Antiangiogenic and apoptotic properties of a novel amphiphilic folate-heparin-lithocholate derivative having cellular internality for cancer therapy

Article abstract of DOI:10.1007/s11095-006-9190-3

Purpose. Anitangiogenic and apoptotic properties of a novel chemically modified heparin derivative with low anticoagulant activity were evaluated on the experimental in vitro and in vivo model. Materials and Methods. Heparin-lithocholate conjugate (HL) wa

Full text of DOI:10.1007/s11095-006-9190-3

#
Pharmaceutical Research, Vol. 24, No. 4, April 2007 ( 2007)
DOI: 10.1007/s11095-006-9190-3
Research Paper
Antiangiogenic and Apoptotic Properties of a Novel Amphiphilic
Folate-Heparin-Lithocholate Derivative Having Cellular
Internality for Cancer Therapy
Mi Kyung Yu,1 ,j Dong Yun Lee,2 ,j Yoo Shin Kim,3 Kyeongsoon Park,4 Soo Ah Park,5 Dai Hyun Son,5
Gee Young Lee,6 Jong Hee Nam,7 Sang Yoon Kim,5 In San Kim,3 Rang Woon Park,3 and Youngro Byun2,8
Received October 16, 2006; accepted November 7, 2006; published online February 21, 2007
Purpose. Anitangiogenic and apoptotic properties of a novel chemically modified heparin derivative
with low anticoagulant activity were evaluated on the experimental in vitro and in vivo model.
Materials and Methods. Heparin-lithocholate conjugate (HL) was initially synthesized by covalently
bonding lithocholate to heparin. Folate-HL conjugate (FHL) was further synthesized by conjugating
folate to HL. Antiangiogenic and apoptotic abilities of HL and FHL were characterized in vitro and in
vivo experimentations.
Results. Compared to unmodified heparin, both HL and FHL represented a low anticoagulant activity (38
and 28%, respectively). HL and FHL maintained antiangiogenic activity even further modification from the
results of Matrigel plugs assay. FHL specifically induced apoptosis on KB cells having highly expressed folate
receptor after cellular internalization. Both administered HL and FHL had similar antiangiogenic activity and
inhibitory effect on tumor growth in vivo although FHL induced higher apoptosis on tumor tissues.
Conclusions. In vivo tumor growth inhibition was possibly due to the decrease of vessel density and
apoptotic cell death, although antiangiogenic effect of FHL seemed more actively affected on growth
inhibition than apoptotic potential in vivo system. Thus, Low anticoagulant FHL having antiangiogenic
and apoptotic properties would provide benefits for the development of a new class of anticancer agent.
KEY WORDS: antiangiogenesis; apoptosis; folate-heparin-lithocholate; tumor growth.
INTRODUCTION
noglycan, has anti-tumor activities (1Y3). In fact, the treat-
ment with heparin in cancer patients was formerly performed
Recent evidences are accumulated that heparin, a highly to treat venous thrombosis, and this has led to the observa-
sulfated polysaccharide belonging to the family of glycosami- tions in large volume of clinical trials that the medical effect
of heparin also prolongs cancer patient_s survival (4). Diverse
effects of heparin on malignant process are mostly based on
its inhibition of angiogenic factors, such as bFGF, or
1 Department of Life Science, Gwangju Institute of Science and interference of metastatic process containing heparanase or
j These authors contributed equally to this work as first authors.
Technology, Gwangju 500-712, South Korea.
selectin mediated cell invasion, and to a lesser extent, on the
activity to impede tumor growth and possibly carcinogenesis
itself. In addition, several other observations for the apopto-
tic activities of heparin have been reported (5,6). For
example, heparin inhibits activator protein-1, which is the
nuclear target of many oncogenic signal transduction path-
ways (7), and it may bind with DNA through charge to
charge interaction; in turn, transcription factors can be
upregulated in both the cytosol and the nucleus (8).
Despite these known benefits of heparin for cancer
treatment, the strong anticoagulant activity of heparin is a
limitation for the clinical applications of heparin at high dose
and for long-term treatment in cancer therapy (9). Further-
more, it is still hard to determine the ultimate effect of
heparin on cancer progression due to the wide variety of its
physiological activities (2,4,10). Although several effective
approaches for using sulfated oligosaccharides (11) or
chemically modified heparins, such as polystyrene bearing
and steroid conjugated heparin (12,13), terminally alkylated
heparin (14), have been investigated as angiogenesis inhib-
2 College of Pharmacy, Seoul National University, San 56-1,
Sillim-dong, Gwanak-gu, Seoul 151-742, South Korea.
3 Department of Biochemistry, School of Medicine, Kyungpook
National University, Daegu 700-422, South Korea.
4 Biomedical Research Center, Korea Institute of Science and
Technology, Seoul 136-791, South Korea.
5 Department of Otolaryngology, Asan Medical Center, College of
Medicine, University of Ulsan, Seoul 138-736, South Korea.
6 Department of Materials Science and Engineering, Gwangju Institute
of Science and Technology, Gwangju 500-712, South Korea.
7 Department of Pathology, School of Medicine, Chonnam National
University, Gwangju 500-757, South Korea.
