Int J Clin Oncol (2005) 10:318327

DOI 10.1007/s10147-005-0508-7

The Japan Society of Clinical Oncology 2005

ORIGINAL ARTICLE
Yutaka Yonemura Yoshio Endo Kayoko Tabata
Taiichi Kawamura Hyo-Yung Yun Etsurou Bandou
Takuma Sasaki Masahiro Miura

Role of VEGF-C and VEGF-D in lymphangiogenesis in gastric cancer

Received: February 3, 2005 / Accepted: May 20, 2005

Abstract
Background. The molecular mechanisms of lymphangiogenesis induced by vascular endothelial growth factor
(VEGF)-C and VEGF-D in gastric cancer were studied.
Methods. VEGF-C and VEGF-D gene expression vectors
were transfected into the gastric cancer cell line KKLS,
which did not originally express VEGF-C and VEGF-D,
and stable transfectants (KKLS/VEGF-C and KKLS/
VEGF-D) were established. The cell lines were inoculated
into the subserosal layer of the stomach and subcutaneous
tissue of nude mice.
Results. VEGF-C and VEGF-D expression in KKLS/
VEGF-C and KKLS/VEGF-D cells was found by reverse
transcription polymerase chain reaction (RT-PCR) and
Western blot analysis. Expression of mouse VEGF receptor
(VEGFR)-2 and mouse VEGFR-3 mRNA was detected in
the KKLS/VEGF-C and KKLS/VEGF-D gastric tumors.
Newly formed lymphatic vessels were detected not only in
the periphery but also in the center of the tumors. The
intratumor lymphatic vessels connected with the preexisting
lymphatic vessels in the muscularis mucosa. The average
numbers of lymphatic vessels in KKLS/VEGF-C (52.0
9.5) and KKLS/VEGF-D (16.4 0.6) gastric tumors were
significantly higher than that in the KKLS/control vector
tumors (4.0 1.4).
Conclusion. VEGF-C and VEGF-D may induce neoformation of lymphatic vessels in experimental gastric tumors by
the induction of VEGFR-3 expression.

Y. Yonemura (*) T. Kawamura E. Bandou

Gastric Surgery Division, Shizuoka Cancer Center, 1007
Shimo-Nagakubo, Nagaizumi-Machi, Shizuoka 411-8777, Japan
Tel. +81-55-989-5235; Fax +81-55-989-5793
e-mail: y.yonemura@scchr.jp
Y. Endo K. Tabata T. Sasaki
Experimental Therapeutics Cancer Research Institute, Kanazawa
University, Kanazawa, Japan
H.-Y. Yun
Chungbuk University, Department of Surgery, Chungbuk, Korea
M. Miura
Department of Anatomy, Oita Medical University, Oita, Japan

Key words Vascular endothelial growth factor (VEGF)-C

VEGF receptor-3 (VEGFR-3) Gastric cancer VEGF-D
Lymphangiogenesis

Introduction
Gastric cancer is the fourth most common malignancy in the
world, and is the second leading cause of death after lung
cancer.1 About 60% of resectable gastric cancers have
lymph node metastasis at the time of diagnosis, and lymph
node metastasis is the strongest prognostic factor. However,
little is known about the molecular mechanisms of lymph
node metastasis in gastric cancer. The multistep process of
lymph node metastasis is initiated by cancer cell invasion
into lymphatic vessels. Proliferation of lymphatic vessels
is a common finding in gastric cancer tissue, and this
favors the permeation of cancer cells into the lymphatic
channels. Accordingly, lymphangiogenesis around cancer
cells may have a crucial role in the initial step of lymph node
metastasis.
Vascular endothelial growth factor (VEGF)-C is wellknown to induce lymphangiogenesis by activating VEGF
receptor (VEGFR)-3, which is expressed on lymphatic
endothelial cells.2
Activation of VEGFR-1 and VEGFR-2 induces mitogenic capacity in endothelial cells of blood vessels.3 VEGF-C
can also activate VEGFR-2 and VEGFR-3, and this activation stimulates both angiogenesis and lymphangiogenesis.3
Recently, we have found that a VEGF-C and VEGFR-3
paracrine loop has a big role in the lymphatic proliferation
of clinical gastric cancer, and that the overexpression of
VEGF-C induces lymph node metastasis.4,5 Induction of
lymphangiogenesis by VEGF-C has been demonstrated in
breast cancer and lung cancer.6,7
Yamada et al.8 identified a new member of the VEGF
family, designated VEGF-D; its amino-acid sequence
shows 23.3% identity with that of VEGF-C. Achen et al.9
reported that VEGF-D binds both VEGFR-2 and VEGFR3, and can activate the receptors similarly to VEGF-C.

319

VEGF-D and VEGF-C thus share two receptors, but

the interrelationships between VEGF-C and VEGF-D in
lymphangiogenesis in gastric cancer remain unknown.
The present study demonstrated the actions of VEGF-C
and VEGF-D in lymphangiogenesis and in metastatic
ability in experimental tumors, using human gastric cancer
cell lines that overexpressed the VEGF-C and VEGF-D
genes.

