American Journal of Transplantation 2006; 6: 2601–2611Blackwell Munksgaard

C© 2006 The AuthorsJournal compilation C© 2006 The American Society of

Transplantation and the American Society of Transplant Surgeons

doi: 10.1111/j.1600-6143.2006.01534.x

Antiangiogenic and Immunomodulatory Effectsof Rapamycin on Islet Endothelium: Relevancefor Islet Transplantation

V. Cantaluppia, L. Bianconea, G. MaurielloRomanazzia, F. Figliolinia, S. Beltramoa,M. S. Ninnirib, F. Galimib, R. Romagnolic,A. Franchelloc, M. Salizzonic, P. Cavallo Perina,C. Ricordid, G. P. Segolonia and G. Camussia,∗

aDepartment of Internal Medicine and Research Centrefor Experimental Medicine (CeRMS), University of Torino,Corso Dogliotti 14, Torino 10126, ItalybDepartment of Biomedical Sciences/INBB, University ofSassari, ItalycLiver Transplantation Centre, University of Torino, CorsoBramante 88, Torino 10126, ItalydDiabetes Research Institute, University of Miami, 1450NW 10 Avenue, Miami, Florida, USA∗Corresponding author: Giovanni Camussi,giovanni.camussi@unito.it

Donor intra-islet endothelial cells contribute to neo-vascularization after transplantation. Several factorsmay interfere with this process and ultimately in-fluence islet engraftment. Rapamycin, a central im-munosuppressant in islet transplantation, is an mTORinhibitor that has been shown to inhibit cancer angio-genesis. The aim of this study was to evaluate the ef-fects of rapamycin on islet endothelium. Rapamycininhibited the outgrowth of endothelial cells fromfreshly purified human islets and the formation ofcapillary-like structures in vitro and in vivo after sub-cutaneous injection within Matrigel plugs into SCIDmice. Rapamycin decreased migration, proliferationand angiogenic properties of human and mouse islet-derived endothelial cell lines with appearance of apop-tosis. The expression of angionesis-related factorsVEGF, a Vb 3 integrin and thrombospondin-1 on isletendothelium was altered in the presence of rapamycin.On the other hand, rapamycin decreased the surfaceexpression of molecules involved in immune processessuch as ICAM-1 and CD40 and reduced the adhesion ofT cells to islet endothelium. Our results suggest thatrapamycin exerts dual effects on islet endothelium in-ducing a simultaneous inhibition of angiogenesis anda down-regulation of receptors involved in lymphocyteadhesion and activation.

Key words: Endothelial cells, engraftment, rapamycin

Abbreviations: EndoGF medium, endothelial growthfactor-enriched medium; VEGF, vascular endothelial

growth factor; bFGF, beta fibroblast growth factor;PDGF, platelet-derived growth factor; HGF, hepato-cyte growth factor; TGFbeta, transforming growthfactor beta; TSP-1, thrombospondin-1; mTOR, mam-malian target of rapamycin; hIECs, human islet en-dothelial cells; mIECs, mouse islet endothelial cells;GFP, green fluorescent protein; vWF, von Willebrandfactor; TNF-alpha, tumor necrosis factor-alpha; IFN-gamma, interferon-gamma; ICAM-1, intracellular ad-hesion molecule-1; OD, optical density

Received 6 June 2006, revised 11 July 2006 andaccepted for publication 17 July 2006

Introduction

Clinical islet transplantation has recently received a strong

impulse from the results obtained by the introduction of a

rapamycin-based glucocorticoid-free immunosuppressive

regimen, leading to insulin independence at 1 year in 90%

of the treated patients (1). However, several problems with

the current procedure still need to be addressed to improve

certain aspects, such as organ procurement and preserva-

tion, islet isolation and culture, modality of transplant and

immunosuppression. Pancreata from multiple donors are

still needed to guarantee a sufficient islet mass since a

substantial number of transplanted islets fails to engraft

into the liver or suffers from poor vascular engraftment. In-

deed, native islets in vivo are richly vascularized and even

though islets make up only 1% of the pancreatic mass,

they receive about 10% of the blood flow (2). After trans-

plantation, islets are revascularized most likely by both the

host and the recipient endothelial cells that form a chimeric

vascular tree (3). However, revascularization process is not

immediate and transplanted islets show the first signs of

angiogenesis (i.e. capillary sprout formation and protrusion)

not earlier than 2 days after transplantation, and the entire

process is completed after 10–14 days. In addition, the vas-

cular density of vascularized transplanted islets is markedly

reduced in comparison to native islets (4,5). Based on these

considerations, the identification and consequent removal

of factors that may impair the angiogenic processes after

islet transplantation may likely increase the success of this

procedure.

2601

Cantaluppi et al.

Rapamycin is widely used as central immunosuppressant

for islet transplantation as part of the original Edmonton

protocol (1). The immunosuppressive mechanism of ra-

pamycin is based on the selective blockade of the mam-

malian target of rapamycin (mTOR) activation, a molecule

known to play a pivotal role in cell cycle progression from

late G1 into S phase in response to T-cell growth fac-

tor stimulation (6). Unfortunately, considering the ubiq-

uitous expression of mTOR in different cell types, the

effects of rapamycin are not restricted to the immune sys-

tem but affect different physiopathological processes in-

volved in cell survival and proliferation, inducing leucopenia

(7), thrombocytopenia (7), delays in wound repair (8) and

tubular regeneration after acute ischemic injury (9). More-

over, it has been recently shown that rapamycin inhibited

metastatic tumor growth and angiogenesis in an in vivo

mouse model (10). The dissection of this phenomenon

revealed that rapamycin exerted antiangiogenic activities

linked to a decrease in the production of vascular endothe-

lial growth factor (VEGF) and to a markedly inhibited re-

sponse of tumor endothelial cells to stimulation by VEGF

itself (10).

The aim of this study was to evaluate the effects of

rapamycin on islet endothelium in the early posttrans-

plantation processes such as islet revascularization and

endothelium–immune system interaction, events that may

strongly affect intrahepatic engraftment of islets.

