role for 1 integrin and its associated 3, 5, and 6 ...€¦ · ptosis via the integrin 1/focal...

Role for �1 Integrin and Its Associated �3, �5, and �6Subunits in Development of the Human Fetal PancreasRennian Wang,

1,2Jinming Li,

1Kristina Lyte,

1Nina K. Yashpal,

1Fraser Fellows,

3and

Cynthia G. Goodyer4

The integrin receptors play a major role in tissue mor-phogenesis and homeostasis by regulating cell interac-tions with extracellular matrix proteins. We haveexamined the expression pattern of integrin subunits inthe human fetal pancreas (8–20 weeks fetal age) andthe relevance of �1 integrin function for insulin geneexpression and islet cell survival. Its subunits �3, �5,and �6 �1 integrins are expressed in ductal cells at 8weeks, before glucagon- and insulin-immunoreactivecells bud off; their levels gradually increase in bothductal cells and islet clusters up to 20 weeks. Colocal-ization of �3, �5 and �6 �1 integrins with endocrine cellmarkers was frequently observed in 8- to 20-week fetalpancreatic cells. When the �1 integrin receptor wasfunctionally blocked in cultured islet-epithelial clusterswith a �1 immunoneutralizing antibody or followingtransient �1 integrin small interfering RNA treatment,there was inhibition of cell adhesion to extracellularmatrices, decreased expression of insulin, and in-creased cell apoptosis. These data offer evidence fordynamic and cell-specific changes in integrin expressionduring human pancreatic islet neogenesis. They alsoprovide an initial insight into a molecular basis forcell-matrix interactions during islet development andsuggest that �1 integrin plays a vital role in regulatingislet cell adhesion, gene expression, and survival.Diabetes 54:2080–2089, 2005

Islets of Langerhans contain a remarkable cellularorganization that is ideal for rapid, yet preciselycontrolled, responses to changes in blood glucoselevels. Any permanent disturbance of this regulatory

system leads to diabetes, one of the most common meta-bolic diseases affecting millions of people throughout theworld. Determining the factors that control islet cell

development and maintain survival and function is essen-tial to help develop viable strategies for any cell-basedapproach toward the repopulation of islets for the treat-ment of diabetes. Therefore, recent efforts have concen-trated on exploring the molecular signals that controlmorphogenesis in the normal human pancreas. One impor-tant research focus has been the integrin receptors, afamily of cell adhesion molecules that mediate cell-celland cell-matrix interactions. They have been shown toregulate the proliferation, maturation, and function ofrodent islets in vitro (1–2). However, the role of integrin-mediated interactions with the extracellular matrix (ECM)on the formation and function of the islets of Langerhansbefore birth, especially in the human, is poorly understood.

Integrins are a large family of heterodimeric transmem-brane adhesion molecules composed of noncovalentlybound �- and �-subunits that possess the unique ability toregulate cell adhesiveness through a process called “in-side-out signaling.” In addition, after binding to theirligands at the cell surface, these receptors integrate thecues from their external environment to the cell by gen-erating specific intracellular signals, in a process termed“outside-in signaling” (3). This results in modifications ofcell structure and functions such as cell adhesion, motility,cell proliferation, differentiation, and gene transcription(3–5).

The �1 integrin family is believed to play a critical rolein morphogenesis (6–7), cell differentiation, and prolifer-ation (8–9) as well as cell survival (10) by binding selec-tively to collagen, fibronectin, and laminin extracellularmatrices (11). The importance of this receptor is evidentfrom the embryonic lethality that ensues in homozygous�1-deficient embryos (7).

Multiple functions of ��1 integrins in a number of organsystems have been described previously; however, re-search on their expression and interactions during pancre-atic development is limited. Thus far, studies have shownthat only a few members of the integrin family affect isletcell survival, maturation, and insulin production (2,12). Inparticular, �3, �5, and �6 integrins have been reported tomediate certain pancreatic developmental events: 1) �3�1mediates the attachment and spreading of primary rat isletcells to ECMs (12) and regulates the migration of CK19�/PDX-1� putative pancreatic progenitors of human fetalpancreatic epithelial cells on netrin-1 (13); 2) �5 expres-sion has been shown to decrease during culture of ratislets, which parallels increased islet apoptosis, implicat-ing this particular integrin in controlling signaling eventsthat protect against cell death (14); 3) Crisera et al. (15)

From the 1Department of Physiology and Pharmacology, University of West-ern Ontario, London, Ontario, Canada; the 2Department of Medicine, Univer-sity of Western Ontario, London, Ontario, Canada; the 3Department ofObstetrics and Gynecology, University of Western Ontario, London, Ontario,Canada; and the 4Department of Pediatrics, McGill University, Montreal,Quebec, Canada.

Address correspondence and reprint requests to Dr. Rennian Wang, VictoriaLaboratory Centre, Room A5-140, 800 Commissioners Rd. E, London, Ontario,N6C 2V5, Canada. E-mail: [email protected].

Received for publication 28 November 2004 and accepted in revised form 4April 2005.

