5, 2, 4, 8, 24, 48 h). After added with 900
5

μL acetonitrile, each sample was vortexed for 2 min and centrifuged at 1500 g for 10
min at 4 °C, after which the supernatant was transferred and evaporated to dryness
under nitrogen stream. Samples were re-dissolved in 100 μL mobile phase, 20 μL of
the solution was injected into an ODS-C18 reversed-phase column (4.6 mm × 250
mm, 5 μm particle size) with a pre-column (4.0 mm × 6.0 mm I.D.) from Agilent. For
Z-GP-Gem analysis, the column was eluted with acetonitrile/0.1% CF3COOH water
(30:70, V/V) (flow rate 1.0 mL/min) at 25 °C and the wavelength was set at 250 nm.
The drug concentrations were calculated from standard curves. Remaining
Z-GP-Gem % = Ct/C0×100%. Similarly, Z-GP-Gem (10 μg/mL) was incubated in
Tris-buffer (pH 7.4) with rhFAPα (5 μg/mL and 2 μg/mL) at 37 °C for 1, 2, 4, 8, 16,
24 h. Subsequently, Samples were collected at scheduled time points and the released
Gem was detected at the wavelength of 268 nm by HPLC.
2.6 Detection of Gem and Z-GP-Gem in tumor tissues by HPLC
The BALB/c mice were divided into two groups (8 mice each group): Z-GP-Gem
group and Gem group. Tumor cells (4T1) were inoculated respectively under the
fourth breast pad until the tumor volume reached 300 mm3, inject single dose Gem
(dissolved in saline) and Z-GP-Gem (dissolved in ethanol: saline (V/V) = 10:90, Gem
equivalent 20 mg/kg) through tail vein. At 4 h, the tumor tissues were removed, 50 μL
saline was added to each 10 mg of tumor tissue and homogenized on ice. Then, 900
μL acetonitrile was added to 100 μL sample, each sample was vortexed for 2 min and
centrifuged at 1500 g for 10 min at 4 °C, after which the supernatant was transferred
and evaporated to dryness under nitrogen stream. Samples were re-dissolved in 100
μL mobile phase, 20 μL of the solution was injected into an ODS-C18 reversed-phase
column (4.6 mm × 250 mm, 5 μm particle sizes) with a pre-column (4.0 mm × 6.0
mm I.D.). For Gem analysis, the mobile phase was consisted of methanol /0.05 M
ammonium acetate (pH 5.7, 9: 91, V/V) and the wavelength was set at 268 nm. The
drug concentrations were calculated from standard curves.
2.7 In vitro cytotoxicity evaluation by MTT
6

In vitro cytotoxicity of Z-GP-Gem, Z-GP-Gem co-cultured with tumor
homogenate and Gem was performed by MTT assay. Briefly, 4T1 and PC-3 cells were
seeded in 96-well plates at a density of 4000 cells per well. After 24 h proliferation,
cells were treated with 200 μL RPIM-1640 containing Z-GP-Gem, Z-GP-Gem
co-cultured with tumor homogenate and Gem at concentration of 100, 50, 10, 1, 0.5,
0.1, 0.01, 0.001 μM and RPIM1640 was used as a control. After 48 h incubation, after
addition of 20 μL MTT (5 mg/mL in PBS) to each well for 4 h, 150 μL DMSO was
added to dissolve the formazan crystals on a gyratory shaker for 10 min. Then,
absorbance values were read on a microplate reader at the wavelength of 490 nm. The
cell viability (%) was calculated according to the following formula: Cell viability (%)
= [OD490 (sample) - OD490 (solvent)/OD490 (control) - OD490 (solvent)] × 100 %,
where OD490 (sample) is the absorbance from the cells treated with various drugs,
OD490 (control) is that from the cells treated with RPIM-1640 and OD490 (solvent) is
that from the PBS. All the results were normalized to the untreated group.
2.8 Wound healing assay
The 4T1 cells were seeded in 6-well plates at a density of 200,000 per well. After
24 h proliferation, cells were treated with 2 mL RPIM-1640 containing Z-GP-Gem,
Z-GP-Gem co-cultured with tumor homogenate and Gem at concentration of 5 μM,
while RPIM-1640 was used as control. After 24 h of culture, wash with PBS and add
2 mL RPIM-1640. The photographs were taken under a microscope and analyze the
data by Imagine J software and calculate migration index, migration index =
(At-A0)/A0.
2.9 Cell uptake of Z-GP-Gem and Gem
4T1 cells were seeded in 6-well plates at 250,000 cells per well. After 24 h
proliferation, each plate of three wells were pre-incubation of dipyridamole (20
μg/mL) for 30 min. RPIM-1640 medium containing Z-GP-Gem (equivalent to 100
μg/mL Gem) was exposed to the cells and continually incubated for 4 h at 37 oC.
Similarly, RPIM-1640 medium containing Gem (100 μg/mL) was used to treat the
7

