Drug Delivery System for the Anticancer Drug Paclitaxel Using Lipocalin-Type Prostaglandin D Synthase Conjugated to a Tumor-Targeting Peptide.

Introduction
Recent drug discovery research has focused
heavily on identifying
new drug candidate compounds using combinatorial chemistry, which can be used to synthesize numerous drug
compounds in a short time, and high-throughput screening,
−

which selects compounds that show high drug efficacy against specific
targets. However, because these methods emphasize pharmacological
activity against a target, the candidate compounds selected tend to
have relatively low water solubility,
,
which is a
major obstacle in the development of new drugs. Furthermore, the number
of receptor proteins with drug-binding pockets that can be targeted
by small-molecule drugs is limited to approximately 400, many of which have already been explored, resulting
in a significant reduction in the efficiency of new therapeutic drug
development. As such, increased attention
has focused on targeting protein–protein interactions (PPIs)
in order to improve the efficiency of drug discovery. However, the physicochemical properties of PPI interfaces
differ from those of conventional protein–ligand binding sites;
thus, candidate compounds tend to have high molecular weight and poor
water solubility. Poorly water-soluble
drugs generally aggregate in aqueous solutions to form a precipitate,
which leads to a decrease in bioavailability and limits drug efficacy
and therapeutic applications. Therefore, in recent years, more research
has been directed toward drug delivery systems (DDSs) using liposomes, micelles, dendrimers, and proteins
,
as carriers.
One example of a PPI-targeting drug is the anticancer agent paclitaxel
(PTX) (Figure
A).
PTX, a diterpene alkaloid with a molecular weight of 854 that was
originally isolated from the bark of the Pacific yew (Taxus brevifolia), induces
apoptosis of cancer cells by inhibiting microtubule depolymerization. PTX exhibits strong antitumor activity against
a wide variety of solid cancers, including those of the breast, lung,
and ovaries, but it causes serious side
effects, such as myelosuppression and peripheral neuropathy. Because PTX is extremely water insoluble, the
commercially available drug Taxol is prepared by dissolving PTX in
Cremophor EL (CrEL)/absolute ethanol (1:1 [v/v]). However, CrEL is also highly toxic and elicits strong acute
hypersensitivity and neurotoxicity. In
addition, CrEL alters the blood distribution of PTX as a result of
the compound being incorporated into circulating CrEL micelles, thereby
reducing the proportion of free drug available for distribution to
cancer cells. Therefore, a novel cancer-targeting
DDS for PTX is urgently needed to maximize the effectiveness of the
drug while minimizing side effects to healthy tissues.
Among protein-based DDS platforms, endogenous carrier
proteins
such as human serum albumin (HSA), ferritin, and transferrin have
been widely explored for the delivery of hydrophobic anticancer agents.
−

Although these systems can improve solubility and circulation, drug
loading often relies on multiple surface binding sites (HSA) or multimeric
assemblies with complex metal/ligand binding (ferritin, transferrin),
which can limit formulation homogeneity, complicate recombinant production,
and lead to variable biodistribution due to competition with endogenous
ligands or receptors. Thus, there remains a need for alternative protein
carriers that provide a well-defined hydrophobic cavity, simple and
reproducible loading, and straightforward engineering of targeting
ligands.
Our research has focused on developing a DDS using
lipocalin-type
prostaglandin D synthase (L-PGDS) as a hydrophobic drug carrier.
−

In contrast to HSA, which presents multiple heterogeneous surface
binding sites, L-PGDS offers a single, structurally well-defined hydrophobic
cavity within a monomeric β-barrel, enabling more homogeneous
complex formation and better control of the drug-to-carrier ratio.
Unlike multimeric carriers such as ferritin and transferrin, L-PGDS
is a small, soluble protein that can be produced recombinantly and
loaded with hydrophobic drugs simply by mixing in aqueous solution
without organic solvents. L-PGDS belongs to the lipocalin family,
a group of secretory lipid-transporting proteins, and has a typical β-barrel structure consisting of
eight antiparallel β-strands (Figure
B). We previously
reported that L-PGDS binds a wide range of hydrophobic small molecules,
with molecular weights in the range of approximately 270 to 780, to
a hydrophobic pocket inside the β-barrel structure. Exploiting this property, we found that L-PGDS
is a useful DDS carrier for a wide variety of poorly water-soluble
drugs, such as diazepam (molecular weight: 285), 7-ethyl-10-hydroxy-camptothecin
(molecular weight: 392), dipyridamole (molecular weight: 505), and
telmisartan (molecular weight: 515).
−