ECM, extracellular matrix; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphe-nyltetrazolium bromide; siRNA, small interfering RNA; TRITC, tetramethyl-rhodamine isothiocyanate; TUNEL, transferase-mediated dUTP nick-endlabeling.

© 2005 by the American Diabetes Association.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked “advertisement” in accordance

with 18 U.S.C. Section 1734 solely to indicate this fact.

2080 DIABETES, VOL. 54, JULY 2005

reported that mouse pancreatic ductal morphogenesisrequires the ECM laminin-1 during embryonic life and isinhibited by the blockade of �6�1 integrin or laminin; 4)

�6�1 is believed to enhance and regulate the insulinsecretory response of rat islets (2); and 5) �1 integrin maybe involved in early motile processes required for the

FIG. 1. Western blot analyses of human fetal pancreatic tissue using anti-�1 (A), anti-�3 (B), anti-�5 (C), and anti-�6 (D) integrin antibodies.Data are expressed as means � SE (n � 3 per age-group) relative to the 8-week group. Representative blots are shown. E: Real-time RT-PCRanalysis of �1 integrin mRNA expression in human fetal pancreatic tissue. Data are normalized to 18S RNA subunit and expressed as means �SE (n � 3 per age-group). *P < 0.05 vs. 8-week group.

R. WANG AND ASSOCIATES

DIABETES, VOL. 54, JULY 2005 2081

formation of new islets by supporting migration of humanfetal �-cells (16). More recently, Hammer et al. (17)determined that 804G matrix protects �-cells against apo-ptosis via the integrin �1/focal adhesion kinase pathwayand that blocking �1 integrin function induces cell death.Thus, these studies suggest a role for �3, �5, �6, and �1integrin receptors in early pancreatic developmentalevents in multiple species.

Based on the above findings, the goal of the currentstudy was to examine the expression pattern of integrinsubunits in situ during islet growth in the human fetalpancreas from 8–20 weeks of fetal age using immunoflu-orescence, Western blot, and real-time RT-PCR. We alsoexamined the role of �1 integrin in cultured islet-epithelialclusters, in mediating islet cell adhesion to extracellularmatrices, insulin gene expression, and cell death, usingimmunoneutralizing antibodies and small interfering RNAs(siRNAs). Here we report that human fetal ductal and isletcells express �1 integrin and its associated �3, �5, and �6subunits during early pancreatic development. Our dataalso provide evidence for a major role for the �1 integrinreceptor in mediating adhesion, insulin gene expression,

and survival of human fetal islet-epithelial clusters. Fur-thermore, this study provides a molecular connectionbetween cultured islets and the ECM that can be manipu-lated and is thus highly useful information for futureinvestigations that seek to improve islet cell–based thera-pies for the treatment of diabetes.

RESEARCH DESIGN AND METHODS

Tissues. Human fetal pancreata (8–20 weeks fetal age) were collectedaccording to protocols approved by the Health Sciences Research EthicsBoard at the University of Western Ontario and the Research Ethics Board ofthe Royal Victoria Hospital at the McGill University Health Centre, inaccordance with the Canadian Council on Health Sciences Research InvolvingHuman Subjects guidelines. Tissues were immediately processed for immu-nohistochemistry, RNA, and/or protein extraction, with at least three pancre-ata per age or experimental group.Immunofluorescence. Pancreata were fixed in 4% paraformaldehyde over-night at 4°C followed by a standard protocol of dehydration and paraffinembedding (18). Sections of 5 �m were cut throughout the length of thepancreas with two sets of six serial sections at 50-�m intervals. The tissuesections were incubated overnight at 4°C with appropriate dilutions of thefollowing primary antibodies: rabbit anti-�3 and anti-�5 integrins (cytoplasmicdomains), mouse anti-�1 and anti-�6�1 integrins (Chemicon, Temecula, CA),rabbit anti-�6 integrin (Santa Cruz Biotechnology, Santa Cruz, CA), mouseanticytokeratin 19 (CK19; Dako, Mississauga, ON, Canada), guinea pig anti-human insulin (Zymed, San Francisco, CA), mouse anti-human glucagon(Sigma, St. Louis, MO), rabbit anti-PDX-1 (gift from Dr. Wright, University ofVanderbilt, Nashville, TN), and antibodies to laminin, fibronectin, and collagenIV (Chemicon) as described previously (19). To identify colocalization ofintegrins with epithelial and endocrine cell markers, double immunofluores-cence staining was performed. Fluorescent secondary antibodies were ob-tained from Jackson Immunoresearch Laboratories (West Grove, PA). Imageswere recorded by a Leica DMIRE2 fluorescence microscope with the Openlabimage software (Improvision, Lexington, MA). Negative controls included theomission of the primary antibodies.Morphometric analysis. Both single- and double-labeled images were re-corded under a high magnification (400�). Endocrine and ductal regions weredefined through staining of consecutive sections with a cocktail of antibodiesfor pancreatic hormones and an antibody for the ductal cell, as previouslydescribed (18–19). The integrin immunoreactive area within the ductal andendocrine cell compartments was traced manually. In each pancreatic section,8–12 random fields were chosen with a minimum of three pancreata per ageor experimental group, and data are expressed as the percentage of integrinimmunoreactivity in both endocrine and duct regions. To determine thepercentage of integrin colocalization with insulin or glucagon, the double-labeled cells are expressed as a percentage of the total number of insulin- orglucagon-positive cells.Western blots. Pancreatic tissues were homogenized in a Nonidet-P40 lysisbuffer (Nonidet-P40, phenylmethylsulfonyl fluoride, sodium orthovanadate[Sigma] and complete protease inhibitor cocktail tablet [Roche, Montreal, QC,Canada]) and centrifuged at 12,000 rpm for 20 min. The supernatant wasrecovered and frozen at �80°C. The protein concentration was measured byBradford protein dye (Bio-Rad, Mississauga, ON, Canada), using bovine serumalbumin (fraction V) as standard. As described previously, 25 �g of pancreaticlysate proteins were separated by 7.5% SDS-PAGE and transferred to anitrocellulose membrane (Bio-Rad) (19–20). The membranes were washed inTris buffer–saline containing 0.1% Tween 20 and blocked with 5% nonfat drymilk overnight at 4°C. Immunoblotting was performed with the integrinantibodies at the concentrations recommended by the manufacturer for 1 h atroom temperature. Secondary antibody was anti-rabbit IgG conjugated tohorseradish peroxidase (Santa Cruz) diluted at 1:1,000. Proteins were de-tected by enhanced chemiluminescence reagents (Amersham, Oakville, ON,Canada) and exposed to BioMax MR Film (Kodak, Rochester, NY). Densito-metric quantification of bands at subsaturating levels was performed using theSyngenetool gel analysis software (Syngene, Cambridge, U.K.) and normalizedto band intensities at 8 weeks of fetal age (19,21). Loading controls of calnexin(BD Biosciences) and �-actin (Sigma) were tested; however, their variabilityduring development precluded their use (19). For negative controls, theprimary antibody was omitted.RT-PCR and real-time RT-PCR. Total RNA was extracted from pancreastissues with TRIZOL reagent (Invitrogen, Burlington, ON, Canada), accordingto the manufacturer’s instructions. Quality of the RNA was verified by agarosegel electrophoresis using ethidium bromide staining. For each RT reaction, 2�g DNA-free RNA were used with oligo(dT) primers and Superscript reverse