8 To whom correspondence should be addressed. (e-mail: yrbyun@
snu.ac.kr)
ABBREVIATIONS: bFGF, basic fibroblast growth factor; FBS, fetal
bovine serum; FHL, folate-heparin-lithocholate; FITC, fluorescein
isothiocyanate; FR, folate receptor; HL, heparin-lithocholate; MTT,
(3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide); PBS,
phosphate buffered saline; PI, propidium iodide; TUNEL, terminal
deoxynucleotidyl transferase-mediated dUTP nick end-labeling;
UFH, unfractionated heparin.
#
705
0724-8741/07/0400-0705/0
2007 Springer Science + Business Media, LLC

706
Yu et al.
itors or anti-tumor agents, most of technologies were was reacted with DCC (1.2 mmol) and NHS (2 mmol) at
addressed on only angiogenesis, and they required high doses 50-C for 6 h. It is generally known that folate has two a-
causing the negative effects discussed above.
(high affinity for the FR) and g-carboxylic acids, but the g-
The folate receptor (FR) is a high-affinity, membrane- carboxylic acid is more selectively activated due to its
anchored protein which mediates the transport of folic acid higher reactivity (23,24). The resulting folate-NHS was
and its conjugates into the cell interior by endocytosis. reacted with ethylenediamine (13 mmol) and pyridine (500
Excessive need of rapidly dividing malignant cells for mg) at room temperature overnight. The folylethylamine
folates may be a reason for the frequent overexpression of (folate-NH2) was precipitated by the addition of excess
FRs in various cancer types such as ovarian, endometrial, acetonitrile, and the precipitate was filtered and washed
breast, nasopharyngeal, renal and colorectal cancers with diethyl ether before drying under vacuum to get yellow
(15Y19).Therefore, folate has been identified as the proto- powder. This was added to the prepared HL (100 mg),
type of a BTrojan horse^ approach to specific tumor dissolved in 20 ml of formamide, and activated by EDAC
targeting for diagnostic and therapeutic purpose (20,21). (3.38 mg) with 5 ml of DIEA for 12 h. The unreacted folate-
Here, we identified the ability of novel heparin derivatives NH2 was removed by dialysis (MWCO 2000). The final
that have antiangiogenic activities combined with apopto- product, FHL, was obtained by freeze-drying at a yield of
tic characteristics for the enhanced tumor target system. 97%. The folate content in FHL was determined by
In this study, we developed to see if the internalization of quantitative UV spectrophotometry at 365 nm. The anti-
heparin could induce apoptosis in cancer cells, and coagulant activities of HL and FHL were measured by
developed low anticoagulant heparin amphiphiles, namely, FXa chromogenic assay (COATEST\Heparin, Milano,
heparin-lithocholic acids conjugate (HL) and folate conju- Italy) (25).
gated HL (FHL). The antiangiogenesis and apoptosis ef-
fects of HL and FHL were evaluated in vitro and in vivo In Vivo Matrigel Plug Assay
experimentations.
In vivo Matrigel plug assay was performed as previously
MATERIALS AND METHODS
Chemicals
described (26). Briefly, Matrigel solution (0.67 ml) containing
bFGF (500 ng/ml) with 100 mg/ml of heparin derivatives
(UFH, HL or FHL, respectively,) was injected subcutaneous-
ly into the flank of male C57BL/6 mice (Samtako Bio Korea;
Unfractionated heparin (UFH, 167 IU/mg), whose aver- Osan, Korea). The injected Matrigel rapidly formed a single
age molecular weight was about 12 kDa, was purchased from gel plug. After 10 days, mice were sacrificed and the Matrigel
Pharmacia Heparin Co. (Franklin, OH). Lithocholic acid, plugs were removed, fixed with 4% paraformaldehyde, and
folic acid, ethylenediamine, anhydrous dimethylformamide embedded in paraffin. The plugs were sectioned and exam-
(DMF), 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide ined with H&E staining for microscopic observation.
(EDAC) and anhydrous formamide were purchased from
To quantify the formation of new blood vessels, the
Sigma Chemical Co. (St. Louis, MO). N-hydroxysuccinimide amount of hemoglobin (Hb) was measured using Drabkin
(NHS), N,N-dicyclohexyl carbodiimide (DCC), dimethyl reagent kit (Sigma). The concentration of Hb was calcu-
sulfoxide (DMSO) and N,N-diisopropylethylamine (DIEA) lated from a known amount of Hb provided by the kit in
were purchased from Aldrich Chemical Co. (Milwaukee, WI). parallel according to the supplier_s protocol. The experi-
Acetone and methanol were obtained from Merck (Darmstadt, ment was repeated four times independently. Investiga-
Germany). Acetonitrile was purchased from Junsei Chemical tions using experimental animals adhered to the
Co. (Tokyo, Japan) and all reagents were of analytical grade BPrinciples of Laboratory Animal Care^ (NIH publication
and were used without further purification.
#85-23, revised in 1985).