Compared with other DDSs such
as liposomes and micelles, L-PGDS can markedly enhance the solubility
of poorly water-soluble drugs simply by mixing the drug suspension
with the L-PGDS solution, without the need for any organic solvents.
However, the inability of L-PGDS to target cells of interest represents
a major impediment to its use as a drug delivery vehicle.
Selection
of an appropriate targeting technology, which delivers
the drug of interest specifically to diseased areas, is a key consideration
in the development of useful DDSs. The use of an optimal targeting
technology reduces the number of necessary administrations, improves
drug efficacy, and minimizes side effects. Many anticancer drugs inhibit
DNA replication and protein synthesis, thus affecting tissues characterized
by active cell division. Consequently, hematopoietic stem cells and
hair follicle cells are often impacted, causing side effects such
as leukopenia and alopecia. Therefore, targeting anticancer drugs
specifically to tumors could potentially reduce these side effects.
Currently, many DDS formulations approved by the Food and Drug Administration
or undergoing clinical trials rely on passive targeting utilizing
the enhanced permeability and retention (EPR) effect.
,
However, the tumor accumulation of anticancer agent using EPR effect
is only ∼10%, with the remaining ∼90% of the drug diffusing
to nontarget sites or accumulating in organs with relatively large
vascular endothelial gaps, such as the liver and spleen. Therefore,
passive targeting is inadequate for anticancer agents. In contrast to passive targeting, active targeting
utilizes low-molecular-weight compounds or peptides that specifically
bind to receptors that are highly expressed on the surface of tumor
vascular endothelial cells and other cancer cells relative to normal
cells. For example, the RGD motif targets
αv integrin, and the NGR motif
targets CD13, and the addition of these
motifs to drug carriers is an area of active research. However, these
motifs only promote the accumulation of anticancer drugs in tumors,
and the internalization of anticancer drugs within cells depends on
the physical properties of the drug carrier. In this context, peptide
screening studies identified an iRGD peptide (CRGDKGPDC) that simultaneously
facilitates both tumor accumulation and cell membrane penetration. The iRGD peptide specifically binds to αvβ3
and αvβ5 integrins highly expressed on tumor vascular
endothelial cells and cancer cells via the RGD motif. Subsequently,
the iRGD peptide is cleaved by proteases, and the exposed CRGDK motif
binds to neuropilin-1 (NRP-1), leading to cancer cell–specific
internalization. Therefore, because the C-terminus of L-PGDS readily
accepts gene fusion of short targeting peptides, we attempted to construct
a tumor-targeting carrier by introducing CRGDK motif into the C-terminus
of L-PGDS.
In this study, PTX was examined as a model drug to
determine whether
the L-PGDS-based DDS is applicable to high-molecular-weight drugs.
Furthermore, to develop a novel PTX formulation with high therapeutic
efficacy relative to the commercially available formulation, the tumor-targeting
peptide CRGDK was incorporated into the C-terminus of L-PGDS (L-PGDS-CRGDK),
and the antitumor activity of the PTX/L-PGDS-CRGDK complex was evaluated
using a mouse xenograft model of human breast cancer-derived MDA-MB-231
tumors. Our results suggest that L-PGDS-CRGDK is a suitable tumor-targeting
DDS for breast cancer and could function as a promising carrier for
PTX.

Materials and Methods

Materials
PTX was purchased from Tokyo Chemical Industry
Co., Ltd. (Tokyo, Japan). HSA and 2-hydroxypropyl-β-cyclodextrin
(HP-β-CD) were purchased from FUJIFILM Wako Pure Chemical Industry
Co., Ltd. (Osaka, Japan). Taxol was purchased from Bristol-Myers Squibb
Co. (New York, NY, USA).

Docking Simulation
The PTX molecule was docked to the
L-PGDS structure using AutoDock Vina (ver. 1.1.2). The structure of PTX and crystal structure of human L-PGDS
were obtained from the ZINC database (ZINC ID: 96006020) and Protein
Data Bank (PDB, PDB ID: 4ORR), respectively. Bound compounds and water molecules
were removed from the PDB file, and all hydrogen atoms were added
to L-PGDS with Asn/Gln/His flip correction using the MolProbity server.
Gasteiger–Marsili charges were added to both structures using
AutoDock tools. The grid box was set to 28 × 26 × 32 Å,
with a spacing of 1.0 Å, and covered the binding pocket of L-PGDS.
In the docking simulation, the PTX molecule and the amino acid residues
in L-PGDS, which were previously predicted to interact with a variety
of poorly water-soluble drugs, were treated
as flexible, whereas the other residues in L-PGDS were treated as
rigid.