FIG. 2. Morphometric analysis of �3, �5, and �6�1 integrin immunore-activity in duct and endocrine regions of the developing human pan-creas. Data are expressed as means � SE (n � 4) *Duct. #Endocrineregion. *#P < 0.05 and ##P < 0.01 vs. respective 12- to 13-week group.

�1 INTEGRIN AND FETAL PANCREAS DEVELOPMENT


transcriptase. PCRs were carried out in a T-gradient Biometra PCR thermalcycler (Montreal Biotech, Kirkland, QC, Canada) to determine the annealingtemperature for each pair of primers (19). The PCR primers used include �1integrin, F, 5�-GACCTGCCTTGGTGTCTGTGC-3� and R, 5�-AGCAACCACACCAGCTACAAT-3� (313 bp); insulin, F, 5�-TCACACCTGGTGG AAGCTC-3� and R,5�-ACAATGCCACGCTTCTGC-3� (179 bp); and 18S, F, 5�-GTAA CCCGTTGAACCCCATT-3� and R, 5�-CCATCCAATC GGTAGTAGCG-3� (131 bp). Controlsinvolved omitting RT, cDNA, or DNA polymerase and showed no reactionbands. Real-time PCR analyses of �1 integrin and insulin were performed on0.1 �g cDNA using the SYBR green qPCR kit in DNA Engine Option (MJResearch, South San Francisco, CA). Data were normalized to the 18S RNAsubunit with at least three pancreata per age or experimental group (19).Similar results were obtained if the data were normalized to glyceraldehyde-3-phosphate dehydrogenase (data not shown). Both housekeeping genesshowed stable mRNA expression in the 8- to 20-week fetal pancreatic tissuesand cultured islets.Cell adhesion assay. To examine integrin function in regulating cell adhe-sion to ECM, human fetal pancreata (14–16 weeks) were digested withcollagenase V (2 mg/ml) for 30 min at 37°C. Islet-epithelial clusters, whichcontained mostly undifferentiated epithelial cells and 2–10% endocrine cells(22), were washed in cold 1� Hanks’ balanced salt solution and recovered inCMRL 1066 supplemented with 10% fetal bovine serum for 2 h at 37°C.Adhesion assays were carried out in 12-well plates (Corning/VWR, Toronto,ON, Canada) coated with fibronectin (50 �g/ml) or laminin (50 �g/ml); rat tailcollagen (1 mg/ml) was also used by applying neutralized collagen onto thesurface of each well to form a thin gel (14). Cell clusters were pretreated for1 h with hamster monoclonal anti-�1-integrin (CD29, 5 �g/ml; Pharmingen,Mississauga, ON, Canada), with hamster IgM isotype (5 �g/ml) or vehicle(control), plated onto coated wells (100 clusters/well) and cultured withCMRL 1066 supplemented with 10% fetal bovine serum for 24 h at 37°C in 5%CO2. At the end of the incubation period, unattached cell clusters werewashed off by repeated rinses in Hanks’ balanced salt solution. The attachedcell clusters were counted using an inverted microscope. The number of cellclusters adhered to coated matrix wells was calculated as a percentage oftotal cell clusters plated; each experiment used triplicate wells/group and wasrepeated five times (14).