cells. After incubation, all samples were lysis in100 μL RIPA lysis buffer for 15 min at
room temperature, 100 μL methanol was added to rinse. The cells were harvested and
centrifuge at 12000 rpm at 4°C for 30 min. 100 μL of supernatant was obtained, after
added with 900 μL acetonitrile, each sample was vortexed for 2 min and centrifuged
at 1500 g for 10 min at 4 °C, after which the supernatant was transferred and
evaporated to dryness under nitrogen stream. The concentrations of Z-GP-Gem and
Gem in the corresponding cells were respectively determined by HPLC method as
described in Section 2.5.
2.10 Pharmacokinetic investigation
Pharmacokinetic studies were carried out on Sprague Dawley rats weighting 200
± 2 g, which were supplied by Experimental Animal Center of Peking University.
Rats were provided with standard food and water ad libitum, maintained on a light/
dark cycle under conditions of 25 ± 3 °C and 50% humidity. All care and handling of
animals were performed in accordance with the principles of Institutional Authority
for Laboratory Animal Care of Peking University. Rats were divided into two groups
(five animals each group) and treated with different drugs via tail vein injection. Rats
received Gem solution (dose of 28.5 μmol//kg, dissolved in saline) and Z-GP-Gem
solution (equivalent dose of 28.5 μmol//kg Gem, dissolved in the mixed solvent of
ethanol and saline (V/V= 1/9), respectively. Blood samples were collected from
retro-orbital plexus after 0.167, 0.5, 1, 2, 4, 6, 8 and 24 h of intravenous
administration. After centrifugation at 1500 g for 10 min, plasma was immediately
separated from blood samples. Plasma samples were treated with acetonitrile and
analyzed by HPLC, as described earlier.
2.11 In vivo anti-tumor study and safety evaluation
BALB/c mice bearing orthotropic 4T1 tumor model was established by
inoculating 2×106 4T1 cells under the fourth breast pad. When tumor size reached 100
mm3, the mice were randomly divided into 5 groups (6 animals for each group): (1)
Saline solution, (2) Gem solution (dissolved in saline, 20 mg/kg), (3) Z-GP-Gem
8

solution of low dose (L-Z-GP-Gem, dissolved in ethanol: saline (V/V) = 10:90, Gem
equivalent 10 mg/kg), (4) Z-GP-Gem solution of medium dose (M-Z-GP-Gem,
dissolved in ethanol: saline (V/V) = 10:90, Gem equivalent 20 mg/kg), (5) Z-GP-Gem
solution of high dose (H-Z-GP-Gem, dissolved in ethanol: saline (V/V) = 10:90, Gem
equivalent 30 mg/kg). Each dose of intravenous injection was given every three day
(total for 4 doses) through the tail vein. When the volume of tumor researched to 100
mm3, mice were injected by veil and the day giving first dose was defined as day 1. A
vernier caliper was used to measure the longest (L) and shortest (S) diameters of
tumors. The tumor size and body weight of each mouse were measured daily. Tumor
size was calculated according to the formula of V=0.5×L×S2. Before first dose and
12th day after the treatment began, 20 μL whole blood samples were collected from
the mice, blood cell analysis and hemoglobin detection were performed by
hemocytometer. Two days after the last injection, the animals were sacrificed to
harvest tumor tissues. Add Bouin’s solution to fix lung tissue for 4 h, and then number
the metastasis points.
2.12 Immunofluorescence assay of TAFs in tumor tissues
4T1 tumors were obtained from 2.11. The tumors of the Saline Gem and
Z-GP-Gem of medium dose were prepared as frozen sections and stored at -80 °C.
The sections were fixed with 4% paraformaldehyde for 15 min at room temperature
and blocked with 5% BSA for 30 min. Then sections were incubated with α-SMA
rabbit monoclonal antibody (1: 200) at 4 ℃ overnight. After overnight incubation,
slides were washed in PBS and incubated in the dark for 2 h at 37 °C, with secondary
antibody Texas Red-labeled goat anti-mouse IgG (H+L) (1:100). After washing in
PBS, tissues were stained with Hoechst 33342 (5 μg/mL) for 15 min at room
temperature. After washing with PBS, the sections were mounted in glycerol: PBS
(V/V = 9:1) solutions, stored in the dark. Observe α-SMA by using STED laser
confocal microscope.
2.13 Statistical analysis
9

All the results were expressed as mean ± standard deviation. For comparisons
between two groups, student’s t-test for independent means was used. For statistical
analysis between multiple groups, one-way analysis of variation (ANOVA) followed
by LSD multiple comparison test were applied. Statistical analysis was performed by
SPSS 16.0 software (SPSS Inc., Chicago). A value of p < 0.05 was considered as
statistically significant.
3. Results
3.1 Synthesis and characterization of Z-GP-Gem prodrug
The Z-GP-Gem prodrug was synthesized by conjugating the Z-GP dipeptide to
4’-amino group of Gem. Briefly, Z-GP dipeptide was firstly prepared by the
amidation between Cbz-protected glycine and L-proline (Liu et al., 2015), and then
the resulted Z-GP dipeptide was conjugated to 4’-amino group of gemcitabine with a
selective protection of 3,5-OH by TBDMS-Cl reagent, this amidation reaction was
used to prepare the protected conjugates, and the final Z-GP-Gem prodrug was
obtained by deprotection. The yield of final product was 57.6%. The chemical
structure and synthetic scheme have been presented in Figure 1. The characteristics of
Z-GP-Gem was confirmed as following: 1HNMR (400 MHz, d6-DMSO) δ 11.20 (s,
1H), 8.30 (d, 1H), 7.38 (m, 4H), 7.35 (d, 1H), 6.35 (m, 1H), 6.21 (m, 1H), 5.34 (m,
1H), 5.04 (s, 2H), 4.22 (m, 1H), 3.71 (m, 1H), 3.59 (m, 2H), 3.21 (m, 2H), 1.96 (m,
3H), 1.65 (m, 2H), 1.35 (m, 2H), 0.98 (t, 3H). 13CNMR (101 MHz, d6-DMSO) δ
173.98, 173.80, 167.89, 167.57, 160.89, 156.91, 137.57, 128.80, 128.22, 128.13,
99.97, 96.07, 95.06, 72.40, 65.80, 59.04, 47.70, 45.86, 44.66, 43.04, 42.87, 29.05,
24.81, 22.29. Besides, the detailed data of all the correlative mediate compounds have
been analyzed by NMR spectrums (seen in Support information).
3.2 Expression levels of FAPα enzyme in different tumor tissues
To select a suitable tumor model for evaluating the biological effects of FAPα
enzyme-activated Z-GP-Gem, the expression levels of FAPα protein in three tumor
tissues came from 4T1, U87 and PC-3 xenografed tumor-bearing mice were
10

