Clinical Science (2001) 101, 439–446 (Printed in Great Britain)
Expression of neuropilin-1 by human
glomerular epithelial cells in vitro and in vivo
Steven J. HARPER*, Chang Ying XING†, Cathy WHITTLE*, Robin PARRY*,
David GILLATT‡, Danielle PEAT§ and Peter W. MATHIESON*
*Academic Renal Unit, University of Bristol, Southmead Hospital, Westbury on Trym, Bristol B10 5NB, U.K., †Department of
Nephrology, The First Affiliated Hospital, Nanjing Medical University, Nanjing 210029, People’s Republic of China, ‡Department
of Urology, Southmead Hospital, Westbury on Trym, Bristol B10 5NB, U.K., and §Department of Pathology, Southmead
Hospital, Westbury on Trym, Bristol B10 5NB, U.K.
A
B
S
T
R
A
C
T
Vascular endothelial growth factor (VEGF) is a potent promoter of endothelial mitogenesis and
of endothelial permeability. Within the kidney it is synthesized primarily in the visceral
glomerular epithelial cells (vGECs) ; however, the role of VEGF in the glomerulus remains
unknown, as does the target cell upon which it acts. Although the target cells may be those of
the glomerular endothelium, there are micro-anatomical reasons why this might not be the case.
This, therefore, led us to consider the possibility that glomerular VEGF may bind to the vGECs
themselves. Since it has been shown that vGECs do not express the main tyrosine kinase VEGF
receptors, we chose to study vGEC expression of the more recently described VEGF isoformspecific receptors, the neuropilins. The expression of mRNAs for neuropilin-1, neuropilin-2 and
soluble neuropilin was studied in whole kidney, sieved glomeruli and cultured podocytes by
reverse transcription–PCR, and neuropilin-1 mRNA expression in isolated single glomeruli was
analysed by nested reverse transcription–PCR. The expression of neuropilin-1 protein was
investigated in cultured vGECs by Western blotting and immunocytochemistry, and in normal
kidney sections by immunohistochemistry. Neuropilin-1 mRNA was detected in whole kidney,
single and sieved glomeruli and cultured vGECs. Neuropilin-1 protein was detected in cultured
vGECs and in vGECs in normal kidney sections by immunohistochemistry. Thus the present
study suggests that vGECs may have the potential to bind the VEGF that they secrete. Functional
studies will be required to address the potential significance of this finding in terms of an
autocrine loop or VEGF sequestration.
INTRODUCTION
Vascular endothelial growth factor (VEGF) is a 34–
42 kDa dimeric glycoprotein which is essential for
normal kidney development [1]. In adulthood, VEGF is
synthesized by visceral glomerular epithelial cells
(vGECs) and by very occasional tubular cells. This is well
established in vivo in humans at the mRNA and protein
levels [2–4]. The major properties of VEGF are the
mediation of increased endothelial permeability (to water
and in some circumstances protein) and endothelial
mitogenesis (reviewed in [5]). Although the glomerular
filtration barrier is highly permeable to water, VEGF
appears to perform neither of these other roles to any
great extent in the normal glomerulus – the vast majority
of individuals are not nephrotic, and the normal glomerulus is not a site of new vessel formation. The specific
role of VEGF in the normal healthy adult glomerulus has
therefore been widely debated. Putative roles in the
maintenance of the glomerular endothelium (including
maintaining fenestration) and permselectivity have been
attractive hypotheses [6,7].
Key words: autocrine factor, glomeruli, neuropilin, podocyte, VEGF.
Abbreviations: GAPDH, glyceraldehyde-3-phosphate dehydrogenase ; RT-PCR, reverse transcription–PCR ; VEGF, vascular
endothelial growth factor ; VEGFR1(2), VEGF receptor 1(2) ; vGEC, visceral glomerular epithelial cell ; WT-1, Wilm ’s tumour-1.
Correspondence: Dr S. J. Harper (e-mail s.harper!bristol.ac.uk).
# 2001 The Biochemical Society and the Medical Research Society
439
440
S. J. Harper and others
Differential exon splicing of the VEGF gene results in
four main mRNA species which code for four major
secreted isoforms : VEGF , VEGF , VEGF
and
")*
"'&
"%&
VEGF
[5,8] (the subscript denotes the number of
"#"
amino acids). Several minor splice variants (VEGF ,
#!'
VEGF
and VEGF ) have been reported, but their
")$
"%)
significance remains uncertain [9–11]. The major isoforms are physico-chemically and functionally distinct.
The longer forms in general have a greater ability to bind
to heparin and extracellular matrix proteins [5].