Transferase-mediated dUTP nick-end labeling assays and insulin mRNA

expression. To analyze for cell death and insulin gene expression, clusters ofthe three experimental groups were cultured in suspension. RNA sampleswere harvested after 2 and 24 h of treatment followed by RT-PCR andreal-time RT-PCR analyses for insulin mRNA (19). For the cell death assays,24-h treated cell clusters embedded in 2% agarose were fixed in 4% parafor-maldehyde followed by paraffin embedding. As described previously, 5-�msections were deparaffinized, pretreated with 0.1% trypsin and incubated withthe transferase-mediated dUTP nick-end labeling (TUNEL) reaction mixture(Roche) for 60 min at 37°C (14,19). The sections were subsequently stainedwith guinea pig anti-human insulin or mouse anti-human glucagon labeledwith rhodamine (tetramethylrhodamine isothiocyanate [TRITC]). The per-centage of total TUNEL-positive islet-epithelial cluster cells, �-cells, and�-cells was determined.�1 integrin siRNA transfections. Freshly isolated clusters, after a 1-hcalibration culture in antibiotic-free medium, were transiently transfected assuspensions for 30 h with 60 nmol/l �1 integrin siRNA (proprietary sequence;accession #NM_002211) or control siRNA (proprietary sequence) commer-cially produced by Santa Cruz Biotechnology using an siRNA transfection kit(Santa Cruz Biotechnology). Islet-epithelial clusters were harvested 72 h aftertransfection and assessed for the expression of �1 integrin and insulin proteinas well as insulin mRNA (19). Transfection efficiency was monitored usingfluorescein-conjugated control siRNA (Santa Cruz Biotechnology) with �60%of the islet-epithelial cluster cells being transfected during each experiment.Cell viability was examined using the MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] assay (19,23) and 100 islet-epithelial clustersfrom both �1 siRNA and control siRNA transfected groups were plated intriplicate and cultured for 72 h. The clusters were harvested in 500 �l ofculture medium, and then 50 �l of stock MTT (5 mg/ml, Sigma) was added for2-h incubation at 37°C. Cells were washed and lysed by 200 �l DMSO (Sigma).The samples were assayed for absorbance at 595 nm using a MultiskanSpectrum spectrophotometer (Thermo Labsystems, Franklin, MA).Statistical analysis. Data are expressed as means SE. Statistical signifi-cance was determined using a two-tailed unpaired Student’s t test or one-wayANOVA followed by the Student-Newman-Keuls group comparison test.Differences were considered to be statistically significant when P 0.05.

FIG. 3. Coexpression patterns of �3/insulin,�5/glucagon, �6�1/insulin, and �6�1/PDX-1 ina 14-week human fetal pancreas. Integrins arelabeled by fluorescein isothiocyanate (green)and insulin, glucagon, or PDX-1 by TRITC(red). Nuclei were counterstained by 4�,6-dia-midino-2-phenylindole (blue). Arrows indicatedouble-labeled cells, with integrins (green)located primarily at cell borders and endocrinecell markers (red) within the cytoplasm. *Non-specific staining of blood cells. Original mag-nification 400�; insert 630�.



RESULTS

Expression of �1 integrin and its associated �3, �5,

and �6 subunits in the developing human pancreas.

To examine when and where integrins appear duringpancreas development, we first screened for several pos-sible integrin receptors in the human pancreas from 8–20weeks of fetal age. Using immunohistochemistry, we ob-served �1, �3, �5, and �6 integrin immunoreactivity in thepancreatic ductal cells at 8 weeks, before glucagon- andinsulin-immunoreactive cells bud off. Western blot analy-sis of the protein level of �1 integrin and its associated�-subunits revealed an increased expression of �1, �3, and�6 integrin that was statistically significant (P 0.05) by16 weeks, whereas the expression of �5 was relativelyconstant from weeks 8–20 (Fig. 1A–D). Quantitative anal-ysis of �1 integrin mRNA by real-time RT-PCR determinedthat its mRNA signal increased significantly by 12 weeks(P 0.05) (Fig. 1E), before the increase observed in �1protein levels (Fig. 1A).

To examine the distribution and colocalization of inte-grin �3, �5, and �6�1 with the epithelial (CK19) andmature endocrine cell (insulin and glucagon) markers,dual immunofluorescence experiments were conducted.Morphometric analysis of the expression patterns of �3,�5, and �6�1 integrins revealed that there was a significantincrease in integrin immunoreactive area at 20 weeks inthe ductal regions (P 0.05) (Fig. 2), except for the �5subunit. Expression in endocrine cells was detected asearly as when single endocrine cells began to bud off fromthe ducts (9 weeks); a gradual increase in their immuno-reactive area in islets was significant for �3 (P 0.05) andfor �6�1 (P 0.005) by week 20 (Fig. 2).