Synthesis and Characterization of HL and FHL
Heparin-lithocholate (HL) was synthesized by coupling
Cell Viability
KB cells (human, nasopharyngeal carcinoma cell line)
heparin with the bile analogue, N-lithocholylethylamine known to have amplified FR-a expressions were cultured
(Litho-NH2), as described previously (22). In brief, heparin in RPMI 1640 medium (Gibco, Grand Island, NY)
(100 mg) was dissolved in 3 ml of formamide by gentle containing 10% FBS and 1% penicillin streptomycin at
heating. EDAC (96 mg) was mixed with the heparin solution, 37-C in a humidified atmosphere of 5% CO2 in air. KB
followed by mixing with Litho-NH2 in DMF (10 ml). This cells were seeded at a density of 5Â104 cells/well in 96-well
reaction was carried out at room temperature under a nitrogen flat-bottomed plates. The cells were folate-starved using a
atmosphere for 24 h. After the mixture was precipitated in folate-deficient special RPMI 1640 medium (modified
excess cold acetone, the precipitate was washed with cold RPMI 1640 without folic acid, vitamin B12 and phenol red;
acetone to remove the unreacted Litho-NH2, and dried in sigma) without FBS. Then, the cells were treated with 100
vacuum. Dried HL was suspended in water and lyophilized to mg/ml of HL, FHL, or FHL plus 2 mM free folate. Following
produce as white powder. The presence of lithocholic acid in 1, 6, 12 and 24 h of incubation under 5% CO2 at 37-C, the
HL and amide linkage were confirmed by 1H NMR (400 cells were washed with PBS and the cell viability was
MHz, CD3OD/D2O=2/1, v/v, JEOL, Tokyo, Japan).
assessed by MTT colorimetric assay using automated micro-
Folate-heparin-lithocholate (FHL) was prepared as plate reader (570 nm; VERSAmaxi, Molecular Devices
follows; folic acid (1 mmol) dissolved in 20 ml DMSO Corp., Sunnyvale, CA).

Antiangiogenic and Apoptotic Properties of FHL
707
Flow Cytometry Analysis
Immunohistochemical Analysis
Tumor tissues of each group were isolated after 2 weeks
KB cells grown in a monolayer were incubated with
indicated concentrations of UFH, HL or FHL. After 12 h, and intratumoral microvessel density was then analyzed using
cells were collected, washed with 1 ml of PBS, and fixed by a rat anti-mouse CD34 monoclonal antibody (Abcam,
1 ml of 70% ethanol with gentle vortexing. The cell pellet Cambrige, UK). Immunoperoxidase staining was done using
was resuspended in PI solution (50 mg/ml) at a concentra- the streptavidin-peroxidase method. As the final step, the
tion of 106 cells/ml, and maintained at room temperature for nuclei of the sections were counterstained with hematoxylin
30 min. DNA content was then analyzed by a flow solution. The image that contained the highest number of
cytometer (Beckton Dickinson, Franklin Lakes, NJ) and microvessels was chosen for each section from an initial scan
sub G1 period, called the apoptotic index, was observed. at Â100 magnification. Then, the vessels were counted in the
Results were calculated as meansTs.d. from at least three selected image at Â200 magnification. At least five fields
individual experiments.
were counted for each section. Two independent investiga-
tors evaluated the number of vessels. TUNEL assay was also
performed with commercial kit (Roche Diagnostics) accord-
ing to the manufacturer_s instructions. The assay was
TUNEL Assay In Vitro
Apoptotic cells were evaluated by double staining for performed in triplicate and repeated three times with five
Annexin-V-Alexa Fluor 488 binding and propodium iodide independent tumors from each treatment group. The per-
incorporation using the Apoptosis Detection Kit (Invitro- centage of apoptosis was calculated by dividing the number
gen detection technologies, Carlsbad, CA). Single stain of TUNEL-positive cells with the total number of cells and
controls were performed before each analysis according to multiplying the result by 100. The counting was performed in
the standard protocol. Flow cytometry analysis was per- the Â100 magnification.
formed on 10,000 cells with a FACS Calibur (Becton
Dickinson) and Cell Quest Pro software. For the terminal Statistics
deoxynucleotidyl transferase-mediated dUTP nick end-la-
beling (TUNEL) assay, KB cells were plated on slide
Data are expressed as means T standard deviation (s.d.).
chamber and cultured in RPMI 1640 containing 10% FBS ANOVA test was used to compare groups and P values of
for 12 h. The culture medium was then replaced with <0.05 were considered significant.
folate-deficient special RPMI 1640 medium without FBS.
Then, 100 mg/ml of UFH or HL were treated to the cells for RESULTS
24 h. In addition, 50, 100 or 300 mg/ml of FHL without or
with 2 mM free folate were also separately treated to the Synthesis and Characterization of HL and FHL
cells for 24 h. Apoptosis was assayed using the TMR Red
in situ Apoptosis Detection Kit (Roche Diagnostic, Indian-
We developed heparin amphiphiles by conjugating
apolis, IN), and cells were counterstained with TOPRO-3 lithocholate as hydrophobic moieties to the hydrophilic
(Molecular Probes, Eugene, OR). Cells were then observed heparin (Fig. 1). HL was initially prepared by covalent
by fluorescence microscopy using an inverted microscope linkage of Lithocholate to heparin, and this resulted in
(LEICA TCS SP2 microscopy, Leica Microsystems GmbH, lower anticoagulant activity (38%) than that produced by
Wetzlar, Germany). Images were obtained with digital the unmodified heparin. The presence of lithocholate in
camera (model DM IRE2, Scientific. Inc., Germany) and HL was confirmed by 1H NMR (400 MHz, CD3OD/
Image-Pro Plus software. TUNEL positive cells were calcu- D2O=2/1, v/v). The characteristic peaks of lithocholate
lated as follows: (number of apoptotic nuclei/number of total appeared at 18-CH3 (0.64 ppm), 19-CH3 (0.91 ppm), 21-
nuclei) Â100. Three measurements were obtained from each CH3 (1.24 ppm), and the new amide linkage between
border zone area by a single investigator blinded to the heparin and lithocholic acid was observed at 8.04 ppm
treatment group.
