close

Вход

Забыли?

вход по аккаунту

?

903

код для вставкиСкачать
  
J. Pathol. 185: 402–408 (1998)
VALIDATION OF ANTI-VASCULAR ENDOTHELIAL
GROWTH FACTOR (ANTI-VEGF) ANTIBODIES FOR
IMMUNOHISTOCHEMICAL LOCALIZATION OF VEGF
IN TISSUE SECTIONS: EXPRESSION OF VEGF IN THE
HUMAN ENDOMETRIUM
 1,2,  . . 1,  3,  4,  . 5,  . 3,
 . 1,  . 2,  . . 2   1*
Molecular Angiogenesis Group, Imperial Cancer Research Fund, Institute of Molecular Medicine, University of Oxford,
Oxford, OX3 9DU, U.K.
2
Nuffield Department of Obstetrics and Gynaecology, University of Oxford, John Radcliffe Hospital, Oxford, OX3 9DU, U.K.
3
Department of Cellular Science, University of Oxford, John Radcliffe Hospital, Oxford, OX3 9DU, U.K.
4
Nuffield Department of Pathology, University of Oxford, John Radcliffe Hospital, Oxford, OX3 9DU, U.K.
5
Department of Pathology, University of Sheffield Medical School, Royal Hallamshire Hospital, Beech Hill Road,
Sheffield, SE10 2JF, U.K.
1
SUMMARY
Transient transfection of COS-1 cells followed by fixation, embedding in paraffin, and immunohistochemistry has identified
anti-vascular endothelial growth factor (anti-VEGF) mouse monoclonal antibodies that efficiently immunostain VEGF in paraffinembedded tissue sections. Immunohistochemical localization of VEGF in 34 specimens of normal human endometrium that had been
collected at different stages of the menstrual cycle was then performed. VEGF was present at all stages of the cycle, but both the pattern
and the intensity of staining varied. Thus, VEGF expression occurred predominantly in the endometrial epithelium and while weak in the
proliferative phase, was strong in the secretory phase. VEGF expression in the stroma was weaker than in the proliferative phase glands
and did not change throughout the cycle. These findings are in agreement with reports of VEGF mRNA expression in the endometrium,
but disagree with previous immunohistochemical studies that employed an immunohistochemically unvalidated antiserum. This study has
shown that the commercially available anti-VEGF monoclonal antibody M293 is excellent for the immunohistochemical localization of
VEGF in paraffin sections. 1998 John Wiley & Sons, Ltd.
KEY WORDS—vascular
endothelial growth factor; endometrium
INTRODUCTION
The endometrium undergoes more extensive physiological angiogenesis in the adult than does any other
tissue. Ultimately under control of the ovarian steroids,
the absence of oestrogen and progesterone receptors in
endometrial endothelium had led to the hypothesis that
the steroids stimulate secretion of angiogenic polypeptides from the receptor-positive epithelium and stroma.1
In accord with this, we have previously shown that
17-â-oestradiol induces expression of VEGF mRNA in
normal endometrial epithelial and stromal cells in vitro.1
VEGF is an endothelial cell-specific growth factor that is
strongly angiogenic in vivo and is a prime candidate to
mediate steroid-induced endometrial angiogenesis. Previous studies have examined the expression of VEGF
mRNA2,3 and protein4 in the uterus. Protein expression
*Correspondence to: Roy Bicknell, Molecular Angiogenesis Group,
Imperial Cancer Research Fund, Institute of Molecular Medicine,
University of Oxford, John Radcliffe Hospital, Oxford OX3 9DU,
U.K.
Contract grant sponsors: Sir Jules Thorn Charitable Trust; Imperial
Cancer Research Fund.
CCC 0022–3417/98/040402–07 $17.50
1998 John Wiley & Sons, Ltd.
was analysed by immunohistochemistry. Immunohistochemical analysis of VEGF has become controversial, in
that the antibodies or antisera that have been employed
in several studies have not been definitively shown to
recognize VEGF in paraffin sections. Despite this,
VEGF antisera have been used for immunohistochemical staining (see, for example, Li et al.4 and HarrisonWoolrych et al.5).