Materials and methods

Cell lines
The human gastric cancer cell lines, AZ-521, TMK-1,
NUGC-4, NUGC-3, KATO III, MKN-45, MKN-28, MKN7, NKPS, and KKLS; the human fibrosarcoma cell line, HT1080; the human prostatic cancer cell line PC-3, and the
human fibroblast cell line KMST-6 were provided by the
Japanese Collection of Research Bioresources (JCRB,
Tokyo, Japan). The cell lines were maintained in RPMI
1640 (Gibco BRL, Grand Island, NY, USA) with 5% fetal
calf serum (Invitrogen, Carlsbad, CA, USA).
Construction of VEGF-C and VEGF-D expression
vectors, cell transfection, and selection
cDNA was produced from the total RNA sample of
human lung cancer tissue by reverse transcriptionpolymerase chain reaction (RT-PCR), using KOD-plus
DNA polymerase (Toyobo, Osaka, Japan) with high
fidelity. The primers were: sense, 5-GGATATTGAAATA
TTCAAAATGTACAGAG-3 and anti-sense, 5-TGTTA
AAAATGACAGGGATGGGGAACTTG-3.
Human
VEGF-D cDNA was cloned into a pEF6/V5-His TOPO
expression vector (Invitrogen, San Diego, CA, USA) that
contains the blasticidin resistant gene, or with the vector
alone (control vector) using Fugene 6 transfection reagent
(Roche, Indianapolis, IN, USA). The full-length VEGF-D
cDNA included the native stop codon to express the native
protein. Transfected cells were selected in growth medium
containing 40 mg/ml blasticidin-S hydrochloride (Funakoshi,
Tokyo, Japan).
To construct a VEGF-C expression vector, human
VEGF-C cDNA was prepared from a total RNA sample of
human malignant glioma U-87MG by RT-PCR, using
KOD-plus DNA polymerase. PCR was performed using the
sense primer, 5-AAGCTGGCTAGTTAAGCTTCCACC
ATGCACTTGCTGGGCTT-3, containing the HindIII
site and the anti-sense primer, 5-TGGATATCTGCAG
AATTCCATAATAGAAAATCGATGAACTG-3, containing the EcoRI site. The RT-PCR product containing the
full-length coding region and the native stop codon of
human VEGF-C mRNA was digested with HindIII and
EcoRI. The fragment was inserted into the pcDNA4-mycHis A vector digested with HindIII and EcoRI. KKLS cells
were transfected either with the pcDNA4-myc-His vector
containing the VEGF-C cDNA (pcDNA4-VEGF-C) or

with the vector alone (control vector) using Fugene 6 transfection reagent. Transfected cells were selected and maintained in growth medium containing 40 mg/ml Zeocin
(Invitrogen).
Stable transfectants of KKLS were designated KKLS/
VEGF-C and KKLS/VEGF-D.
Enzyme-linked immunosorbent assay (ELISA)
Culture medium and anti-VEGF-C antibody (diluted 1 : 200
with bovine serum albumin-phosphate buffered saline
[BSA-PBS], C-20, Santa Cruz Biotech, Santa Cruz, CA,
USA) or anti-VEGF-D antibody (1 : 500; C-18, Santa Cruz
Biotech) were added to a 96-well plate and incubated at
room temperature for 60 min. Wells were washed with a
TBS-Tween buffer. Alkaline phosphatase-conjugated antigoat secondary antibody (diluted 1 : 1000 with BSA-PBS)
was added to each well. After 60-min incubation, the plate
was washed, and p-nitrophenyl phosphate (Chemicon,
Temecula, CA, USA), was dispensed to each well, followed
by incubation for 60 min. The yellow color of nitrophenol
was measured at 405 nm. Serial dilutions (four fold) of
human recombinant VEGF-C (Santa Cruz Biothech) and
VEGF-D (Santa Cruz Biothech), from 20 mg/ml to 5 pg/ml,
were used to make a calibration curve.
Reverse transcription (RT)-polymerase chain
reaction (PCR)
RT-PCR analysis was performed using a modification of
the method of Conboy et al.10 After the isolation of RNA
by Isogen (Nippon Gene, Tokyo, Japan),11 the RNA was
mixed with oligo-dT primer and reverse-transcribed with
AMV reverse transcriptase (Life Sciences, St. Petersburg,
FL, USA), followed by PCR amplification (Perkin-Elmer
Cetus, Norwalk, CT, USA) with primers specific for human
genes and mouse genes specific for lymphangiogenesis.4
The primers specific for human and mouse genes are shown
in Table 1.
Western blot analysis
Protein samples were mixed with sample buffer and
dithiothreitol (DTT) prior to separation with the NuPAGE
Electrophoresis System (Novex, San Diego, CA, USA).4
After the completion of electrophoresis, samples were
transferred to polyvinglidene (PVDF) membranes
(Immobilon; Millipore, Bedford, MA, USA). The membranes were incubated with antibodies (anti-human VEGFC [H190], anti-human VEGF-D [C-18], anti-mouse CD-31
[M-20; Dako, Copenhagen, Denmark], anti-mouse LYVE1 [Santa Cruz Biotech], and anti-mouse VEGFR-3 (Ab
3127; Chemicon) for 2 h. Then the membranes were incubated with the secondary antibodies specific for each
primary antibody for 40 min, rinsed, and treated with
an electrochemi luminescence (ECL) detection reagent
(Amersham Life Science, London, UK).