It was readily apparent that many single insulin- or

FIG. 4. The percentage of cells demonstrating coexpression of �3, �5,and �6�1 integrins with insulin (A) or glucagon (B) in 12- and 18-weekhuman fetal pancreata. Data are expressed as means � SE (n � 4).**P < 0.01 for 18 vs. 12 weeks.

FIG. 5. Localization of �3/collagen IV, �5/fibronectin, and �6�1/laminin in a 14-week human fetal pancreas. Integrins are labeled by fluoresceinisothiocyanate (green) and extracellular matrices by TRITC (red). Arrows indicate integrin-positive cells associated with labeled ECM.*Nonspecific staining. Original magnification 400�; insert 630�.



glucagon-positive cells or newly formed small isletscostained with the three integrins (Fig. 3). A high propor-tion of �3 or �5 integrin and insulin costaining wasobserved at 12 weeks with slightly decreased coexpres-sion by week 18 (Fig. 4A). In contrast, �6�1 integrincolocalization with insulin increased slightly from 12–18

weeks (63 9% vs. 78 12%) (Fig. 4A). The percentage ofcells coexpressing �3 or �6�1 with glucagon showed nodifference during development; however, there was asignificant increase in frequency of coexpression of �5with glucagon at 18 weeks (P 0.01) (Fig. 4B), indicatingthat �5 may play a prominent role in glucagon cell

FIG. 6. Effect of �1 integrin loss on adhesion of islet-epithelial clusters on different matrices. A: Phase-contrast micrographs of clusters in cultureafter 24 h in the absence (Ctrl) or presence of monoclonal anti–�1-integrin or IgM antibodies on collagen I–, fibronectin-, or laminin-coateddishes. B: Quantitative analysis of cluster adherence rate. Data are expressed as means � SE (n � 5). **P < 0.01, ***P < 0.001 relative tocontrols. C: Representative RT-PCR and real-time RT-PCR analyses of preproinsulin expression in the three experimental groups after 2 and/or24 h of culture. Quantitative RT-PCR data are normalized to 18S RNA subunit and expressed as means � SE (n � 4). *P < 0.05 vs. controls.



maturation and functional maintenance. Colocalization of�6�1 integrin with PDX-1, an early pancreatic develop-mental transcription factor, was frequently observed inboth duct and islet regions (Fig. 3).Correlation of ECM proteins and integrin receptors

in the developing human pancreas. Collagen IV, fi-bronectin, and laminin are important ECM components inbasement membranes as well as ligands for �3�1 integrin(11). Fibronectin is a known adhesive site for �5�1 and�6�1 integrin adheres to laminin (11,14). To correlateintegrin expression and their ECM ligands in the develop-ing human pancreas, dual immunofluorescent staining wasperformed on paraffin-embedded fetal pancreatic sections.Figure 5 shows �3 integrin-positive cells colocalizing withcollagen IV, which highlights the basal membrane ofpancreatic ducts and islets. There was also localization of�3 integrin–positive cells at sites where fibronectin andlaminin were observed in the developing human pancreas(data not shown). Expression of �5 integrin was localizedto cells that were surrounded by a matrix of fibronectin(Fig. 5) but not collagen IV and laminin (data not shown).Staining of �6�1 integrin was associated with cells in closeproximity to a matrix of laminin (Fig. 5). These observa-tions suggest that the developing human pancreas hasparallel expression patterns of integrins and their respec-tive ligands.

�1 integrin receptor mediates fetal islet-epithelial

cluster adhesion, insulin gene expression, and cell

survival. We subsequently tested the effects of blocking�1 integrin activity in islet-epithelial cell clusters using a�1 integrin immunoneutralizing antibody. The percentageof clusters in both control and IgM antibody–treatedgroups that adhere to collagen I, fibronectin, or lamininwas high (up to 80%) and clusters attached to fibronectinand collagen I began to spread after 24 h of culture (Fig. 6A

and B). In contrast, islet-epithelial clusters treated withanti–�1-integrin had a significantly reduced ability to ad-here to ECM, such that the percentages of adhesiondropped to 12 1, 15 2, and 31 6%, respectively, after24 h (P 0.001) (Fig. 6A and B). Decreasing adhesion wasassociated with lost �1 integrin expression: quantitativeanalysis demonstrated a 37% reduction in �1-integrin–expressing cells after 24 h of anti–�1-antibody treatment(35 5% vs. 55 4% of controls, P 0.001). Cellscoexpressing �1 integrin with insulin were also signifi-cantly decreased (3.3 1% vs. 8.6 3% of controls, P 0.05), whereas the total percentage of insulin-positive cellswas slightly reduced (8.5 4% vs. 11.4 3% of controls).We investigated whether blocking the �1 integrin receptoraffects insulin gene expression using quantitative RT-PCRassays (Fig. 6C). There was no change in preproinsulinexpression after 2 h of �1 integrin blockade. However,

FIG. 7. A: The apoptotic index of islet-epi-thelial clusters cultured in the absence(Ctrl) or presence of anti–�1-integrin immu-noneutralizing or IgM antibodies for 24 h.Data are expressed as means � SE (n � 4).*P < 0.05, **P < 0.01, ***P < 0.001 relativeto controls. Double labeling for TUNEL(green) and insulin (red) in islet-epithelialclusters B: Arrowheads indicate TUNEL-pos-itive cells, arrow indicates an apoptotic�-cell (green nucleus, red cytoplasm), andasterisks indicate nonspecific staining. Orig-inal magnification 400�; insert 630�.



after 24 h, there was an 18% reduction in insulin mRNA(P 0.05) compared with the control group (Fig. 6C).