VEGF binds to a number of receptors. These are also
functionally distinct. Its major targets are two cell surface
tyrosine kinase receptors, VEGFR1 (flt-1) and VEGFR2
(KDR). The latter initiates angiogenesis, cell migration
and permeability changes. The role of VEGFR1 is less
well defined. Although it can stimulate cellular migration,
it has also been implicated as an inhibitor of VEGF
activity, by acting as a decoy for VEGFR2 [12]. VEGFR1
also exists in a soluble form (sVEGFR1 ; sFlt) which is
inhibitory when bound to free VEGF. These receptors
are expressed by the vascular endothelium, including that
within the glomeruli [13,14]. A paracrine action for
VEGF has therefore been proposed [13]. These findings
would seem to indicate that, for vGEC-secreted VEGF
to bind to glomerular VEGF receptors, it would have to
move against the gradient of filtration (although this
clearly would be down a marked concentration gradient).
This apparent paradox has been the subject of much
conjecture, since, for some VEGF isoforms at least,
movement ‘ upstream ’ would appear difficult – particularly VEGF , which has no heparin-binding proper"#"
ties and would therefore tend to be washed rapidly into
Bowman’s space. Furthermore, binding studies also
suggest that there are no VEGF receptors in the distal
nephron. At first glance, therefore, the production of
VEGF by vGECs would seem to be redundant.
"#"
The neuropilins have recently been identified as
potential co-receptors for VEGFR2. They are present in
many tissues [15,16] and appear to be VEGF isoform
specific [17,18]. Neuropilin-1 enhances the binding of
VEGF
to VEGFR2 [12,15], promoting endothelial
"'&
proliferation and changes in permeability [19–21] ; indeed, the soluble extracellular neuropilin-1 appears to be
sufficient for this augmentation [12]. Furthermore,
VEGFR1 may bind to neuropilin-1 to inhibit the above
[12]. Neuropilin-1 and neuropilin-2 are receptors for
semaphorin class 3 axon guidance receptors (these
receptors are involved in directing axonal growth in the
developing nervous system). Both receptors also bind
VEGF
(and placental growth factor). In addition,
"'&
neuropilin-2 has been shown to bind VEGF [18], an
"%&
isoform which, until recently, had only been identified in
tissue derived from the female reproductive tract [11].
The paradoxes described above led us to reconsider the
potential expression of VEGF receptors by vGECs.
vGECs do not express the tyrosine kinase VEGF
# 2001 The Biochemical Society and the Medical Research Society
receptors [13,14]. We were therefore prompted to investigate the expression of neuropilin-1, neuropilin-2 and
the recently described soluble neuropilin in normal
human renal tissue and in cultured vGECs.
METHODS
Nephrectomy tissue
Nephrectomy tissue was supplied by the Department of
Urology, Southmead Hospital, from patients undergoing
nephrectomy for polar renal tumour (age range 47–78
years). All patients were non-diabetic and normotensive,
with normal excretory renal function and no urinary
sediment. The research was carried out in accordance
with the Declaration of Helsinki (1989) of the World
Medical Association. In line with the local Ethical
Committee policy at the time, and previous studies [11],
no formal Ethical Committee approval was required for
the study of truly discarded tissue ; however, verbal
consent was obtained from the patients. In line with
recent guidelines, formal Ethical Committee approval
has been granted.
Glomerular harvest and cell culture
vGECs were isolated by sieving, and characterized by
immunofluorescence and reverse transcription–PCR
(RT-PCR) studies for cytokeratin, WT-1 (Wilm’s
tumour-1), VEGF, synaptopodin, von Willebrand factor,
CD45 and smooth muscle myosin, as detailed previously
[22]. Single glomeruli were retrieved as described previously [11]. Pre-prepared adult human kidney library
cDNA was obtained commercially from Invitrogen
(A550-43).
Molecular studies
Primers
All primer sequences and predicted amplicon lengths are
given in Table 1.
RT-PCR
RNA was extracted from cultured vGECs and whole
kidney tissue with guanidinium isothiocyanate following
a recognized protocol [23]. cDNA was generated using
an RT system kit (Promega, Southampton, U.K.) according to the manufacturer’s instructions. Each PCR
reaction consisted of 1 µl of 10 µM forward and reverse
primer stock (Interactiva), 2.5 µl of 2 mM dNTP, 2.5 µl
of 10i buffer (Hybaid), 0.2 µl of gold Taq polymerase
(Hybaid) and 100 ng of cDNA, made up to a final
volume of 25 µl with nano-pure water.
