Blood Barrier Testis

8/22/2019 Blood Barrier Testis

1/10

T he n e w e n g l a n d j o u r n a l o f medicine

n engl j med 363;17 nejm.org october 21, 2010 1621

original article

Expression of Follicle-Stimulating Hormone

Receptor in Tumor Blood VesselsAurelian Radu, Ph.D., Christophe Pichon, Ph.D., Philippe Camparo, M.D.,

Martine Antoine, M.D., Yves Allory, M.D., Anne Couvelard, M.D.,Galle Fromont, M.D., Mai Thu Vu Hai, Ph.D.,

and Nicolae Ghinea, Ph.D.

From Mount Sinai School of Medicine,

New York (A.R.); and INSERM Unit 753,Villejuif (C.P.), Val-de-Grce Hospital, Par is(P.C.), Tenon Hospital, Paris (M.A.),INSERM Unit 955-Eq 07, UniversitParis-Est, Crteil (Y.A., M.T.V.H., N.G.),Beaujon Hospital, Clichy (A.C.), and Cen-tre Hospitalier Universitaire de Poitiers,Poitiers (G.F.) all in France. Addressreprint requests to Dr. Ghinea at INSERMUnit 955-Eq 07, 8 rue du Gnral Sarrail,Universit Paris-Est, Crteil, France, or [email protected].

N Engl J Med 2010;363:1621-30.Copyright 2010 Massachusetts Medical Society.

A b s t ra c t

Background

In adult humans, the follicle-stimulating hormone (FSH) receptor is expressed only

in the granulosa cells of the ovary and the Sertoli cells of the testis. It is minimallyexpressed by the endothelial cells of gonadal blood vessels.

Methods

We used immunohistochemical and immunoblotting techniques involving four sepa-

rate FSH-receptor-specific monoclonal antibodies that recognize different FSH re-

ceptor epitopes and in situ hybridization to detect FSH receptor in tissue samples

from patients with a wide range of tumors. Immunoelectron microscopy was used

to detect FSH receptor in mouse tumors.

Results

In all 1336 patients examined, FSH receptor was expressed by endothelial cells in

tumors of all grades, including early T1 tumors. The tumors were located in the pros-

tate, breast, colon, pancreas, urinary bladder, kidney, lung, liver, stomach, testis, and

ovary. In specimens obtained during surgery performed to remove tumors, the FSH

receptor was not expressed in the normal tissues located more than 10 mm from the

tumors. The tumor lymphatic vessels did not express FSH receptor. The endothelial

cells that expressed FSH receptor were located at the periphery of the tumors in a

layer that was approximately 10 mm thick; this layer extended both into and outside

of the tumor. Immunoelectron microscopy in mice with xenograft tumors, after

perfusion with antiFSH-receptor antibodies coupled to colloidal gold, showed that

the FSH receptor is exposed on the luminal endothelial surface and can bind and

internalize circulating ligands.

Conclusions

FSH receptor is selectively expressed on the surface of the blood vessels of a wide

range of tumors. (Funded by INSERM.)

The New England Journal of Medicine

Downloaded from nejm.org by SIANIPAR MANGARA on August 9, 2013. For personal use only. No other uses without permission.

Copyright 2010 Massachusetts Medical Society. All rights reserved.


2/10

Th e n e w e n g l a n d j o u r n a l o f medicine

n engl j med 363;17 nejm.org october 21, 20101622

The follicle-stimulating hormone

(FSH) receptor is a glycosylated transmem-

brane protein that binds FSH and belongs

to the family of G-protein-coupled receptors. FSH,

a key hormone in mammalian reproduction, is

produced mainly in the anterior pituitary gland,

and the target organs are the ovary and testis. In

females, FSH stimulates follicular maturation andestrogen production through aromatization of an-

drogens.1 In males, FSH stimulates Sertoli-cell pro-

liferation in immature testes and maintains nor-

mal spermatogenesis in adults.2

In adult humans and animals, the FSH recep-

tor is known to be expressed only in the testicu-

lar Sertoli cells and the ovarian granulosa cells,3,4

and it is expressed in low levels in the endothe-

lial cells of the ovary5 and testis.6 In the testis,

the FSH receptor mediates the translocation of

FSH across the blood-testis endothelial barrier

by a process of receptor-mediated transcytosis.6We conducted a study to assess endothelial-cell

expression of the FSH receptor in blood vessels in

a wide range of human cancers.