Materials and Methods

Reagents and antibodiesDithizone, Hoechst 33258, propidium iodide (PI), TNF-alpha, IFN-gamma,

mouse monoclonal anti-human thrombospondin 1 (TSP-1) antibody and ra-

pamycin were from Sigma Chemical Company (St. Louis, MO). Rapamycin

was solubilized in DMSO and stored at −20◦C in the dark as previously

described (11). In all experimental procedures, the same amount of solvent

was used as control. Polyclonal rabbit anti-human insulin, anti-human CD31

(PECAM-1), anti-human tie-2 and anti-human VEGFR2 (KDR) antibodies

were from Santa-Cruz Biotechnology (Santa Cruz, CA). Mouse mono-

clonal anti-human VEGF was from US Biological (Swampscott, MA). Mouse

monoclonal anti-human aVb3 integrin was from Chemicon International

(Temecula, CA). Rabbit polyclonal anti-human von Willebrand factor (vWF),

Alexa Fluor-conjugated acetylated-LDL and Alexa Fluor-conjugated sec-

ondary antibodies were from Invitrogen (Carlsbad, CA). Mouse monoclonal

anti-human CD40 and anti-human CD54 (ICAM-1) antibodies were from

Serotec (Oxford, UK).

CellsFour different preparations of freshly purified human islets discarded from

transplant use for insufficient islet mass were prepared using the Ricordi

method (12) and used in this study. Purified islets (>90% pure) were cul-

tured in CMRL medium (Mediatech Inc., Herndon, VA) containing 5 mg/mL

albumin (Kedrion Spa, Lucca, Italy) and 2 mM glutamine (GIBCO BRL,

Gaithersburg, MD). Human (hIECs) and mouse (mIECs) pancreatic islet en-

dothelial cell lines were generated as previously described (13). Briefly, cells

outgrowing from islets were removed by trypsin/EDTA treatment and trans-

fected with 4 lg pBR322 plasmid vector containing SV40-T large antigen

gene at 250 mV and 960 lF in 4-mm electroporation cuvettes in an Invitro-

gen electroporator II (Invitrogen Corp., Carlsbad, CA). Clones were selected

for 1 mg/mL G418 resistance and screened for immunofluorescence ex-

pression of vWF. Positive clones were further subcloned by limiting dilu-

tion method and cultured in RPMI (Sigma), containing 10% FCS (Hyclone,

Logan, Utah) and 2 mM glutamine (GIBCO BRL). Where specified, cells

were cultured in RPMI enriched with endothelial growth factors (named

EndoGF medium and containing 10 ng/mL VEGF, 10 ng/mL bFGF, 10 ng/mL

PDGF and 0.5 U/mL heparin from 10% FCS (Hyclone, Logan, Utah) and

2 mM glutamine (GIBCO BRL).

Assessment of islet viability and functionViability assay of freshly purified islets was performed by UV light mi-

croscopy after double-staining the cells with 0.46 lM fluorescein diacetate

(FDA, Sigma) and 14.34 lM PI (Sigma). Islet function was evaluated by

ELISA-assay insulin secretory response (ALPCO Windham, NH) by prein-

cubation for 1 h in 2.8 mM glucose medium followed by 2-h incubation in

25 mM glucose medium. The stimulation indices were calculated by divid-

ing data of insulin secretion (mU/L/IEQ) in the presence of high glucose

medium by mean basal insulin secretion levels using a spectrophotometric

plate reader at 590-nm wave length.

Figure 1: Rapamycin inhibited the outgrowth of endothelialcells from freshly purified islets. (A and B) Representative im-

ages of dithizone-stained islets observed 5 days after plating on

tissue culture dishes in EndoGF medium in the absence (A) or

presence (B) of 10 ng/mL rapamycin (magnification 100×). (C

and D) Representative images of endothelial cells forming tubular

structures derived from islets 5 days after plating on Matrigel-

coated surfaces and cultured in EndoGF medium in the absence

(C) or presence (D) of 10 ng/mL rapamycin (magnification 100×).

(E and F) Phenotypic characterization of cells outgrowing from

islets in the presence of EndoGF medium. Endothelial pheno-

type was assessed by positive immunofluorescence staining with

anti-VEGFR2 (KDR) (panel E, magnification 400×, nuclei counter-

stained by 0.5 lg/mL Hoechst) and with anti-vWF antibodies (panel

F, magnification 400×, nuclei counterstained by 1 lg/mL propid-

ium iodide).

2602 American Journal of Transplantation 2006; 6: 2601–2611

Rapamycin and Islet Endothelium

Endothelial outgrowth from freshly purified isletsFreshly purified islets (500 IEQ) were plated on tissue culture dishes and

incubated with EndoGF medium in the presence or absence of rapamycin.

Endothelial outgrowth from islets was studied under a Nikon microscope

system for living cell analysis (14). The same experimental procedures were

performed on four different preparations of freshly purified islets.

Migration of hIECs and mIECsHIECs and mIECs were plated and rested for 12 h with RPMI containing 1%

FCS and then incubated with RPMI and different agonists. Cell migration

was studied with a 10× phase-contrast objective under the above-men-

tioned Nikon system. The net migratory speed (velocity straight line) was

calculated by the MicroImage software based on the straight line distance

between the starting and ending points divided by the time of observa-

tion (14). Migration of at least 30 cells for each experimental point was

analyzed.

Figure 2: Inhibition of hIECs and mIECs motility induced byrapamycin. (A and B) Analysis of scattered hIECs (A) and mIECs

(B) motility studied by time-lapse microscopy. Rapamycin induced

a time-dependent inhibition of spontaneous and growth factor-

induced migration of hIECs and mIECs. Migration of at least

30 cells for each experimental point was analyzed. All experi-

mental points were conducted in triplicate. Data are expressed

as speed averages (lm/h) ± SD. ANOVA with Newmann-Keuls

multicomparison test was performed: rapamycin induced a signif-

icant decrease of both hIECs and mIECs motility in comparison

to stimulation with normal or EndoGF medium at every period of

time analyzed (p < 0.005 at 2, 6 and 12 h after incubation). Three

independent experiments were performed with similar results.

In vitro angiogenesis assay of islets and hIECsIn vitro formation of vessel-like tubular structures was studied on 500 IEQ

human islets or 5000 hIECs seeded on growth factor-reduced Matrigel

(Becton Dickinson Labware, Bedford, MA) under an inverted microscope

in a plexiglass Nikon NP-2 incubator at 37◦C. After cells had attached, the

medium was removed and 0.5 mL medium containing different stimuli was

added. Image analysis was performed at 1-h intervals.