The general amplification protocol was 94 mC for 5 min
(one cycle) ; 94 mC for 20 s, 55–60 mC for 30 s and 72 mC
for 30 s (30–35 cycles) ; and finally 72 mC for 5 min (one
cycle). The specific number of cycles and annealing
temperature for each primer set varied slightly, as
Neuropilin expression by podocytes
Table 1
Primer sequences and amplicon lengths for PCR and nested PCR
Gene
Forward primer
Reverse primer
Amplicon length
GAPDH (external)
GAPDH (internal)
CD45 (external)
CD45 (internal)
VEGF isoforms
Neuropilin-1 (external)
Neuropilin-1 (internal)
Neuropilin-2
Soluble neuropilin
VEGFR1
VEGFR2
Soluble VEGFR1
CACCCATGGCAAATTCCATG
GCCAAAAGGGTCATCATCTC
AGCTCGAAAGCCCTTTAACC
AGAATACTGGCCGTCAATGG
GTGAATGCAGACCAAAGAAAG
TCAACTTCAACCCTCACTT
TATTCCCAGAACTCTGCCC
CCTGTGGTTGGATGTATGAC
CAACTATGATACACCTGAGCT
ATCAGAGATCAGGAAGCACC
GACTTCAACTGGGAATACCC
ACAATCAGAGGTGAGCACTG
GCAGGTTTTTCTAGACGGCA
GTAGAGGCAGGGATGATGTTC
ACATCCACTTTGTTCTCGGC
GCTGAAGGCATTCACTCTCC
AAACCCTGAGGGAGGCTC
TGATGGTGATGAGGATGG
TGTCATCCACAGCAATCCCA
GCAGTTCTCCAGTGGGACAT
ACATTTCGTATTTTATTTGATACC
GGAACTTCATCTGGGTCCAT
CATGGACCCTGACAAATGTG
CTGCTATCATCTCCGAACTC
524
287
629
237
96, 193, 228, 300
2378
253
439
316
441
208
219
Figure 1
RT-PCR gene products
Lane 1, whole kidney ; lane 2, soup of sieved glomeruli ; lanes 3 and 4, cultured
podocytes from two different kidneys ; lane 5, kidney library cDNA ; lane 6, controls
lacking reverse transcriptase ; lane 7, negative controls.
determined by preliminary experiments used for
optimization. Primer pairs were designed to span different exons to discriminate from genomic DNA. Primer
sequences were specific for the target gene. Concurrent
RT-PCR studies for VEGF isoforms were undertaken as
positive controls. RT-PCR studies of the tyrosine kinase
VEGF receptors were also undertaken on cultured
podocytes. Positive and negative (no reverse transcriptase) controls were incorporated into each experiment.
Single-glomerular harvest and single-glomerular nested RT-PCR
Single-glomerular harvest and single-glomerular nested
RT-PCR were adapted from well established previously
published methods [11,24].
mRNA was extracted from isolated glomeruli using a
poly(T) Dynabead kit (Dynal RNA Direct) and following the manufacturer’s instructions. Briefly, each
glomerulus was lysed and then incubated with poly(T)
beads. The beads were washed to remove cell debris prior
to a reverse transcription step, which was carried out
while the mRNA was still attached to the Dynabeads.
Following reverse transcription the cDNA-linked
Dynabeads were incubated in 50 µl of TE buffer (10 mM
Tris, 1 mM EDTA, pH 8) and heated to 95 mC for 1 min.
The cDNA-linked Dynabeads were then resuspended in
50 µl of TE buffer for storage or PCR.
cDNA-linked Dynabeads were washed in 10i PCR
buffer (Hybaid Proof 2 buffer) before addition to a 25 µl
PCR reaction mixture containing 400 nM each of external
forward and reverse primers (Figure 1, Table 1), 1i
Proof 2 buffer (including 2.5 mM MgCl ) and 200 µM
#
dNTPs (Hybaid). The reaction was ‘ hot started ’ by
addition of 0.75 unit of Taq polymerase (PWO proof
mix ; Hybaid) in a first-round synthesis as follows : 95 mC
for 7 min (Hot start after 3 min), 95 mC for 1 min, 55 mC
for 1.5 min and 72 mC for 2 min. The cDNA-linked
Dynabeads were then removed by pelleting after 2 min at
95 mC. The supernatant containing the first strand of
cDNA was then amplified using 18 cycles of 95 mC for
30 s, 55 mC for 50 s and 72 mC for 50 s, and finished off
with a final extension of 5 min at 72 mC.
A 1 µl sample from the first round of PCR was used as
template in the second round of amplification. The
reaction mixture was as above, except that it contained
internal forward and reverse primers for neuropilin-1.
The final MgCl concentration was 1.5 mM, and 0.75 unit
#
of Taq polymerase was added per reaction. Conditions
for the second round were as follows : 95 mC for 30 s,
55 mC for 40 s and 72 mC for 40 s (25 cycles), followed by
a 5 min extension at 72 mC. All PCRs were carried out in
a Hybaid thermal cycler (Hybaid, Ashford, U.K.).
Nested PCR for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was as described previously [11].
Optimal cycle numbers in the first and second rounds of
PCR were determined by preliminary experiments.
CD45 controls were used to exclude trafficking cells of
bone marrow origin in single glomeruli, as described
previously [11].