individually detected by the immunofluorescence assay. As seen from qualitative
fluorescence images (Figure 2A) and quantitative analysis (Figure 2B), it was shown
that a relative higher fluorescence intensity of FAPα protein was observed in 4T1
tumor tissues as those in U87 and PC-3 tissues. Also, compared to the content of
FAPα protein in PC-3 tissues detected by the western-blot assay, a relative higher
FAPα protein content was displayed in the 4T1 tumor tissues (Figure 2C). Based on
the above results, 4T1 tumor model was selected for evaluating the biological effects
of FAPα enzyme-activated Z-GP-Gem.
3.3 Degradation of Z-GP-Gem by rhFAPα enzyme
The stabilities of Z-GP-Gem incubated in PBS, blood plasma and tumor
homogenate for 48 h were investigated, and the remained concentrations of
Z-GP-Gem at predetermined time points were detected by HPLC analysis (Figure
3A). Obviously, the fastest degradation rate of Z-GP-Gem was found in the tumor
homogenate group, while the lowest degradation shown in PBS group. At 4 h point,
the remained Z-GP-Gem in 4T1 tumor tissue was rapidly decreased to 50%,
remarkably lower than other two groups (73% in plasma and 93% in PBS) at the same
time. Moreover, 96% of Z-GP-Gem was degraded when it was incubated in 4T1
tumore tissue at 24 h point and no detectable Z-GP-Gem was remained at 48h point,
while the remained Z-GP-Gem were of 68% in plasma and 53% in PBS at 48h point,
respectively. Moreover, the released Gem from Z-GP-Gem activated by rhFAPα
enzyme was also detected at different time points by HPLC method (Figure 3B).
With time extension, more and more Gem was released from Z-GP-Gem activated by
rhFAPα enzyme with a concentration-dependent manner, at 24h point, the
concentrations of released Gem were reach to 50% and 63% under the concentration
of 2 μg/mL and 5 μg/mL rhFAPα, respectively.
3.4 Cytotoxicities of Z-GP-Gem before and after FAPα enzyme-activation
The cytotoxicities of Gem, Z-GP-Gem before and after FAPα enzyme-activated
in both 4T1 and PC-3 cell lines were assessed by using MTT assay. As seen in Figure
11

4A, it was shown that Z-GP-Gem exhibited a lower cytotoxicity against 4T1 cells
than that of Gem at all equivalent concentrations, even at a high concentration of 10
μM the survival rate was still to 70% above, while Gem displayed obvious
cytotoxicity against 4T1 cells. After pretreating with 4T1 tumor tissue homogenate,
the FAPα enzyme-activated Z-GP-Gem (+FAPα) displayed a dramatic increase of
cytotoxicity as similar as that of parent Gem. In addition, a similar trend of
cytotoxicity (IC50 value: untreated Z-GP-Gem > FAPα enzyme-activated Z-GP-Gem >
Gem) was observed in PC-3 cell line (Figure 4B).
3.5 In vitro anti-migration activity of FAPα enzyme-activated Z-GP-Gem
In vitro anti-migration abilities of Gem, Z-GP-Gem and FAPα enzyme-activated
Z-GP-Gem against 4T1 cells was investigated by wound healing assay (Figure 5).
From the quantitative results in Figure 5B, it was shown that the Z-GP-Gem prodrug
(migration index: 0.40) exhibited a relatively weaker anti-migration ability against
4T1 cells compared to that of Gem (migration index: 0.31), while FAPα
enzyme-activated Z-GP-Gem (migration index: ~0.32) had the similar anti-migration
ability as Gem.
3.6 Cell uptake of Z-GP-Gem
The cell uptake of Gem, Z-GP-Gem and FAPα enzyme-activated Z-GP-Gem in
4T1 cells before and after nucleoside transporter inhibition by dipyridamole were
detected by HPLC method. As seen in Figure 6, there was no obvious difference of
intracellular concentration between Z-GP-Gem and Gem at 4h time point. In the cells
pre-treated with dipyridamole as the hENT inhibitor (Takano et al., 2010), the
intracellular concentration of Gem was sharply decreasing to ~10% of that untreated
with dipyridamole, while the concentration of Z-GP-Gem was still remaining about
70% of the corresponding group untreated with dipyridamole. These results
indicated that most of Z-GP-GEM prodrug enters 4T1 cells in a nucleoside
transporter-independent manner, in contrast with nucleoside transporter-dependent
cellular uptake of Gem (Gesto et al., 2012).
12