320
Table 1. Primers specific for anglogenesis- and lymphanglogenesis-related human and mouse genes
Target gene

Product size (bp)

Sequence

Human VEGF-A

737, 541, 409

Sense: 5-TCGGGCCTCCGAAACCATGAACTTTCT-3
Antisense: 5-GGTTTCTGGATTAAGGACTGTTCTGTC-3

Human VEGF-C

536

Sense: 5-ATTTGCTGCAGCACATTATAATACAGAGAT-3
Antisense: 5-TCACTATATGAAAATCCTGGCTCACAAGCC-3

Human VEGF-D

514

Sense: 5-TGAGAGTGGTCTTCTGTTCCAGCAAGTGGA-3
Antisense: 5-GCTCAGCATCCCATCGGTCCACTAGGTTTG-3

Human TFR

415

Sense: 5-ACAGACTCTACATGTAGGAT-3
Antisense: 5-AAACCTTGAAGTTGCTGGTA-3

Human VEGFR-1

545

Sense: 5-GCTGCAAATATCTAGCTGTACCT-3
Antisense: 5-GGAATTGCTTTGGTCAATTCGTC-3

Human VEGFR-2

311

Sense: 5-AGACTTTGAGCATGGAAG-3
Antisense: 5-CCATTCCACCAAAAGATG-3

Human VEGFR-3

298

Sense: 5-AGCCATTCATCAACAAGCCT-3
Antisense: 5-GGCAACAGCTGGATGTCATA-3

Human b-actin

591

Sense: 5-GAAAATCTGGCACCACACCTT-3
Antisense: 5-TTGAAGGTAGTTTCGTGGAT-3

Mouse VEGF-A

269

Sense: 5-GGGTTTCGGGAACCAGACCTCTCAC-3
Antisense: 5-CAACCCTAATCTTCCGGGCTTGGCG-3

Mouse VEGFR-1

397

Sense: 5-ATGGCTCGCCTGCAACATTGAAGTC-3
Antisense: 5-CTCTCAATTCTGTTTCCTAAGTTGCTGC-3

Mouse VEGFR-2

700

Sense: 5-TCTGTCAGTGACCAGCATGGCATCGT-3
Antisense: 5-GGAGCTGGGTAACTGAGATACTTCACA-3

Mouse VEGFR-3

699

Sense: 5-CTCCATGACCCCTCCAACCCTGAA-3
Antisense: 5-CTCAGCCCACACTGTACAGTTCAAA-3

Mouse CD31

322

Sense: 5-TCAGAATTTCAGCAAGATCGCCGAGG-3
Antisense: 5-GCACTGGTATTCCATGTCTCTGGTGG-3

Mouse b-actin

540

Sense: 5-GTGGGCCGCTCTAGGCACCAA-3
Antisense: 5-CTCTTTGATGTCACGCACGATTTC-3

Animal experiments

Evaluation of lymphatic vessel count

For implantation, subconfluent KKLS/VEGF-C, KKLS/

VEGF-D, and KKLS/control vector cells were harvested
with trypsin ethylenediamine tetraacetic acid (EDTA) and
resuspended to a final concentration of 1 108/ml PBS.
Using a 30-gauge needle attached to a 1-ml syringe, cells
(1 107 cells/0.1 ml) were injected orthotopically into the
subserosal layer of the body of the stomach and subcutaneous tissue of the back in 5-week-old female BALB/c-nu/nu
mice (Charles River, Hamamatsu, Japan).

The lymphatic vessel count was assessed by light microscopy in the area of the tumor containing the highest numbers of lymphatic vessels. The areas containing the highest
numbers of lymphatic vessels were identified by scanning
tumor sections. After the area with highest number of lymphatic vessels was identified, a vessel count was performed
in a 100 field.
Evaluation of metastatic ability of cancer cells in chick
embryos by PCR

Staining of lymphatic vessels and blood capillaries

On days 14 and 28 after inoculation, gastric tumors and
subcutaneous tumors were removed and fixed with formaldehyde-CaCl2 fixative (2% paraformaldehyde, 1% Ca Cl2)
in 0.1 M cacodylate buffer containing 7% sucrose for 24 h at
4. Then the tissues were embedded in OCT compound
(Miles, Diagnostic Division, Elkhart, IN, USA), and the
sections were cut at 15-mm thickness with a cryostat.12
Lymphatic vessels showing positive staining for 5-Nase
activity revealed the lead sulfide reaction product, colored
dark brown. In contrast, no 5-Nase activity was recognized
in the blood vessels. After staining for alkaline phosphatase
(ALPase), a blue reaction product was restricted to blood
vessels.