In this study we have examined two polyclonal antisera and three monoclonal antibodies for their ability to
immunostain COS-1 cells that have been transfected
with either empty or VEGF-containing plasmid vector,
followed by fixation and embedding in paraffin. This
study has shown that both antisera non-specifically
stained mock transfectants. All three monoclonal antibodies stained VEGF; however, the sensitivity of the
commercially available M293 monoclonal antibody
was greater than the other two and this was the only
antibody to stain tissue sections efficiently. We conclude that the M293 monoclonal antibody is
eminently suitable for immunohistochemical localization of VEGF expression in paraffin-embedded tissue
sections.
Received 22 April 1997
Accepted 2 March 1998
403
IMMUNOHISTOCHEMISTRY OF VEGF
MATERIALS AND METHODS
Tissue was obtained from women (aged 30–45 years)
undergoing hysterectomy for a subjective complaint
of menorrhagia. No pelvic pathology was seen at operation. This was confirmed by subsequent histological
examination by an independent histopathologist. All
patients had a history of regular 26- to 30-day menstrual
cycles and had not used either oral or intrauterine
contraception, nor taken any hormones for at least 6
months prior to surgery. The stage of the menstrual
cycle at which the tissue was obtained was determined
from the patient’s menstrual history and endometrial
histology.6–8
Antibodies were as follows: rabbit polyclonal antiserum A-20, Santa Cruz Biotechnology; goat polyclonal
antiserum P293, R & D Systems, Abingdon, U.K.;
mouse monoclonal antibody M293, R & D Systems,
Abingdon, U.K.; mouse monoclonal antibody A.4.6.1, a
gift from Dr Napoleone Ferrara, Genentech Inc., South
San Francisco, CA, U.S.A. Mouse monoclonal antibody
VG76e/D9 was raised in our laboratory.9
Transient transfection of COS-1 cells
COS-1 cells were grown to 40 per cent confluence in a
15 cm tissue culture dish, at which time the medium was
replaced with 10 ml of DMEM/10 per cent fetal calf
serum to which had been added 400 ìl of DEAEdextran (10 mg/ml in PBS), 10 ìl of chloroquine
(100 m), and 10 ìg of plasmid pcDNA1Neo containing
the full length gene for VEGF165 diluted in 200 ìl of
PBS. The VEGF expression construct was prepared by
polymerase chain reaction (PCR) amplification of the
full length VEGF165 sequence from a puc18 plasmid
containing the sequence (a gift of Dr Judith Abraham,
Mountain View, CA, U.S.A.). The fragment was cloned
into TA vector and sequenced prior to subcloning into
pcDNA1Neo. The VEGF secretion signal was included
to ensure that the COS-1 cells produced a protein that
could be secreted as it is in vivo. After 3 h at 37C, the
mixture was aspirated and 10 per cent DMSO in PBS
was added for 2 min. The medium was aspirated and the
cells were washed once with PBS prior to incubation in
DMEM/10 per cent fetal calf serum. The following day,
the cells were removed by exposure to trypsin for
formaldehyde fixation and embedding in paraffin. The
two polyclonal antisera used in this study were titrated
against mock and VEGF transfected COS-1 cells and
found to stain both with equal intensity at concentrations of 1, 5, 10, 25, and 50 ìg total IgG/ml and 2·08,
6·25. 12·5, 25, and 50 ìg total IgG/ml for the P293 and
A-20 antisera, respectively. The secondary detection
systems were biotinylated rabbit anti-goat (at a dilution
of 1:300) and biotinylated goat and anti-rabbit (at a
dilution of 1:500) for the P293 and A-20 antisera,
respectively. Each of the monoclonal antibodies was
examined for staining of the COS-1 transfectants at the
following concentrations: 3·125, 6·25, 12·5, 25, and
50 ìg IgG/ml. A concentration of 12·5 ìg IgG/ml was
found to give optimal staining for each of the three
antibodies.