Methods

Tissue Specimens

Tumor specimens (see Table 1 in the Supplemen-

tary Appendix, available with the full text of this

article at NEJM.org) were obtained from 1336 pa-

tients immediately after surgery. The patients did

not receive cytotoxic agents or hormones before

surgery. The specimens were fixed in formalin

and embedded in paraffin. Five study investiga-

tors performed histologic analysis of each tumor

specimen. Gleason scores and stages for prostate

tumors were assigned according to the World

Health Organization guidelines.7 (The Gleason

score is the sum of the two most common histo-

logic patterns or grades in a prostate tumor, each

of which is graded on a scale of 1 to 5, with 5 be-

ing the most cytologically aggressive.) The other

tumor types were graded histologically accord-ing to the standards of the American Joint Com-

mittee on Cancer.8

The protocol was approved by the institutional

review board or ethics committee at each study

site. Written informed consent was obtained at the

time of surgery from all living donors from whom

samples were obtained. Control samples consist-

ed of normal tissue that routinely accompanies

tumors removed by surgery. Donors of normal

tissue also provided written informed consent.

Antibodies

Two FSH-receptor antibodies, FSHR18 and

FSHR323, were purified from hybridomas (Amer-

ican Type Culture Collection numbers CRL-2688

and CRL-2689). Two other FSH-receptor antibodies,FSHR190 and FSHR225, were provided by anoth-

er researcher. The antibodies were shown to be

monospecific for the FSH receptor by the follow-

ing three independent methods: immunoblotting

of cell extracts from ovarian specimens and of ly-

sates of cell cultures transfected with FSH-receptor

complementary DNA (cDNA), coimmunoprecipi-

tation of radiolabeled FSH from cells that express

the FSH receptor, and immunohistochemical de-

tection of the FSH receptor in cells known to be

targets for the FSH in human tissues (granulosa

cells in the ovary and Sertoli cells in the testis).5

In Situ Hybridization and Confocal

Microscopy

In situ hybridization was performed with the use

of a biotinylated cDNA antisense oligonucleotide

probe9 (for details, see the Supplementary Ap-

pendix).Indirect immunofluorescence confocal

microscopy was performed on paraformaldehyde-

fixed cryostat sections obtained from unfixed

frozen prostate specimens or on paraffin-embed-

ded sections. The FSH receptor was detected with

the use of the FSHR323 antibody (for details, see

the Supplementary Appendix).

Immunoprecipitation, Electrophoresis,

and Western Blotting

Frozen tissue was solubilized in TRIS buffer

(pH 7.4) containing 0.4 M sodium chloride and

1.2% Triton X-100 (for details, see the Supplemen-

tary Appendix), and the FSHR323 antibody was

added to the supernatant. The complexes were cap-

tured on protein-A Sepharose beads and analyzed

by means of sodium dodecyl sulfatepolyacryl-amide-gel electrophoresis (SDS-PAGE), followed

by immunoblotting with the use of the FSHR18

antibody.

Immunoelectron Microscopy

Tumors were generated by injecting nude mice

with LNCaP human prostate-cancer cells (for de-

tails, see the Supplementary Appendix). The blood





3/10

FSH-Receptor Expression in Tumor Blood Vessels


was removed by systemic perfusion, and anti-FSH-

receptor antibodies coupled to colloidal gold were

systemically introduced by means of perfusion for

20 minutes. The vasculature was washed and pro-

cessed for electron microscopy.6

Results

Expression of FSH Receptor by Endothelial

Cells in Prostate Tumors

Our initial analysis was performed in human pros-

tate tumors with the use of the antiFSH-receptor

monoclonal antibody FSHR323. Immunohisto-

chemical analysis of paraff in-embedded sections

revealed strong staining of endothelial cells in tu-

mors (Fig. 1A). The FSH-receptor-positive blood

vessels were located at the periphery of the tu-

mors, as detailed below. No staining of endothe-

lial cells was visible in normal prostate tissue

(Fig. 1B), which was located more than 10 mmoutside the tumors in specimens obtained by total

prostatectomy. A faint FSH-receptor signal was

visible occasionally in the tumor cells (Fig. 1A).