# 2001 The Biochemical Society and the Medical Research Society
441
442
S. J. Harper and others
Detection of PCR products
PCR products were mixed with 3 µl of loading dye
(0.25 % Bromophenol Blue\50 % glycerol) and electrophoresed in a 2 % agarose (Sigma) gel. Bands were
visualized by ethidium bromide staining on a UVP dualintensity transilluminator.
Protein expression
Western blotting
Podocytes were cultured in a humidified atmosphere of
5 % CO \95 % air at 37 mC for 14 days, and solubilized
#
with RIPA buffer [consisting of 0.1 % SDS, 0.5 % sodium
deoxycholate, 1 % Nonidet P-40, 150 mM NaCl, 50 mM
Tris and 10 µl\ml proteinase inhibitor cocktail (SigmaAldrich Co.)]. Insoluble material was removed by centrifugation at 12 000 g for 10 min. Soluble material was
subjected to SDS\PAGE on an 8 % (w\v) acrylamide gel
using reducing conditions, and transferred to a nitrocellulose membrane (Sigma-Aldrich Co.). After blocking
with 1 % (w\v) BSA (for immune staining ; SigmaAldrich Co.) plus 1 % (w\v) dried milk (Sigma-Aldrich
Co.) and goat IgG, the membrane was incubated with a
variety of primary antibodies as for immunohistochemistry detailed below, and then with horseradish
peroxidase-conjugated pre-adsorbed secondary antibodies. After the membrane had been incubated in
Chemiluminescence Luminol Reagent (Santa Cruz Biotechnology, Santa Cruz, CA, U.S.A.), X-ray film
(Amersham Pharmacia Biotech) was exposed to the
membrane, developed and scanned. Subsequently the
filter was stripped and re-probed with secondary antibody only.
Immunocytochemistry and immunohistochemistry
Immunocytochemistry was performed on cultured
podocytes after incubation and fixation as described
previously [22], using a goat polyclonal anti-neuropilin1 antibody (Santa Cruz ; sc-7239) diluted 1 : 50 (in Trisbuffered saline) and a secondary FITC-labelled donkey
anti-(goat IgG) (Santa Cruz ; sc-2024) diluted 1 : 100 (in
Tris-buffered saline). Cells were grown on chamber
slides (Lab-Tek II ; Life Technologies, Paisley, Scotland,
U.K.). The cells were fixed for 5 min in 1 : 1 (v\v) mixture
of ice-cold ethanol (BDH) and acetone (BDH) and
blocked in 10 % (v\v) fetal calf serum in PBS, pH 7.4, for
20 min. Appropriate negative controls were used : preincubation of primary antibody with control peptide,
isotype control, omission of primary antibody, omission
of secondary antibody.
Immunohistochemistry was performed on ‘ normal ’
renal tissue derived from the normal pole of nephrectomy
specimens. Tissue was stored frozen and in wax. Since we
were unaware of any previously published protocols for
the detection of neuropilin-1 protein by immuno# 2001 The Biochemical Society and the Medical Research Society
histochemistry, a variety of storage, fixation and antigen
retrieval methods were used.
Frozen tissue was put directly into liquid nitrogen.
Frozen sections were then fixed using a variety of
methods, including paraformaldehyde, acetone and alcohol. Wax-embedded tissue was fixed in neutral pH
formalin for 4–48 h overnight at room temperature, or
microwave-fixed for 20 s. A variety of antigen retrieval
methods were used, including trypsin and proteinase K
digestion, and pressure cooker\citrate buffer treatment.
Furthermore, each combination of fixation\antigen
retrieval for frozen\wax tissue was investigated with all
five commercially available anti-neuropilin-1 antibodies
of which we are aware. All are polyclonal : (i) rabbit anti(mouse\rat neuropilin-1) PC343 (Oncogene, Calbiochem, Nottingham, U.K.) ; this antibody was
generated using a 14-amino-acid peptide that had one
mismatch with the human sequence ; (ii) rabbit anti(human neuropilin-1) 52-0107 (Zymed, San Francisco,
CA, U.S.A.) ; (iii) rabbit anti-(human neuropilin-1) sc5541 (Santa Cruz) ; (iv) goat anti-(human neuropilin-1)
sc-7239 (Santa Cruz), raised against the C-terminus ; and
(v) goat anti-(human neuropilin-1) sc-7240 (Santa Cruz),
raised against the N-terminus.
Secondary antibodies included : swine anti-rabbit, alkaline phosphatase labelled (Dako DO306) ; donkey antigoat, alkaline phosphatase labelled (Santa Cruz ; sc-2022) ;
and swine anti-rabbit, peroxidase labelled (Dako PO217).
Alkaline phosphatase was visualized with Nitro Blue
Tetrazolium\5-bromo-4-chloro-3-indoyl phosphate as
described previously [25]. Immunoperoxidase was
detected with a StrAvigen Multilink kit (Biogenex) and
visualized with diaminobenzidine chromogen solution
(Biogenex).