1998 John Wiley & Sons, Ltd.
Immunohistochemical staining
Normal endometrial and carcinoma tissue was
obtained from the archival files of the Histopathology
Department of the John Radcliffe Hospital; 5–6 ìm
sections were cut, mounted onto silane-coated slides,
dried overnight at 37C or for 1 h at 60C, and stored at
room temperature. The sections were dewaxed using
Citroclear (HS Supplies, Aylesbury, U.K.) and rehydrated sequentially in absolute, 95 per cent, 70 per cent,
20 per cent ethanol, and finally distilled water. Slides
were incubated in a trough of doubly distilled water
(DDW) at 37C for 10 min and in 200 ml of PBSA
containing 25 mg of protease type 24 (Sigma) at 37C for
another 10 min. Alternatively, pressure cooking or
microwave treatment of sections could be used for
antigen retrieval, but these procedures resulted in greater
damage to tissue than did use of a protease. Slides were
then left in DDW at room temperature for 30 min. The
staining used was the streptavidin–biotin–alkaline phosphatase (ABC) method; prior to application of the
primary antibody, sections were incubated with 10 per
cent normal human serum (NHS) to block non-specific
protein binding sites. Primary antibodies diluted to
12·5 ìg/ml in TBS were added to slides for 20 min. Slides
were washed in TBS twice for 5 min and incubated for
30 min with biotinylated rabbit anti-mouse IgG diluted
to 1/400 in TBS. The sections were washed again in TBS
and incubated with alkaline phosphatase conjugated to
streptavidin, at a dilution of 1:200 in TBS for a further
30 min. The ‘New Fuchsin’ substrate system was used to
visualize sections. Chromogen development was performed using a ‘New Fuchsin’ substrate system according to the manufacturer’s instructions by incubation for
5–20 min. Levamisole (1 m) was added to the ‘New
Fuchsin’ system to quench endogenous alkaline phosphatase. Slides were washed in TBS and then tap water,
counter-stained with haematoxylin, and mounted with
Apathy’s mounting medium. Negative controls involved
replacement of the primary antibody with an equal
concentration of mouse IgG2b (Sigma).
RESULTS
Identification of an antibody or antiserum that permits
immunohistochemical localization of VEGF in tissue
sections
The most direct means by which to verify the
specificity of antibody staining of cells and tissues is to
utilize transiently transfected cells expressing the gene of
interest. Thus, to identify an anti-human VEGF antibody that would immunostain paraffin-embedded
sections, the cDNA of VEGF165 was cloned into
pcDNA1Neo and transiently transfected into COS-1
cells. COS-1 cells were transfected with the empty vector
to provide a negative control.
Formalin-fixed and paraffin-embedded VEGF165
transfectant sections were stained with anti-VEGF antibodies using the streptavidin–biotin–alkaline phosphatase method combined with proteinase retrieval of
antigen. Microwave retrieval of antigen was found to be
J. Pathol. 185: 402–408 (1998)
404
L. ZHANG ET AL.
Table I—The anti-human VEGF antibodies that were evaluated in this study
Abbreviation of antibody
M293
VG76e/D9
A.4.6.1
A-20
P293
Type
Source
Mouse monoclonal antibody
Mouse monoclonal antibody
Mouse monoclonal antibody
Rabbit polyclonal antiserum
Goat polyclonal antiserum
R & D Systems
ICRF
Genentech
Santa Cruz Biotechnology
R & D Systems
Table II—Determination of the specificity and sensitivity of anti-human VEGF monoclonal antibodies and
polyclonal sera by immunohistochemical staining of VEGF165 transient transfectants, empty vector
pcDNA1Neo transfectants (negative control), and endometrial tissue sections. The percentage of stained
cells is given in each case
Monoclonal antibody
Polyclonal antisera
Cells or tissue section
M293
VG76e/D9
A.4.6.1
A-20
P293
VEGF165 transfected COS-1 cells
pcDNA1Neo transfected COS-1 cells
Endometrial sections
90–95
0
Strong
5–10
0
0
50
0
0
90–95
90–95
ND
90–95
90–95
ND
High
High
High
Absent
Absent
Specificity
ND=not determined.
equally effective, but resulted in greater tissue damage.
The antibodies and antisera examined are listed in
Table I and the results are summarized in Table II.
Background VEGF expression in COS-1 cells was barely
detectable. It is noted that virtually 100 per cent of the
transfected cells stain positive for VEGF (Fig. 1B). This
is surprising, in that transfection of other genes in the
same plasmid using an identical protocol has never
occurred, in our hands, with more than a 10–20 per cent
efficiency. However, closer inspection of the figure
clearly shows 10–20 per cent of the cells staining much
more intensely than the others. In view of the known
binding of VEGF to heparin-like molecules, it is postulated that 10–20 per cent of the cells are positive
transfectants (stain intensely) and that these secrete
VEGF165 which then binds to the untransfected cells,
leading to weak staining of the other 80–90 per cent of
cells present.
The two polyclonal antisera examined gave strongly
positive staining in both positive and control transfected
cells over the concentration range of 1–50 ìg total
IgG/ml for P293 and 2·08–50 ìg total IgG/ml for A-20.