No staining occurred when tissue sections ob-

tained from patients with prostate cancer were in-

cubated with nonimmune mouse IgG2a of the

same isotype as FSHR323 (Fig. 1A in the Supple-

mentary Appendix), or when the primary antibody

was omitted (Fig. 1B in the Supplementary Ap-

pendix). The endothelial cells were identified as

belonging to blood vessels by costaining them

with an antibody against the vascular endothelial

marker, von Willebrand factor (Fig. 1F, 1G, and

1H). The lymphatic vessels in tumors, identified

with the use of the monoclonal antibody D2-40,

did not express the FSH receptor (Fig. 1D and 1E).

The identity of the antigen recognized in tumor

endothelial cells by the FSHR323 antibody was

confirmed by immunohistochemical analysis with

two other well-characterized monoclonal anti-

bodies, FSHR190 and FSHR225, which bound epi-

topes of the FSH receptor that are different from

those recognized by FSHR3235

and revealed astaining pattern identical to that of FSHR323

(Fig. 2 in the Supplementary Appendix).

The identity of the antigen was further con-

firmed by immunoprecipitation followed by de-

tection with an independent monoclonal antibody.

Prostate-tumor extract was immunoprecipitated

with the use of the FSHR323 antibody, subjected

to SDS-PAGE, and analyzed by means of immu-

noblotting with the use of FSHR18, which rec-

ognizes an epitope that is different from that

recognized by FSHR323.5 An 87-kD band corre-

sponding to the known molecular weight of the

mature glycosylated FSH receptor5 was detected

in the prostate-cancer tissue (Fig. 2A). No FSH-

receptor signal was visible in extracts from

normal-appearing prostate tissue obtained frompatients with prostate cancer (Fig. 2A). It is ex-

tremely improbable that a protein other than the

FSH receptor contains both epitopes recognized

by the two antibodies and has the same molecu-

lar weight as the FSH receptor. In situ hybridiza-

tion confirmed the expression of FSH-receptor

RNA in the tumor endothelial cells (Fig. 2B and

2C). Thus, all three techniques used (immuno-

histochemical analysis with the use of three in-

dependent antibodies, immunoblotting, and in

situ hybridization) identified the FSH receptor.

The same consistent expression of the FSHreceptor by the endothelial cells of tumor blood

vessels was detected in tissue specimens ob-

tained from all 773 patients with prostate can-

cer who were evaluated, and the absence of ex-

pression was confirmed in all associated normal

tissues. The lowest Gleason score in the tumors

analyzed was 5; this score characterizes tumors

of low malignant potential. Among the 773 tu-

mors analyzed, 41% had Gleason scores of 5 or

6. The most advanced tumors analyzed had the

highest Gleason score, 10. FSH-receptorpositive

blood vessels were also detected in approximately

20% of benign prostatic hyperplasia specimens, in

vessels surrounding hyperplastic glands (Fig. 1C).

In contrast to the findings in the tumors, the FSH-

receptorpositive vessels were present throughout

the hyperplastic areas, not only at the periphery.

The anatomical location of benign prostatic hy-

perplasia differed from the sites of most prostate

tumors10 and thus, FSH-receptorbased imaging

could be used to distinguish the two conditions.

Expression of FSH Receptor by EndothelialCells in Other Cancers and Nonmalignant

Tissues

Immunohistochemical studies similar to those

described above were performed for 10 other tu-

mor types in a total of 563 patients (for details,

see Table 1 in the Supplementary Appendix). Ap-

proximately 70% of the tumors were grade I or II,

approximately 25% were grade III, and the re-





4/10



A

D

F G H

E

B C

*

Blood Vessel

Blood Vessel

LymphaticVessel

LymphaticVessel

Figure 1. Expression of Follicle-Stimulating Hormone (FSH) Receptor by Vascular Endothelial Cells in Human ProstateTumors.