Purification of Recombinant Human L-PGDS
The recombinant
human C65A/C167A-substituted L-PGDS mutant (ε280 =
25,900 M–1 cm–1) was expressed
as a glutathione-S-transferase (GST) fusion protein
in Escherichia coli BL21 (DE3; TOYOBO,
Osaka, Japan), as described previously. The fusion protein was bound to glutathione Sepharose 4B (Cytiva,
Tokyo, Japan) and incubated overnight at room temperature with 165
units of thrombin (Sigma-Aldrich, St. Louis, MO, USA) to cleave the
GST-tag. The released L-PGDS was eluted using phosphate-buffered saline
(PBS [pH 7.4]), and further purified by gel filtration chromatography
using HiLoad 26/600 Superdex 75 pg (Cytiva) in 5 mM Tris-HCl buffer
(pH 8.0). Purified L-PGDS solution was dialyzed against PBS (pH 6.0).

Purification of HSA
HSA was dissolved in PBS (pH 6.0).
To remove contaminant proteins, the HSA solution was purified by gel
filtration chromatography using HiLoad 26/600 Superdex 200 pg (Cytiva)
in PBS (pH 6.0). The purity of HSA was confirmed by SDS-PAGE analysis
(Figure S1). The concentration of HSA was
calculated using the molar absorption coefficient at 280 nm (34,445
M–1 cm–1).

Determination of PTX Solubility
An excess amount of
PTX was suspended in PBS (pH 6.0) by sonication and added to each
protein solution (final concentration of protein: 1 μM). After
stirring at 4 °C for 1 h, the mixtures were concentrated using
Amicon Ultra Centrifugal Filter Devices (Millipore Corp., Bedford,
MA, USA). Additionally, an excess amount of PTX was suspended in PBS
or PBS containing 1 mM HP-β-CD and rotationally mixed at 4 °C
for 1 h. Each sample was filtered through a 0.22 μm filter (Millipore)
to remove insoluble PTX, and the concentration of PTX was determined
by HPLC analysis.

HPLC Analysis
Each sample was mixed with acetonitrile
and immediately vortexed for 30 s. After 15 min of standing on ice,
the solution was centrifuged at 18,800g for 15 min
at 4 °C. The resulting supernatant was filtered through a 0.22
μm filter (Millipore) and diluted with ultrapure water to adjust
the acetonitrile concentration to 60%, and then 10 μL (L-PGDS
or L-PGDS-CRGDK) or 50 μL (HSA) of sample was injected onto
a high-performance liquid chromatography system (Waters Alliance 2795
Separations module; Waters Co., Milford, CT, USA) equipped a COSMOSIL
5C18-MS-II column (4.6 mm ID × 150 mm, 5 μm, Nacalai Tesque,
Kyoto, Japan). The mobile phase consisted of a mixture of trifluoroacetic
acid, acetonitrile, and water (0.1/60/40 [v/v/v]) at a flow rate of
1.0 mL/min. The eluate was monitored at 227 nm to determine the amount
of PTX.

Tryptophan Fluorescence-Quenching Assay
Various concentrations
of PTX in dimethyl sulfoxide (DMSO) were added to solutions of L-PGDS
or L-PGDS-CRGDK, and the final protein concentration was adjusted
to 2 μM in PBS (pH 7.4) containing 5% (v/v) DMSO. After incubation
at 37 °C for 20 min, the intrinsic tryptophan fluorescence was
measured using an F-7000 fluorescence spectrophotometer (Hitachi Co.,
Tokyo, Japan) at an excitation wavelength of 290 nm and emission wavelength
of 342 nm. The value of the apparent dissociation constant (K
d) was calculated as previously reported. The n value, representing the
number of binding sites of the ligand, was fixed at 1.0.

Construction and Purification of Recombinant L-PGDS-CRGDK
The nucleotide sequence encoding L-PGDS-CRGDK was amplified by
PCR using the nucleotide sequence encoding L-PGDS as a template along
with the appropriate primers listed in Table S1. The resulting PCR product was digested with BamHI and EcoRI restriction enzymes and inserted into
the pGEX-4T2 vector (Cytiva). After sequence validation, E. coli BL21 (DE3; TOYOBO) was transformed with the
L-PGDS-CRGDK expression vector. L-PGDS-CRGDK expression and purification
were performed as described above (Purification of Recombinant Human
L-PGDS). The concentration of L-PGDS-CRGDK was determined spectroscopically
based on the molar absorption coefficient of ε280 (27,055 M–1 cm–1). The purity
of L-PGDS-CRGDK was confirmed by SDS-PAGE analysis (Figure S1).