3.7 In vivo pharmacokinetics
Pharmacokinetic behaviors of both Z-GP-Gem and Gem in plasma were
investigated after a single intravenous injection of Z-GP-Gem or unmodified Gem at a
respectively dose of 28.5 μmol/kg in male Sprague Dawley rats. As shown in Figure
7, the concentration of Gem released from Z-GP-Gem prodrug in plasma was much
lower than that of parent Gem within 8 hours in this experiment (p < 0.01). At 24 h
time point, there still has the detectable level of whether Z-GP-Gem or Gem released
from Z-GP-Gem group, while no detectable drug was seen in parent Gem group. The
pharmacokinetic parameters were shown in Table 1. The area under the curve
(AUC0-24) of Z-GP-Gem prodrug was 169.00 μmol/L*h, significantly higher than that
of parent Gem group (101.00 μmol/L*h). Also, the t1/2α and t1/2β of Z-GP-Gem
prodrug were 1.53 h and 5.25 h respectively, significantly longer than those of
unmodified Gem (0.77 h and 3.24 h). Moreover, due to the lower releasing
concentration of Gem from Z-GP-Gem group, the AUC0-24 (35.70 μmol/L*h) and t1/2α
(0.34 h) values were obviously lower than those of parent Gem, which implied that
the stability in vivo of Z-GP-Gem was higher than that of Gem. These
pharmacokinetic results suggested that a prolonged plasma circulation of Gem is
beneficial to improve tumor accumulation and therapeutic efficiency of gemcitabine
in vivo.
To verify the different accumulation of Z-GP-Gem prodrug and parent Gem in
4T1 tumor tissue, two groups of BALB/c mice bearing 4T1 tumor model were given
single dose intravenous injection of Z-GP-Gem and Gem at an equivalent Gem
concentration of 20 mg/kg, respectively. The tumor tissues were extracted at 4 h of
post-injection and homogenized for HPLC analysis of gemcitabine. As shown in
Figure 9C, it was displayed that the intra-tumoral concentration of gemcitabine in the
tumor tissue treated with Z-GP-Gem group was 2.69 μmol/g, about 2.7-fold higher
than that of Gem group (1.009 μmol/g).
3.8 In vivo tumor growth inhibition
13

The anti-tumor activities of Z-GP-Gem and Gem were evaluated on BALB/c mice
orthotropic implanted mouse breast 4T1 tumor model, following intravenous injection
of different formulations via tail vein for a total four treatments (each of 3 days
interval) starting from the tumors size of approximately 100 mm3. As shown in
Figure 8A, the medium dose of Z-GP-Gem prodrug (41.9 mg/kg, equivalent to 20
mg/kg of Gem) displayed a stronger inhibition of tumor growth than equivalent dose
of parent Gem (**p < 0.01). Among three doses of Z-GP-Gem groups, the inhibition
activities on tumor growth were in dose-dependent manner as follows: H-Z-GP-Gem >
M-Z-GP-Gem > L-Z-GP-Gem. Moreover, the enhanced inhibition effects on tumor
growth of Z-GP-Gem were further confirmed by the images of ex vivo tumor tissues
(Figure 8B) and average tumor weight at the end-point (Figure 8C).
3.9 In vivo tumor metastasis inhibition
As seen from the images in Figure 9A, the visible tumor metastasis points
(marked by black arrow) in the lung of mouse treated with the Z-GP-Gem (medium
dose, equivalent to 20 mg/kg of Gem) were significantly less than those treated with
the parent Gem group. The quantitative data (Figure 9B) further confirmed the tumor
metastasis inhibition activities of Z-GP-Gem were in a dose-dependent manner as
follows: H-Z-GP-Gem > M-Z-GP-Gem > L-Z-GP-Gem.
3.10 In vivo depletion of TAFs in tumor
α-smooth muscle actin (α-SMA) has been known as an important biomarker of
TAFs. To investigate the inhibiton effects of Z-GP-Gem prodrug on TAFs in tumor,
α-SMA monoclonal antibody labeled by Texas Red was used to trace the TAFs by
immunofluorescence method. As seen in Figure 11, the fluorescence intensity of
α-SMA in tumor tissue treated with Z-GP-Gem prodrug was significantly lower than
that of Gem group. It was suggested that the enhanced tumor accumulation of
Z-GP-Gem could exhibit stronger depletion ability on TAFs than Gem.
3.11 In vivo safety
14