Cancer cells (106 cells) were injected into the vein of the
chorioallantoic membrane (CAM) of 11-day chick embryos
in fertilized eggs.32,33 After incubation for 7 days, the embryonic livers and lungs were dissected and DNA was extracted
using a rapid DNA preparation.32 The 576-bp product of the
human b-globin gene segment was amplified by PCR. The
sense and anti-sense primers for the human b-globin gene
were: Hu b1 (5-AGAGCCATCTATTGCTTACA-3) and
Hu b8 (5-TATGACATGAACTTAACCAT-3). DNA
from the embryonic organs was amplified by 25 cycles of
PCR with the primers Hu b1 and Hu b8 and analyzed by
agarose gel electrophoresis. The relative density of the
DNA fragments of the b-globin and b-actin genes on a
scanned gel image were quantified using NIH image.

321
Fig. 1. Vascular endothelial growth factorA (VEGF-A), VEGF-C, and VEGF-D
mRNA expression in gastric cancer cell
lines, a fibroblast cell line (KMST-6), an
osteosarcoma cell line (HT1080), and a prostatic cancer cell line (PC-3), TFR, transferrin receptor

Fig. 2. Protein expression in KKLS control vector cells and stable clones of KKLS cells transfected with VEGF-C or VEGF-D expression
vector

Statistical analysis
All statistical calculations were performed using SPSS
statistical software. Data values are expressed as means
plus or minus SDs. The c2 test and Students t-test
were used to determine the significance of intergroup
differences.

Results
VEGF-A, VEGF-C, and VEGF-D mRNA expression in
human cell lines
As shown in Fig. 1, VEGF-A was expressed in all cell lines.
In contrast, VEGF-C mRNA was expressed in two gastric

cancer cell lines (AZ-521 and TMK-1), and in PC-3,

HT-1080, and KMST-6 (a human fibroblast cell line), but no
cell line expressed VEGF-D mRNA.
VEGF-C, and VEGF-D protein expression in
KKLS/VEGF-C and KKLS/VEGF-D cells
Figure 2 shows the protein expression of VEGF-C
and VEGF-D in KKLS/VEGF-C and KKLS/VEGF-D
cells. Western blot analysis detected precursors of 58-kDa
VEGF-C and VEGF-D in each cell line. Figure 3
shows the results of Western blot analyses for the conditioning medium of cell lines after 1-h treatment with
plasmin. After plasmin treament, 29-kDa C-terminal
peptides of VEGF-C and VEGF-D were detected in the
conditioning medium, but the 58-kDa and 53-kDa

322
Fig. 3. Western blot analysis of VEGFC and VEGF-D in conditioning
medium of stable clones. After 1-h
treatment with plasmin, 29-kDa Cterminal peptides of VEGF-C and
VEGF-D were detected, but the 58kDa and 53-kDa C-terminal peptides
of VEGF-C and VEGF-D precursor
disappeared

C-terminal peptides of VEGF-C and VEGF-D precursor

disappeared.
The concentrations of VEGF-C and VEGF-D in the
conditioning medium of KKLS/VEGF-C and KKLS
VEGF-D cells were 22 and 113 ng/ml, but VEGF-C and
VEGF-D were not detected in the conditioning medium of
KKLS/control vector.
Tumorigenesis of KKLS, KKLS/VEGF-C,
and KKLS/VEGF-D cells in the stomach of nude mice
The average weight of gastric tumors in the KKLS/control
vector group (35.4 26.3 mg) on day 28 was significantly
lower than that in the KKLS/VEGF-C group (1160 54 mg)
or the KKLS/VEGF-D group (660 568 mg; P < 0.05).
mRNA and protein expression of VEGF-related genes in
stomach tumors
The expressions of human VEGF-C and human VEGF-D
mRNA; mouse VEGFR-1, -2, and -3 mRNA; and mouse
CD31 mRNA are shown in Fig. 4. In KKLS/VEGF-C gastric tumors, the expression of mouse VEGFR-1, VEGFR-2,
and VEGFR-3 mRNA was detected. In contrast, KKLS/
VEGF-D tumors expressed mouse VEGFR-2 and mouse
VEGFR-3 mRNA. Mouse CD31 expression was detected
in all cell lines.
Figure 5A shows the results of Western blot analysis
with anti-human VEGF-C and VEGF-D specific antibodies
using protein extracted from gastric tumors. VEGF-C and
VEGF-D were detected in KKLS/VEGF-C and KKLS/

VEGF-D tumors, but not in the KKLS/control vector

tumor. The expression of mouse VEGFR-3, mouse LYVE1, and mouse CD31 in gastric tumors was also examined by
Western blot analysis (Fig. 5B). In the KKLS/VEGF-C and
KKLS/VEGF-D gastric tumors, the expression of the three
proteins was higher than that in the KKLS/control vector
tumors.