Blockade of �1 integrin was also associated with asignificant increase in the number of cells in the islet-epithelial clusters undergoing apoptosis, as assessed bythe TUNEL assay (P 0.001) (Fig. 7). There was also asignificant increase in the total number of apoptotic �- and�-cells in the anti–�1-antibody group (P 0.01) (Fig. 7).

Furthermore, we examined the effects of suppressing �1integrin expression in islet-epithelial clusters using a spe-cific �1 integrin siRNA. After the transfection and 72 h ofculture, a significant downregulation of �1 integrin protein

expression was observed (43 3.5% vs. 72 1.2%, P 0.01) (Fig. 8A). Knockdown of �1 integrin was associatedwith a decrease in preproinsulin expression and reducedcell viability as determined by the MTT assay (P 0.02)(Fig. 8B). Decreasing insulin gene expression was corre-lated with a significant loss in the number of insulin-positive cells in the clusters (8 1% vs. 17 3%, P 0.05)(Fig. 8A).

DISCUSSION

The present study demonstrates that �1 integrin and itsassociated �-subunits are expressed in the early to mid-

FIG. 8. Transfection of islet-epithe-lial clusters with �1 integrin siRNAresulted in decreases in �1 integrinprotein, the number of immunoreac-tive �1 integrin and insulin express-ing cells (A), �1 integrin and insulinmRNA expression, and cell viability(RT omitted) (B). Cell viability wasassessed by MTT assay. Data areexpressed as means � SE (n � 3).*P < 0.05, **P < 0.01 relative tocontrols.



gestation human fetal pancreas in a dynamic, temporallyregulated fashion. Furthermore, blockade of �1 activityresults in loss of islet-cluster attachment to various extra-cellular matrices, diminished insulin gene expression, anddecreased cell viability. Taken together, the data from thisdescriptive and functional investigation demonstrate thatintegrin-ECM interactions in the developing human pan-creas are critical for normal islet formation and endocrinefunction.

Immunohistochemical, morphometrical, RNA, and pro-tein analyses showed a specific temporal and spatialpattern for �1 integrin expression associated with �3, �5,and �6 subunits during fetal development. They are de-tectable within the ducts as early as 8 weeks of fetal ageand gradually increase in expression from 12 to 20 weeks.After 9–11 weeks, newly forming single endocrine cells orsmall islets frequently expressed �3, �5, and �6�1, sug-gesting that these receptors are involved in regulatingdifferentiation and migration of endocrine cell types bud-ding from and in close proximity to ductal structures. Inaddition, within the larger islets observed at mid-gestation,integrin subunit expression is high, supporting a role forthese receptors in the formation and function of matureislets. Interestingly, the integrin �6�1 subunit has beenpreviously reported to mediate �-cell differentiation andthe �-cell secretory response in dissociated fetal mousepancreatic epithelium on a laminin-1 matrix (1–2). Ourresults indicate that ��1 integrin receptors are also likelyto play an important role during ontogeny of the humanfetal pancreas.

Proteins such as type IV collagen, laminin, and fibronec-tin have been previously described as major componentsof the basement membrane in the postnatal human pan-creas (24). Studies from our laboratory as well as others(13,16,25) show that these ECM molecules are also com-ponents of the human fetal pancreatic basement mem-brane and are expressed within the developing pancreas ina specific spatial pattern. The integrin subunits �3, �5, and�6�1 are expressed in cells that localize in proximity toimmunoreactive areas for these matrix molecules. Dissect-ing the functional significance of the individual � subunit–ECM interactions will be important in determining theirrole in mediating pancreatic development. For example,studies of fetal mouse pancreatic epithelia on a commer-cially available basement membrane gel, Matrigel, havedemonstrated that laminin and the �6 subunit mediate themorphological events of ductal formation (15). Further-more, Jiang et al. (1) have shown that dissociated pancre-atic cells from the 13.5d mouse fetus have increased �-celldifferentiation mediated by �6 integrin when placed on amatrix of laminin-1.

To examine the functional role of �1 in developingislets, we used an immunoneutralizing monoclonal anti-body. Blockade of �1 integrin resulted in impairment ofislet-epithelial cluster adhesion on several ECM, highlight-ing that the �1 subunit plays a critical role during pancre-atic development. Treatment with an equal amount of IgMantibody had no effect, indicating that the interferencewith �1 function by a blocking antibody is the result of aspecific interaction. In support of these data, Kaido et al.(16) recently reported that the �V�1 integrin may be

responsible for early motile processes that regulate humanfetal islet formation.