The indications of body weight, hemoglobin and red blood cells were measured
for evaluating the in vivo safety of Z-GP-Gem during the period of administration. As
shown in Figure 10A, no obvious change of body weight was observed in all
Z-GP-Gem groups, while a mildly decrease of body weight was observed in Gem
group started from 5th day post-injection. Moreover, there was no obvious changes
happened in the level of RBC (Figure 10B) and HGB (Figure 10C) in low and
medium doses of Z-GP-Gem groups, except for a weak change of RBC and HGB in
the high-dose Z-GP-Gem group. These results demonstrated the Z-GP-Gem exhibited
a better biosafety in vivo.
4. Discussions
Although gemcitabine was used as the first-line therapeutic agent for various
tumors, the rapid deamination by CDA enzyme in plasma and the limited uptake in
hENT-deficient patients still impeded its clinical application (Gesto et al., 2012).
Furthermore, the non-selective tissue accumulation of Gem further limited its
anti-tumor activity and in vivo biosafety (Moysan et al., 2013). In our study, a
FAPα-activated prodrug of Gem (Z-GP-Gem), prepared by conjugating Z-GP
dipeptide to 4-amino group of Gem, was firstly developed for improving tumor
therapy and biosafety (Figure 1).
By the conjugation of Z-GP dipeptide to 4-amino group of Gem, it was expected to
promote the in vivo stability and tumor accumulation of gemcitabine during clinical
application. From the results of body weights (Figure 10A) and hemoglobin analysis
(Figure 10B and 10C), it was shown that Z-GP-Gem prodrug displayed a well in vivo
safety compared to Gem. Besides, in vitro cytotoxicity of Z-GP-Gem was
significantly lower than that of Gem in both 4T1 (Figure 4A) and PC-3 (Figure 4B)
cells. Intrinsically, the superior biosafety was probably owing to the chemical stability
of Z-GP-Gem in plasma. Similar as previous reports (Dasari et al., 2013; Maiti et al.,
2013), the amidation of 4-amino group of Gem by conjugating the Z-GP peptide
could protect deamination from the recognition of CDA enzyme in plasma.
Furthermore, in vivo pharmacokinetic results showed a prolonged plasma circulation
15

of Gem released from Z-GP-Gem group compared to parent Gem group (Figure 7). A
relative higher accumulation of gemcitabine was found in the tumor tissues treated
with Z-GP-Gem prodrug than that of parent Gem (Figure 9C), suggesting that the
Z-GP-Gem prodrug is beneficial to improve tumor accumulation of released
gemcitabine in vivo.
More interestingly, most of Z-GP-Gem prodrug enters 4T1 cells in a nucleoside
transporter-independent manner, in contrast with nucleoside transporter-dependent
cellular uptake of Gem (Figure 6), it was probably due to the improved lipophilicity
of gemcitabine by the modification of Z-GP dipeptide, which could enhance
intracellular uptake and improve tumor accumulation of gemcitabine. The lipophilic
nature of Z-GP-Gem prodrug has been confirmed by its logP value (Table S5)
through traditional shake-flask experiments, which can support the capability of the
prodrug passively crossing the lipid membrane of the cell.
An ideal prodrug should not only exhibit good chemical stability but also be
enzyme-activated back to parent drug to exhibit its biological effects. In this study,
compared to the stabilities of Z-GP-Gem prodrug in PBS and plasma medium, a
relative fast degradation of prodrug was found in the tumor homogenates, the
remained Z-GP-Gem in 4T1 tumor tissue was rapidly decreased to 50% and more
than 96% after 4 h and 24 h incubation with tumor homogenate, and no detectable
data at 48 h (Figure 3A); Besides, the increased Gem release from Z-GP-Gem by
rhFAPα cleavage was also observed with time extension (Figure 3B); Additionally, in
vitro cytotoxicity evaluation on both 4T1 and PC-3 cells (Figure 4) indicated that
Z-GP-Gem exhibited a lower cytotoxicity against cells than that of Gem at all
equivalent concentrations, while the FAPα enzyme-activated Z-GP-Gem (pretreated
with 4T1 tumor homogenates) displayed a dramatic increase of cytotoxicity as similar
as that of parent Gem. It is worthy believing that FAPα enzyme-activated
Z-GP-Gem showed a remarkable tumor cell growth inhibition after 48 h incubation at
a concentration of 10 μM (Figure 4). This significant difference between the
cytotoxcities of Z-GP-Gem before and after pretreated with 4T1 tumor homogenates
was probably due to the improved release of gemcitabine from prodrugs by FAPα
16

enzyme-activated cleavage and thus improve cell growth inhibition effects as same as
parent drug. Similar trend was also seen in the evaluation of in vitro anti-migration
effects, FAPα enzyme-activated Z-GP-Gem also showed a similar anti-migration
ability as Gem (Figure 5). Moreover, after the mice given single dose intravenous
injection of Z-GP-Gem and Gem at an equivalent Gem concentration of 20 mg/kg
respectively, the significantly higher tumor accumulation of gemcitabine was found in
Z-GP-Gem prodrug than that of parent Gem indicated that enhanced permeation and
retention of Z-GP-Gem prodrugs in tumor tissues were mostly transferred to free
gemcitabine by the FAPα enzyme-activated release (Figure 9C).
Actually, improving in vivo anti-tumor activity and biosafety are always the
ultimate goal of prodrug modification. In this study, the stronger inhibition effects of
Z-GP-Gem prodrugs on tumor growth and lung metastasis were observed in the mice
bearing 4T1 tumor model (Figure 8 and 9). For tumor growth inhibition, it was
demonstrated that medium-dose Z-GP-Gem exhibited significantly higher antitumor
activity in the 4T1 breast cancer compared with parent Gem at same dose (Figure 8).
The enhanced therapeutic efficiency of Z-GP-Gem prodrug could be explained as
follows: (1) the enhanced cellular uptake of Z-GP-Gem with a nucleoside
transporter-independent manner; (2) the improved stability of Z-GP-Gem by
protecting gemcitabine from deamination; (3) enhanced tumor accumulation of
gemcitabine from Z-GP-Gem prodrug by FAPα enzyme-activated released.
As well known, tumor-associated fibroblasts (TAFs) are the important component
of the tumor microenvironment (TME) (Wang et al., 2018). They contribute to tumor
growth and invasiveness by remodeling the extracellular matrix, secreting various
soluble factors and suppressing immune responses. Therapies directed at TAFs can be
considered as a potential approach to anti-cancer treatment (Duyverman et al., 2012).
α-smooth muscle actin (α-SMA) has been known as an important biomarker of TAFs.
In this study, the inhibition effects of Z-GP-Gem prodrug on TAFs in tumor were
investigated by the immunofluorescence method with α-SMA monoclonal antibody.
As seen in Figure 11, the fluorescence intensity of α-SMA in tumor tissue treated
with Z-GP-Gem prodrug was significantly lower than that of Gem group. It was
17

