doi:10.1006/cyto.2000.0692, available online at http://www.idealibrary.

com on

SHORT COMMUNICATION

IDENTIFICATION OF POLYMORPHISMS WITHIN

THE VASCULAR ENDOTHELIAL GROWTH
FACTOR (VEGF) GENE: CORRELATION WITH
VARIATION IN VEGF PROTEIN PRODUCTION
Carolyn J. Watson,1 Nicholas J. A. Webb,2 Martyn J. Bottomley,1
Paul E. C. Brenchley1

Dysregulated vascular endothelial growth factor (VEGF) expression has been implicated as a
major contributor to the development of a number of common disease pathologies. The aim of
this study was to establish the extent of genetic variability within the VEGF gene and to
determine whether this genetic variation influenced levels of VEGF protein expression. The
promoter region and exon 1 of the VEGF gene were screened for polymorphisms using
single-stranded conformation (SSCP) polymorphism analysis and direct PCR-sequencing. We
identified 15 novel sequence polymorphisms most of which were rare. Eleven of these polymor-
phisms were single base substitutions, three were single base insertions and one was a two base
deletion. Thirteen of the polymorphisms were located within the promoter and two in the 5
untranslated region (5 UTR) of the gene. We established PCR-RFLP typing systems for ten of
the polymorphisms. For the two common polymorphisms at
460 and +405, we developed a
combined sequence specific priming (SSP) PCR typing system to determine the cis/trans
orientation of each allele and hence, ascertain haplotypes. A significant correlation was observed
between lipopolysaccharide (LPS) stimulated peripheral blood mononuclear cell (PBMC)
VEGF protein production and genotype for the +405 polymorphism.
 2000 Academic Press

Vascular endothelial growth factor (VEGF) is
an endothelial cell-specific mitogen, which promotes
angiogenesis and is a potent mediator of vascular
permeability.1 VEGF gene expression is modulated by
a variety of effectors including cytokines,1,2 lipopoly-
saccharide (LPS),3 hormones1 and hypoxia.2 Dysregu-
lated VEGF expression is implicated in a number of
disease pathologies. Increased VEGF expression
resulting in inappropriate VEGF-induced angiogenesis
is linked with tumour growth and metastasis,4 rheuma-
From the 1Department of Medicine, Manchester University,
Manchester M13 9PT, UK; 2Department of Paediatric Nephrol-
ogy, Royal Manchester Children’s Hospital, Pendlebury,
Manchester M27 4HA, UK
Correspondence to: Dr C. J. Watson, Immunology Research Lab-
oratories, Manchester Institute of Nephrology and Transplan-
tation, B Floor, Manchester Clinic, Manchester Royal Infirmary,
Oxford Road, Manchester M13 9WL, UK
Received 13 January 2000; received in revised form 11 February
2000; accepted for publication 7 March 2000
 2000 Academic Press
1043–4666/00/081232+04 $35.00/0
KEY WORDS: genotype/LPS/PBMCs/polymorphism/VEGF
1232
toid arthritis5 and diabetic retinopathy.6 Conversely,
the ability to produce VEGF in response to hypoxia is
linked to the development of collateral vessels and
protection against myocardial disease.7
Our own data3,8 and those of others7 suggest that
there is, in response to specific stimuli, considerable
variation between individuals in their levels of VEGF
expression. Our aim was to determine whether the
variation in VEGF protein production observed was
due to the presence of polymorphic sequences within
the promoter or 5 untranslated region (5 UTR) of the
VEGF gene.
RESULTS
Identification of novel polymorphisms
1262 bp of the promoter and all of exon 1 of the
VEGF gene were screened for sequence variations
using PCR-SSCP analysis and sequencing in a normal
population of 115 mixed race healthy individuals.
A total of 15 different sequence polymorphisms
CYTOKINE, Vol. 12, No. 8 (August), 2000: pp 1232–1235
 VEGF gene polymorphisms / 1233