Circular Dichroism Measurements
Circular dichroism
(CD) spectra of L-PGDS and L-PGDS-CRGDK in PBS (pH 7.4) were acquired
using a J-820 spectropolarimeter (JASCO, Tokyo, Japan) equipped with
a Peltier PTC-423L thermos-unit (JASCO) operated at 37 °C. The
optical quartz cuvette path length was 1 mm for the far-UV region
(200–250 nm) and 10 mm for the near-UV region (250–340
nm). The concentration of each protein was adjusted to 5 μM
(far-UV) or 50 μM (near-UV).

Cell Culture
MDA-MB-231 and MDA-MB-468 human breast
cancer cells were purchased from the American Type Culture Collection
(Manassas, VA, USA). The cells were cultured at 37 °C without
CO2 gas in Leibovitz’s L-15 Medium (Life Technologies,
Carlsbad, CA, USA) containing 10% fetal bovine serum.

Cellular Internalization Assay
MDA-MB-231 or MDA-MB-468
cells were seeded into a 4-well cell culture slide at 8 × 104 cells/well and cultured for 24 h. The medium was replaced
with L-15 containing 50 nM enhanced GFP (EGFP), EGFP-L-PGDS, or EGFP-L-PGDS-CRGDK.
After incubation for 1 h at 37 °C, the cells were washed three
times with PBS (pH 7.4) on ice and fixed in 4% paraformaldehyde for
10 min at 25 °C. After staining nuclei with 4′,6-diamidino-2-phenylindole
(DAPI; Dojindo, Kumamoto, Japan), images were obtained using an LSM700
confocal laser scanning microscope (Carl Zeiss, Jena, TH, Germany).

Cell Viability Assay
MDA-MB-231 cells were seeded into
96-well plates at a density of 4 × 103 cells per well
and cultured for 24 h. The medium was then replaced with L-15 containing
PTX, PTX/L-PGDS or PTX/L-PGDS-CRGDK, and the cells were incubated
for an additional 72 h. Cell viability was assessed using the Cell
Count Reagent SF (Nacalai Tesque).

Animal Study
All mice used in this study were purchased
from Japan SLC Inc. (Shizuoka, Japan). The mice were housed on a 12
h/12 h light/dark schedule with food and water available ad libitum
for 1 week to allow for recovery from the stress of transportation.
All animal experimental procedures were approved by the Osaka Prefecture
University Animal Care and Use Committee (permit number: 19–4,
25–42).

In Vivo Tumor Growth Inhibition Assay
Five weeks-old
female BALB/c-nu/nu mice were injected subcutaneously in the right
flank with 5 × 106 MDA-MB-231 cells. When the tumor
volume reached 300 mm3, the mice were randomly divided
into 4 groups. Taxol (4 mg/kg/day, n = 6), the PTX/L-PGDS
complex (equivalent to 4 mg PTX/kg/day, n = 6), the
PTX/L-PGDS-CRGDK complex (equivalent to 4 mg PTX/kg/day, n = 4), or PBS alone (n = 7) was administered intravenously
every other day for 2 weeks. Tumor length (a) and width (b) and body
weight were measured every day for 30 days, and the tumor volume was
calculated as (a × b
2)/2.

Anaphylaxis Test
Six-week-old male ddY mice were subcutaneously
administered either chicken ovalbumin (OVA; Sigma, Tokyo, Japan; 2
mg/mL), L-PGDS (2 mg/mL), or L-PGDS-CRGDK (2 mg/mL), each suspended
in an equal volume of aluminum hydroxide (13 mg/mL; Sigma) as an adjuvant.
After 14 days, the mice were intravenously administered OVA, L-PGDS,
or L-PGDS-CRGDK (1 mg/mL), and the body temperature was monitored
using a noncontact infrared thermometer (THERMOFOCUS, TEC- NIMED Srl,
Varese, Italy).

Statistical Analyses
Statistical analyses of in vivo
tumor growth inhibition assay data were performed using one-way analysis
of variance followed by the Tukey–Kramer HSD test. Differences
were considered statistically significant at p <
0.05.