Experimental Tumor Graft Growth Models
FHL
Seven-week old female athymic BALB/C nude mice
weighing 20Y25 g were purchased from Charles River
Laboratories (Wilmington, MA) and maintained on a
folate-free rodent diet. Then, KB cells (1Â107 cells/100 ml)
were inoculated subcutaneously at the left flank of the
animals. When the tumor size were reached to 70Y100
mm3, the mice were received following treatments: (a) 0.9%
NaCl solution (normal saline) for control group, (b) HL
(5 mg/kg) and (c) FHL (5 mg/kg), respectively. The drugs
were given by intravenous injection on every third day for 2
weeks. Tumor sizes were daily monitored and tumor
volumes were calculated as a2ÂbÂ0.52, where a=width and
b=tumor length.
O
H
N
C
O
heparin
Heparin
NH
O
HN
NH
O
HO
O
Lithocholate
HL
N
OH
HO
H
N
N
O
N
N
H
N
NH2
Folate
Fig. 1. Chemical structure of heparin-lithocholate (HL) and folate-
heparin-lithocholate (FHL) conjugates.

708
Yu et al.
(data not shown). This amide linkage was further confirmed 24 h showed 78% apoptotic cells. Cells in FHL also showed
at the wavelength of 1,665 cmj1 in the FT-IR spectra (data increased necrosis (11%) at 24 h, but it was not very high
not shown). In addition, the peak that appeared near 2,360 compared to the amount of increased apoptotic cells.
cmj1 represented the stretching vibration of intrinsic
sulfonamide (jNHSO3) in heparin as previously reported Cellular Internalization of FHL in KB Cells
(27). Folate conjugated HL (FHL) was then synthesized as
shown in Fig. 1. This FHL also represented a low
We investigated that the intracellular delivery, which
anticoagulant activity (28%). The average particle size of is the comparative cellular uptake, of FITC-labeled FHL
heparin amphiphiles in distilled water was about 170 nm occurred in a FR-mediated specific manner and that the
(there was about 10 nm size change after the folate apoptosis in cells have taken up FITCYFHL. Cells were
conjugation), and zeta potential values of HL and FHL incubated with FITCYFHL or FITCYHL (100 mg/ml) for
were j50 and j25 mV, respectively.
24 h to evaluate whether these molecules were taken up
inside cells, and the TUNEL assay and nuclei staining
using TOPRO-3 were carried out simultaneously to
monitor apoptosis (Fig. 4). The cells incubated with
In Vivo Matrigel Plugs Assay
In vivo antiangiogenic activities of HL and FHL were FITCYFHL showed higher cellular uptake, especially in
investigated by measuring the extent of blood vessel invasion the nuclei, than those with HL, indicating that folate
into Matrigel plugs. As shown in Fig. 2, blood vessels were conjugation provide with the ability of cellular endocytosis.
formed in the order of positive control (with bFGF only), TUNEL-positive cells precisely matched those with FITC
UFH, FHL, HL and negative control (with PBS). Further image and TOPRO-3 staining. These results imply that
estimation of hemoglobin content in the Matrigel plugs cells incubated with FHL were scored as apoptotic because
showed that HL or FHL significantly inhibited the vascular- they were affected by internalized heparin amphiphiles.
ization inside the plugs with 29 and 43% of hemoglobin
content, respectively, (P<0.01 vs. positive control). The Inhibitory Effects of HL and FHL on Tumor Graft Growth
inhibition effect of UFH was lower than that of HL or FHL
although UFH significantly inhibited the vascularization
induced by bFGF (P<0.05 vs. positive control).