TABLE 1. The 15 VEGF gene sequence polymorphisms ated within each of the forward primers to create a
identified and their corresponding allele frequencies in a normal BstUI site when the
460C allele was present and
population of 115 individuals
a BsrSI site when the
417T allele was present.
Sequence Allele frequencies in the normal
polymorphism Positionb healthy population (n=230)
460 C<T/ +405 G<C SSP typing and
haplotype analysis
G<C +405 0.709 0.291 Combination of each of the forward SSP primers
T insertiona +183/+191 0.996 0.004
T insertiona
26/
35 0.996 0.004
specific for the two alleles at
460 with each of the
C<A
37 0.983 0.017 reverse SSP primers specific for the two alleles at +405
A<C
141 0.935 0.065 allowed for detection of the four possible haplotypes
G<Ac
152 0.983 0.017
(see Fig. 1). Expected and observed haplotype fre-
C<Tc
160 0.983 0.017
C<Ta
165 0.996 0.004 quencies are shown in Table 3. The
460C/+405G
C<A
172 0.987 0.013 haplotype was found to be the most commonly
AG deletion
247/
250 0.991 0.009 observed haplotype in the normal population. The
T<C
417 0.974 0.026
C<T
460 0.526 0.474
460C/+405C haplotype was very rare only being
C<Aa
914 0.996 0.004 observed in one out of the 230 chromosomes analysed
G insertiond
1045/
1051 0.987 0.013 (frequency=0.004).
C<Ad
1117 0.987 0.013

a
Only observed in one individual, btranscription start= +1 (from ref. 9), c and
d
polymorphisms inherited together.
were identified. The 15 polymorphisms and the
corresponding allele frequencies are listed in Table 1.
Eight of the polymorphisms identified either cre-
ated a new restriction endonuclease recognition site or
destroyed an existing site and could be detected using
an RFLP typing approach, as detailed in Table 2. For
the C<T base change at
460 and the T<C base
change at
417, a single base alteration was incorpor-
Associations between PBMC VEGF protein
production, gender and genotype
PBMCs from 21 individuals were cultured for 72 h
alone, with LPS or with cobalt ions and supernatant
VEGF protein levels were measured. Individuals were
genotyped for the
460 and +405 polymorphisms
and levels of PBMC VEGF protein production (stimu-
lated minus control values) were compared with regard
to gender and genotype for each polymorphism. No
differences were observed between males (n=11) and
females (n=10) for LPS- or cobalt-stimulated VEGF

TABLE 2. PCR primer pairs and reaction conditions used for RFLP typing, with the corresponding polymorphic restriction
enzyme sites, PCR product size and the restriction fragment sizes observed for each allele for ten of the polymorphisms. PCR
amplification was carried out in 20
l reactions as described previously10 using 100 ng of genomic DNA with 3–5 mMa MgCl2 and
samples were denatured at 94C for 3 min followed by 32 cycles at 94C for 1 min, 55–65C* for 1 min 72C for 1 min, with a final
step of 72C for 5 min

PCR DNA
Sequence product size PCR conditionsa Restriction fragment sizes
polymorphism Primers (bp) (annealing temp. MgCl2) enzyme site Alleles (bp)