Generation of lentiviral vectors and islet endothelial celllines infectionViral production was performed as previously described (15,16). Briefly,

293T cells were transfected with a four-plasmid lentiviral system by the

CaCl2 precipitation method. Supernatants were collected 48 and 72 h af-

ter transfection, filtered and concentrated by two successive ultracentrifu-

gations. Viral preparation titers were determined by p24 ELISA (Alliance,

PerkinElmer, Wellesley, MA), by GFP titer determination on 293T cells and

Figure 3: Rapamycin-induced inhibition of migration of hIECsand mIECs was dose dependent. (A and B) Rapamycin de-

creases in a dose-dependent manner the EndoGF growth factor-

induced motility of hIECs (A) and mIECs (B). Migration of at least

30 cells for each experimental point was analyzed. All experimental

points were conducted in triplicate. Data are expressed as speed

averages (lm/h) ± SD. ANOVA with Newmann-Keuls multicom-

parison test was performed: rapamycin induced a significant de-

crease of both hIECs and mIECs motility in comparison to stimula-

tion with vehicle alone at the dose of 0.1 ng/mL, reaching a plateau

at 50 ng/mL (∗p < 0.005 EndoGF medium + different doses of ra-

pamycin vs. EndoGF alone). Three independent experiments were

performed with similar results.

American Journal of Transplantation 2006; 6: 2601–2611 2603

Cantaluppi et al.

by TaqMan real-time PCR determination of transduced proviral genomes.

For gene marking, hIECs and mIECs were plated on 24-well tissue culture

plates in EndoGF medium plus 10% FCS and then transduced overnight

with a lentivector carrying a CMV-GFP expression cassette, at MOI about

20. HIECs and mIECs infected with lentiviral vectors were termed hIECs-

GFP and mIECs-GFP, respectively.

Matrigel implants in miceSubcutaneous (s.c.) implantation of islets or islet-derived endothelial cell

lines in Matrigel plugs was performed to evaluate the antiangiogenic effects

of rapamycin in vivo. Briefly, Matrigel was mantained at −20◦C until use

and thawed at 4◦C O/N immediately before implant (17). Freshly purified

islets (2000 IEQ), hIECs-GFP or mIECs-GFP (104 cells) were resuspended

in 250 lL of fresh medium without FCS and mixed to 500 lL of Matrigel

on ice using cooled pipette tips in the absence or presence of 10 ng/mL

Figure 4: Rapamycin inhibited in vitro hIECs angiogenesis. (A)

Inhibitory effect of rapamycin on spontaneous and growth factor-

induced formation of capillary vessel-like tubular structures by

hIECs as shown by capillary connection counts per field. (B–E) Rep-

resentative images of angiogenesis inhibition of hIECs induced by

rapamycin challenge. When plated on Matrigel-coated wells, un-

stimulated hIECs show the formation of tubular structures within

12 h (B). EndoGF medium accelerates this angiogenic process (D),

whereas 10 ng/mL rapamycin induced a decrease of capillary con-

nections on both unstimulated (C) and growth factor-stimulated

(E) hIECs. All experiments were conducted in triplicate and

10 fields/each were analyzed. Data are expressed as average

number of capillary connections ± SD. ANOVA with Newmann-

Keuls multicomparison test was performed: (∗p < 0.005 rapamycin

or EndoGF medium vs. vehicle alone); (#p < 0.005 EndoGF

medium + rapamycin vs. rapamycin). Three independent exper-

iments were performed with similar results.

rapamycin and s.c. implanted into the scruff region of the neck of SCID

mice (human islets and hIECs-GFP) or of C57Bl/6 mice (mIECs-GFP). After 2

weeks mice were sacrificed and Matrigel plugs were retrieved for histology

and immunohistochemistry.

Immunofluorescence studiesCells outgrown from islets or hIECs cultured in chamber slides were fixed

with 1% paraformaldehyde, permeabilized with 0.1% Triton-X-100 (Sigma)

and stained with antibodies directed to endothelial antigens. All samples

were incubated with appropriate Alexa Fluor-conjugated secondary antibod-

ies. Matrigel implants containing human islets were fixed in formaldehyde

and embedded in paraffin prior to staining. All samples were counterstained

with 1 lg/mL PI or with 0.5 lg/mL Hoechst, mounted with antifade mount-

ing medium (Vector Laboratories, Burlingame, CA), and examined under a

UV light microscope. The evaluation of intraislet revascularization was per-

formed by confocal microscopy (Leica TCS SP2 Heidelberg, Germany) after

costaining for insulin and the endothelial marker CD31. The MicroImage

software was used to determine the number and the total area/section of

CD31-positive cells in neoformed vessels within islets and in surrounding

tissue.

HIECs and mIECs cytotoxicity assayHIECs and mIECs were cultured on 24-well plates (Falcon Labware, Oxnard,

CA) at a concentration of 5 × 104 cells/well, starved for 12 h without FCS

and then incubated with different doses of rapamycin (0.1–100 ng/mL) in

the presence or absence of endothelial growth factors in a medium without

phenol red containing 250 lg/mL XTT (Sigma). The absorption values at

450 nm were measured. All experiments were done in triplicate.

Detection of apoptosis of hIECs and mIECsHIECs and mIECs were subjected to TUNEL assay (terminal deoxynu-

cleotidyltransferase (TdT)-mediated dUTP nick end labeling) (ApopTag, On-

cor, Gaithersburg, MD) after starving for 12 h without FCS and incubation

for 48 h with different doses of rapamycin (ranging from 0.1 ng/mL to

100 ng/mL) in the presence of EndoGF medium. After stimulation, both

endothelial cell lines were fixed in 1% paraformaldehyde, postfixed in pre-

cooled ethanol-acetic acid 2:1, incubated with TdT enzyme in a humidified

chamber at 37◦C for 1 h and counterstained with antidigoxigenin-FITC an-

tibody and with PI (1 lg/mL). Samples were analyzed under a UV light mi-

croscope with an appropriate mounting medium. Green-stained apoptotic

cells were counted in different microscopic fields (magnification 100×).

Gene array technologyHuman GEarray kit for the study of angiogenesis markers (SuperArray Inc.,

Bethesda, MD) was used to characterize the gene expression profiles of un-

stimulated or stimulated freshly purified islets with 10 ng/mL of rapamycin

for 48 h. Hybridization was performed according to the manufacturer’s in-

structions as previously described (18).