Immunohistochemical analysis was performed on paraffin-embedded sections of human prostate tissues with theuse of the anti-FSH-receptor monoclonal antibody 323, followed by a secondary peroxidase-coupled antibody visual-

ized with the use of the red-brown peroxidase-reaction product. Sections were also stained with hematoxylin (Pan-els A and B). A prostate tumor with a Gleason score of 6 showed strong staining of vascular endothelial cells (Panel A,

arrows), whereas blood vessels of normal prostate tissue were negative for the FSH receptor (Panel B, arrows). A faintFSH-receptor signal was visible occasionally in the tumor cells (Panel A, asterisk). The FSH receptor was expressed

in the blood vessels of hyperplastic tissue in 20% of samples from patients with benign prostatic hyperplasia (Panel C,arrows). In serial sections of tumor tissue, the blood vessels showed a strong presence of FSH receptor (Panel D),

whereas the lymphatic vessels, identified with the use of the monoclonal antibody D2-40 (Panel E), did not expressthe FSH receptor. Double immunofluorescence on prostate-tumor tissue confirmed the identity of the cells express-

ing the FSH receptor (Panels F through H). An antibody against the vascular endothelial-cell marker von Willebrand

factor, followed by a green-labeled secondary antibody (Panel G), overlapped with the signal from the anti FSH-receptor antibody detected by a secondary red-labeled antibody (Panel F). Merging of the two antibody signals with

a Nomarski image is shown in Panel H. The scale bar represents 25 m in all panels.





5/10



maining tumors were grade IV. Examples of oth-

er types of tumors are shown in Figure 3. In each

of the tumors analyzed, without exception, we de-

tected consistent expression of the FSH receptor

by endothelial cells. As we found with the pros-

tate tumors, occasional cancer cells were also

faintly stained in breast tumors and exocrine

pancreatic tumors (data not shown).In all normal control samples, which were ob-

tained from the same sections as the tumor sam-

ples, the endothelial cells of blood vessels were

negative for the FSH receptor (Fig. 3 in the Sup-

plementary Appendix), with the exception of a

faint FSH-receptor signal that was barely visible

in the blood vessels of testes and ovaries (Fig. 4 in

the Supplementary Appendix), which are known to

express the FSH receptor.5,6

We also analyzed nonmalignant inflammatory,

regenerative, and proliferative tissues and found

that expression of the FSH receptor is not a gen-eral feature of such tissue responses. In samples

from patients with rheumatoid arthritis (Fig. 3B,

panel e), chronic pancreatitis (Fig. 3B, panel f),

Crohns disease, and wound healing (Fig. 5A, 5B,

and 5C in the Supplementary Appendix), no FSH-

receptor expression was detected in the endothe-

lial cells. In placental endothelial cells, high lev-

els of FSH receptor were expressed in all vessels

(Fig. 5D in the Supplementary Appendix).

Localization of FSH Receptor at Tumor

Periphery

A general characteristic of the vessels with endo-

thelial cells that expressed the FSH receptor in

prostate tumors was that they were located at

the periphery of the tumors, in shells that had a

thickness of approximately 10 mm (range, 7 to 15)

and extended a few millimeters both inside and

outside the tumor in the apparently normal tissue.

No FSH receptor-expressing vessels were detected

in the deeper areas of the tumors. Figure 4 shows

the distribution of the vessels expressing the FSH

receptor.The same shell-type distribution of endothelial

cells expressing the FSH receptor was observed in

all 11 tumor types examined, with the exception

of renal-cell carcinomas; in the latter type, in ap-

proximately 30%, the FSH receptor was expressed

uniformly in the vessels throughout the tumor,

and in approximately 40%, the FSH receptor was

expressed only at the exterior of the tumor. The

percentage of vessels expressing the FSH recep-

tor reached a maximum of 40 to 100% at the de-

marcation line between the tumor and the normal

tissue and decreased gradually to zero both to-

ward the interior and away from the tumor. The

values for both shell thickness and maximum per-

centage of FSH-receptorpositive vessels were low-

est in tumors of the exocrine pancreas and hepato-

BA

C

Molecular

Mass

Cancerou

sTiss

ue

Normal

Tiss

ue

220

97.4

87

66

46

30

Figure 2. Immunoblotting and In Situ Hybridization Confirmation of the

Identity of the Antigen Recognized in Tumors by the FSHR323 Antibody.