Results

Docking Simulation
Estimation of the binding of PTX
to L-PGDS was performed by docking simulations using AutoDock Vina
(Figure
). The structural
model of the PTX/L-PGDS complex showed that the PTX molecule was located
in the upper region of the hydrophobic cavity of L-PGDS (Figure
A). To identify the
L-PGDS residues that interact with PTX, we analyzed the top 20 models
with the highest docking scores using the LigPlot + program and mapped these residues onto the structure
of L-PGDS (Figure
B). Eleven residues (Leu55, Trp54, Leu62, Phe83, Met94, Tyr107, Trp112,
Tyr116, Ser133, Phe143, and Met145) located in the upper region of
the binding cavity of L-PGDS formed hydrophobic interactions with
PTX in 75% of the structural models of the PTX/L-PGDS complex. Among
these residues, Leu55, located in the H2 helix, interacted with PTX
in all docking models, indicating that the region around Leu55 is
important for the binding of PTX and L-PGDS. Most docking models indicated
there were no hydrogen bonds between L-PGDS and PTX. These results
suggest that PTX binds to L-PGDS primarily via hydrophobic interactions.

Enhancement of PTX Solubility by L-PGDS
To investigate
the effect of L-PGDS on the solubility of PTX in aqueous solutions,
the PTX concentration in PBS containing L-PGDS was measured (Table
). PTX was soluble
in PBS only up to a concentration of 0.12 ± 0.01 μM; however,
in the presence of 1 mM L-PGDS, the solubility of PTX increased to
434 ± 6 μM (3617-fold higher than in PBS). The solubility
of PTX in the presence of 1 mM HSA or HP-β-CD, which are used
clinically as solubilizers for poorly water-soluble drugs, was 3.9
± 0.74 μM and 0.78 ± 0.04 μM, respectively.
These results demonstrate that L-PGDS significantly enhances the solubility
of PTX compared with widely used pharmaceutical solubilizers.
A tryptophan fluorescence quenching assay was then
performed to
examine the binding affinity of PTX to L-PGDS. The intrinsic tryptophan
fluorescence of L-PGDS was quenched in a concentration-dependent manner
by the addition of PTX (Figure
), indicating that the binding site of PTX in L-PGDS is located
near tryptophan residues. Based on quenching curve fitting, the calculated K
d value of PTX for L-PGDS was 8.7 ± 1.5
μM.

Construction and Characterization of L-PGDS with Tumor-Targeting
Peptide
To develop a tumor-targeting carrier, a tumor-targeting
peptide (CRGDK) was conjugated to the C-terminal region of L-PGDS
(Figure
A). The purified
product, L-PGDS-CRGDK, showed a single band in SDS-PAGE analysis (Figure S1). The far-UV and near-UV CD spectra
of L-PGDS-CRGDK were similar to those of L-PGDS (Figure
B), indicating that conjugation
of the tumor-targeting peptide to the C-terminal region did not significantly
affect the secondary or overall structure of the protein. Furthermore,
the solubility of PTX in 1 mM L-PGDS-CRGDK (446 ± 7 μM)
was comparable to that in L-PGDS (434 ± 6 μM) (Table
). The K
d value for PTX binding to L-PGDS-CRGDK (10.9 ± 1.1
μM) was highly similar to that for PTX binding to L-PGDS (8.7
± 1.5 μM) (Figure
C). These data thus confirmed that conjugation of the tumor-targeting
peptide (CRGDK) to the C-terminal region of L-PGDS did not affect
the secondary or tertiary structure of the protein nor did it affect
the solubility-enhancing properties of L-PGDS or its ability to bind
PTX.

Cellular Internalization Assay
EGFP-conjugated L-PGDS
(EGFP-L-PGDS) and L-PGDS-CRGDK (EGFP-L-PGDS-CRGDK) were constructed
by conjugating EGFP to the N-terminal region of L-PGDS and L-PGDS-CRGDK
to evaluate the cellular uptake of L-PGDS and L-PGDS-CRGDK. No green
fluorescent signal was observed upon treatment of MDA-MB-231 cells,
an NRP-1 expressing human breast cancer cell line (Figure S2), with EGFP or EGFP-L-PGDS (Figure
A). By contrast, strong green fluorescent
spots were observed in cells treated with EGFP-L-PGDS-CRGDK, indicating
that L-PGDS was taken up by these breast cancer cells via the introduced
tumor-targeting peptide (Figure
A). No green fluorescent spots were observed in MDA-MB-468
human breast cancer cells, which do not express NRP-1 (Figures
B and S2). These results indicate that conjugation of a tumor-targeting
peptide to the C-terminal region significantly enhances NRP-1–mediated
cellular uptake of L-PGDS.