Negative controls included isotype controls, preincubation with control peptide, omission of primary
antibody and omission of secondary antibody. The
Zymed control peptide consisted of an 11-amino-acid
synthetic peptide derived from the C-terminus of human
neuropilin-1. Endogenous tubular alkaline phosphatase
was inhibited by pretreatment with 0.2 M HCl, and
peroxidase by immersion in 3 % H O .
# #
RESULTS
vGEC culture and characterization
Cultured vGECs demonstrated a typical polyhedral
shape with a cobblestone appearance on confluence.
They were positive for cytokeratin and WT-1 by
immunofluoresence. By RT-PCR the cells were positive
for VEGF, WT-1 and synaptopodin, and negative for
von Willebrand factor, CD45 and smooth muscle myosin, excluding contamination by endothelial cells, leucocytes and mesangial cells respectively.
Neuropilin expression by podocytes
Figure 2 Nested RT-PCR for neuropilin-1 and GAPDH on
individual isolated glomeruli
Each lane represents the RT-PCR product of adjacent single glomeruli isolated
within seconds of each other from full-thickness needle biopsies of the normal pole
of nephrectomy specimens. Lanes 1 and 2, kidney 1 ; lanes 3 and 4, kidney 2 ; lane
5, kidney 3 ; lanes 6–8, kidney 4.
predicted VEGF isoforms were seen in all kidney tissue
examined (Figure 1). Neuropilin-1 mRNA (specific
receptor for VEGF ) was seen in total kidney, a soup of
"'&
sieved glomeruli, cultured vGECs from a number of
kidneys and a commercially available kidney library
(Figure 1, lanes 1–5 respectively). Neuropilin-2 mRNA
was seen in cultured vGECs and was only just detectable
in total kidney. It was not detected in sieved glomeruli or
in the kidney library (Figure 1). mRNA for soluble
neuropilin was demonstrated in cultured vGECs and
total kidney, and weakly in sieved glomeruli (Figure 1).
All these results were reproducible in tissue from four
different kidneys.
Nested RT-PCR
Neuropilin-1 products of the predicted size were
identified in six of eight single glomeruli isolated within
30 min of nephrectomy (Figure 2). CD45 controls were
all negative by nested RT-PCR of single glomeruli
(results not shown).
Protein studies
Figure 3 Neuropilin-1 protein expression in cultured
podocytes
The positions of molecular mass markers are shown on the left (K l kDa). The
neuropilin-1 band is anticipated at 130 kDa.
Molecular studies
RT-PCR
Cultured vGECs did not express mRNAs for the
tyrosine kinase VEGF receptors (results not shown),
although mRNAs for both were present in whole kidney.
All renal tissue expressed mRNAs for VEGF , VEGF ,
")*
"'&
VEGF and VEGF , as described previously [11].
"%)
"#"
In contrast, PCR products of the predicted size were
seen for neuropilin-1, neuropilin-2 and soluble neuropilin in a variety of kidney tissues (Figure 1). mRNAs for
A
Figure 4
These studies concentrated on the expression of
neuropilin-1 protein, because neuropilin-2 mRNA was
not seen in any fresh tissue [whole kidney or fresh sieved
(non-cultured) glomeruli], and we are unaware of any
commercially available antibody that specifically
identifies soluble neuropilin.
The best Western blot and immunohistochemistry
results were seen with the rabbit anti-(human neuropilin-1)
antibody (Zymed ; 52-0107) plus an appropriate
secondary antibody. The best immunocytochemistry
results were obtained with goat anti-(human neuropilin-1)
(Santa-Cruz ; sc-7239) plus appropriate secondary
antibody.
Western blotting
A band of the predicted size for neuropilin-1 protein
(130 kDa) was detected in primary cultured podocytes
B
Immunocytochemistry for neuropilin-1 expression in cultured vGECs
(A) Intense immunofluorescence has a membrane distribution (arrows). (B) Isotype control.
# 2001 The Biochemical Society and the Medical Research Society
443
444
S. J. Harper and others
A
C
B
D
E
Figure 5
C
Immunohistochemistry for neuropilin-1 in normal adult kidney
(A) Low power (original magnification i100), showing staining of vessels and glomeruli. (B) Medium power (i200), demonstrating strong glomerular staining in
one glomerulus and modest staining in its neighbour (predominantly endothelial labelling). (C) Medium power (i400). Intense glomerular staining is maximal at the
periphery of the glomerular tufts in the vGECs. (D) High power (i1000 ; oil immersion). Staining is clearly seen within the cytoplasm of the vGECs (arrows). A
haematoxylin counterstain was used. (E) Typical example of a negative control (preincubation with control peptide).