Panel A shows the results of immunoblotting to detect the follicle-stimulat-

ing hormone (FSH) receptor. Equal amounts of Triton X-100 extracts fromcancerous and normal prostate tissues from the same donor were immu-

noprecipitated with the FSHR323 antibody, subjected to sodium dodecylsulfatepolyacrylamide-gel electrophoresis under reducing conditions, and

transferred to nitrocellulose membranes. The samples were probed withthe FSHR18 monoclonal antibody detected by ECL. FSHR18 recognized an

epitope that was different from the epitope recognized by FSHR323. The ar-row in the column of molecular-mass markers indicates the known size of

the mature glycosylated FSH receptor, 87 kD. The faint bands correspond

to the mouse IgG used for immunoprecipitation. When the experiment wasrepeated with tissues obtained from four other patients, the results were

identical. Panels B and C show FSH-receptorpositive blood vessels (arrows)detected by in situ hybridization. FSH-receptor messenger RNA was revealed

with the use of a biotinylated probe detected by a streptavidinalkaline phos-phatase conjugate, visualized by the purple phosphatase-reaction product.

The sections were also stained with methyl green. The specimens wereobtained from a patient with prostate cancer (Panel B) and a patient with

clear-cell renal-cell carcinoma (Panel C). The scale bars represent 25 m.





6/10



a

b

c

d

e

f

a

b

c

d

e

f

A

B

Figure3.

TumorTissuesfromO

therTypesofHumanCancersandInflam

matoryTissues.

Theperoxidase-coupledantibo

dywasvisualizedwiththeuseofthered-b

rownperoxidase-reactionproduct.Sectio

nswerealsostainedwithhematoxylin.Allarrowsshownpointto

bloodvessels.PanelAshowss

pecimensobtainedfrompatientswithbre

astcancerinsitu(a),nonsmall-celllung

cancer(b),liveradenocarcinoma(c),colo

nadenocarcinoma(d),

pancreaticadenocarcinoma(e)

,andstomachadenocarcinoma(f)(scale

bar,25m).PanelBshowsspecimensob

tainedfrompatientswithatestistumor(seminoma)(a),ovarian

cancer(b),urinarybladdercan

cer(c),andclear-cellrenal-cellcarcinoma

(d).TheFSHreceptorwasnotexpressedbytheendothelialcellsinspecimensfrom

patientswithrheuma-

toidarthritis(inthesynovialtissue)(e)andchronicpancreatitis(inthep

ancreaticparenchyma)(f)(scalebar,25

m).





7/10



carcinomas; intermediate in tumors of the urinary

bladder, ovary, lung, and stomach; and highest

in tumors of the prostate, kidney, colon, breast,

and testis. The thickness of the shell did not ap-

pear to be related to the size of the tumor.

Exposure of FSH Receptor on the Luminal

Surface of the Tumor Epithelium

Tumor-specific endothelial receptors could be can-

didates for tumor imaging and therapy because

they may be directly accessible to intravenously

delivered agents, which may also accumulate in

the tumors after crossing the endothelial cells.

In Situ Animal Model

To assess whether endothelial FSH receptors are

accessible to intravenously delivered ligands, we

used an in situ animal model: mice that carried

LNCaP human xenograft tumors. The tumor en-

dothelial cells in these mice express the FSH re-ceptor (Fig. 6 in the Supplementary Appendix).

After perfusion of these mice for 20 minutes with

anti-FSH-receptor antibodies coupled to colloidal

gold, immunoelectron microscopy showed that in

tumors, gold particles were attached to the lumi-

nal aspect of the endothelium, on the plasma-

lemma proper (Fig. 5A), on the diaphragms of

fenestrae (Fig. 5B), and in the coated pits (Fig. 5C).

Particles were also visible in the interior of the

endothelial cells in coated vesicles (Fig. 5D), en-

docytic vesicles (Fig. 5E), and infrequently in mul-

tivesicular bodies (Fig. 5F). No tracer was detected

in the intercellular junctions of the endothelial

layer (Fig. 5G), indicating that no intercellular

transport occurred. Particles were also visible in

the lumen of channels that connect the luminal

and abluminal fronts of the endothelium and in

the subendothelial space adjacent to these struc-

tures (Fig. 7A in the Supplementary Appendix).

Accumulations of gold particles were infrequent-

ly visible in the interstitial space (Fig. 7B in the

Supplementary Appendix).