KB cells were subcutaneously implanted into the flanks
of BALB/C nude mice, and subsequent tumor growth,
neovascularization and apoptosis induction were monitored
after intravenous injection of 5 mg/kg/3 day of HL or FHL
for 14 days. As shown in Fig. 5, mice that received HL or
Cytotoxicity of FHL Via Folate Receptor on KB Cells
In vivo cytotoxicity of FHL was evaluated in KB cells by FHL exhibited higher antitumor activity than control. Tumor
MTT assay. After 24 h incubation with FHL (100 mg/ml), growth suppression in FHL indicated improved therapeutic
viable cells were only 25% (P<0.001), (Fig. 3A). However, efficacy (69% of growth inhibition versus control) compared
the cytotoxicity of FHL in KB cells was not completely to the mice in non-folate HL (56% of growth inhibition
shown by treating free folate to block the binding of FHL to versus control). However, there was not significant difference
the FR. Furthermore, non-folate-conjugated HL did not between FHL and HL in inhibiting tumor growth. To
affect cell viability. We examined the DNA content of cells determine whether the reduced size of FHL- or HL-treated
to determine whether the cellular cytotoxicity of FHL is tumors coincides with reduction in neovascularization or
related with the induction of apoptosis. Cells populated at a apoptosis induction, we initially used five representative
sub-G1 phase, which is typically indicative of apoptosis (6), FHL-, HL-treated or saline-treated tumors to quantify the
were evaluated by flow cytometry analysis (Fig. 3B and 3C). density of microvessels after immunostaining with anti-CD34
The treatment of 100 mg/ml of UFH or HL for 12 h did not antibody. The reduced size of FHL- or HL-treated tumors
induce the sub-G1 accumulation, whereas, cells in FHL was consistent with a decrease in CD34-positive microvessels,
revealed 3.2-fold higher sub-G1 population than in UFH. In and the reduction ratio was 60 and 48% versus control,
addition, FHL treated cells in the sub-G1 period markedly respectively, (Fig. 6A and B). Interestingly, FHL-treated
were increased from 4.8 to 38.2% with increase in the culture tumors significantly showed 3.2-fold higher induction of
time up to 24 h. In contrast, when we added 2 mM free folate apoptosis compare with HL-treated tumors (P<0.001) (Fig.
to the medium, FHL did not significantly affect cell cycle as 6C and D). However, this apoptotic level might not induce
exhibited in cell viability.
noticeably increased in vivo tumor regression as we con-
We further evaluated the apoptotic effect of FHL to firmed in tumor volume result, demonstrating that the
distinguish it from necrotic effect. It can be also used to antiangiogenic effect of heparin amphiphiles might dominant
quantify three populations of cells with drug treatment: in vivo, compared to the apoptotic effect.
normal viable cells, apoptotic cells, and populations of
necrosis using Annexin V and PI staining. As shown in DISCUSSION
Fig. 3D, normal viable cells (AVj/PIj) are shown in the
bottom left window, apoptotic cells (AV+/PIj) in the bottom
We prepared the heparin-lithocholate conjugate (HL)
right window, and the population of necrosis (AV+/PI+) in comprised of covalently bound heparin, as a hydrophilic
the top right window. Untreated cells yielded only 3% segment, and lithocholic acid as a hydrophobic segment.
apoptotic cells; cells treated with UFH for 24 h showed 7% The amide formation between carboxyl groups of glucu-
apoptotic cells; and cells treated with HL for 24 h showed ronic acid and iduronic acid residues in heparin and amine
18% apoptotic cells. However, cells treated with FHL for groups of Litho-NH2 was confirmed by 1H NMR and IR.

Antiangiogenic and Apoptotic Properties of FHL
709
Fig. 2. In vivo antiangiogenic activities. A Matrigel plugs containing bFGF alone (+control), bFGF plus UFH, bFGF plus FHL, bFGF plus
HL, or no bFGF (PBS,-control) were photographed. B Relative hemoglobin contents in Matrigel plugs determined by the Drabkin_s method.
Data were expressed as meansTs.d. (N=5). (*P<0.05 versus + control, **P<0.001 versus + control). C Sections of each Matrigel plugs stained
with H&E were examined by light microscopy (Â200).

710
Yu et al.
Fig. 3. Cellular cytotoxicity of HL and FHL in KB cells. A Cell viability of the KB cells treated with 100 mg/ml of HL (P), FHL (h) or FHL
plus 2 mM free folate (r) was measured by MTT assay. Data were expressed as means T s.d. (N=5). (*P<0.001 versus FHL plus 2 mM free
folate). B The sub-G1 population in KB cells treated with UFH ( , 100 mg/ml), HL ( , 100 mg/ml) or different doses of FHL(Ì) alone or plus 2
mM free folate ( ) for 12 h. Data were expressed as means T s.d. (N=3). (*P<0.05, **P<0.001 versus UFH). C The sub-G1 population in KB
cells treated with 100 mg/ml of FHL (P) or FHL plus 2 mM free folate (°) at different treating times. Data were expressed as means T s.d.
(N=4). (*P<0.005, **P<0.002). D Flow cytometry analysis for apoptotic and necrotic KB cells stained with Annexin V (AV) and propidium
iodide (PI) after treatment with 100 mg/ml of UFH, HL or FHL for 24 h. Viable cells (AVj/PIj) are in the lower left-hand quadrant. Early
apoptotic cells (AV+/PIj) are in the lower right-hand quadrant. Terminal apoptotic/necrotic cells (AV+/PI+) are in the upper right-hand
quadrant. Sizes of cell sub-populations are given as percentage of total populations.

Antiangiogenic and Apoptotic Properties of FHL
711
sulfate groups, which play an important role in the binding
with endothelial cells (28) or in blocking of FGF-heparan
sulfate binding (11), were not modified.