indicated that Z-GP-Gem could exhibit a stronger depletion ability on TAFs. So why
could the prodrug damage TAFs stronger more than Gem Group? We presumed a
hypothesis as below: a relative higher accumulation of Z-GP-Gem in tumor tissue
than that of parent Gem was contributed by its distinct pharmacokinetic behavior in
circulation and permeability in tumor tissues (Figure 9C). As the Z-GP-Gem was
recognized and cleaved by FAPα proteins which were overexpressed on the surface of
TAFs (Lai et al., 2012), more Gem released from Z-GP-Gem exerted its cytotoxic on
TAFs and tumor cells, thereby displayed the distinctive depletion ability on TAFs. The
damage of tumor environment formed by TAFs could further facilitate Z-GP-Gem and
the released Gem entering into the deep of tumor tissue and improving tumor cell
growth inhibition. Therefore, it was noteworthy that the improved inhibition of tumor
growth and depletion of TAFs caused by the large accumulation of Z-GP-Gem
prodrugs in tumor tissue probably resulted in the inhibition effects on lung metastasis.
5. Conclusions
In this study, the FAPα-activated prodrug of Gem was firstly designed and
synthesized with a convenient operation. The modification of Z-GP dipeptide at
4-amino group of Gem not only prevented fast metabolism from recognition of CDA
enzyme but also improved cellular uptake with a hENT-independent manner.
Moreover, Z-GP-Gem exhibited the distinct depletion ability of TAFs and improved
the released Gem accumulation in 4T1 tumor tissue with a specific FAPα
enzyme-activated. All these results demonstrated the FAPα-activated prodrug
Z-GP-Gem would be a promising and potential prodrug for tumor therapy.
6. Acknowledgements
The authors acknowledge financial support from the projects of National Natural
Science Foundation of China (Grant No. 81473158, 81690264 and 81773650), the
New Drug R&D program of China (Grant No. 2018ZX09721003-004) and the
18

Opening Project of Key Laboratory of Drug Targeting and Drug Delivery System,
Ministry of Education (Sichuan University).
7. References
Abdayem, P., Ghosn, M., Valero, V., Walters, R., Arun, B., Murray, J.L., 2014. Phase I and II study
of gemcitabine and vinorelbine in heavily pretreated patients with metastatic breast cancer
and review of the literature. J. Cancer 5(5), 351-359.
Aertgeerts, K., Levin, I., Shi, L., Snell, G.P., Jennings, A., Prasad, G.S., 2005. Structural and
kinetic analysis of the substrate specifcity of human fbroblast activation protein alpha. J. Biol.
Chem. 280, 19441-19444.
Collins, P.J., McMahon, G., O'Brienm P., 2004. Purifcation, identifcation and characterisation of
seprase from bovine serum. Int. J. Biochem. Cell Biol. 36, 2320-2333.
Dasari, M., Acharya, A.P., Kim, D., Lee, S., Rhea, J., 2013. H-gemcitabine: a new gemcitabine
prodrug for treating cancer. Bioconjug. Chem. 24(1), 4-8.
Dreiocker, F., Muller, M.Q., Sinz, A., Schafer, M., 2010. Collision-induced dissociative chemical
cross-linking reagent for protein structure characterization: applied Edman chemistry in the
gas phase. J. Mass Spectrom. 45(2), 178-189.
Dubey, R.D., Saneja, A., Gupta, P.K., Gupta, P.N., 2016. Recent advances in drug delivery
strategies for improved therapeutic efficacy of gemcitabine. Eur. J. Pharm. Sci. 93, 147-162.
Duyverman, A.M., Steller, E.J., Fukumura, D., Jain, R.K., Duda, D.G., 2012. Studying primary
tumor-associated fibroblast involvement in cancer metastasis in mice. Nat. Protoc. 7(4),
756-762.
Fan, M., Liang, X., Li, Z., Wang, H., Yang, D., Shi, B., 2015. Chlorambucil gemcitabine conjugate
nanomedicine for cancer therapy. Eur. J. Pharm. Sci. 79, 20-26.
Gesto, D.S., Cerqueira, N.M., Fernandes, P.A., Ramos, M.J., 2012. Gemcitabine: a critical
nucleoside for cancer therapy. Curr. Med. Chem. 19(7), 1076-1087.
Hoang, N.T., Kadonosono, T., Kuchimaru, T., Kizaka-Kondoh, S., 2016. Hypoxia-inducible
factor-targeting prodrug TOP3 combined with gemcitabine or TS-1 improves pancreatic
cancer survival in an orthotopic model. Cancer Sci. 107(8), 1151-1158.
Hou, M., Xue, P., Gao, Y.E., Ma, X., Bai, S., Kang, Y., 2017. Gemcitabine-camptothecin
conjugates: a hybrid prodrug for controlled drug release and synergistic therapeutics.
Biomater Sci. 5(9), 1889-1897.
Huang, S., Fang, R., Xu, J., Qiu, S., Zhang, H., Du, J., 2011. Evaluation of the tumor targeting of a
FAP alpha-based doxorubicin prodrug. J. Drug Target. 19(7), 487-496.
Hui, L., Chen, Y., 2015. Tumor microenvironment: Sanctuary of the devil. Cancer Lett. 368(1),
7-13.
Karampelas, T., Skavatsou, E., Argyros, O., Fokas, D., Tamvakopoulos, C., 2017. Gemcitabine
based peptide conjugate with improved metabolic properties and dual mode of efficacy. Mol.
Pharm. 14(3), 674-685.
Lai, D., Ma, L., Wang, F., 2012. Fibroblast activation protein regulates tumor-associated
fibroblasts and epithelial ovarian cancer cells. Int. J. Oncol. 41(2), 541-550.
Lee, S., Lee, Y., Kim, W., Nam, J., Jeong, S., Yoo, J.W., 2015. Evaluation of glycine-bearing
19

