[CANCER RESEARCH58. 315-321. January 15. 19981

ABSTRACT

The human epithelial mucin, MUd, is a large transmembrane glycoprotein that is expressed on most simple epithelia. It is overexpressed andaberrantly glycosylated on many human epithelial tumors, including morethan 90% of human breast cancers. MUC1 is of interest as an immunotherapy target because patients with breast, ovarian, and pancreaticcancers have T lymphocytes in their tumor-draininglymph nodes that canbe induced to recognize and lyre MUC1-expressing tumor cells. We have

produced a transgenic mouse model that expresses the human MUC1molecule on an inbred CS7BI/6 background to Investigate the effect ofendogenous expression of MUC1 on the ability of mice to generate antitumorimmunityto MUC1-expressingtumors.Transgenicmiceexpressedthe humantransgeneIna patternand levelconsistentwiththatobservedIn humans. Transgenlc mice were tolerant to stimulation by MUC1 asevidencedby theabilityof MUC1-expressingtumorcellsto growin thesemice, whereas MUC1-expressing cells were eliminated from wild-typemice. Moreover, transgenlc mice Immunized with MUC1 peptides failed toexhibit Immunoglobulin class switching to the IgG subtypes. These datasuggest that endogenous expression of MUC1 protein by MUC1 tram

genic mice Induces T-cell tolerance to stimulation by MUC1. The tramgenic mice will provide a useful model to Investigate the mechanisms thatregulate Immunological tolerance to tumor antigens and will facilitate theInvestigation of anti-MUC1 immunotherapy formulations.

INTRODUCTION

The human epithelial mucin MUC1 is a type I membrane-associated glycoprotein expressed on most simple epithelia. It has an extensive extracellular domain that is composed primarily of a variablenumber of TRs3 (1—3).The TR domain is a heavily O-glycosylatedregion rich in serine, threonine, and proline residues and consists ofrepeating units of 20 amino acids with the sequence PDTRPAPGSTAPPAHGVTSA MUC1 is frequently overexpressed in humantumors, including carcinoma of the breast, lung, pancreas, colon, andovaries, Expression of MUC1 by tumors often results in loss of apicallocalization and alterations in posttranslational modifications cornparedto correspondingnormalcelltypes(reviewedin Ref.4).

Lymph nodes from patients with breast, ovarian, and pancreaticadenocarcinornas contain lymphocytes that recognize and lyse

MUC 1-expressing tumor cells (5, 6), Recognition of MUC 1 by lyrnphocytes may be due to exposure of epitopes in the TR region of theprotein backbone that normally are not exposed due to extensiveO@glycosylation, These findings provide evidence for MUC 1-specificstimulation of the immune system (@—7);however, cellular responsesto MUC1 in humans have only been demonstrated by in vitro assaysfollowing extensive stimulation of lymphocytes with antigen andcytokines (5, 6),

Received 7/16/97; accepted I 1/1 1/97.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 10 accordance with18 IJS.C. Section 1734 solely to indicate this fact.

I This work was supported by the Nat1on@l Cancer Institute, NIH, Grant R01-CA64389

and fundsfromthe MayoFoundationfor liducationand Research(to S.J. 0.); and byNational Cancer Institute, NIH. Grants ROl-CA57362, R30 CA36727, and T32 CA09476andGrantLB595fromthe NebraskaDepartmentof Health(to M.A.H.).

2 To whom requests for reprints should be addressed, at Department of Biochemistry

and MolecularBiology,MayoClinicScottsdale,13400li SheaBoulevard,Scottsdale,AZ 85259.

3 The abbreviations used are: TR, tandem repeat; MUC1.Tg, MUCI transgenic; wt,

wild-type.

Studies have shown that MUC1 is highly immunogenic in mice(8—10);this is most likely due to the low degree of homology in theTR region between MUC1 and its murine homologue, Mud (1 1).Mice immunized with peptides corresponding to the MUC1 TR produce strong humoral immune responses and mice immunized withtumor cells that express MUC1 or with injections of MUC1 cDNA arein most cases protected from subsequent tumor challenge by stronghumoral and cellular responses to MUC1 (12—14).The degree ofdissimilarity between the sequence of human MUC1 and murineMuc 1 makes it likely that human MUC 1 is recognized as a foreignantigen by mice. Hence, it is not surprising that the murine immuneresponse to MUC1 provides effective protection against MUC1-expressing murine tumors because these responses are not produced inthe context of immunological tolerance.