In vivo Matrigel plugs assay showed that both HL and
FHL had the lower hemoglobin content in the Matrigel than
unmodified heparin in the bFGF induced vessel formation,
implies that Both HL and FHL is more closely associated with
bFGF than unmodified heparin, and decrease the activity of
bFGF. Some literatures reported that suramin, the efficient
angiosuppressive agent, blocked the receptor binding loop of
the bFGFs through the hydrophobic interaction between the
naphtyl or phenyl rings of the bound suramin and Pro, Cys,
Gly, Arg, Gln, or Lys residues of bFGF (29), in addition to
ion-pair, hydrogen bonding interactions, and van der Waals_
contacts (30). These interactions provided the aggregates of
suramin-bound bFGF, and made them to lose the activity of
the bFGF for receptors, since hydrophobic residues of bFGF
in receptor binding site have a crucial role for high affinity
receptor binding (31). We speculate that the pronounced
vessel formation inhibition of HL and FHL observed in vivo
results from an additional coordinated mechanism, which
efficiently inhibit the activity of growth factors, in particular
basic fibroblast growth factor, through the hydrophobic
interaction as well as ionic interaction.
In addition to mediating angiogenesis, FHL internalized
into the cancer cells can be cytotoxic, and cause cancer cell
death. We have shown that FHL treatment causes apoptosis
in KB cells. FR-mediated delivery offers the physiological
uptake of folate conjugated molecules by endocytosis, and
provides them the chance to affect inside the cancer cells.
Interestingly, FHL showed high cytotoxicity against KB cells,
but neither HL nor FHL did so in the presence of excess free
folate. Furthermore, high concentrations of HL (100 mg/ml)
did not affect the viability of tumor cells up to 24 h. The cell
cycle arrest within the sub-G1 phase was observed only in the
FHL treated cells. UFH or HL, however, did not induce
apoptotic cell population. Correspondingly, AV+ and PI-cells,
one of the earliest markers of apoptotic death, were the most
abundant in the exposure of cells to FHL. HL also represented
early stages of apoptosis only by 18%, implying that the slight
3000
2500
2000
1500
1000
Fig. 4. Cellular internalization of FHL in KB cells microscope after
treatment with 100 mg/ml of FITC-HL or FITC-FHL for 24 h. KB
cells co-stained with TUNEL and TOPRO-3 stain were visualized by
using confocal laser scanning. Arrows: FITC/TUNEL double positive
cells.
In the aqueous condition, heparin amphiphiles were self-
aggregated to make nanoparticles. Since negatively charged
heparin might cover the core hydrophobic lithocholic acids
as previously discussed (22), the x potential of self-
aggregates were still negative (j50 mV). Folate conjugated
HL (FHL) was synthesized by coupling folate to HL in order
to bind to folate receptors in the tumor cell membrane. Both
HL and FHL with a low anticoagulant characteristic are
useful in long term delivery or at high dose because they can
overcome several limitations noted in the administration of
heparin; hemorrhage, thrombocytopenia, or osteoporosis etc.
had been cited as problems of heparin treatment. It is
important, however, that HL and FHL should retain their
*
500
0
0
2
4
6
8
10
12
14
Time (day)
Fig. 5. Tumor graft growth curve during treating saline (P), HL
(h; 5 mg/kg) or FHL (r; 5 mg/kg) by intravenous injection Balb/c
mice bearing KB cells via the lateral tail vein on every 3rd day for 2
weeks. Data were expressed as means T s.d. (N=5). (*P<0.05 versus
antiangiogenic ability and antitumoral efficacy since the control).

712
Yu et al.
Fig. 6. Immunohistochemical analysis of tumor grafts. A Anti-CD34 immunostaining for microvessels in the tumor tissue after treatment of
HL or FHL for 14 days. Brown color: Anti-CD34 positive microvessels. B The number of CD34 positive microvessels counted from the
randomly selected fields of each tumor section in the Â100 magnification. Data were expressed as means T s.d. (N=5) (*P<0.001 versus
control). C TUNEL staining for apoptotic KB cells in the tumor tissue after treatment of HL or FHL for 14 days. Brown color: apoptotic cells.
D The number of TUNEL positive cells counted from the randomly selected fields of each tumor section in the Â100 magnification. Percent
apoptosis was calculated by dividing the total number of TUNEL-positive cells in each field (Â100) by the total number of cells and
multiplying by 100. Data were expressed as means T s.d. (n=5). (*P<0.05 versus control, **P<0.001 versus control).

Antiangiogenic and Apoptotic Properties of FHL
713
3. M. Hejna, M. Raderer, and C. C. Zielinski. Inhibition of metastases
by anticoagulants. J. Natl. Cancer Inst. 91:22Y36 (1999).
4. L. R. Zacharski and D. L. Ornstein. Heparin and cancer.
Thromb. Haemost. 80:10Y23 (1998).
5. E. Erduran, Y. Tekelioglu, Y. Gedik, and A. Yildiran. Apopto-
tic effects of heparin on lymphoblasts, neutrophils, and mono-
nuclear cells: results of a preliminary in vitro study. Am. J.
Hematol. 61:90Y93 (1999).
6. H. L. Li, K. H. Ye, H. W. Zhang, Y. R. Luo, X. D. Ren, A. H.
Xiong, and R. Situ. Effect of heparin on apoptosis in human
nasopharyngeal carcinoma CNE2 cells. Cell Res. 11:311Y315
(2001).