+405 G<C F5 -ATTTATTTTTGCTTGCCATT-3 304 55C, 4 mM BsmFI G 193, 111

R5 -GTCTGTCTGTCTGTCCGTCA-3 C 304

141 A<C F5 -CCCCTGCCCCCTTCAATA-3 263 60C, 3 mM HhaI A 159, 104
R5 -AGCCTCAGCCCCTCCACA-3 C 124, 104, 35

152 G<A F as for
141 A<CF 456 60C, 3 mM BsrSI G 241, 215
R5 -CACCTAACCGCTGCGCCT-3 A 456

160 C<T F as for
141 A<CF 263 60C, 3 mM MwoI C 105, 53, 46, 36, 23
R as for
141 A<CR T 141, 53, 46, 23

165 C<T F as for
141 A<CF 456 60C, 3 mM BstNI C 252, 204
R as for
152 G<AR T 456

172 C<A F as for
141 A<CF 263 60C, 3 mM FokI C 137, 126
R as for
141 A<CR A 137, 65, 61

417 T<C F5 -AGGGGTCACTCCAGGATTCCAG-3 134 55C, 3 mM BsrSIb T 115, 19
R5 -TACGTGCGGACAGGGCCTGA-3 C 134

460 C<T F5 -TGTGCGTGTGGGGTTGAGCG-3 175 60C, 5 mM BstUIb C 155,20
R as for
417 T<CR T 175

914 C<A F5 -TTTTGCCAGACTCCACAG-3 253 55C, 3 mM TfiI C 253
R5 -CAGTGTGTCCCTCTGACAA-3 A 170, 83

1117 C<A F5 -AACCCCCATTTCTATTCAG-3 278 55C, 4 mM AlwNI C 146, 132
R5 -CTGTGGAGTCTGGCAAAA-3 A 278

b
Restriction site created by base change in the forward primer. G and C denote bases changed to create restriction enzyme sites.
1234 / Watson et al. CYTOKINE, Vol. 12, No. 8 (August, 2000: 1232–1235)

(A) (B) (n=8; 660.7 pg/ml114.5), intermediate production

1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8
was observed for the GC genotype (n=10; 537.4 pg/
ml73.4) and lowest production for the CC genotype
(n=3; 262.6 pg/ml99.5). There was no correlation
between the +405 genotype and cobalt-stimulated
901–903 bp production.
297 bp

DISCUSSION

(C) (D) We have identified 15 novel sequence polymor-

1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8
901–903 bp
297 bp
Figure 1. Examples of the PCR results obtained from
460/+405
SSP typing.
Combination of each of the
460 forward with each of the +405
reverse SSP primers allowed for detection of the four possible
haplotypes. PCR reactions were carried out as described in Table 2
using 3.5 mM MgCl2, 12 pmoles of each SSP and 8 pmoles of each
control primer and an annealing temperature of 65C. PCR products
were electrophoresed through a 1% agarose gel. (A)
460T and
+405G primers; (B)
460C and +405G primers; (C)
460C and
+405C primers and (D)
460T and +405C primers. Sample order
was as follows: 1, 1
g 100 bp ladder (Life Technologies, Paisley,
UK); 2,
460C/+405G homozygote; 3,
460T/+405C homozy-
gote; 4,
460T/+405G homozygote; 5,
460C/+405C,
460C/
+405G heterozygote; 6,
460T/+405C,
460T/+405G hetero-
zygote; 7,
460T/+405G
460C/+405G heterozygote; 8, negative
PCR control. The 901–903 bp band corresponds to the SSP PCR
product and the 297 bp band corresponds to the PCR control
product.
TABLE 3. The expected and observed frequencies for each of
the
460/+405 polymorphism haplotypes within the popu-
lation of 115 normal healthy individuals. The expected haplo-
type frequencies were calculated from the observed allele
frequencies for each polymorphism assuming no linkage
Expected Observed
Haplotype haplotype frequency haplotype frequency