Lymphovascular neogenesis in gastric and

subcutaneous tumors
In the stomach tumors on day 14, blood vessels positive for
ALPase reaction were detected diffusely in all the tumors
(Fig. 6). However, 5-Nase-positive lymphatic vessels were
scarce on day 14. Lymphatic vessels in the normal stomach
(Fig. 7E) were found in the laminal propria and intramuscular layer of the stomach wall, but were scarce in the submucosal layer. On day 28, lymphatic vessels stained with
5-Nase in KKLS/control vector tumor were detected only
in the periphery of tumors (Fig. 7A). In contrast, lymphatic
vessels were detected not only in the peripheral area but
also in the center of the tumors in KKLS/VEGF-C (Fig. 7B)
and KKLS/VEGF-D tumors, The newly formed central
lymphatic vessels connected with the peripheral lymphatic
vessels. In serial sections, newly formed central lymphatic
vessels (Fig. 7G; hatch symbol), connected with the peripheral lymphatic vessels (Fig. 7F, hatch symbol), and drained
into the preexisting collecting lymphatic vessels in the upper
layer of the muscularis mucosa (Fig. 7G; asterisk). In the
subcutaneous tumors of KKLS/VEGF-D cells, marked neoformation of 5Nase-positive lymphatic vessels was found

323
Fig. 4. mRNA expression of angiogenesis- and
lymphangiogenesis-related factors in gastric
tumors bearing KKLS/control vector, KKLS/
VEGF-C, and KKLS/VEGF-D cell stable
clones. mVEGFR-1, mouse VEGF receptor-1

Fig. 5. A Protein expression of

human VEGF-C and VEGF-D in
gastric tumors bearing the stable
clones. VEGF-C and VEGF-D,
of human origin, were detected
in KKLS/VEGF-C and KKLS/
VEGF-D tumors, but not in the
KKLS/control vector tumor. B
Protein expression of mouse CD31, VEGFR-3, and LYVE-1 in
gastric tumors bearing the stable
clones. These molecules, of
mouse origin, were detected in
the gastric tumors bearing stable
clones

324

(Fig. 7D). The number of 5-Nase positive-vessels in the

gastric tumors of KKLS/control vector cells was 4.0 1.4,
and the numbers in the KKLS/VEGF-C and KKLS/VEGFD tumors were 52.0 9.5 and 16.4 0.6, respectively. There
was a statistically significant difference in the number of
lymphatic vessels between the KKLS/control vector group
and the KKLS/VEGF-C and KKLS/VEGF-D tumors (P <
0.05), and the number in the KKLS/VEGF-C tumor was
significantly higher than that in the KKLS/VEGF-D tumor.
The numbers of lymphatic vessels in the subcutaneous
tumor-bearing KKLS control vector, KKLS/VEGF-C, and
KKLS/VEGF-D cells were 3.6 0.6, 16.3 0.5 and 69.7
17.8, respectively. A significant difference was found
between the control vector group and the VEGF-C and
VEGF-D group (P < 0.05). However, the numbers of blood
vessels in the gastric tumor-bearing KKLS control vector,
KKLS/VEGF-C, and KKLS/VEGF-D cells were 41.1
12.5, 45.3 5.7 and 54.0 12.0, respectively. There was no
significant difference in blood vessel numbers between the
three groups.

Fig. 6. Enzyme histochemical evaluation of blood and lymphatic vessels with enzyme-histochemical staining for alkaline phosphatase
(ALPase) and 5-Nase activity in gastric tumor-bearing KKLS/VEGFC cells on day 14. Arterial blood vessels are stained blue. Lymphatic
vessels are scarce. 10

Metastatic ability of cancer cells in chick embryos,

studied by PCR
Figure 8 shows human b-globin gene expression on day 7
in embryonic chick liver and lung. The densities of the
b-globin gene bands in chick liver and lung inoculated with
KKLS/VEGF-C and KKLS/VEGF-D were significantly
higher than those of the KKLS/control vector (Table 2).

Discussion
VEGF-C is known to have a major role in lymphangiogenesis in cancer tissue, but the relationship between
VEGF-D and lymphangiogenesis is controversial.1315 In the
embryo, VEGF-D is expressed by melanocytes, lung connective tissue, and fibroblasts. In normal adult tissue, it is
detected in lymph nodes, heart, and small bowel.24,25
The present study aimed to clarify the mechanisms of
lymphangiogenesis and the metastatic ability of VEGF-C
and VEGF-D in gastric cancer, using KKLS cells and the
stable transfectants of VEGF-C and VEGF-D genes.
Among the ten gastric cancer cell lines tested, VEGF-C
mRNA was expressed in two cell lines, and it was also
expressed in a fibroblast cell line, KMST-6. Accordingly,
not only cancer cells but also fibroblasts may have a role in
lymphangiogenesis in gastric cancer tissue. We have already
reported that VEGF-C mRNA was expressed in 47% of
primary gastric cancers.4
VEGF-D mRNA, in contrast was not expressed in any
gastric cancer cell lines or in fibroblasts. Ishikawa et al.33
reported that VEGF-D mRNA expression was detected
in only 17% of clinical gastric cancers.23 In colon tissue,
VEGF-D is expressed in the normal gland rather than
in colon cancer or adenoma.31 In contrast, VEGF-D expression is found in 60%80% of breast cancers and oral
cancers.21,22
From these results, it seems that VEGF-D may not have
a main role in lymphangiogenesis in gastric cancer, because
of the low expression rate of VEGF-D in gastric cancer,
although, as shown in the present study, once gastric cancer
cells express VEGF-D, lymphangiogenesis is induced.