(Figure 3). Subsequent stripping of the filter and reprobing with the secondary antibody alone revealed no
band.
Immunocytochemistry and immunohistochemistry
Immunocytochemistry revealed intense membrane
fluorescence for neuropilin-1 in cultured vGECs (Figure
4A). Controls were negative on all occasions (example in
Figure 4B).
All antibodies produced a positive signal with wax
specimens ; however, the best immunohistochemistry
# 2001 The Biochemical Society and the Medical Research Society
results were seen with the Zymed antibody using 36 h
formalin-fixed tissue and pressure cooker\citrate buffer
antigen retrieval (Figure 5). All other combinations of
fixation, storage, antigen retrieval and antibodies resulted
in very faint staining.
Within normal kidney, glomeruli demonstrated
varying intensities of staining, with the vGECs showing
consistent granular positivity (Figures 5A, 5C and 5D).
Basement membranes were weakly positive, and in many
afferent arterioles and glomerular tufts positivity could
be seen in endothelial cells. Mesangial cells demonstrated
no convincing staining. Some proximal tubules on some
Neuropilin expression by podocytes
sections demonstrated very weak staining. Not all
glomeruli demonstrated staining of equal intensity (Figure 5B). This mirrors the nested RT-PCR experiment, in
which neuropilin-1 mRNA was detected only in some
glomeruli. Negative controls gave no signal on any
occasion ; an example is shown in Figure 5(E).
DISCUSSION
This is the first report of the expression of neuropilins in
human vGECs. Cultured vGECs synthesize and express
neuropilin-1 at the mRNA and protein level in vitro. In
addition, neuropilin-1 mRNA is expressed in fresh
kidney and in sieved and isolated glomeruli, and
neuropilin-1 protein is identified in vGECs in normal
human kidney sections.
Although VEGF was initially thought to act on the
glomerular endothelial cells within the kidney, recent
work has raised the possibility that both mesangial and
tubular cells may express VEGF tyrosine kinase or
isoform-specific receptors [16,26]. Thomas et al. [26]
have described the expression of VEGFR1 and VEGFR2
in the mesangium of renal biopsies from proliferative
renal diseases (in contrast with normal tissue), but
neuropilin-1 expression was not investigated. However,
all three receptors were present on cultured mesangial
cells [26]. A soluble (and truncated) form of neuropilin1 has been reported to be synthesized by tubular cells, at
least at the mRNA level [16]. In mice, in situ
hybridization studies for neuropilin-1 have shown a
localization of signal to the glomeruli, but further cellular
localization was not possible because of relatively poor
resolution of isotopic in situ hybridization [25,27]. Nonisotopic in situ hybridization for soluble neuropilin
revealed a tubular signal only [16]. Our immunohistochemistry for neuropilin-1 revealed a signal in the
afferent arterioles and glomerular capillaries, with the
strongest signal in the vGECs. It is therefore possible that
all three major glomerular cell types have the potential to
synthesize neuropilin-1.
vGECs have not been shown to express VEGFR1 or
VEGFR2. However, the data that we present here
highlight the possibility that vGECs may bind VEGF via
membrane-bound neuropilin-1.
In situ binding studies have confirmed the presence of
VEGF binding sites within the glomerulus ; it has been
proposed that these are present principally on the
endothelial cells. However, Simon and colleagues [13]
were unable to exclude binding of VEGF to vGECs
because of the intensity of glomerular staining with
exogenous radiolabelled VEGF.
The available data on the expression of VEGF and
VEGF receptors on vGECs mirror the findings with a
variety of tumour-derived cells (e.g. PC3 prostate carcinoma cells, 231 breast carcinoma cells), which, like
vGECs, express VEGF and neuropilin-1, but do not
express VEGFR1 or VEGFR2 [15]. In this context it has
been proposed that VEGF may have an autocrine role to
enhance tumour cell survival, differentiation and motility, or that tumour cell neuropilin-1 may have a VEGF
storage or sequestration role [15]. Similar roles clearly
should be considered for vGEC-derived VEGF. An
autocrine cycle may contribute to vGEC survival or state
of differentiation. However, we are not aware of any
functional data to support this, and certainly we have
been unable to demonstrate VEGF-induced phosphorylation of neuropilin-1 in cultured vGECs, in line with the
investigations of Soker and colleagues [15] in tumour
cells, although, as these authors note, there are many
other potential signalling pathways. We suggest, therefore, that vGEC neuropilin-1 expression may have a
simpler role, for example delaying the passage of VEGF
into the urinary space by its sequestration on to the
podocyte cell surface.
There is increasing evidence that neuropilin-1 can act
as a co-receptor for VEGFR2 ; indeed, the presence of
neuropilin-1 mediates a 4-fold increase in the binding of
VEGF to VEGFR2 [15]. A similar co-receptor role is
"'&
hypothesized for neuropilin-2 when binding to VEGF .