7. S. J. Busch, G. A. Martin, R. L. Barnhart, and R. L. Jackson.
Trans-repressor activity of nuclear glycosaminoglycans on Fos
and Jun/AP-1 oncoprotein-mediated transcription. J. Cell Biol.
116:31Y42 (1992).
8. D. Berry, D. M. Lynn, R. Sasisekharan, and R. Langer.
Poly(beta-amino ester)s promote cellular uptake of heparin
and cancer cell death. Chem. Biol. 11:487Y498 (2004).
9. J. Hasan, S. D. Shnyder, A. R. Clamp, A. T. McGown, R.
Bicknell, M. Presta, M. Bibby, J. Double, S. Craig, D. Leeming,
K. Stevenson, J. T. Gallagher, and G. C. Jayson. Heparin
octasaccharides inhibit angiogenesis in vivo. Clin. Cancer Res.
11:8172Y8179 (2005).
uptake might be possible due to its amphiphilic structure. In
contrast, most of KB cells were not affected by UFH, because
heparin internalization without any helper was unlikely; only
20 pg of internalized heparin was observed out of 1 mg/ml of
heparin in 5Â104 cells (8). Thus, the internalization of
heparin amphiphiles into the cell seems to be an important
factor for cellular apoptosis. TUNEL-FITC double staining
clearly verified this mechanism. Notably, the uptake was
concentrated on the nucleus, and this suggests that FHL
escaped from endosome and might enter the nucleus. Several
possible pathways involved in apoptosis of internalized
heparin are rarely reported, such as interfering with
transcription factor activity or activating the caspasej3/j7
(8), the inhibition of DNA binding by casein kinase II (32),
and interfering with the polymerase chain reaction, which
may involve DNA polymerase.
HL resulted in superior suppression of tumor growth as
well as tumor angiogenesis compared with UFH in a mouse
KB tumor model. The results of significantly diminished
vessels in HL treated tumor were consistent with the result of
Matrigel plugs assay, and HL leads to the tumor growth
inhibition. In our previous study, we speculated that the
pronounced tumor growth inhibition of heparin-bile acid
conjugates observed from efficient inhibition of the activity
of growth factors, in particular basic fibroblast growth factor,
through the hydrophobic interaction as well as ionic inter-
action (33,34). Also, the antitumoral effect of FHL was
similar to that of HL, even though it retains antiangiogenic
properties and has additional apoptotic characteristics for the
KB cells. Indeed, it is possible that the antitumorigenic effect
of FHL is possibly due to both the inhibition of angiogenesis
and the induction of apoptotic tumor cell death, however, the
vascularization modulation may predominantly affect on in
vivo system than the induction cancer cell death.
10. H. Engelberg. Actions of heparin that may affect the malignant
process. Cancer 85:257Y272 (1999).
11. C. R. Parsish, C. Freeman, K. J. Brown, D. J. Francis, and W. B.
Cowden. Identification of sulfated oligosaccharide-based inhib-
itors of tumor growth and metastasis using novel in vitro assays
for angiogenesis and heparanase activity. Cancer Res.
59:3433Y3441 (1999).
12. K. Ono, M. Ishihara, K. Ishikawa, Y. Ozeki, H. Deguchi, M.
Sato, H. Hashimoto, Y. Saito, H. Yura, A. Kurita, and T.
Maehara. Periodate-treated, non-anticoagulant heparin-carrying
polystyrene (NAC-HCPS) affects angiogenesis and inhibits
subcutaneous induced tumour growth and metastasis to the
lung. Br. J. Cancer 86:1803Y1812 (2002).
13. P. E. Thorpe, E. J. Derbyshire, S. P. Andrade, N. Press, P. P.
Knowles, S. King, G. J. Watson, Y. C. Yang, and M. Rao-
Bette. Heparin-steroid conjugates: new angiogenesis inhibi-
tors with antitumor activity in mice. Cancer Res. 53:3000Y3007
(1993).
In this study, we firstly demonstrated that the internal-
ization of heparin could induce apoptosis of cancer cells and
report a novel anticancer therapeutic agent called FHL, which
is an angiogenesis inhibitor combined with the apoptosis of
tumor cell. Therefore, we expect that the exploitation of the
effective internalization method of heparin like the folate
conjugated heparin amphiphiles would provide a good
opportunity for the development of a new class of anticancer
agents.
14. M. Gohda, T. Magoshi, S. Kato, T. Noguchi, S. Yasuda, H.
Nonogi, and T. Matsuda. Terminally alkylated heparin. 2.
Potent antiproliferative agent for vascular smooth muscle cells.
Biomacromolecules 2:1178Y1183 (2001).
15. S. Miotti, S. Canevari, S. Menard, D. Mezzanzanica, G. Porro,
S. M. Pupa, M. Regazzoni, E. Tagliabue, and M. I. Colnaghi.
Characterization of human ovarian carcinoma-associated anti-
gens defined by novel monoclonal antibodies with tumor-
restricted specificity. Int. J. Cancer 39:297Y303 (1987).
16. L. R. Coney, A. Tomassetti, L. Carayannopoulos, V. Frasca, B. A.
Kamen, M. I. Colnaghi, and V. R. Zurawski Jr.. Cloning of a
tumor-associated antigen: MOv18 and MOv19 antibodies recog-
nize a folate-binding protein. Cancer Res. 51:6125Y6132 (1991).