Tolerance is a major hindrance to the development of effectiveimmune responses to tumors (15, 16). Tolerogenic mechanisms protect against untoward autoimmune reactions under nonpathologicalcircumstances (17—19)but may preclude effective immunotherapy fortumors. In this report, we describe the production of an inbred syngeneic C57BI/6 mouse strain transgenic for human MUC1(MUC1.Tg) and its initial characterization as a model system toinvestigate parameters of immunological tolerance to MUC 1. Micethat are transgenic for a foreign protein develop B- and T-cell compartment tolerance and are refractory to immunization with the proteinencoded by the transgene (20—22). Such immune tolerance is predicted to be similar to that seen with endogenous self proteins;consequently, the correct temporal and spatial expression of MUC Iduring development should have profound effects on the ability of themunne immune system to provide effective anti-MUC 1 tumor immunity. A comparison of the immune response of the MUC1.Tg micewith that of wt C57B1/6 mice will provide insight into the effect ofendogenous expression of the MUCJ gene on immune responses andimmune tolerance and facilitate the evaluation of anti-MUC1 vaccineformulations in vivo within the context of an animal that expressesMUC1

315

Tolerance and Immunity to MUC1 in a Human MUC1 Transgenic Murine Model1

Gerald J. Rowse, Richard M. Tempero, Michelle L. VanLith, Michael A. Hollingsworth, and Sandra J. Genffler@

Department of Biochemistry and Molecular Biology, Mayo Clinic Scottsdale. Scottsdale, Arizona 85259 (G. J. R., S. J. G.J. and Eppley Institute for Research in Cancer and AlliedDiseases, University ofNebraska Medical Center, Omaha. Nebraska 68198 (R. M. T.. M. L V.. M. A. H.J

MATERIALSAND METHODS

Generationof the MUC1,TgMice. Micewerehousedandmaintainedinmicroisolator cages under specific pathogen-free conditions. A 10.6-kbgenomic SacII fragment containing the entire MUCI gene sequence, as well as1.5 kb of 5' sequence and 800 bases of 3' sequence, was isolated (23). DNAwas diluted to a final concentration of I mg/mi for injection into the pronucleiof fertilized CS7B1/6mouse eggs, which were then transferred to pseudopregnant CD1 females.

Southern illot Analysis, Founder mice were identified by Southern blotanalysis. Ten-gil aliquots (approximately 5 @.tgeach) of tail DNA were digestedwith I@coRI(New England Biolabs, Beverly, MA) at 37°Cin a total volume of25 @1.The DNA digestswereelectrophoresedthrougha0.7%agurosegel andtransferred to nylon membrane (Hybond N, Amersham Corp., ArlingtonHeights, IL.). Subsequent hybridization and washing procedures were carriedout according to the manufacturer's specifications. The human probe pMUC7(2),correspondingto —@500bpof theTRdomain,waslabeledbyrandompriming in the presence of Ea-32PIdCTP (Amersham) to a specific activity of

>1 X 10@dpm/@.Lg.PCR Screening,PCRwas usedto routinelyidentifyMUC1.Tg positive

mice in the colony. PCR was carried out in a total volume of 50 @lin

on June 5, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TOLERANCE IN MUC1 TRANSOENIC MICE

Perkin-Elmer Corp. Gene Amp tubes with the following reagents: 5 @llOX

PCR buffer (Life Technologies, Inc., Grand Island, NY), 0.02% formamide, 1mM MgCl2, 0.2 mM dNTPs, 100 nM 5'-CTTGCCAGCCATAGCACCAAG-3'(bp 745-765; Ref. 1), and 100 flM 5'-CTCCACGTCGTGGACATfGATG-3'

(bp 1086—1065; Ref. I) primers, 10.25 units of Taq polymerase (Life Technologies, Inc.), I @lof tail DNA (approximately 500 ng), and reagent-qualityH20.The amplification program consisted of one cycle of 10 mm at 94°Cand40 cycles of 30 s each at 94, 61, and 72°C.The PCR product of each reactionwas analyzed by size fractionation through a I% agarose gel. Amplification ofMUC1-positive DNA resulted in a 500-bp fragment.

Cell Culture.Cell culturereagentswerepurchasedfromLifeTechnologies, Inc., unless noted otherwise. The B16-BL6 (Bl6) murine melanoma cellline (24), kindly provided by J. Talmadge (University of Nebraska MedicalCenter, Omaha, NE), was cultured in DMEM supplemented with 10% fetalbovine serum, nonessential amino acids, vitamins, penicillin (50 units/mI), andstreptomycin (50 @xg/ml)cocktail and 1mMsodium pyruvate (Sigma ChemicalCo., St. Louis, MO).

Expression of Epitope-tagged MUC1. Epitope-tagged MUCI (MUC1F)cDNA was subcloned into the expression vector pH(3APr-l-neo (25) as described previously (26). Bl6 cells were transfected with linearized plasmidusing the Lipofectin reagent according to the manufacturer's specifications.

After transfection, single colonies were selected and expanded for MUC1Fscreening. Clonal cell lines were evaluated for MUC1F expression by immunoblouing, flow cytometry, and analysis of surface immunofluoresence withthe M2 antibody (International Biotechnologies Inc., Eastman Kodak Co., NewHaven, CT).