Adhesion of Jurkat to hIECs monolayersJurkat cells (ATCC, Rockville, MD) were labeled overnight with 10 lm Vy-

brant Cell Tracer kit (Invitrogen) according to manufacturer’s in RPMI and

10% FBS (13). Labeled-Jurkat cells were counted, resuspended to 50 ×106 /mL RPMI without FCS and added to confluent monolayer of hIECs

cultured on six-well plates and previously incubated with vehicle alone or

inflammatory cytokines (10 ng/mL TNF-alpha and 10 ng/mL IFN-gamma) in

the presence or absence of 10 ng/mL rapamycin. Experiments were carried

out in triplicate for 1 h at 37◦C in conditions of slight agitation. At the end

of incubation, plates were filled with medium and aspirated three times to

remove unbound Jurkat cells. All samples were fixed with 1% paraformalde-

hyde and observed under a UV light microscope. Green fluorescent-Jurkat

cells were counted on 10 different fields at 200× magnification.

2604 American Journal of Transplantation 2006; 6: 2601–2611

Rapamycin and Islet Endothelium

Figure 5: Reduced vascularization offreshly purified human islets after sub-cutaneous transplantation into SCIDmice in the presence of rapamycin.(A and B) Representative images of im-

munofluorescence studies on freshly pu-

rified islets implanted into Matrigel plugs

in the scruff region of the neck of SCID

mice in the absence (A) or presence (B)

of 10 ng/mL rapamycin. Rapamycin in-

duced a marked decrease of endothelial

cells (arrow heads) outgrowing from the

transplanted islets (indicated by ‘I’) as de-

tected by immunofluorescence analysis

of tie-2 (blue staining in A and B), a marker

of endothelial phenotype. All sections

were counterstained by 1 lg/mL propid-

ium iodide and evaluated at 100× mag-

nification. (C and D) Rapamycin-induced

significant inhibition of neovasculariza-

tion of transplanted islets evaluated by

counts of number (C) and total area (D,

expressed in lm2/section) of neoformed

vessels within Matrigel. Student’s t-test

was performed: (∗p < 0.005 rapamycin

vs. vehicle alone). (E and F) Confocal mi-

croscope analysis of costaining for in-

sulin (red) and for the endothelial marker

CD31 (green) showed a reduced mi-

crovascular density within islets treated

with rapamycin (F) in comparison to vehi-

cle (E). All sections were counterstained

by 0.5 lg/mL Hoechst 33258 and eval-

uated at 200× magnification. (G and

H) Rapamycin-induced significant inhibi-

tion of intraislet vascularization of trans-

planted islets evaluated by counts of

number (G) and total area (H, expressed

in lm2/section) of vessels. Student’s

t-test was performed: (∗p < 0.005 ra-

pamycin vs. vehicle alone). Six mice per

group were used in this experiment.

FACS analysisUnstimulated or stimulated hIECs were detached from tissue culture plates

with EDTA and stained for 45 min at 4◦C with appropriate antibodies. Cells

were then fixed in 1% paraformaldehyde and subjected to FACS analysis

(Becton Dickinson, Mountain View, CA).

Statistical analysisAll data of different experimental procedures are expressed as av-

erage ± SD. Statistical analysis was performed by Student’s t-test

or ANOVA with Newmann-Keuls multicomparison test where

appropriate.

American Journal of Transplantation 2006; 6: 2601–2611 2605

Cantaluppi et al.

Figure 6: Inhibition of in vivo hIEC angiogenesis induced byrapamycin. (A–D) Matrigel plugs implanted for 2 weeks into the

neck scruff region of SCID mice and containing hIECs-GFP in the

presence of vehicle alone (A and C) or 10 ng/mL rapamycin (B

and D). Histologic analysis showed the abundant presence of ne-

oformed capillaries in control Matrigel plugs (A) in comparison to

their paucity in rapamycin-containing implants (B). By fluorescence

analysis patent vessels were formed by GFP-positive cells in con-

trol Matrigel plugs (C), whereas in the presence of rapamycin

mainly scattered hIECs-GFP were detectable (D). (E and F) Ra-

pamycin induced a significant inhibition of in vivo hIECs angiogen-

esis evaluated by counts of number (E) and total area (F, expressed

in lm2/section) of neoformed vessels in tissue sections. Student’s

t-test was performed: (∗p < 0.005 rapamycin vs. vehicle alone).

Six mice per group were used in this experiment.

Figure 7: Antiproliferative effect of rapamycin on hIECs andmIECs. (A and B) Time-course analysis of spontaneous and growth

factor (EndoGF)-induced inhibition of proliferation exerted by ra-

pamycin on hIECs (A) and mIECs (B), evaluated by XTT-based as-

say. A significant decrease of proliferation of both endothelial cell

lines after 24 and 48 h of stimulation was observed (p < 0.005

rapamycin or EndoGF medium vs. vehicle alone and EndoGF +rapamycin vs. rapamycin). (C and D) Dose-dependent growth ar-

rest induced by rapamycin challenge on hIECs (C) and mIECs (D)

incubated with EndoGF. A significant decrease of proliferation of

endothelial cell lines after 48 h of stimulation starting from a dose

of rapamycin of 0.1 ng/mL for hIECs and 1 ng/mL for mIECs. This

effect reached a peak with a rapamycin concentration of 50 ng/mL

in both cell lines: (∗p < 0.005 different doses of rapamycin vs. ve-

hicle alone). Data are expressed as average OD intensity of three

different experimental points ± SD. ANOVA with Newmann-Keuls

multicomparison test was performed. Three independent experi-

ments were performed with similar results.