We conclude that the use of these polyclonal antisera for
the immunolocalization of VEGF in paraffin sections is
questionable. All three monoclonal antibodies showed
cytoplasmic staining in paraffin-embedded VEGF165
transfectants, but not of the negative control cells that
had been transfected with empty vector (Fig. 1). The
percentage of transfected cells staining strongly positive
with each antibody was M293 90–95 per cent, A4.6.1
50 per cent, and VG76e/D9 only 5–10 per cent. Thus,
it appears that all three antibodies recognize VEGF
in paraffin-embedded cells. When used on paraffin 1998 John Wiley & Sons, Ltd.
embedded human endometrial sections, only the M293
antibody showed positive staining. No staining was
detected in a negative control endometrial tissue section
in which the primary antibody was replaced by mouse
IgG2b. The above results clearly show that the mouse
monoclonal anti-VEGF antibody M293 is the most
specific anti-VEGF antibody for immunohistochemical
localization of VEGF in human tissue.
Expression of VEGF protein in normal human
endometrium throughout the menstrual cycle
VEGF protein was present in human endometrium
throughout the menstrual cycle and showed menstrual
cycle-related expression in all 34 specimens examined.
VEGF was principally detected in glandular epithelial
cells and, to a lesser extent, in the stromal cell. The
highest intensity of immunostaining for VEGF was seen
in the menstrual, early proliferative (Fig. 2A and late
secretory ( Fig. 2C) phases, with a substantial reduction,
particularly in epithelial cell staining, during the late
proliferative to early/mid-secretory phases (Fig. 2B).
Figure 2B is a section from near the endometrial surface,
whereas Fig. 2C is from the myometrial/endometrial
junction; this tends to make Fig. 2B appear comparatively advanced next to Fig. 2C. The basal glandular
epithelium near the myometrial/endometrial junction
(top left in Fig. 2C) showed a higher immunostaining
intensity for VEGF than that nearer the endometrial
surface in the late secretory phase of the cycle (Fig. 2C).
This section was so chosen to illustrate this. In contrast,
a higher immunostaining intensity was seen in glandular
epithelium in the functionalis in the early proliferative
J. Pathol. 185: 402–408 (1998)
IMMUNOHISTOCHEMISTRY OF VEGF
405
Fig. 1—Immunohistochemical staining of COS-1 cells after fixation, embedding in paraffin,
and sectioning. (A) Control cells, (B) VEGF165 transfected cells, and (C) mock (empty
vector) transfected cells. All sections are shown at a magnification of 10
1998 John Wiley & Sons, Ltd.
J. Pathol. 185: 402–408 (1998)
406
L. ZHANG ET AL.
Fig. 2—Immunohistochemical localization of VEGF expression in normal human endometrium during the menstrual
cycle. (A) Early proliferative phase (10), (B) late proliferative phase (10), (C) late secretory phase (10), and (D)
late secretory phase (40)
phase. Menstrual cycle-independent moderate staining
of small blood vessels was seen (Fig. 2D), and of some
endothelial cells in arterioles as well as cells in the
muscular wall of the arterioles at the endometrial and
myometrial junction. Endothelial cells and cells in the
vascular smooth muscle wall of arterioles and venules in
myometrium were also positive. Again, the intensity of
staining did not change during the menstrual cycle. No
immunostaining was seen in controls in which the primary antibody was replaced with mouse IgG2b. In view
of the lack of specificity of staining of transfectants seen
with the polyclonal antisera, these were not examined on
paraffin-embedded sections.
Expression of VEGF protein in human endometrial
adenocarcinoma
Uterine endometrial tissues from 32 patients with
endometrial adenocarcinoma showed VEGF immunostaining similar to that of normal endometrium, but of
greater intensity, especially in the glandular epithelium
(Figs 3A and 3C). Weaker staining was detected in the
stroma and blood vessels (Fig. 3B). Once again, all
controls were negative (Fig. 3D).
1998 John Wiley & Sons, Ltd.
DISCUSSION
Transient transfection of COS-1 cells in vitro followed
by fixation, paraffin embedding, and immunohistochemistry has identified the R & D Systems monoclonal
anti-VEGF (M293) as a useful antibody with which to
perform immunohistochemical localization of VEGF in
paraffin-embedded tissue sections. This analysis has also
shown that the A-20 polyclonal VEGF antiserum stains
non-specifically; its use in immunohistochemistry is thus
questionable.