We also demonstrated that blockade of �1 integrinreceptor in islet-epithelial clusters is associated with anincrease in the number of cells undergoing apoptosis, witha specific increase in �- and �-cell death. These data are inline with the previously described role for the �1 receptorin offering protection from cell death (26). Adherent cellsrequire integrin signaling for survival; otherwise theyundergo a process termed anoiksis (27), evidenced bydisengagement of epithelial and fibroblast cells from theirmicroenvironment components.

The perturbation of �1 integrin in the developing fetalislet clusters was also associated with a decrease in insulinmRNA and protein expression. This is not an unexpectedresult given that several studies have shown that adultrodent and human islets cultured on or embedded invarious ECM have improved insulin secretion and glucose-stimulated secretory responses, potentially mediated byintegrin-ECM interactions (2,28–29). Whether maturationof the glucose-induced insulin response will occur if fetalcells are cultured in the presence of �1 integrin and itsassociated � subunits is yet unknown. However, given thedata from our laboratory as well as others, describing animportant role for these integrins during development ofthe human fetal pancreas (13,16,25), such studies arelikely to have positive results.

The siRNA silencing systems are extremely useful toolsfor studying the functional importance of genes (30–32).Most siRNA studies have been carried out on cell lines,with limited information on the effects of gene silencing onprimary islets (30). Our recent study of neonatal rat islets,using a nonadenoviral transient transfection of �1 integrinsiRNA, demonstrated a significant decrease in islet cellsurvival (19). We therefore examined the effect of �1integrin siRNA transfection on human fetal islet-epithelialclusters. The results were similar to what we observedwith the immunoneutralizing antibody: a significant de-crease in �1 integrin protein correlated with a reduction ininsulin mRNA and protein as well as cell viability. Thesedata support the hypothesis that �1 integrin may be animportant regulator of pancreatic endocrine neogenesis aswell as being involved in cellular resistance to apoptoticstimuli.

In summary, the present study provides insight into theexpression of integrin receptors in the human fetal pan-creas and sheds light on how the �1 receptor, in conjunc-tion with its binding partners �3, �5, and �6, may playmultiple roles in islet cell biology, including adhesion,function, and survival. Identifying such factors is a criticalfirst step in developing new islet cell–based therapies forthe treatment of �-cell destruction in insulin-dependentdiabetes.

ACKNOWLEDGMENTS

This work was supported by grants from the Departmentof Medicine at the University of Western Ontario, theLawson Health Research Institute, and the Canadian Insti-tutes of Health Research. R.W. is supported by a UniversityFaculty Award from the Natural Sciences & EngineeringResearch Council of Canada.

We thank the Department of Pathology at London



Health Science Centre for allowing us to access the TissueBank and providing the human fetal pancreas tissuesections.

REFERENCES

1. Jiang FX, Cram DS, DeAizpurua HJ, Harrison LC: Laminin-1 promotesdifferentiation of fetal mouse pancreatic beta-cells. Diabetes 48:722–730,1999

2. Bosco D, Meda P, Halban PA, Rouiller DG: Importance of cell-matrixinteractions in rat islet beta cell secretion in vitro: role of �6�1 integrin.Diabetes 49:233–243, 2000

3. Hynes RO: Integrins: a family of cell surface receptors. Cell 48:549–554,1987

4. Schwartz MA, Ingber DE: Integrating with integrins. Mol Biol Cell 4:389–393, 1994

5. Clark EA, Brugge JS: Integrin and signal transduction pathways: the roadtaken. Science 268:233–239, 1995

6. Bagutti C, Wobus AM, Fassler R, Watt FM: Differentiation of embryonalstem cells into keratinocytes: comparison of wild-type and �1 integrin-deficient cells. Dev Biol 179:184–196, 1996

7. Fassler R, Pfaff M, Murphy J, Noegel AA, Johansson S, Timpl R, AlbrechtR: Lack of �1 integrin gene in embryonic stem cells affects morphology,adhesion, and migration but not integration into the inner cell mass ofblastocysts. J Cell Biol 128:979–988, 1995

8. Streuli CH, Bissell MJ: Mammary epithelial cells, extracellular matrix, andgene expression. Cancer Treat Res 53:365–381, 1991

9. Carroll JM, Romero MR, Watt FM: Suprabasal integrin expression in theepidermis of transgenic mice results in developmental defects and aphenotype resembling psoriasis. Cell 83:957–968, 1995

10. Howlett A, Bailey N, Damsky C, Petersen O, Bissell M: Cellular growth andsurvival are mediated by �1 integrins in normal human breast epitheliumbut not in breast carcinoma. J Cell Sci 108:1945–1957, 1995

11. Bouvard D, Brakebusch C, Gustafsson E, Aszodi A, Bengtsson T, Berna A,Fassler R: Functional consequences of integrin gene mutations in mice.Circ Res 89:211–223, 2001

12. Kantengwa S, Beatens D, Sadoul K, Buck CA, Halban PA, Rouiller DG:Identification and characterization of �3 �1 on primary and transformedrat islet cells. Exp Cell Res 237:394–402, 1997