In contrast to the findings in the tumor en-dothelial cells, extremely rare gold particles were

observed in endothelial cells of normal organs

(lung and prostate) obtained from the same ani-

mals. Most microscopical f ields did not show any

particles in the lung tissue (Fig. 5H) or the pros-

tate tissue. The most likely explanation for their

presence is that they were the few unbound par-

ticles in the blood vessels that had not been re-

moved by washing, and they were subsequently

cross-linked by glutaraldehyde to the endothelial-

cell surface. In the tumors, the gold particles were

visible on all examined endothelial cells in all

blood-vessel profiles; by contrast, in testes, the

particles were present in only approximately 10%

of the vessels, on the endothelial-cell plasma mem-

branes, and in coated pits, and at much lower

density than in the tumor blood vessels. Isolated

gold particles were also visible in rare cases in

the subendothelial space of the testes.

Discussion

We report FSH-receptor expression by endotheli-

al cells of blood vessels in a wide range of tumors

in all 1336 patients examined. Two previous stud-

ies showed immunohistochemical detection of the

FSH receptor in the tumor cells of prostate ade-

nocarcinomas.11,12 The presence of the FSH recep-

tor in the endothelial cells was not mentioned,

possibly because specimens from the peripheral

areas had not been examined. This supposition

VesselsExpressingFSH

Receptor(%)

No.ofVesselspermm2

100

75

50

25

0

50

40

30

10

20

06 4 2 0 642 8 10

Distance from Tumor Border (mm)

Tumor Normal Tissue

FSH-expressing vessels No. of vessels

Figure 4. FSH-Receptor Expression According to Vessel Location.

The blood vessels were visualized with the use of antivon Willebrand fac-

tor antibodies followed by Alexa-488 dye secondary antibodies, and FSH-receptorstained vessels were visualized by the FSHR323 antibody followed

by Alexa-555 dyelabeled secondary antibodies. The vessels were countedon 148 microscopical digital images of tumors obtained from five patients.

Zero indicates the border of the tumor, the negative numbers indicate theinterior of the tumor, and the positive numbers indicate the exterior of the

tumor. The red circles and dashed line represent the percentage of FSH-receptorexpressing vessels. The blue squares indicate the total number

of vessels per square millimeter; the mean number was higher in the interi-

or of the tumor than in the exterior (372 vs. 251 vessels per square milli-meter, P


8/10



cannot be verified because the antibodies are no

longer available.

The location of the FSH-receptorpositive ves-

sels in the normal tissue immediately adjacent tothe tumor is consistent with the view that the tu-

mor cells in the invasive front attract surround-

ing blood vessels toward the tumor, and during

this process, FSH-receptor expression is activated.

Another possibility is suggested by the observa-

tion that in breast tumors in humans, endothelial

cells proliferate at the tumor periphery but not in

the interior.13 Accordingly, FSH-receptor expres-

sion by endothelial cells may be associated with

their proliferation in this particular location.

If it becomes possible to exploit FSH-receptor

expression for imaging purposes, the location of

the FSH-receptor signal at the boundary between

the tumoral and the normal tissues should makeit useful for defining the target volume for ra-

diation therapy or surgery.

If it can be shown that intravenously adminis-

tered antibodies can detect tumor endothelial

cells, the presence of FSH receptor on the surface

of these cells in a wide range of tumors makes it

a potential target for both tumor imaging and

therapy. As shown in Figure 4 in the Supplemen-

tary Appendix, FSH-receptor density on the go-

nadal endothelial cells is much lower than that

BasalMembrane

A B C

D E F

G H

Lumen

Lumen Lumen

Lumen Lumen

PlasmaMembrane

Lumen

Lumen Lumen

Plasma Membrane

Fenestra

Fenestra

JunctionJunction

Luminal CaveolaLuminal Caveola

Coated Pit

CoatedVesicle

LuminalCaveola

Endosome

Endosome

MultivesicularBody

Figure 5. AntiFSH-Receptor Antibodies Coupled to Perfused Colloidal Gold in the Vasculature of Tumor-Bearing Mice.

Tumors generated by the prostate-cancer line LNCaP were excised after a 20-minute perfusion of the mice with5-nm gold particles conjugated to FSHR323 and were processed for electron microscopy. Ultrathin sections were

stained with uranyl acetate and lead citrate. The conjugate particles were bound to the plasma membrane (Panel A),to the diaphragms of fenestrae (Panel B), and to luminal coated pits (Panel C), and they were internalized in coated

vesicles (Panel D), endosomes (Panel E), and multivesicular bodies (Panel F). The particles did not cross the endo-thelial layer through the junctions (Panel G). No binding was visible on either the luminal plasma membrane or the

luminal caveolae associated with the surface of capillaries in the lung, the first organ of the mouse that was accessi-ble by perfusion to the FSHR323-gold particles (Panel H). Scale bars in Panels A through F and Panel H represent

50 nm, and the scale bar in Panel G represents 100 nm.