17. S. D. Weitman, R. H. Lark, L. R. Coney, D. W. Fort, V. Frasca,
V. R. Zurawski Jr, and B. A. Kamen. Distribution of the folate
receptor GP38 in normal and malignant cell lines and tissues.
Cancer Res. 52:3396Y3401 (1992).
18. G. Toffoli, C. Cernigoi, A. Russo, A. Gallo, M. Bagnoli, and M.
Boiocchi. Overexpression of folate binding protein in ovarian
cancers. Int. J. Cancer 74:193Y198 (1997).
19. N. Parker, M. J. Turk, E. Westrick, J. D. Lewis, P. S. Low, and
C. P. Leamon. Folate receptor expression in carcinomas and
normal tissues determined by a quantitative radioligand binding
assay. Anal. Biochem. 338:284Y293 (2005).
ACKNOWLEDGEMENTS
This study was supported by the grant from Next
Generation New Technology Development Program of the
Korean Ministry of Commerce, Industry, and Energy (Grant
no. 10011353).
20. C. P. Leamon and P. S. Low. Folate-mediated targeting: from
diagnostics to drug and gene delivery. Drug Discov. Today
6:44Y51 (2001).
21. J. Sudimack and R. J. Lee. Targeted drug delivery via the folate
receptor. Adv. Drug Deliv. Rev. 41:147Y162 (2000).
22. K. Park, K. Kim, I. C. Kwon, S. K. Kim, S. Lee, D. Y. Lee, and
Y. Byun. Preparation and characterization of self-assembled
nanoparticles of heparin-deoxycholic acid conjugates. Langmuir
20:11726Y11731 (2004).
REFERENCES
1. R. Sasisekharan, Z. Shriver, G. Venkataraman, and U.
Narayanasami. Roles of heparan-sulphate glycosaminoglycans
in cancer. Nat. Rev., Cancer 2:521Y528 (2002).
2. S. M. Smorenburg and C. J. NoordenVan. The complex effects
of heparins on cancer progression and metastasis in experimen-
tal studies. Pharmacol. Rev. 53:93Y105 (2001).

714
Yu et al.
23. S. Wang, R. J. Lee, C. J. Mathias, M. A. Green, and P. S. Low.
Synthesis, purification, and tumor cell uptake of 67Ga-
deferoxamine-folate, a potential radiopharmaceutical for tumor 29. P. A. Raj, E. Marcus, and R. Rein. Conformational require-
ity and role in heparin neutralization. Biochem. J. 238:847Y854
(1986).
imaging. Bioconjug. Chem. 7:56Y62 (1996).
24. C. J. Mathias, S. Wang, P. S. Low, D. J. Waters, and M. A.
ments of suramin to target angiogenic growth factors.
Angiogenesis 2:183Y199 (1998).
Green. Receptor-mediated targeting of 67Ga-deferoxamine- 30. R. J. Linhardt. Heparin-induced cancer cell death. Chem. Biol.
folate to folate-receptor-positive human KB tumor xenografts.
Nucl. Med. Biol. 26:23Y25 (1999).
25. T. Chandy, G. S. Das, R. F. Wilson, and G. H. Rao. Use of
plasma glow for surface-engineering biomolecules to enhance
blood compatibility of Dacron and PTFE vascular prosthesis.
Biomaterials 21:699Y712 (2000).
11:420Y422 (2004).
31. O. Filhol, C. Cochet, and E. M. Chambaz. DNA binding activity
of casein kinase II. Biochem. Biophys. Res. Commun.
173:862Y871 (1990).
32. E. Beutler, T. Gelbart, and W. Kuhl. Interference of heparin
with the polymerase chain reaction. Biotechniques 9:166Y170
(1990).
26. Y. Lee, H. T. Moon, and Y. Byun. Preparation of slightly
hydrophobic heparin derivatives which can be used for solvent 33. K. Park, G. Y. Lee, Y. S. Kim, M. Yu, R. W. Park, I. S. Kim, S.
casting in polymeric formulation. Thromb. Res. 92:149Y156 (1998).
27. J. O. Nam, J. E. Kim, H. W. Jeong, S. J. Lee, B. H. Lee, J. Y.
Choi, R. W. Park, J. Y. Park, and I. S. Kim. Identification of the
Y. Kim, and Y. Byun. Heparin-deoxycholic acid chemical
conjugate as an anticancer drug carrier and its antitumor
activity. J. Control. Release 114:300Y306 (2006).
anb3 integrin-interacting motif of big-h3 and its anti-angiogenic 34. K. Park, Y. S. Kim, G. Y. Lee, J. O. Nam, S. K. Lee, R. W. Park,
effect. J. Biol. Chem. 278:25902Y25909 (2003).
28. T. Barzu, J. L. RijnVan, M. Petitou, P. Molho, G. Tobelem,
and J. P. Caen. Endothelial binding sites for heparin. Specific-
S. Y. Kim, I. S. Kim, and Y. Byun. Antiangiogenic effect of bile
acid acylatedfs heparin derivative. Pharm. Res. 24:176Y185
(2007).