460C/+405G 0.373 0.522

460C/+405C 0.153 0.004

460T/+405G 0.336 0.187

460T/+405C 0.138 0.287
protein production. For the
460 polymorphism no
relationship was seen between genotype and LPS- or
cobalt-stimulated production. For the +405 polymor-
phism there was a significant correlation between
genotype and LPS-stimulated PBMC VEGF protein
production [Spearman correlation (two-tailed)
r=0.4951, P=0.0225]. Highest VEGF production
(meanSEM) was observed for the GG genotype
phisms within the promoter and 5 UTR of the VEGF
gene. We have developed PCR-RFLP typing strategies
for ten of the polymorphisms. For the two most
common polymorphisms at
460 and +405, we have
developed an SSP-PCR typing system to determine the
cis/trans orientation of each allele and hence allow
determination of
460/+405 haplotypes.
Using the software programme TFSEARCH
ver.1.3 (Yutaka Akiyama 1995) and the TRANSFAC
database,11 which identifies transcription factor bind-
ing motifs, the polymorphism at +405 was predicted to
lie within a potential myeloid zinc finger protein
(MZF1) binding site, with the less frequent C allele
possibly reducing the binding specificity of this motif.
The MZF1 transcription factor is involved in
transcriptional regulation during myelopoiesis.12 It
represses transcription in non-haematopoietic cells
and activates transcription in haematopoietic cells13
and has been shown to regulate transcription of
myeloid-specific genes.
In this study, we observed a significant correlation
between +405 genotype and LPS-stimulated PBMC
VEGF protein production with lowest VEGF protein
production observed for CC homozygotes and highest
production for GG homozygotes. This correlation
would suggest that LPS may be working through the
MZF1 transcription factor and that disruption of the
MZF1 binding site due to the presence of a C allele
may reduce VEGF gene transcription and protein
production.
Future studies need to be conducted using
reporter gene constructs to determine if the alleles of
the +405 polymorphism have a direct effect on the
level of gene transcription and ultimately protein
production in PBMCs and in other cell types.
Dysregulated VEGF expression has been impli-
cated in the pathogenesis of a number of diseases.
Detailed disease association studies are in progress to
determine if any of these polymorphisms are linked
with a particular pathology. Our initial description of
eight of the rare polymorphisms showed no associ-
ations with proteinuric renal disease.14 However, the
newly described common polymorphisms at
460 and
+405 may prove to be useful genetic markers of
VEGF-linked pathologies in arthritis, cancer, diabetes,