The only previous immunohistochemical study of
VEGF expression in human endometrium was with the
A-20 antiserum that we have shown here to have a high
non-specific component of staining. Immunolocalization
of VEGF with antibody M293 disagrees with that study,
but is in close agreement with reports of in situ hybridization of VEGF mRNA expression.2,3 Thus, although
VEGF expression was detected in the endometrium at
all stages of the menstrual cycle, we found expression
predominantly in the glandular epithelium rather than
the stroma. Expression in the epithelium was particularly strong in the late secretory, menstrual and early
proliferative phases, but weaker in the late proliferative
J. Pathol. 185: 402–408 (1998)
IMMUNOHISTOCHEMISTRY OF VEGF
407
Fig. 3—(A) Immunohistochemical localization of VEGF expression in endometrial carcinoma. B shows staining of endometrial blood vessels
and C carcinoma. D shows a negative control in which the primary antibody M293 was replaced by an antibody to IgG2b
and early secretory phases. In all cases, staining was
much stronger in the epithelium of the basalis than in
that in the functionalis. VEGF was present in the
stroma, but was much weaker and constant throughout
the cycle.
1998 John Wiley & Sons, Ltd.
Expression in glandular epithelium does not show a
clear correlation with either in vivo oestrogen or progesterone levels and we conclude that other factors may
also be involved. It has long been postulated that the late
secretory and menstrual endometrium are profoundly
J. Pathol. 185: 402–408 (1998)
408
L. ZHANG ET AL.
hypoxic. When the oxygen concentration falls below 2
per cent, VEGF mRNA is induced in many cell types.
Hypoxia could account for the strong VEGF expression
in the late secretory and menstrual phases of the cycle.
However, this does not account for expression in the
early proliferative phase, where hypoxia is thought to
have ceased. Here, stimulation of VEGF expression by a
rising level of oestradiol is most probably occurring.
We have previously shown that amongst several
endothelial growth factors, only VEGF stimulates
growth of decidual endothelium. This, taken together
with the in situ and immunohistochemical data presented here, points to a pivotal role for VEGF in
endometrial angiogenesis.
ACKNOWLEDGEMENTS
This work was supported by the Sir Jules Thorn
Charitable Trust and the Imperial Cancer Research
Fund
1998 John Wiley & Sons, Ltd.
REFERENCES
1. Zhang L, Rees MCP, Bicknell R. The isolation and long-term culture of
normal human endometrial epithelium and stroma. J Cell Sci 1995; 108:
323–331.
2. Shweicki D, Itin A, Neufeld G, Gitay-Goren H, Keshet E. Patterns of
expression of vascular endothelial growth factor (VEGF) and VEGF
receptors in mice suggest a role in hormonally regulated angiogenesis. J Clin
Invest 1993; 91: 2235–2243.
3. Charnock-Jones DS, Sharkey A, Rajputwilliams J, et al. Identification and
localization of alternately spliced messenger-RNAs for vascular endothelial
growth factor in human uterus and oestrogen regulation in endometrial
carcinoma cell-lines. Biol Reprod 1993; 48: 1120–1128.
4. Li XF, Gregory J, Ahmed A. Immunolocalisation of vascular endothelial
growth factor in human endometrium. Growth Factors 1994; 11: 277–282.
5. Harrison-Woolrych ML, Sharkey DS, Charnock-Jones DS, Smith SK.
Localisation and quantification of vascular endothelial growth-factor
messenger-ribonucleic-acid in human myometrium and leiomyomata. J Clin
Endocrinol Metab 1995; 80: 1853–1858.
6. Noyes RW, Hertig AI, Rock J. Dating the endometrial biopsy. Fertil Steril
1950; 1: 3–25.
7. Ferenczy A. Anatomy and histology of the uterine corpus. In: Blaustein’s
Pathology of the Female Genital Tract. 3rd edn. New York: SpringerVerlag, 1987; 257–291.
8. Buckley CH, Fox H. Biopsy Pathology of the Endometrium. London:
Chapman & Hall, 1989.
9. Scott PAE. Vascular endothelial growth factor in breast carcinoma. D.Phil.
Thesis, University of Oxford 1996.
J. Pathol. 185: 402–408 (1998)
Документ
Категория
Без категории
Просмотров
2
Размер файла
1 099 Кб
Теги
903
1/--страниц
Пожаловаться на содержимое документа