2606 American Journal of Transplantation 2006; 6: 2601–2611

Rapamycin and Islet Endothelium

Figure 8: Rapamycin increased hIECs and mIECs celldeath induced by serum deprivation and impairedgrowth factor-mediated rescue from apoptosis. (A and

B) Evaluation of proapoptotic activity by TUNEL assay in-

duced by rapamycin on hIECs (A) and mIECs (B) after

48 h incubation. Rapamycin significantly increased the

number of apoptotic hIECs and mIECs in culture with or

Results

Effects of rapamycin on angiogenesis in freshlypurified islet endothelium and inhIECs and mIECsendothelial cell linesEndothelial cells outgrew from freshly purified human

islets cultured in the presence of endothelial growth

factors-rich medium (EndoGF medium) (Figure 1A). As

shown in Figure 1B, the addition of rapamycin blunted

this phenomenon. The antiangiogenic effect of rapamycin

on freshly purified islet endothelium was confirmed by

the marked decrease of capillary-like structures in vitro

when islets were seeded on Matrigel-coated surfaces

(Figure 1C,D). Cells outgrowing from islets were charac-

terized as endothelial cells by staining with typical en-

dothelial markers such as VEGFR2 (KDR, Figure 1E), vWF

(Figure 1F), CD31 and tie-2 and by the uptake of acetylated-

LDL (not shown). Notably, the dose of rapamycin used to

obtain the inhibition of endothelial outgrowth (10 ng/mL)

did not alter islet viability (comparable PI/FDA staining

>90% in each sample) and function, evaluated as insulin

response after high glucose challenge (stimulation indices

3.42 ± 0.85 for vehicle alone and 3.59 ± 0.97 for rapamycin-

treated islets, respectively) after 5 days of culture.

To further investigate the inhibition of endothelial out-

growing from islets, we studied the effects of rapamycin

on hIECs and mIECs motility by time-lapse recording mi-

croscopy. The baseline migration rate of hIECs and mIECs

corresponding to the spontaneous motility of resting, un-

stimulated cells was first measured and found to remain

steady for the whole period of observation never ex-

ceeding 5–6 lm/h. Rapamycin induced a significant de-

crease of spontaneous migration of both hIECs and mIECs

(Figure 2A,B). As expected, the incubation with EndoGF

medium induced a marked acceleration of spontaneous

cell motility, peaking as early as 1 h after stimulation and

remaining significantly higher compared to unstimulated

cells throughout all the observation period (Figure 2A,B).

The addition of rapamycin to EndoGF medium significantly

without coincubation with Endo GF medium. Data are expressed

as average number of apoptotic cells/field ± SD (magnification

100×). ANOVA with Newmann-Keuls multicomparison test was

performed: (∗p < 0.005 rapamycin or EndoGF medium vs. ve-

hicle alone); (#p < 0.005 Endo GF + rapamycin vs. rapamycin).

(C and D) Dose-dependent proapoptotic effect of rapamycin on

hIECs (C) and mIECs (D) after 48 h incubation in the presence

of EndoGF medium. Data are expressed as average number of

apoptotic cells/field ± SD. ANOVA with Newmann-Keuls multi-

comparison test was performed, showing a significant increase

of apoptosis of both endothelial cell lines after 48 h of stimula-

tion with a dose of rapamycin of 0.1 ng/mL. This effect reached a

peak with a rapamycin concentration of 50 ng/mL: (∗p < 0.005 dif-

ferent doses of rapamycin vs. vehicle alone). Three independent

experiments were performed with similar results.

American Journal of Transplantation 2006; 6: 2601–2611 2607

Cantaluppi et al.

reduced the migration of both endothelial cell lines

(Figure 2A,B) in a dose-dependent manner (Figure 3A,B).

We then studied the in vitro angiogenic ability of hIECs that

form capillary-like structures when seeded on Matrigel-

coated plates. The addition of EndoGF medium acceler-

ated the angiogenic process, resulting in enhanced forma-

tion of a capillary network (Figure 4A,D) in comparison to

vehicle alone (Figure 4A,B). Rapamycin significantly inhib-

ited spontaneous (Figure 4A,C) and growth factor-induced

(Figure 4A,E) formation of capillary-like structures.

The inhibition induced by rapamycin on islet angiogenesis

was also evaluated in vivo by s.c. injection of freshly pu-

rified islets within Matrigel plugs into the scruff region of

the neck of SCID mice. Matrigel s.c. implant in mice is

a well-established model of angiogenesis (19) and it has

been recently demonstrated as effective in transplanting

islets (20). In the presence of rapamycin, implants showed

a marked reduction of vascularization as detected by im-

munofluorescence staining for the endothelial marker tie-2

(Figure 5A,B) and by counts of number and total area of ne-

oformed vessels within Matrigel (Figure 5C,D). In addition,

rapamycin also induced a marked down-regulation of the

vascular density within the islets (Figure 5E–H).

We then performed xenografts of hIECs-GFP and isografts

of mIECs-GFP by injection into the scruff region of the neck

of SCID or C57Bl/6 mice, respectively. Consistently with

the in vitro angiogenesis studies, in the presence of ra-

pamycin hIECs showed a marked inhibition of their abil-

ity to proliferate and to form neovessels as detected by

histologic (Figure 6A,B) and immunofluorescence studies

(Figure 6C,D). Moreover, rapamycin induced a significant

reduction of number and total area of vessels (Figure 6E,F).

Immunofluorescence analysis also showed a marked al-

teration of normal structure of mIECs-GFP isografts (not

shown).

We then studied the effect of rapamycin on islet endothe-

lium growth. HIECs and mIECs were starved overnight

without FCS and subsequently incubated with normal or

EndoGF medium containing increasing doses of rapamycin

(0.1–100 ng/mL). As expected, endothelial growth factors

induced an increase of proliferation of endothelial cells in

comparison to vehicle alone. Coincubation with rapamycin

resulted in a significant inhibitory effect on spontaneous

and growth factor-stimulated proliferation of both hIECs

and mIECs (Figure 7A,B). This effect was detectable at

doses of 0.1–1 ng/mL and reached a plateau at the dose

of 50 ng/mL for both cell lines (Figure 7C,D).

In the transplantation setting, the efficiency of islet revas-

cularization results from the net balance between opposing

factors that may affect endothelial cell survival. For this rea-

son, we evaluated the effect of rapamycin on endothelial

growth factor-dependent rescue from apoptosis induced

by serum deprivation. As shown in Figure 8, incubation

Figure 9: Modulation of angiogenic factors induced by ra-pamycin on hIECs. Immunofluorescence analysis of hIECs cul-

tured in the absence (control) or presence of 10 ng/mL rapamycin.