"%&
However, there is, as yet, no evidence that the neuropilins
can act as functional VEGF receptors in their own right
in endothelial cells. Both neuropilin-1 and neuropilin-2
lack cytoplasmic signal transduction domains.
In nervous tissue, neuropilin-1 interacts with a cytoplasmic protein known as neuropilin-interacting protein
(NIP), which is identical with two other recently
identified proteins, GIPC (GAIP-interacting protein Cterminus) and SEMCAP (M-SEMF cytoplasmic domainassociated protein). This protein is a member of the RGS
(regulators of G-protein signalling) family of proteins
that act as GTPase-activating proteins [28–30]. More
recently, neuropilin-1 has been shown to form stable
complexes with another semaphorin-binding and transducing protein called plexin-1 [31].
We are unaware of any other tyrosine kinase or Gprotein-linked podocyte receptors that interact with
neuropilin-1. The presence or absence in vGECs of
intracellular signalling mechanisms similar to those present in neuronal tissue remains speculative, but requires
further investigation.
In summary, we submit that vGECs express at least
one membrane molecule capable of binding VEGF.
Further functional studies are required to address the
potential role of vGEC-bound neuropilin-1.
ACKNOWLEDGMENTS
S. J. H. is supported by Wellcome Trust Grant
057936\Z\99. C. Y. X. is supported by the China Scholarship Council No. 97037. The work presented here was
# 2001 The Biochemical Society and the Medical Research Society
445
446
S. J. Harper and others
supported in part by National Kidney Research Fund
Grant R35\1\98. We are grateful to Mrs Tracey Perry for
technical help with immunochemistry.
15
16
REFERENCES
17
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Kitamoto, Y., Tokunaga, H. and Tomita, K. (1997)
Vascular endothelial growth factor is an essential
molecule for mouse kidney development :
glomerulogenesis and nephrogenesis. J. Clin. Invest. 99,
2351–2357
Brown, L. F., Berse, B., Tognazzi, K. et al. (1992)
Vascular permeability factor mRNA and protein
expression in human kidney. Kidney Int. 42, 1457–1461
Bailey, E., Bottomley, M. J., Westwell, S. et al. (1999)
Vascular endothelial growth factor mRNA expression in
minimal change, membranous and diabetic nephropathy
demonstrated by non-isotopic in situ hybridisation.
J. Clin. Pathol. 52, 735–738
Simon, E., Gro$ ne, H. J., Johren, O. et al. (1995)
Expression of endothelial growth factor and its receptors
in human renal ontogenesis and in adult kidney. Am. J.
Physiol. 268, 240–250
Ferrara, N. and Davis-Smith, T. (1997) The biology of
vascular endothelial growth factor. Endocr. Rev. 18, 4–25
Roberts, W. G. and Palade, G. E. (1995) Increased
microvascular permeability and endothelial fenestration
induced by vascular endothelial growth factor. J. Cell Sci.
108, 2369–2379
Brenchley, P. (1996) VEGF\VPF : A modulator of
microvascular function with potential roles in glomerular
pathophysiology. J. Nephrol. 9, 10–17
Poltorak, Z., Cohen, T., Sivan, R. et al. (1997) VEGF145,
a secreted vascular endothelial growth factor isoform that
binds to extracellular matrix. J. Biol. Chem. 272,
7151–7158
Houck, K. A., Ferrara, N., Winer, J., Cachianes, G., Li,
B. and Leung, D. W. (1991) The vascular endothelial
growth factor family : Identification of a fourth molecular
species and characterization of alternative splicing of
RNA. Mol. Endocrinol. 5, 1806–1814
Jingying, L., Xue, Y., Agarwal, N. and Roque, R. S.
(1999) Human muller cells express VEGF 183, a novel
splice variant of vascular endothelial growth factor.
Invest. Opthalmol. Vis. Sci. 40, 752–759
Whittle, C., Gillespie, K., Harrison, R., Mathieson, P. W.
and Harper, S. J. (1999) Heterogenous vascular
endothelial growth factor (VEGF) isoform mRNA and
receptor mRNA expression in human glomeruli, and the
identification of VEGF mRNA, a novel truncated
splice variant. Clin. Sci."%)
97, 303–312
Fuh, G., Garcia, K. C. and de Vos, A. M. (2000) The
interaction of neuropilin-1 with vascular endothelial
growth factor and its receptor Flt-1. J. Biol. Chem. 275,
26690–26695
Simon, M., Ro$ ckl, W., Hornig, C. et al. (1998) Receptors
of vascular endothelial growth factor\vascular
permeability factor (VEGF\VPF) in fetal and adult
human kidney : Localization and ["#&I]VEGF binding
sites. J. Am. Soc. Nephrol. 9, 1032–1044
Feng, D., Nagy, J. A., Brekken, R. A. et al. (2000)
Ultrastructural localization of the vascular permeability
factor\vascular endothelial growth factor (VPF\VEGF)
receptor-2 (FLK-1, KDR) in normal mouse kidney and in
the hyperpermeable vessels induced by VPF\VEGFexpressing tumors and adenoviral vectors. J. Histochem.