celecoxib derivatives as a colon-specific mutual prodrug acting on nuclear factor-kappaB, an
anti-inflammatory target. Drug Des. Devel. Ther. 9, 4227-4237.
Liu, G.Q., Mo, E.P., Wang, X.X., Wu, N., Liu, F.L., Yuan, W.Q., 2015. Plasma pharmacokinetic
and heart distribution studies of Z-GP-EPI, a hypocardiotoxic prodrug of epirubicin. Tropic J.
Pharm. Res. 14(5), 899-905.
Maiti, S., Park, N., Han, J.H., Jeon, H.M., Lee, J.H., Bhuniya, S, 2013.
Gemcitabine-coumarin-biotin conjugates: a target specific theranostic anticancer prodrug. J
Am. Chem. Soc. 135(11), 4567-4572.
May, J.P., Undzys, E., Roy, A., Li, S.D., 2016. Synthesis of a gemcitabine prodrug for remote
loading into liposomes and improved therapeutic effect." Bioconjug. Chem .27(1), 226-237.
Moysan, E., Bastiat, G., Benoit, J.P., 2013. Gemcitabine versus modified gemcitabine: a review of
several promising chemical modifications. Mol. Pharm. 10(2), 430-444.
Muggia, F., Diaz, I., Peters, G.J., 2012. Nucleoside and nucleobase analogs in cancer treatment:
not only sapacitabine, but also gemcitabine. Expert Opin. Investig. Drugs 21(4), 403-408.
Pratt, S.E., Durland-Busbice, S., Shepard, R.L., Heinz-Taheny, K., Iversen, P.W., Dantzig, A.H.,
2013. Human carboxylesterase-2 hydrolyzes the prodrug of gemcitabine (LY2334737) and
confers prodrug sensitivity to cancer cells. Clin. Cancer Res. 19(5), 1159-1168.
Takano, M., Kimura, E., Suzuki, S., Nagai, J., Yumoto, R., 2010. Human erythrocyte nucleoside
transporter ENT1 functions at ice-cold temperatures. Drug Metab. Pharmacokinet. 25(4),
351-360.
Tansi, F.L., Ruger, R., Bohm, C., Kontermann, R.E., Teichgraeber, U.K., Fahr, A., 2016. Potential
of activatable FAP-targeting immunoliposomes in intraoperative imaging of spontaneous
metastases. Biomaterials 88, 70-82.
Tao, X.M., Wang, J.C., Wang, J.B., Feng, Q., Gao, S.Y., Zhang, L.R., 2012. Enhanced anticancer
activity of gemcitabine coupling with conjugated linoleic acid against human breast cancer in
vitro and in vivo. Eur. J. Pharm. Biopharm. 82(2), 401-409.
Teichgraber, V., Monasterio, C., Chaitanya, K., Boger, R., Gordon, K., Dieterle, T., 2015. Specific
inhibition of fibroblast activation protein (FAP)-alpha prevents tumor progression in vitro.
Adv. Med. Sci. 60(2), 264-272.
Van Rite, B.D., Lazrak, Y.A., Pagnon, M.L., Palwai, N.R., Neves, L.F., McFetridge, P.S., 2011.
Enzyme prodrug therapy designed to target L-methioninase to the tumor vasculature. Cancer
Lett. 301(2), 177-184.
Wang, J.J., Lei, K.F., Han, F., 2018. Tumor microenvironment: recent advances in various cancer
treatments. Eur. Rev. Med. Pharmacol. Sci. 22(12), 3855-3864.
Zhao, C., Li, Y., Qin, Y., Wang, R., Li, G., Sun, C., 2013. Pharmacokinetics and metabolism of
SL-01, a prodrug of gemcitabine, in rats. Cancer Chemother. Pharmacol. 71(6), 1541-1550.
Figures and Tables
20