Rapamycin induced a reduction of VEGF and aVb3 integrin on

hIECs. In addition, an overexpression of TSP-1, an angiogenesis

inhibitor, was detected on rapamycin-treated hIECs (100× magni-

fication, antibody staining in green and nuclear counterstaining in

red with 1 lg/mL propidium iodide).

with rapamycin worsened hIECs (Figure 8A) and mIECs

(Figure 8B) apoptosis induced by serum deprivation while

EndoGF medium significantly reduced cell death. In addi-

tion, rapamycin significantly inhibited the antiapoptotic ef-

fect of EndoGF-medium on hIECs (Figure 8A,C) and mIECs

(Figure 8B,D) in a dose-dependent manner.

We then investigated the modulation of expression of se-

lected molecules involved in angiogenesis after rapamycin

challenge on cultured hIECs by immunofluorescence stud-

ies. Rapamycin-treated hIECs showed a decreased expres-

sion of VEGF and aVb3 integrin and a marked increase of

TSP-1 (Figure 9). Consistently with these results, we also

observed a down-regulation of VEGF, PDGF, betaFGF, HGF,

aVb3 integrin genes, and up-regulation of the TSP-1 gene

in human islets exposed to rapamycin (not shown).

Effects of rapamycin on lymphocyte-islet endothelialinteractionRapamycin has been shown to prevent up-regulation of

immune cell surface molecules involved in cell adhesion

and cross-talk during immune cell activation. To determine

the role of rapamycin in intraislet lymphocyte recruitment,

we conducted experiments aimed at evaluating the adhe-

sion of Jurkat cells, a T-cell line, to hIECs monolayers. The

2608 American Journal of Transplantation 2006; 6: 2601–2611

Rapamycin and Islet Endothelium

addition of rapamycin inhibited the adhesion of Jurkat cells

induced by 10 ng/mL TNF-alpha and 10 ng/mL IFN-gamma

(Figure 10A). On cytokine-stimulated hIECs, the addition

of rapamycin prevented the overexpression of the lympho-

cyte costimulatory molecule CD40 and of the adhesion re-

ceptor CD54 (ICAM-1) (Figure 10B).

Discussion

A prominent role of islet vasculature conceivably takes

place in the setting of islet transplantation. Recently,

it was demonstrated that freshly purified islets retain

viable native endothelial cells that may actively participate

to the revascularization process by forming chimeric

donor/recipient-derived vessels (3). This event may have

direct consequences in facilitating islet engraftment

and function (3) and may promote strategies aimed at

preserving islet endothelium before transplantation by

minimizing the pretransplant culture time (21) or by

using endothelial growth factors such as VEGF (5,22).

Indeed, several factors may positively or negatively

influence islet endothelium viability and function in the

setting of islet purification, culture and transplantation.

In the past, immunosuppressive drug combinations have

been indicated as a limiting factor for transplanted islets

function due to the diabetogenic effect of steroids and

to the direct toxic effect of calcineurin inhibitors (23).

Clinical islet transplantation has recently received a strong

impulse from the results obtained by the introduction of a

rapamycin-based glucocorticoid-free immunosuppressive

regimen leading to insulin independence at 1 year in 90%

of the treated patients (1,24). However, an inhibitory effect

of rapamycin on tumor angiogenesis has been recently

described (10,25). In light of the evidences on the impor-

tance of neovascularization of the transplanted islets, we

evaluated whether rapamycin may have a direct impact

on the angiogenic ability of intraislet endothelial cells.

We observed that rapamycin markedly inhibited the out-

growth of endothelial cells from freshly purified islets in

vitro and in vivo after s.c. injection into Matrigel plugs in

mice. These results were paralleled by the inhibition of hu-

man and murine islet endothelial cell line migration, prolif-

eration and angiogenesis induced by this immunosuppres-

sant.

The antiangiogenic effect of rapamycin that we observed

both on purified islets and on islet endothelial cell lines was

exerted at rapamycin concentrations that correspond to the

through blood rapamycin levels currently recommended

for islet transplant recipients (1). Interestingly, the antian-

giogenic effect was detectable at rapamycin doses that did

not impair beta-cells viability or function in vitro. Moreover,

it is conceivable that islet endothelium after transplantation

may be exposed to even higher doses of rapamycin due to

the peak rapamycin concentration in the portal circulation

obtained after gastrointestinal absorption of this drug (26).

Figure 10: Effects of rapamycin on T-cell adhesion on hIECsand on modulation of vascular surface receptors involvedin lymphocyte-endothelium interaction. (A) In vitro adhesion

assay of green fluorescent Jurkat T-cell line to hIEC monolay-

ers. Labeled Jurkat cells were added to hIECs confluent mono-

layers incubated with vehicle alone or 10 ng/mL TNF-alpha plus

10 ng/mL IFN-gamma (cytokines) in the presence or absence of

10 ng/mL rapamycin. After extensive washing, all samples were

fixed and analyzed under a UV light microscope. Fluorescent cells

in 10 fields/well were counted at 200× magnification and data

are representative of average number of adherent cells/field ±SD. Three independent experiments were performed with similar

results. ANOVA with Newmann-Keuls multicomparison test was

performed: (∗p < 0.005 rapamycin or cytokines vs. vehicle alone);

(#p < 0.005 cytokines + rapamycin vs. cytokines). (B) FACS anal-

ysis showed a slight basal expression of CD40 and ICAM-1 (bold

lines) on hIECs in culture. Exposure of hIECs to cytokines resulted

in a marked increase of both CD40 and ICAM-1 expression that

was prevented by coincubation with 10 ng/mL rapamycin. Isotype-

matched antibody controls are shown as dotted lines.

American Journal of Transplantation 2006; 6: 2601–2611 2609

Cantaluppi et al.

Furthermore, we analyzed the potential effect of rapamycin

on apoptosis of islet endothelial cells. At supratherapeu-

tic concentrations rapamycin was shown to have delete-

rious effects on the viability of rat and human islets (27).

Other cell types such as monocytes (28), macrophages

(28), BxPC3 and Panc-1 human pancreatic adenocarcinoma

cell lines (29) were not susceptible to apoptosis after ra-

pamycin challenge. In our study, rapamycin prevented the

rescue effect of growth factors when hIECs and mIECs

were exposed to detrimental culture conditions such as

prolonged serum deprivation. This phenomenon may have

relevance right after islet implant in the sinusoid vessels

where islets are exposed to multiple mutual-conflicting sig-

nals that ultimately affect their survival and functional en-

graftment or their death.