9/10



in tumors, and therefore, it may be possible to

find a therapeutic window so that the gonadal

vessels are not substantially affected. The Sertoli

and granulosa cells, which express high levels of

FSH receptor, could be spared if the toxic agents

targeted at the FSH receptor were delivered in a

particulate form that would not cross the blood-

gonadal endothelial barrier. The volume that canbe targeted by FSH receptor represents a substan-

tial fraction of the volume of small tumors (e.g.,

a 3-mm-thick peripheral layer inside a tumor with

a 2-cm diameter accounts for 66% of its volume,

leaving an unstained core with a 14-mm diame-

ter). The functional relevance of the volume that

can be targeted may be even greater than its

geometric contribution because of a more in-

tense proliferation of cancer cells at the tumor

periphery.14-19

Our in situ experiments cannot be considered

to be a proof-of-principle demonstration that theFSH receptor expressed on tumor-associated blood

vessels can be exploited clinically. We removed

the blood from the mice before perfusing gold-

labeled antibodies into vasculature. Our model

does not mimic a clinical application in humans.

The binding of FSH to FSH receptor in ovarian

granulosa cells induces an increase in hypoxia-

inducible factor 1 protein levels under normoxic

conditions, which in turn leads to up-regulation

of vascular endothelial growth factor (VEGF).20

This observation provides support for the specu-

lation that FSH-receptor expression could in-

duce VEGF and VEGF receptor 2 (VEGFR-2) signal-

ing in tumor endothelial cells and thus promote

angiogenesis. The fact that FSH and FSH-receptor

signaling is known to generate activated Gq/11

protein21 suggests yet another biologic role for

the presence of the FSH receptor in tumor en-dothelial cells. Gq/11 has been shown to induce

VEGFR-2 signaling in human umbilical-vein

endothelial cells, even in the absence of VEGF.22

This effect may substantially enhance the prolif-

eration and migration of endothelial cells in

cancer independently of VEGF availability. For

both mechanisms, we speculate that blocking

FSH-receptor signaling could be a new antitumor

strategy.

In conclusion, we found that the FSH receptor,

which was present on the endothelial surface of

blood vessels at the periphery of a wide range oftumors, was accessible to agents injected intra-

venously.

Supported by INSERM.Disclosure forms provided by the authors are available with

the full text of this article at NEJM.org.We thank Sylvette Reposo and Pascale Soyeux-Porte for excel-

lent technical assistance, Drs. Karen Leroy and Jeanne Tran Van

Nhieu for some of the liver and colon antibodies, Dr. HuguesLoosfelt for the FSHR190 and FSHR225 specimens, Dr. Xavier

Decrouy for assistance with confocal microscopy, Dr. ChristoChristov for assistance with electron microscopy, and Dr. Fran-

cis Vachereau for the LNCaP xenografts.

References

1. Macklon NS, Fauser BC. Follicle de-velopment dur ing the normal menstrual

cycle. Matur itas 1998;30:181-8.

2. Plant TM, Marshall GR. The function-

al signif icance of FSH in spermatogenesis

and the control of its secretion in maleprimates. Endocr Rev 2001;22:764-86.

3. Sprengel R, Braun T, Nikolics K, Sega-loff DL, Seeburg PH. The testicular recep-

tor for follicle-stimulating hormone: struc-ture and functional expression of cloned

cDNA. Mol Endocrinol 1990;4:525-30.

4. Simoni M, Gromoll J, Nieschlag E. Thefollicle stimulating hormone receptor: bio-

chemistry, molecular biology, physiology,and pathophysiology. Endocr Rev 1997;

18:739-73.

5. Vannier B, Loosfelt H, Meduri G, PichonC, Milgrom E. Anti-human FSH receptor

monoclonal antibodies: immunochemicaland immunocytochemical characterization

of the receptor. Biochemistry 1996;35:1358-

66.6. Vu Hai MT, Lescop P, Loosfelt H,

Ghinea N. Receptor-mediated transcytosisof follicle stimulating hormone through

the rat testicular microvasculature. BiolCell 2004;96:133-44.