Table 2. Metastatic ability of KKLS cells overexpressing VEGF-C and VEGF-D in chick embryo
Clone

Liver metastasis

Lung metastasis

Control vector

35.4 26.3

5.7 8.6

KKLS/VEGF-C

75.1 16.2
P = 0.006 (vs control vector)

93.2 37.8
P = 0.001 (vs control vecctor)

Control vector

14.8 8.8

17.7 14.4

KKLS/VEGF-D

37.0 14.9
P = 0.007 (vs control vecctor)

58.8 11.4
P = 0.0002 (vs control vecctor)

Values are means SD

Tumor cells (106 cells/embryo) were inoculated into the CAN vein of a 10-day-old chick embryo.
After 7 days, livers and lungs were dissected and DNA was extracted. The sample DNA was
analyzed by PCR, using specific primer pairs of the human b-globin gene. After electrophoresis,
the density of the 576-bp DNA fragment on the scanned gel image was quantified using NIH
image

325
Fig. 7AG. Enzyme histochemical
evaluation of lymphatic vessels with
enzyme-histochemical staining for
5-Nase activity. Lymphatic vessels
are stained brown. A Gastric tumorbearing KKLS/control vector cells
on day 28. Lymphatic vessels are not
found in the center of the tumor.
B Gastric tumor-bearing KKLS/
VEGF-C cells on day 28. Lymphatic
vessels are rich in the periphery of
the tumor, and are also found in the
central part of the tumor. C Normal
skin stained for 5-Nase. Lymphatic
vessels are found (asterisks). D Subcutaneous tumor stained for 5-Nase.
Widespread lymphangiogenesis was
detected in the subcutaneous tumorbearing KKLS VEGF-D cells on day
28. E ALP-ase-5-Nase double-staining of normal mouse stomach. Lymphatic vessels are stained brown
(arrows in E, F, G). E, F, G Two
slices from serial sections of KKLS/
VEGF-C gastric tumor. Lymphatic
vessel (hatch symbol in F) connects
with an intratumoral lymphatic
vessel (hatch symbol in G) and a collecting lymphatic vessel in the muscularis mucosa (mm; asterisk in G).
lp, lamina propria; ss, subserosa;
BV, blood vessel; L, lymphatic
vessel; mp, proper muscle

E
A

Fig. 8. Detection of human b-globin gene (576 bp) in the livers and lungs of chick embryos by polymerase chain reaction (PCR), using a specific
primer pair (Hub-1 and Hub-8). Six embryos (lanes 16) for each group were studied for the metastatic ability of cancer cells. M, marker; HT,
HT-1080

326

However, in contrast, VEGF-D may have a major role in

lymphangiogenesis in breast and oral cancers.2123
VEGF-C and VEGF-D precursors are processed to the
active form by plasmin, which is activated by plasminogen activator. The precursors of VEGF-C and VEGF-D
(58 kDa and 53 kDa, respectively) were detected in the protein extracted from the stable transfectants. The present
study showed that the precursors of VEGF-C and VEGF-D
in the medium were cleaved by plasmin, and finally became
active forms, of 21 kDa. The homodimer of the 21-kDa
active form activates VEGFR-2 or VEGFR-3.3,26 Because
KKLS cells express plasminogen activator,35 the activation
of VEGF-C or VEGF-D by the plasmin/plasminogen
activator cascade could effectively affect experimental gastric tumors bearing KKLS/VEGF-C and KKLS/VEGF-D
cells.
The difficulties in clarifying the mechanisms of lymphangiogenesis in tumors could be related to the absence of good
markers for lymphatic vessels.19 Recently, specific markers
for lymphatic vessels such as LYVE-1 and VEGFR-3 have
been found, and studies using specific antibodies against
these molecules have clarified the mechanisms of experimental lymphangiogenesis.20
Skobe et al.6 reported that VEGF-C-overexpressing
breast cancer cells induced intratumoral lymphangiogenesis
and lymph node metastasis. Stacker et al.27 described that
VEGF-D induced lymphangiogenesis in a mouse tumor
model. In these studies, LYVE-1 and VEGFR-3 were used
as specific markers for lymphatic vessels.
In the present study, lymphatic and blood vessels were
identified by a double-staining method for 5-Nase and
ALPase; 5-Nase is an enzyme that is involved in the metabolism of nucleotides. 5-Nase activity is very high on the
cell surface of lymphatic endothelial cells, but the activity is
not found on the endothelial cells of blood vessels.28 In
contrast, ALPase activity is high on arterial endothelial
cells, but is low on the lymphatic endothelium. The doublestaining method for 5-Nase and ALPase activities enables
the easy discrimination of lymphatic vessels and blood
vessels. Recent work of Parr and Jiang29 has demonstrated
that 5-Nase expression was significantly correlated with
the expression of VEGFR-3, LYVE-1, and Prox-1. Furthermore, 5-Nase staining enables us to visualize
three-dimensional lymphatic distribution by using thick
slices from specimens.
The present study clearly demonstrated that VEGF-C
and VEGF-D induced lymphangiogenesis in stomach and
subcutaneous tumors. These findings were also supported
by the findings that the overexpression of mouse VEGFR3 mRNA and LYVE-1 protein was found in the KKLS/
VEGF-C and VEGF-D gastric tumors.
Padera et al.16 reported that there was no functional
lymphatic vessel in the center of an experimental tumor,
and that lymph node metastasis started from the invasion of
tumor cells from the lymphatic vessels in the peripheral
zone of a tumor. The intratumoral pressure generated by
rapid tumor growth and vascular pressure may induce the
collapse of lymphatic vessels, and it was presumed that the
intratumoral lymphatic vessels had no function.1618