Cytotoxicity Assay
The cytotoxic effects of PTX/L-PGDS
and PTX/L-PGDS-CRGDK were evaluated against MDA-MB-231 cells. L-PGDS
alone exhibited no detectable cytotoxicity up to 10 μM in MDA-MB-231
cells (Figure S3). In contrast, PTX/L-PGDS
and PTX/L-PGDS-CRGDK reduced cell viability in a PTX concentration-dependent
manner, which was nearly identical to that observed in cells treated
with PTX alone (Figure
). The IC50 values of PTX alone, PTX/L-PGDS and PTX/L-PGDS-CRGDK
against MDA-MB-231 cells were 2.3 ± 0.3, 4.4 ± 0.9, and
5.7 ± 0.9 nM, respectively. These results indicate that the cytotoxic
effects of PTX/L-PGDS and PTX/L-PGDS-CRGDK were comparable to those
of PTX alone in vitro.

In Vivo Antitumor Study
The in vivo antitumor activity
of PTX/L-PGDS and PTX/L-PGDS-CRGDK was investigated in nude mice implanted
with MDA-MB-231 tumors. When the tumor volume reached approximately
300 mm3, mice were intravenously injected with PBS, PTX/L-PGDS,
PTX/L-PGDS-CRGDK, or the commercial PTX formulation Taxol (Figure
A). Tumors in PBS-treated
mice grew rapidly, reaching a mean volume of 1610 ± 81.5 mm3 by 30 days after the first injection. During the first 14
days, tumor growth was significantly inhibited in the Taxol-treated
group compared with the PBS-treated group (day 14, p < 0.05). However, from the end of treatment (∼day 14),
the tumor volume rapidly increased, reaching a volume of 1540 ±
160 mm3, comparable to that of the PBS-treated control
group by day 30. By contrast, tumor growth in mice treated with PTX/L-PGDS
complexes was similar to that observed in Taxol-treated mice until
day 14, but tumor growth was suppressed thereafter compared with Taxol-treated
mice. In PTX/L-PGDS–treated mice, the tumor volume on day 30
was 1060 ± 76.0 mm3, which was 68.8% of the tumor
volume in Taxol-treated mice (p < 0.05 compared
with Taxol-treated mice). The antitumor activity of the PTX/L-PGDS-CRGDK
complex was significantly greater than that of Taxol (p < 0.01) and tended to be significantly greater than that of PTX/L-PGDS
as well. The tumor volume in mice treated with the PTX/L-PGDS-CRGDK
complex was 831 ± 89.1 mm3 on day 30, corresponding
to 54.0% and 78.4% of the tumor volume of mice treated with Taxol
or the PTX/L-PGDS complex, respectively (p < 0.01
compared with Taxol-treated mice).
The body weight of mice was monitored daily during
the experiment
(30 days) as an indicator of potential side effects (Figure
B). No weight loss was observed
in mice administered PBS, PTX/L-PGDS, PTX/L-PGDS-CRGDK, or Taxol.
These results indicate that the PTX/L-PGDS-CRGDK complex exerts no
acute toxicity and is more effective at inhibiting tumor growth than
Taxol or the PTX/L-PGDS complex. These results demonstrate the effectiveness
of introducing tumor-targeting peptides into L-PGDS.

Immunogenic Evaluation of L-PGDS-CRGDK
We previously
reported that human L-PGDS does not induce anaphylaxis responses in
mice. To assess whether conjugation of
a tumor-targeting peptide to human L-PGDS affects its immunogenicity,
we performed an anaphylaxis test in mice. OVA-sensitized mice showed
a marked decrease in body temperature following OVA administration,
confirming the high immunogenicity of OVA (Figure
). In contrast, no changes in body temperature
were observed in mice sensitized with L-PGDS or L-PGDS-CRGDK after
administration of the corresponding proteins (Figure
). These results indicate that L-PGDS-CRGDK
does not elicit immunogenic or anaphylactic responses.