13. Yebra M, Montgomery AM, Diaferia GR, Kaido T, Silletti S, Perez B, JustML, Hildbrand S, Hurford R, Florkiewicz E, Tessier-Lavigne M, Cirulli V:Recognition of the neural chemoattractant Netrin-1 by integrins �6�4 and�3�1 regulates epithelial cell adhesion and migration. Dev Cell 5:695–707,2003

14. Wang R, Rosenberg L: Maintenance of �-cell function and survival follow-ing islet isolation requires re-establishment of the islet-matrix relationship.J Endocrinol 163:181–190, 1999

15. Crisera CA, Kadison AS, Breslow GD, Maldonado TS, Longaker MT, GittesGK: Expression and role of laminin-1 in mouse pancreatic organogenesis.Diabetes 49:936–944, 2000

16. Kaido T, Perez B, Yebra M, Hill J, Cirulli V, Hayek A, Montgomery AM:�v-Integrin utilization in human �-cell adhesion, spreading, and motility.J Biol Chem 279:17731–17737, 2004

17. Hammer E, Parnaud G, Bosco D, Perriraz N, Maedler K, Donath M, RouillerG, Halban PA: Extracellular matrix protects pancreatic �-cells againstapoptosis: role of short- and long-term signaling pathways. Diabetes

53:2034–2041, 200418. Wang R, Bouwens L, Kloppel G: �-Cell proliferation in normal and

streptozotocin-treated newborn rats: sites, dynamics and capacity. Diabe-

tologia 37:1088–1096, 199419. Yashpal NK, Li J, Wheeler MB, Wang R: Expression of �1 integrin receptors

during rat pancreas development—sites and dynamics. Endocrinology

146:1798–180720. Paraskevas S, Aikin R, Maysinger D, Lakey JR, Cavanagh TJ, Hering B,

Wang R, Rosenberg L: Activation and expression of ERK, JNK, and p38MAP-kinases in isolated islets of Langerhans: implications for culturedislet survival. FEBS Lett 455:203–208, 1999

21. Lubman RL, Zhang XL, Zheng J, Ocampo L, Lopez MZ, Veeraghavan S,Zabaki SM, Danto SI, Borok Z: Integrin �3-subunit expression modulatesalveolar epithelial cell monolayer formation. Am J Physiol Lung Cell Mol

Physiol 279:L183–L193, 200022. Beattie GM, Levine F, Mally MI, Otonkoski T, O’Brien JS, Salomon DR,

Hayek A: Acid �-galactosidase: a developmentally regulated marker ofendocrine cell precursors in the human fetal pancreas. J Clin Endocrinol

Metab 78:1232–1240, 199423. Aikin R, Maysinger D, Rosenberg L: Cross-talk between phosphatidylino-

sitol 3-kinase/AKT and c-jun NH2-terminal kinase mediates survival ofisolated human islets. Endocrinology 145:4522–4531, 2004

24. Van Deijnen JH, Van Suylichem PT, Wolters GH, Van Schilfgaarde R:Distribution of collagens type I, type III and type V in the pancreas of rat,dog, pig and man. Cell Tissue Res 277:115–121, 1994

25. Cirulli V, Beattie GM, Klier G, Ellisman M, Ricordi C, Quaranta V, FrasierF, Ishii JK, Hayek A, Salomon DR: Expression and function of �v�3 and�v�5 integrins in the developing pancreas: roles in the adhesion andmigration of putative endocrine progenitor cells. J Cell Biol 150:1445–1460,2000

26. Farrelly N, Lee YJ, Oliver J, Dive C, Streuli CH: Extracellular matrixregulates apoptosis in mammary epithelium through a control on insulinsignaling. J Cell Biol 144:1337–1348, 1999

27. Ruoslahti E: Integrins. J Clin Invest 87:1–5, 199128. Lucas-Clerc C, Massart C, Campion JP, Launois B, Nicol M: Long-term

culture of human pancreatic islets in an extracellular matrix: morpholog-ical and metabolic effects. Mol Cell Endocrinol 94:9–20, 1993

29. Thivolet CH, Chatalain P, Nicoloso H, Durand A, Bertrand J: Morphologicaland functional effects of extracellular matrix on pancreatic islet cellcultures. Exp Cell Res 159:313–322, 1985

30. Bain JR, Schisler JC, Takeuchi K, Newgard CB, Becken TC: An adenovirusvector for efficient RNA interference-mediated suppression of target genesin insulinoma cells and pancreatic islets of Langerhans. Diabetes 53:2190–2194, 2004

31. Elbashir SM, Harborth J, Weber K, Tuschl T: Analysis of gene function insomatic mammalian cells using small interfering RNAs. Methods 26:199–213, 2002

32. Morris KV, Chan SW, Jacobsen SE, Looney DJ: Small interfering RNA-induced transcriptional gene silencing in human cells. Science 305:1289–1292, 2004



role for 1 integrin and its associated 3, 5, and 6 ...€¦ · ptosis via the integrin 1/focal...

Documents