The antiangiogenic effect of rapamycin has been par-

tially elucidated and it is thought to depend on the in-

hibitory effect on VEGF-mediated activation of Akt, a ser-

ine/threonine protein kinase directly up-stream of mTOR

in the rapamycin-sensitive signaling pathway and on de-

creased production of VEGF (10). It has been recently

shown that rapamycin treatment inhibits tumor progres-

sion in renal transplant patients bearing Kaposi’s sarcoma,

whose lesions are characterized by abundant expression

of VEGF and increased phosphorylation of Akt (25).

We performed experiments in the presence of endothe-

lial growth factors that were shown to be expressed after

islet implants within sinusoids (30–32), thus mimicking the

proangiogenic milieau where islet neovascularization takes

place. We showed that rapamycin challenge strongly inhib-

ited VEGF production by islet endothelium and induced an

altered expression of other molecules involved in tumor

and normal tissue angiogenesis such as aVb3 integrin and

TSP-1. Overexpression of aVb3 integrin on tumor vascula-

ture has been associated with an aggressive phenotype of

several solid tumor types (33) and we recently showed the

up-regulation of this molecule on islet endothelium induced

by platelet-activating factor (13). TSP-1 is a potent angio-

genesis inhibitor that triggers apoptosis in activated en-

dothelial cells (32,34). Interestingly, TSP-1−/− mice showed

islet hyperplasia due to an increased density of blood ves-

sels, suggesting a key role for this protein in vascularization

of native and transplanted islets (35).

Taken together, our findings showed that rapamycin affects

islet vascularization interfering with the production of dif-

ferent pro- and antiangiogenic factors within islet endothe-

lium. Furthermore, the reduced release of these molecules

by endothelial cells may affect beta cells surviving due to

the deprivation of growth factors (36,37).

On the other hand, we provided data on the effect of

rapamycin in modulating cell-to-cell receptor interaction

among islet endothelial cells and lymphocytes that may

potentially occur during allograft rejection. In addition, ra-

pamycin also reduced the expression of CD40 and ICAM-1

in hIECs stimulated with TNF-alpha and IFN-gamma. In

particular, endothelial CD40, a critical costimulatory recep-

tor for lymphocyte activation, has been identified as a key

molecule for the interaction with immune system in vascu-

lar processes involved in transplant rejection (38,39). More-

over, treatment with anti-CD40 ligand antibodies has been

proven to be beneficial for the prevention of allograft re-

jection in nonhuman primate models of islet transplanta-

tion (40). ICAM-1 is a major surface molecule involved in

leukocytes adhesion to endothelial cells and subsequent

recruitment in sites of immune injury (41). These effects

induced by rapamycin were previously observed also in

antigen presenting cells such as dendritic cells (42).

Based on our results, a potential limitation of the current

immunosuppressive protocols may be represented by the

antiangiogenic activity of rapamycin, particularly detrimen-

tal in the early engraftment phase. However, its vascu-

lar immunomodulatory properties may be relevant in the

following phases. In conclusion, new therapeutic sequen-

tial protocols that consider the whole activity profile of ra-

pamycin may be highly advantageous to the field of islet

transplantation.

Acknowledgment

This work was supported by Italian Ministry of University and Research

(MIUR) FIRB project (RBNE01HRS5-001) to G.C. and COFIN 01 to L.B.,

G.C. and P.C.P., by Italian Ministry of Health (Ricerca Finalizzata 02) to G.C.

and G.P.S. and by ‘Ricerca Finalizzata—Regione Piemonte’ to L.B.

References

1. Shapiro AM, Lakey JR, Ryan EA et al. Islet transplantation in seven

patients with type 1 diabetes mellitus using a glucocorticoid-free

immunosuppressive regimen. N Engl J Med 2000; 343: 230–238.

2. Carlsson PO, Mattsson G. Oxygen tension and blood flow in rela-

tion to revascularization in transplanted adult and fetal rat pancre-

atic islets. Cell Transplant 2002; 11: 813–820.

3. Nyqvist D, Kohler M, Wahlstedt H, Berggren PO. Donor islet en-

dothelial cells participate in formation of functional vessels within

pancreatic islet grafts. Diabetes 2005; 54: 2287–2293.

4. Biancone L, Ricordi C. Pancreatic islet transplantation: An update.

Cell Transplant 2002; 11: 309–311.

5. Lai Y, Schneider D, Kidszun A et al. Vascular endothelial growth

factor increases functional beta-cell mass by improvement of an-

giogenesis of isolated human and murine pancreatic islets. Trans-

plantation 2005; 79: 1530–1536.

6. Kirken RA, Wang YL. Molecular actions of sirolimus: Sirolimus and

mTor. Transplant Proc 2003; 35(3 Suppl): 227S–230S.

7. Hering BJ, Wijkstrom M. Sirolimus and islet transplants. Trans-

plant Proc 2003; 35(3 Suppl): 187S–190S.

8. Valente JF, Hricik D, Weigel K et al. Comparison of sirolimus vs.

mycophenolate mofetil on surgical complications and wound heal-

ing in adult kidney transplantation. Am J Transplant 2003; 3: 1128–

1134.

9. Lieberthal W, Fuhro R, Andry CC et al. Rapamycin impairs recovery

from acute renal failure: Role of cell-cycle arrest and apoptosis of

tubular cells. Am J Physiol Renal Physiol 2001; 281: F693–F706.

2610 American Journal of Transplantation 2006; 6: 2601–2611

Rapamycin and Islet Endothelium

10. Guba M, von Breitenbuch P, Steinbauer M et al. Rapamycin in-

hibits primary and metastatic tumor growth by antiangiogenesis:

Involvement of vascular endothelial growth factor. Nat Med 2002;

8: 128–135.

11. Butzal M, Loges S, Schweizer M et al. Rapamycin inhibits prolif-

eration and differentiation of human endothelial progenitor cells

in vitro. Exp Cell Res 2004; 300: 65–71.

12. Ricordi C, Lacy PE, Scharp DW. Automated islet isolation from

human pancreas. Diabetes 1989; 38(Suppl 1): 140–142.

13. Biancone L, Cantaluppi V, Romanazzi GM et al. Platelet-activating

factor synthesis and response on pancreatic islet endothelial cells:

Relevance for islet transplantation. Transplantation 2006; 81: 511–

518.