However, the present study demonstrated that lymphatic

vessels were detected not only in the peripheral zone but
also in the center of tumors, and the newly formed lymphatic vessels in the tumor center connected with the
peripheral lymphatic vessels.
In the KKLS/VRGF-D gastric tumors, newly formed
lymphatic vessels were detected, but prominent lymphangiogenesis was found in the subcutaneous tumors of KKLS/
VEGF-D cells. In the tumors, mouse VEGFR-2 and
VEGFR-3 mRNA and mouse LYVE-1 protein were
overexpressed. Stacker et al.26 described the induction of
lymphangiogenesis in skin by VEGF-D-overexpressing
tumor cells in the severe combined immunodeficient
(SCID) mouse. These results indicate that VEGF-C and
VEGF-D may have different roles in lymphatic endothelial
cell proliferation in the stomach and skin.
Interestingly, lymphangiogenesis was more prominent
on day 28 than on day 14, but angiogenesis was already
found on day 14. These results indicate that lymphangiogenesis occurs later than angiogenesis.
The newly formed blood vessels were found diffusely in
the tumor, but the new lymphatic vessels were found focally. The newly formed lymphatic vessels connected with
the peripheral lymphatic vessels, and drained into the preexisting collecting lymphatic vessels in the muscularis mucosa. These results strongly suggest that the newly formed
lymphatic vessels may bud from the preexisting lymphatic
vessels. The reason for the delay of lymphangiogenesis relative to angiogenesis may be related to the low proliferative
activity of lymphatic endothelial cells or may be due to the
smaller number of preexisiting lymphatic vessels with
proliferative activities in the stomach wall. In contrast,
there was no relationship in the number of blood vessels
between the control tumors and the VEGF-C- or VEGF-Doverexpressing tumors. Three cell lines expressed VEGFA, and VEGF-A may have a major role in the angiogenesis
of tumors. Accordingly, no relationship in the number of
blood vessels was found between KKLS/control vector
tumors and KKLS/VEGF-C or KKLS/VEGF-D tumors.
A significant correlation between VEGF-C/VEGF-D
expression and lymph node metastasis has been reported.
The present study demonstrated that cancer cells overexpressing VEGF-C or VEGF-D increased hematogenous
metastasis in the chick embryo. The increased permeability
of blood vessels induced by VEGF-C may be related to
hematogenous metastasis.34 Furthermore, the growth of the
VEGF-C/VEGF-D-overexpressing cells in the stomach of
nude mice was significantly higher than that of control cells.
From these results, it seems that VEGF-C or VEGF-D
may have increased metastatic potential and proliferative
activity by an autocrine or paracrine mechanism in this
experimental metastatic model.30

Conclusion
The present study demonstrated an intimate relationship
between VEGF-C/VEGF-D and lymphangiogenesis in ex-

327

perimental gastric tumors. We expect the development of

novel drugs to control the VEGF-C pathway in gastric
cancer.

References
1. Ohgaki H, Matsukura N (2003) Stomach cancer. In: Stewart
BW, Kleihues P (eds). World cancer report. IARC, Lyon; pp 197
197
2. Jeltsch M, Kaipainen A, Joukov V, et al. (1997) Hyperplasia of
lymphatic vessels in VEGF-C transgenic mice. Science 276:1423
1425
3. Joukov V, Pajusolo K, Kaipainen A, et al. (1996) A novel vascular
endothelial growth factor, VEGF-C, is a ligand for the Flt4
(VEGFR-3) and KDR (VEGFR-2) receptor tyrosine kinases.
EMBO J 15:290298
4. Yonemura Y, Endo Y, Fujita H, et al. (1999) Role of VEGF-C
expression on the development of lymph node metastasis in gastric
cancer. Clin Cancer Res 5:18231829
5. Yonemura Y, Fushida S, Bando E, et al. (2001) Lymphangiogenesis and vascular endothelial growth factor receptor-3 in
gastric cancer. Eur J Cancer 37:918923
6. Skobe M, Hawighorst T, Jackson DG, et al. (2001) Induction of
tumor lymphangiogenesis by VEGF-C promotes breast cancer
metastasis. Nature Med 7:197198
7. Ohta Y, Shridhar V, Bright RK, et al. (1999) VEGF and VEGF
type C play an important role in angiogenesis and lymphangiogenesis in human malignant mesothelioma tumors. Br J Cancer
81:5461
8. Yamada Y, Nezu J, Shimane M, et al. (1997) Molecular cloning of
a novel vascular endothelial growth factor, VEGF-D. Genomics
42:483488
9. Achen MG, Jeltsch M, Kukk, E, et al. (1998) Vascular endothelial
growth factor D (VEGF-D) is a ligand for the tyrosine kinase
VEGF receptor 3 (Flk-1) and VEGF receptor 3 (Flt4). Proc Natl
Acad Sci USA 95:548553
10. Conboy JG, Chan J, Mohandas N, et al. (1988) Multiple protein 4.1
isoforms produced by alternative splicing in human erythroid cells.
Proc Natl Acad Sci USA 85:90629065
11. Chomczynsky P, Sacchi N (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156159
12. Shimoda H, Kato S, Kudo T (1997) Enzyme-histochemical demonstration of the intramural lymphatic network in the monkey jejunum. Arch Histol Cytol 60:215224
13. Koyama Y, Kaneko K, Akazawa K, et al. (2003) Vascular endothelial growth factor-C and vascular endothelial growth factor-d
messenger RNA expression in breast cancer: association with
lymph node metastasis. Clin Breast Cancer 4:354360
14. O-charoenrat P, Rhys-Evans P, Eccles SA (2001) Expression of
vascular endothelial growth factor family members in head
and neck squamous cell carcinoma correlates with lymph node
metastasis. Cancer 92:556568
15. Niki T, Iba S, Tokunou M, et al. (2000) Expression of vascular
endothelial growth factors A, B, C, and D and their relationships to
lymph node status in lung adenocarcinoma. Clin Cancer Res
6:24312439