Cytochem. 48, 545–556
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Soker, S., Takashima, S., Miao, H. Q., Neufeld, G. and
Klagsbrun, M. (1998) Neuropilin-1 is expressed by
endothelial and tumor cells as an isoform-specific
receptor for vascular endothelial growth factor. Cell 92,
735–745
Gagnon, M. L., Bielenberg, D. R., Gechtman, Z. et al.
(2000) Identification of a natural soluble neuropilin-1 that
binds vascular endothelial growth factor : In vivo
expression and antitumor activity. Proc. Natl. Acad. Sci.
U.S.A. 97, 2573–2578
Soker, S., Fidder, H., Neufeld, G. and Klagsbrun, M.
(1996) Characterization of novel vascular endothelial
growth factor (VEGF) receptors on tumor cells that bind
VEGF via its exon 7-encoded domain. J. Biol. Chem.
"'&
271, 5761–5767
Gluzman-Poltorak, Z., Cohen, T., Herzog, Y. and
Neufeld, G. (2000) Neuropilin-2 and neuropilin-1 are
receptors for the 165-amino acid form of vascular
endothelial growth factor (VEGF) and of placenta growth
factor-2, but only neuropilin-2 functions as a receptor for
the 145-amino acid form of VEGF. J. Biol. Chem. 275,
18040–18045
Neufeld, G., Cohen, T., Gengrinovitch, S. and Poltorak,
Z. (1999) Vascular endothelial growth factor (VEGF) and
its receptors. FASEB J. 13, 9–22
Millauer, B., Shawver, L. K., Plate, K. H., Risau, W. and
Ullrich, A. (1994) Glioblastoma growth inhibited in vivo
by a dominant-negative Flk-1 mutant. Nature (London)
367, 576–579
Waltenberger, J., Claesson-Welsh, L., Siegbahn, A.,
Shibuya, M. and Heldin, C. H. (1994) Different signal
transduction properties of KDR and Flt1, two receptors
for vascular endothelial growth factor. J. Biol. Chem. 269,
26988–26995
Parry, R. G., Gillespie, K. M. and Mathieson, P. W.
(2001) Effects of type 2 cytokines on glomerular
epithelial cells. Exp. Nephrol. 9, 275–283
Chomczynski, P. and Sacchi, N. (1987) Single step
method of RNA isolation by acid guanidium thiocyanate
phenol chloroform extraction. Anal. Biochem. 162,
156–159
Bicknell, G. R., Shaw, J. A., Pringle, J. H. and Furness,
P. N. (1996) Amplification of specific mRNA from a
single human renal glomerulus, with an approach to the
separation of epithelial cell mRNA. J. Pathol. 180,
188–193
Bailey, E., Harper S. J., Pringle, J. H. et al. (1998) Visceral
glomerular epithelial cell DNA synthesis in experimental
and human membranous disease. Exp. Nephrol. 6,
352–358
Thomas, S., Vanuystel, J., Gruden, G. et al. (2000)
Vascular endothelial growth factor receptors in human
mesangium in vitro and in glomerular disease. J. Am. Soc.
Nephrol. 11, 1236–1243
Robert, B., Zhao, X. and Abrahamson, D. R. (2000)
Co-expression of neuropilin-1,Flk1, and VEGF in
developing and mature mouse kidney glomeruli."'%
Am. J.
Physiol. Renal Physiol. 279, F275–F282
De Vries, L., Lou, X., Zhao, G., Zheng, B. and Farquhar,
G. (1998) GIPC, a PDZ domain containing protein,
interacts specifically with the C terminus of RGS-GAIP.
Proc. Natl. Acad. Sci. U.S.A. 95, 12340–12345
Cai, H. and Reed, R. R. (1999) Cloning and
characterisation of neuropilin-1 interacting protein : A
PSD\dig\ZO-1n domain containing protein that interacts
with the cytoplasmic domain of neuropilin-1. J. Neurosci.
19, 6519–6527
Wang, L., Kalb, R. G. and Strittmatter, S. M. (1999) A
PDZ protein regulates the distribuation of the
transmembrane semaphorin, M-SemF. J. Biol. Chem. 274,
14137–14146
Takahashi, T., Fournier, A., Nakamura, F. et al. (2000)
Plexin-neuropilin-1 complexes form functional
semaphorin-3A receptors. Cell 99, 59–69
Received 17 January 2001/12 April 2001; accepted 14 June 2001
# 2001 The Biochemical Society and the Medical Research Society