7. Eble JN, Sauter G, Epstein JI, Sester-henn IA, eds. World Health Organization

classification of tumours: pathology andgenetics of tumours of the urinary system

and male genital organs. Lyon, France:

IARC Press, 2004.8. Green FL, Compton CC, Fritz AG, Shah

JP, Winchester DP, eds. American JointCommittee on Cancer: AJCC cancer stag-

ing atlas. Heidelberg, Germany: Springer-Verlag, 2006.

9. Bckers TM, Nieschlag E, Kreutz MR,

Bergmann M. Localization of follicle-stim-ulating hormone (FSH) immunoreactivity

and hormone receptor mRNA in testicu-lar tissue of infert ile men. Cell Tissue Res

1994;278:595-600.

10. De Marzo AM, Platz EA, Sutclif fe S, etal. Inflammation in prostate carcinogen-

esis. Nat Rev Cancer 2007;7:256-69.

11. Mariani S, Salvatori L, Basciani S, et

al. Expression and cellular localization of

follicle-stimulating hormone receptor innormal human prostate, benign prostatic

hyperplasia and prostate cancer. J Urol2006;175:2072-7.

12. Ben-Josef E, Yang SY, Ji TH, et al. Hor-mone-refractory prostate cancer cells ex-

press functional follicle-stimulating hor-mone receptor (FSHR). J Urol 1999;161:

970-6.

13. Fox SB, Gatter KC, Bicknell R, et al.

Relationship of endothelial cell prolifera-

tion to tumor vascularity in human breastcancer. Cancer Res 1993;53:4161-3.

14. Begum R, Douglas-Jones AG, MorganJM. Radial intratumoral increase and cor-

relation of microvessels and proliferationin solid breast carcinoma. Histopatholo-

gy 2003;43:244-53.

15. Belin JA, van Diest PJ, Baak JP. Rela-tionships between vascularization and pro-

liferation in invasive breast cancer. J Pathol1999;189:309-18.

16. Shoji M, Dobashi Y, Morinaga S, Jiang

SX, Kameya T. Tumor extension and cellproliferation in adenocarcinomas of the

lung. Am J Pathol 1999;154:909-18.

17. Rodins K, Cheale M, Coleman N, Fox

SB. Minichromosome maintenance pro-

tein 2 expression in normal kidney andrenal cell carcinomas: relationship to tu-

mor dormancy and potential clinical util-ity. Clin Cancer Res 2002;8:1075-81.

18. Fernebro J, Engellau J, Persson A, Ryd-holm A, Nilbert M. Standardizing evalua-





10/10



tion of sarcoma proliferation: higher Ki-67 expression in the tumor periphery than

the center. APMIS 2007;115:707-12.

19. Ohashi K, Nemoto T, Eishi Y, Matsu-

no A, Nakamura K, Hirokawa K. Prolifer-

ative activity and p53 protein accumula-tion correlate with early invasive trend, and

apoptosis correlates with differentiationgrade in oesophageal squamous cell car-

cinomas. Virchows Arch 1997;430:107-15.20. Alam H, Weck J, Maizels E, et al. Role

of the phosphatidylinositol-3-kinase andextracellular regulated kinase pathways

in the induction of hypoxia-inducible fac-tor (HIF)-1 activity and the HIF-1 target

vascular endothelial growth factor in ovar-

ian granulosa cells in response to follicle-stimulating hormone. Endocrinology 2009;

150:915-28.

21. Castro-Fernndez C, Maya-Nez G,

Mndez JP. Regulation of follicle-stimulat-ing and luteinizing hormone receptor sig-

naling by regulator of G protein signalingproteins. Endocrine 2004;25:49-54.

22. Zeng H, Zhao D, Yang S, Datta K, Muk-hopadhyay D. Heterotrimeric G alpha q/G

alpha 11 proteins function upstream of

vascular endothelial growth factor (VEGF)receptor-2 (KDR) phosphorylation in vas-

cular permeability factor/VEGF signaling.J Biol Chem 2003;278:20738-45.

Copyright 2010 Massachusetts Medical Society.



Documents

Blood Barrier Testis