14. Biancone L, Cantaluppi V, Boccellino M et al. Motility induced by

human immunodeficiency virus-1 Tat on Kaposi’s sarcoma cells

requires platelet-activating factor synthesis. Am J Pathol 1999;

155: 1731–1739.

15. Dull T, Zufferey R, Kelly M et al. A third-generation lentivirus vec-

tor with a conditional packaging system. J Virol 1998; 72: 8463–

8471.

16. Galimi F, Noll M, Kanazawa Y et al. Gene therapy of Fanconi ane-

mia: Preclinical efficacy using lentiviral vectors. Blood 2002; 100:

2732–2736.

17. Montrucchio G, Lupia E, Battaglia E et al. Tumor necrosis factor

alpha-induced angiogenesis depends on in situ platelet-activating

factor biosynthesis. J Exp Med 1994; 180: 377–382.

18. Biancone L, Cantaluppi V, Duo D, Deregibus MC, Torre C, Camussi

G. Role of L-selectin in the vascular homing of peripheral blood-

derived endothelial progenitor cells. J Immunol 2004; 173: 5268–

5274.

19. Jain RK, Schlenger K, Hockel M, Yuan F. Quantitative angiogene-

sis assays: Progress and problems. Nat Med 1997; 3: 1203–1208.

20. Bharat A, Benshoff N, Olack B, Ramachandran S, Desai NM, Mo-

hanakumar T. Novel in vivo murine model to study islet potency:

Engraftment and function. Transplantation 2005; 79: 1627–1630.

21. Olsson R, Carlsson PO. Better vascular engraftment and function

in pancreatic islets transplanted without prior culture. Diabetologia

2005; 48: 469–476.

22. Zhang N, Richter A, Suriawinata J et al. Elevated vascular en-

dothelial growth factor production in islets improves islet graft

vascularization. Diabetes 2004; 53: 963–970.

23. Pileggi A, Ricordi C, Alessiani M, Inverardi L. Factors influencing

Islet of Langerhans graft function and monitoring. Clin Chim Acta

2001; 310: 3–16.

24. Ryan EA, Paty BW, Senior PA et al. Five-year follow-up after clinical

islet transplantation. Diabetes 2005; 54: 2060–2069.

25. Stallone G, Schena A, Infante B et al. Sirolimus for Kaposi’s sar-

coma in renal-transplant recipients. N Engl J Med 2005; 352:

1317–1323.

26. Desai NM, Goss JA, Deng S et al. Elevated portal vein drug lev-

els of sirolimus and tacrolimus in islet transplant recipients: Local

immunosuppression or islet toxicity? Transplantation 2003; 76:

1623–1625.

27. Bell E, Cao X, Moibi JA et al. Rapamycin has a deleterious effect

on MIN-6 cells and rat and human islets. Diabetes 2003; 52: 2731–

2739.

28. Woltman AM, de Fijter JW, Kamerling SW et al. Rapamycin in-

duces apoptosis in monocyte- and CD34-derived dendritic cells

but not in monocytes and macrophages. Blood 2001; 98: 174–

180.

29. Shah SA, Potter MW, Ricciardi R, Perugini RA, Callery MP. FRAP-

p70s6K signaling is required for pancreatic cancer cell prolifera-

tion. J Surg Res 2001; 97: 123–130.

30. Vasir B, Jonas JC, Steil GM et al. Gene expression of VEGF and

its receptors Flk-1/KDR and Flt-1 in cultured and transplanted rat

islets. Transplantation 2001; 71: 924–935.

31. Linn T, Schneider K, Hammes HP et al. Angiogenic capacity of

endothelial cells in islets of Langerhans. FASEB J 2003; 17: 881–

883.

32. Welsh M, Claesson-Welsh L, Hallberg A et al. Coexpression of the

platelet-derived growth factor (PDGF) B chain and the PDGF beta

receptor in isolated pancreatic islet cells stimulates DNA synthe-

sis. Proc Natl Acad Sci U S A 1990; 87: 5807–5811.

33. Jin H, Varner J. Integrins: Roles in cancer development and as

treatment targets. Br J Cancer 2004; 90: 561–565.

34. Jimenez B, Volpert OV, Crawford SE, Febbraio M, Silverstein RL,

Bouck N. Signals leading to apoptosis-dependent inhibition of neo-

vascularization by thrombospondin-1. Nat Med 2000; 6: 41–48.

35. Johansson M, Mattsson G, Andersson A, Jansson L, Carlsson

PO. Islet endothelial cells and pancreatic beta-cell proliferation:

Studies in vitro and during pregnancy in adult rats. Endocrinology

2006; 147: 2315–2324.

36. Cheng K, Fraga D, Zhang C et al. Adenovirus-based vascular en-

dothelial growth factor gene delivery to human pancreatic islets.

Gene Ther 2004; 11: 1105–1116.

37. Garcia-Ocana A, Vasavada RC, Cebrian A et al. Transgenic over-

expression of hepatocyte growth factor in the beta-cell markedly

improves islet function and islet transplant outcomes in mice. Di-

abetes 2001; 50: 2752–2762.

38. Kirk AD, Burkly LC, Batty DS et al. Treatment with humanized

monoclonal antibody against CD154 prevents acute renal allograft

rejection in nonhuman primates. Nat Med 1999; 5: 686–693.

39. Biancone L, Cantaluppi V, Camussi G. CD40-CD154 interaction in

experimental and human disease (review). Int J Mol Med 1999;

3: 343–353.

40. Koulmanda M, Smith RN, Qipo A, Weir G, Auchincloss H, Strom

TB. Prolonged survival of allogeneic islets in cynomolgus mon-

keys after short-term anti-CD154-based therapy: Nonimmuno-

logic graft failure? Am J Transplant 2006; 6: 687–696.

41. Park SY, Kim HW, Moon KC, Hong HK, Lee HS. mRNA expression

of intercellular adhesion molecule-1 and vascular cell adhesion

molecule-1 in acute renal allograft rejection. Transplantation 2000;

69: 2554–2560.

42. Monti P, Mercalli A, Leone BE, Valerio DC, Allavena P, Piemonti

L. Rapamycin impairs antigen uptake of human dendritic cells.

Transplantation 2003; 75: 137–145.

American Journal of Transplantation 2006; 6: 2601–2611 2611