Discussion
To help meet the strong demand for DDSs
for hydrophobic drugs,
our research has focused on L-PGDS as a potential drug carrier.
−

In the present study, we sought to develop an effective DDS based
on L-PGDS using PTX as a model drug, which is characterized by a high
molecular weight (M
w = 854), extremely
low water solubility, and high antitumor activity.
The docking
simulation shown in Figure
indicated that L-PGDS interacts with PTX
via hydrophobic interactions in the upper region of the β-barrel.
In addition, a concentration-dependent partial fluorescence quenching
of tryptophan residues in L-PGDS was observed with respect to PTX
(Figure
). L-PGDS
contains three tryptophan residues (Figure
B). Inui et al. and Kume et al. previously reported
that Trp43, located at the bottom of the cavity, showed more intense
fluorescence than Trp54 and Trp112, which are located in the upper
region of the cavity. Furthermore, ligand binding near Trp43 caused
stronger fluorescence quenching of L-PGDS. The partial fluorescence
quenching of tryptophan observed in the present study suggests that
PTX interacts with residues Trp54 and/or Trp112 in the upper region,
which is distant from Trp43, supporting the prediction from the docking
simulation that PTX binds in the upper region of the L-PGDS cavity.
Based on the fluorescence quenching observed in the present study,
the calculated K
d value for the interaction
between L-PGDS and PTX was 8.7 ± 1.5 μM, indicating high
binding affinity. The solubility of PTX in 1 mM L-PGDS was 434 ±
6 μM, reflecting the high binding affinity of L-PGDS for PTX.
This solubility is approximately 110-fold and 560-fold greater than
the solubility of PTX in HSA and HP-β-CD, respectively, two
commonly used solubilizers for poorly water-soluble drugs (Table
).
We previously
reported that L-PGDS binds lipophilic compounds with
molecular weights in the range of 270 to 780. The present study extended the upper molecular weight limit of compounds
that can bind to L-PGDS to 854, the molecular weight of PTX. Given
the increasing trend toward higher-molecular-weight drugs targeting
PPIs, the ability of L-PGDS to bind larger compounds is a significant
advantage for its use as a DDS carrier.
One of the key challenges
in the development of DDSs is how to
deliver drugs specifically to the target sites. To address this challenge,
we conjugated the tumor-targeting peptide CRGDK, which binds to NRP-1,
to the C-terminus of L-PGDS. CD spectra revealed that conjugation
of this peptide did not alter the overall structure of L-PGDS (Figure
B). In addition,
L-PGDS-CRGDK exhibited a K
d value of 10.9
± 1.1 μM for PTX binding, and the solubility of PTX in
1 mM solution was 446 ± 7 μM, values comparable to those
observed for L-PGDS lacking CRGDK. These results indicate that peptide
conjugation did not affect the fundamental function of L-PGDS. In
vitro tumor cell uptake experiments demonstrated that L-PGDS-CRGDK
was efficiently internalized into the cytoplasm of MDA-MB-231 cells,
which express NRP-1, but not into MDA-MB-468 cells, which do not express
NRP-1 (Figure
). By
contrast, L-PGDS lacking CRGDK was not internalized by either cell
type (Figure
). These
findings clearly demonstrate that L-PGDS-CRGDK is an effective drug
carrier capable of specifically delivering anticancer drugs into cancer
cells. Furthermore, tumor targeting by this carrier can be easily
achieved by simply introducing a suitable targeting sequence onto
L-PGDS. The introduction of targeting sequences into conventional
DDS carriers, such as liposomes and micelles, typically requires multiple
and complex chemical conjugation steps. Thus, L-PGDS offers a clear
advantage as a tumor-targeting DDS carrier. However, the in vitro
cytotoxic effect of PTX/L-PGDS-CRGDK against MDA-MB-231 cells was
nearly identical to those of PTX/L-PGDS and PTX alone (Figure
). Since cell viability was
measured after 72 h of exposure to each sample, free PTX may have
been readily taken up by MDA-MB-231 cells and exerted its cytotoxic
effect. To further clarify the impact of L-PGDS-CRGDK–mediated
cellular uptake, additional cytotoxicity assays with shorter exposure
times are warranted.