16. Padera TP, Kadambi A, di Tomaso E, et al. (2002) Lymphatic

metastasis in the absence of functional intratumor lymphatics.
Science 296:18831886
17. Stacker SA, Baldwin ME, Achen MG (2002) The role of tumor
lymphangiogenesis in metastatic spread. FASEB J 16:922934
18. Leu AJ, Berk DA, Lymboussaki A, et al. (2000) Absence of
functional lymphatics within a murine sarcoma: a molecular and
functional evaluation. Cancer Res 60:43244327
19. Pepper MS (2002) Lymphangiogenesis and tumor metastasis: myth
or reality? Clin Cancer Res 7:462468
20. Nisato RE, Tille JC, Pepper MS (2003) Lymphangiogenesis and
tumor metastasis. Thromb Haemost 90:591597
21. Nakamura Y, Yasuoka H, Tsujimoto M, et al. (2003) Flt-4-positive
vessel density correlates with vascular endothelial growth factor-d
expression, nodal status, and prognosis in breast cancer. Clin
Cancer Res 9:53135317
22. Shintani S, Li C, Ishikawa T, et al. (2004) Expression of vascular
endothelial growth factor A, B, C, and D in oral squamous cell
carcinoma. Oral Oncol 40:1320
23. Ishikawa M, Kitayama J, Kazama S, et al. (2003) Expression of
vascular endothelial growth factor C and D (VEGF-C and -D) is an
important risk factor for lymphatic metastasis in undifferentiated
early gastric carcinoma. Jpn J Clin Oncol 33:2127
24. Achen MG, Williams RA, Minekus MP, et al. (2001) Localization
of vascular endothelial growth factor-D in malignant melanoma
suggests a role in tumour angiogenesis. J Pathol 193:147154
25. Partanen TA, Paavonen K (2001) Lymphatic versus blood vascular
endothelial growth factors and receptors in humans. Microsc Res
Tech 55:108121
26. Stacker SA, Stenvers K, Caesar C, et al. (1999) Biosynthesis of
vascular endothelial growth factor-D involves proteolytic processing which generates non-covalent homodimers. J Biol Chem
274:3212732136
27. Stacker SA, Caesar C, Baldwin ME, et al. (2001) VEGF-D promotes the metastatic spread of tumor cells via the lymphatics. Nat
Med 7:186191
28. Zimmmermann H (1992) 5-Nucleotidase molecular structure and
functional aspects. Biochem J 285:345365
29. Parr C, Jiang WG (2003) Quantitative analysis of lymphangiogenic
markers in human colorectal cancer. Int J Oncol 23:533539
30. Kajita T, Ohta Y, Kimura K, et al. (2001) The expression of vascular endothelial growth factor C and its receptors in non-small cell
lung cancer. Br J Cancer 85:255260
31. Hanrahan V, Currie MJ, Gunningham SP, et al. (2003) The angiogenic switch for vascular endothelial growth factor (VEGF)-A,
VEGF-B, VEGF-C, and VEGF-D in the adenoma-carcinoma
sequence during colorectal cancer progression. J Pathol 200:183
194
32. Tuchiya Y, Sato H, Endo Y, et al. (1993) Tissue inhibitor of
metalloproteinase 1 is a negative regulator of the metastatic ability
of a human gastric cancer cell line, KKLS, in the chick embryo.
Cancer Res 15:13971402
33. Endo Y, Seiki M, Uchida H, et al. (1992) Experimental metastasis
of oncogene-transformed NIH3T3 cells in chick embryo. Jpn J
Cancer Res 83:274280
34. Joukov V, Sorsa T, Kumar V, et al. (1997) Proteolytic processing
regulates receptor specificity and activity of VEGF-C. EMBO J
16:38983911
35. Noguchi-Takino M, Endo Y, Yonemura Y, et al. (1995) Relationship between expression of plasminogen activator system and
metastatic ability in human cancers Int J Oncol 8:97105