ischaemic heart disease and proteinuric nephropathies.
Furthermore, the correlation of VEGF production,
initiated by different physiological stimuli, with a
specific genotype may allow identification of
individuals who exhibit high or low levels of VEGF
production predisposing them to VEGF-mediated
pathologies.
MATERIALS AND METHODS
SSCP analysis
PCR product (5
l) and 2
l of formamide dye were
denatured at 95C for 5 min, electrophoresed through 10%
39:1 polyacrylamide gels in 1Tris-borate-EDTA at 10C
for 16 h at 10–14 mA then visualized by silver staining of
the gels.
Fluorescent direct PCR-cycle sequencing
PCR fragments were purified then directly cycle-
sequenced using the Cy5 Thermo Sequenase dye terminator
kit and reactions were analysed on the ALFexpress auto-
mated DNA sequencer (Amersham Pharmacia Biotech,
Amersham, UK).
Restriction fragment length polymorphism
(RFLP) typing
PCR product (15
l) was digested with 2–5 U of appro-
priate restriction enzyme for 3 h at the recommended tem-
perature and the digested DNA fragments were analysed on
3% 1TBE agarose gels.
SSP haplotyping of the
460 C<T and the
+405 G<C polymorphisms
Forward SSP primers were designed to distinguish
between the two alleles at
460 (
460C 5 -GCGTG
TGGGGTTGAGGGC-3 ,
460T 5 -TGCGTGTGGGGT
TGAGGGT-3 ) and reverse primers were designed to distin-
guish between the two alleles at +405 (+405G 5 - GTC
ACTCACTTTGCCCCTGTCC-3 , +405C 5 -TCACTCAC
TTTGCCCCTGTCG-3 ). Amplification control primers
were designed which amplified a part of the first exon of the
human transforming growth factor beta-1 (TGF-1) gene
(TGFF 5 -GCTACCGCTGCTGTGGCTACT-3 nucleo-
tide 2012-2032, TGFR 5 -ACGCGGGTGACCTCC
TTG-3 nucleotide 2308-2291 Genbank accession no.
X05839) and cycled using the same PCR conditions as the
SSP primers.
VEGF production from cultured peripheral blood
mononuclear cells (PBMCs)
PBMCs were isolated from the whole blood of healthy
individuals (n=21, ten females and 11 males) and cultured, in
triplicate, for 72 h either alone or with 1 ng/ml of LPS
(Quadratech, Epsom, UK) or 50
M cobalt chloride as
previously described.15 PBMC culture supernatants were
harvested after 72 h and protein levels measured using a
VEGF gene polymorphisms / 1235
VEGF Quantikine ELISA kit (R and D Systems, Abingdon,
UK).
Acknowledgement
This work was supported by grant B0626 from the
Arthritis Research Campaign, UK.
REFERENCES
1. Ferrara N (1999) Molecular and biological properties of
vascular endothelial growth factor. J Mol Med 77 (7):527–543.
2. Neufeld G, Cohen T, Gengrinovitch S, Poltorak Z (1999)
Vascular endothelial growth factor (VEGF) and its receptors.
FASEB J 13 (1):9–22.
3. Watson CJ, Webb NJA, Bottomley MJ, Brenchley PEC
(1996) Peripheral mononuclear cells express VEGF and the VEGF
receptor flt-1. J Amer Soc Nephrol 7 (9):A2390.
4. Claffey KP, Robinson GS (1996) Regulation of VEGF/VPF
expression in tumour cells: consequences for tumour growth and
metastasis. Cancer Metastasis Rev 15 (2):165
176.
5. Koch AE, Harlow LA, Haines GK, Amento EP, Unemori
EN, Wong WL, Pope RM, Ferrara N (1994) Vascular endothelial
growth factor. A cytokine modulating endothelial function in
rheumatoid arthritis. J Immunol 152 (8):4149
4156.
6. Miller JW, Adamis AP, Aiello LP (1997) Vascular endothe-
lial growth factor in ocular neovascularization and proliferative
diabetic retinopathy. Diabetes Metab Rev 13 (1):37–50.
7. Schultz A, Lavie L, Hochberg I, Beyar R, Stone T, Skorecki
K, Lavie P, Roguin A, Levy AP (1999) Interindividual heterogeneity
in the hypoxic regulation of VEGF: significance for the development
of coronary artery collateral circulation. Circulation 100:547–552.
8. Webb NJA, Bottomley MJ, Watson CJ, Brenchley PEC
(1998) Vascular endothelial growth factor (VEGF) is released
from platelets during clotting : implications for measurement of
circulating levels in clinical disease. Clin Sci 94 (4):395–404.
9. Tischer E, Mitchell R, Hartman T, Silva M, Gospodarowicz
D, Fiddes JC, Abraham JA (1991) The human gene for vascular
endothelial growth factor: multiple protein forms are encoded
through alternative exon splicing. J Biol Chem 266 (18):11947–
11954.
10. Webb NJA, Watson CJ, Roberts ISD, Bottomley MJ, Jones
CA, Lewis MA, Postlethwaite RJ, Brenchley PEC (1999) Circulating
vascular endothelial growth factor is not increased during relapses of
steroid-sensitive nephrotic syndrome. Kidney Int 55 (3):1063–1071.
11. Heinemeyer T, Wingender E, Reuter I, Hermjakob H,
Kel AE, Kel OV, Ignatieva EV, Ananko EA, Podkolodnaya OA,
Kolpakov FA, Podkolodny NL, Kolchanov NA (1998) Databases
on transcription regulation: TRANSFAC, TRRD and COMPEL.
Nucl Acids Res 26:364–370.
12. Lenny N, Westendorf JJ, Hiebert SW (1997) Transcrip-
tional regulation of myelopoiesis. Mol Biol Rep 24 (3):157–168.
13. Hromas R, Davis B, Rauscher III FJ, Klemsz M, Tenen D,
Hoffman S, Xu D, Morris JF (1996) Hematopoietic transcriptional
regulation by the myeloid zinc finger gene, MZF-1. Curr Top
Microbiol Immunol 211:159–164.
14. Watson CJ, Webb NJA, Bottomley MJ, Brenchley PEC
(1998) Polymorphisms in the vascular endothelial growth factor gene
(VEGF) promoter are not associated with the development of
proteinuria in minimal change nephrotic syndrome (MCNS) and
focal segmental glomerular sclerosis (FSGS). J Amer Soc Nephrol
9:A2497.
15. Bottomley MJ, Webb NJA, Watson CJ, Holt L, Bukhari
M, Denton J, Freemont AJ, Brenchley PEC (2000) Placenta growth
factor (PlGF) induces vascular endothelial growth factor (VEGF)
secretion from mononuclear cells and is co-expressed with VEGF in
synovial fluid. Clin Exp Immunol 119 (1):182–188.