Figure
A shows
the tumor-inhibitory effects of the PTX/L-PGDS and PTX/L-PGDS-CRGDK
complexes in nude mice bearing MDA-MB-231 tumors, relative to the
commercial PTX formulation Taxol. PTX/L-PGDS and Taxol showed similar
tumor-inhibitory effects during the 14 day treatment period. However,
the tumor-suppressing effect of PTX/L-PGDS persisted beyond the treatment
period, whereas the inhibitory effect of Taxol diminished after treatment
cessation. PTX exerts its cytotoxicity through selective binding to
the β subunit of microtubule proteins. Therefore, the comparable tumor-inhibitory effects of Taxol and
PTX/L-PGDS observed during the 14 day treatment period suggest that
PTX released from PTX/L-PGDS was present in the cells at effective
concentrations similar to those achieved with free PTX (Taxol). The
fluorescence quenching assay revealed that L-PGDS binds PTX in an
equilibrium system with a dissociation constant of 8.7 ± 1.5
μM. These findings indicate that the affinity of PTX for L-PGDS
is much weaker than that for β-tubulin, and that, upon binding
to the β-tubulin subunit, the equilibrium between PTX and L-PGDS
rapidly shifts toward PTX release. Thus, the rate of PTX release from
L-PGDS is unlikely to be a rate-limiting factor for its antitumor
activity. On the other hand, PTX/L-PGDS exhibited a sustained antitumor
effect even after the treatment period, suggesting that sufficient
amounts of PTX/L-PGDS remained intracellularly, providing a stable
supply of PTX, and/or that its pharmacokinetic profile differed from
that of Taxol. Furthermore, the inhibitory effect of PTX/L-PGDS-CRGDK
was consistently stronger than those of both Taxol and PTX/L-PGDS
throughout the 30 day experimental period. Given that the dissociation
constant (10.9 ± 1.1 μM) of L-PGDS-CRGDK for PTX was comparable
to that of L-PGDS, the enhanced anticancer activity of PTX/L-PGDS-CRGDK
is likely attributable to its higher intracellular uptake. However,
the detailed biodistribution profile of PTX/L-PGDS and PTX/L-PGDS-CRGDK
remains to be elucidated. Additional studies, including single-dose
pharmacokinetic analyses (noncompartmental analysis of t
1/2, AUC, and CL) and near-infrared–labeled biodistribution
studies (evaluating liver, spleen, kidney, and tumor), are required
to fully clarify the therapeutic mechanisms of PTX/L-PGDS and PTX/L-PGDS-CRGDK.
We plan to report these findings in our next publication. At the end
of the experiment on day 30, tumor volumes in the control and Taxol-treated
groups had increased approximately 5-fold relative to the initial
volume. Conversely, the tumor volumes in the PTX/L-PGDS and PTX/L-PGDS-CRGDK
groups were suppressed to 66% and 52% of the volume of the control
group, respectively. These results suggest that the L-PGDS-based DDS
formulation for PTX could reduce the frequency of administration by
allowing longer administration intervals, as the tumor-suppressing
effect of PTX is sustained. By contrast, the antitumor effect of Taxol
rapidly wanes after treatment is discontinued, necessitating continuous
administration. Taxol is also associated with significant side effects,
such as histamine-mediated hypersensitivity reactions and sensory
neuropathy due to the dissolving agent CrEL. PTX itself is also associated with adverse effects, such as bone
marrow suppression and peripheral neuropathy. Therefore, the ability to reduce the frequency of administration
represents a significant advantage of PTX/L-PGDS-CRGDK. Furthermore,
although PTX/L-PGDS inhibited tumor growth, the inhibitory effect
of PTX/L-PGDS-CRGDK was even more pronounced. This enhanced antitumor
effect is likely due to the more efficient and specific internalization
of PTX/L-PGDS-CRGDK into tumor cells. Moreover, L-PGDS-CRGDK did not
induce any anaphylaxis responses similar to L-PGDS (Figure
). Given that PTX/L-PGDS-CRGDK
exhibits specific and potent antitumor activity and that conjugation
of the tumor-targeting peptide to L-PGDS does not alter its immunogenicity,
this conjugate holds strong potential as a DDS carrier capable of
inhibiting tumor growth while minimizing side effects.

Conclusions
L-PGDS is capable of binding the poorly
water-soluble anticancer
drug PTX, which has a molecular weight of 854, probably through hydrophobic
interactions with amino acid residues in the upper region of the β-barrel.
To enhance the specificity of L-PGDS for cancer cells, the peptide
CRGDK, which binds to NRP-1, was conjugated to the C-terminus of L-PGDS.
L-PGDS-CRGDK was successfully internalized into the cytoplasm of NRP-1–expressing
MDA-MB-231 breast cancer cells, whereas L-PGDS lacking CRGDK showed
no such translocation. Furthermore, PTX/L-PGDS-CRGDK significantly
suppressed tumor growth compared with both PTX/L-PGDS and the commercially
available PTX formulation Taxol. Given that many drugs currently in
development are poorly water-soluble and of high molecular weight,
the findings of this study suggest that L-PGDS-CRGDK is capable of
binding relatively large and poorly water-soluble molecules, highlighting
its potential as an effective DDS carrier for targeted tumor therapy.

Supplementary Material