Figure 1. The chemical structure of Z-GP-Gem prodrug and its synthetic route.
Figure 2. Expression of FAPα in different tumor tissues. (A) The expression of FAPα in 4T1,
PC-3 and U87 tumor tissues by CLSM image (green: FAPα antibody labeled by FITC; red:
Cytoskeleton labeled by Rhodamine-modified Phalloidin; blue: nuclei labeled by Hoechst 33342).
(B) Fluorescence intensity of FAPα in 4T1, PC-3 and U87 tumor tissues. **p < 0.01. (C)
Expression of FAPα by western blot in both 4T1 and PC-3 tumor tissues.
21

Figure 3. Degradation of Z-GP-Gem in different media and release of Gem in rhFAPα buffer. The
remaining Z-GP-Gem ratio of Z-GP-Gem (10 μg/mL) incubated with 4T1 tumor tissue
homogenate, plasma and PBS within 48 h analyzed by HPLC (A); and the released Gem from
Z-GP-Gem incubated in Tris-buffer (pH 7.4) with rhFAPα protein within 24 h detected by HPLC
(B). Data are presented as mean ± SD (n = 3).
Figure 4. Survival rates (%) of both 4T1 (A) and PC-3 (B) cells after incubated with Gem,
Z-GP-Gem and Z-GP-Gem co-cultured with 4T1 tumor homogenate for 4 hours at 48 h points.
Data are presented as mean ± SD (n = 4).
22

Figure 5. (A) Evaluation of anti-migration ability of Gem, Z-GP-Gem and Z-GP-Gem co-cultured
with tumor homogenate for 4 hours in 4T1 cells by wound healing assay (magnification: 100×).
(B) Quantitative analysis of 4T1 cell migration ability at 24 h point using Image J software. The
data was shown as mean ± SD (n = 3). *p < 0.05.
Figure 6. Uptake pathway of Z-GP-Gem and Gem in 4T1 cells. Intracellular concentration of
Z-GP-Gem and Gem after treating with Z-GP-Gem and Gem of 381 μmol/L at 4 h in 4T1 cells
with or without pre-incubation of dipyridamole (20 μg/mL) for 30 min. The data was shown as
mean ± SD (n = 3). ***p <0.001, *p <0.05.
23

5 0
4 0
3 0
2 0
1 0
0
G e m o f Z -G P -G e m G ro u p
Z -G P -G em
G em
0
5
1 0
1 5
2 0
2 5
T im e (h )
Figure 7. Plasma concentration-time curves of Gems and Z-GP-Gem in rats after intravenous
injection at a dose of 28.5 μmol/kg. The values are the mean ± SD (n = 5). Gem was dissolved in
saline, while Z-GP-Gem in the mixture of 10% ethanol/saline.
24

Figure 8. In vivo anti-tumor effects of Z-GP-Gem and Gem on BALB/c mice bearing in 4T1
tumor. (A) Tumor growth curve after treated with different groups for a total 4 doses via tail vein
injection (n = 6). (Dosage: Gem: 20 mg/kg; L-Z-GP-Gem: Gem equivalent 10 mg/kg;
M-Z-GP-Gem: Gem equivalent 20 mg/kg; H-Z-GP-Gem: Gem equivalent 30 mg/kg). (B and C)
The weight of ex vivo tumor tissues at the experimental endpoint. ***p < 0.001, **p < 0.01.
25

Figure 9. (A) Metastasis points of lung at the experimental endpoint (n = 3). (B) Quantitative
analysis of metastasis point of lung at the experimental endpoint. (n = 6). The data was shown as
mean ± SD. *P < 0.05. (C) Gemcitabine concentration (n = 4) in BALB/c mice tumor tissue after
4 h of one dose (20 mg/kg) tail vein injection of Gem and Z-GP-Gem. **p < 0.01.
26

Figure 10. (A) Changes of body weight (n = 6). (B)The changes of red blood cell (RBC) (n = 5)
(C) and hemoglobin (HGB) in blood samples collected on day 1 and day 12 (n = 5). The data was
shown as mean ± SD. **p < 0.01, *p < 0.05.
27

Figure 11. TAF depletion of Z-GP-Gem and Gem groups. Immunofluorescence of TAF on the
tumor sections after treated with different groups for a total 4 doses via tail vein injection. (red:
α-SMA monoclonal antibody labeled by Texas Red; blue: nuclei labeled by Hoechst 33342).
(Dosage: Gem: 20 mg/kg; Z-GP-Gem: Gem equivalent 20 mg/kg).
28

Graphical Abstract. Schematic illustration of the enhanced anti-tumor efficiency of Z-GP-Gem
prodrug by FAPα-activation.
Table 1. Pharmacokinetic parameters of Z-GP-Gem (equal to 28.5μmol/kg Gem) and parent Gem
(28.5μmol/kg) after single dose intravenous injection in rats (n=5)
Parameter
t1/2α (h)
Z-GP-Gem
1.63
Released Gem
0.34
Parent Gem
0.77
t1/2β (h)
5.25
7.56
3.24
AUC0-24 (μmol/L*h)
169.00
35.70
101.00
29