Immunoblotting. Cell lysates were prepared by scraping cells into 1 ml oflysis buffer (10 mM Tris, 150 mM NaCI, I mrviphenylmethylsulfonyl fluoride,1% Triton X-lO0). After a 30-mm incubation at 4°C,lysates were centrifugedat 1000 X g (4°C)for 5 mm to remove cell debris. Supernatants weretransferred to fresh tubes and stored at —20°C.Protein content of lysates wasdetermined by using the Bio-Rad Protein Assay (Bio-Rad, Hercules, CA) withBSA as standard.Cell lysates wereresolvedon 6%denaturingpolyacrylamidegels, transferred to polyvinylidene difluonde membranes electrophoretically,

and blocked in Blouo [5% dry milk in TBS (0.9% NaCl, 10 mMTris, pH 7.4,0.5% MgCl2)] overnight. M2 or isotype control antibody (2 @gIml) in Blouo

were incubated for I h at 25°C.Following three 10 mm washes with Blotto,horseradish peroxidase-conjugated secondary antibodies (Zymed LaboratoriesInc., San Francisco, CA) were diluted 1:2000 in Blotto and incubated for 1 hat 25°Cfollowed by three washes with Blono. ECL reagents were applied asper the manufacturer's instruction, and the blots were exposed to ECLsensitive film (Amersham).

Tumor Challenges.wt (TheJacksonLaboratory,BarHarbor,ME)andMUC1.Tg C57B1I6 mice were anesthetized with Metofane (Pitman-Moore,Mandelein, IL), and MUC1F-expressing B16 or control transfectants wereinjected s.c. between the scapulae. Tumor length (1) and width (2r) weredetermined at the indicated days postinjection. Tumor volume was calculatedusing the formulas 4/3irr@ for spherical tumors and irr@lfor cylindrical tumors.

Immunohistochemistry. Tumors and normal tissues were obtained fromwt and MUC1.Tg mice. Tissues were fixed in methacarn (60% methanol, 30%chloroform, 10% glacial acetic acid), paraffin embedded, and sectioned forimmunohistochemical analysis. Briefly, endogenous peroxidase activity was

decreased by incubating the tissue sections with 0.6% H202 in PBS (pH 7.4) for

30 mm at 25°C.After washing with PBS, pH 7.4, sections were blocked with50% FCS and stained with HMFG-2 (IgGl) murine hybridoma (which recognizes the human MUC1 TR sequence but is not reactive with the mouse Mud

protein; Ref. 2), CT1 rabbit polyclonal antibody to the cytoplasmic tail ofMUC1 (27), or antibody previously blocked with 5 mg/ml synthetic peptide(undiluted) for 1 h at 25°C.Following three 5-mm washes with PBS, pH 7.4,the sections were incubated with peroxidase-conjugated rabbit antimouseantibody (Dako Corp., Carpinteria, CA; diluted 1:100) for 1 h at 25°C.Following three 5-mm washes, the substrate solution, consisting of 0.03%hydrogen peroxide in PBS and I mg/mI diaminobenzoate (Sigma), was added,and the color was allowed to develop for 5—8mm. Following color development, the reaction was quenched by washing the slide in distilled water. Thetissue sections were counterstained lightly with hematoxylin.

SyntheticPeptides.Thepeptidesusedin thisstudywereTAPPA,a 24-amino acid peptide containing one complete 20-amino acid TR and 4 residuesof the adjoining repeat, and MUC1 105-mer (a generous gift of Dr. 0. Finn,

University of Pittsburgh School of Medicine, Pittsburgh, PA), which contains5.25 copies of the TR. Peptides were synthesized by manual solid-phasesynthesis using 9-fluorenylmethyloxycarbonyl NH2-terminal protected amino

acids. Synthesis, purification, and characterization by mass spectroscopy of the

products are described elsewhere in detail (28, 29).PeptideImmunization.Freeze-driedpeptideswerereconstitutedin PBS,

pH 7.4, and mixed with equal volumes of RIB! Detox adjuvant (RIB! ImmunoChem Research Inc., Hamilton, MT). The mixture was vortexed vigorouslyfor 10 mm. Mice were injected i.p. with 50 @&gof peptide and 50 @gofadjuvant in a total volume of 200 @l.Control mice were immunized with 50,Lg of adjuvant alone. The immunizations were repeated at 3 and 6 weeks, andblood samples were obtained by tail vein bleeding 1week following the secondand third immunizations.

ELISA. Detection of MUC1-specific antibodies was carried out by ELISA

using synthetic peptides corresponding to 1.2 or 5.25 copies of the MUC1 TR.Briefly, 96-well microtiter plates (Immunlon 4, Dynatech, Chantilly, VA) werecoated with 100 id/well of peptide (5 @g/mlin PBS, pH 7.4) and incubatedovernight at room temperature. The plates were blocked with 200 s.d/well of5% skim milk in PBS for 2 h at room temperature. The plates were washed

once with 200 pAof buffer (PBS + 0.05% Tween 20). To determine antibodytiters, 50 @lof dilute sera (1/100 to 1/12,800) was added per well in triplicate

to the 96-well plates and incubated at 37°Cfor 1 h. The plates were washedfour times with 200 @lof buffer per well. Secondary antibodies specific fordifferent immunoglobulin subtypes (rabbit antimouse; Zymed) were added at1/100 dilution, 50 s.d/well,and the plates were incubated for 1 h at 37°C.Theplates were washed four times with 200 @lof buffer per well. The tertiaryantibody (peroxidase-conjugated goat antirabbit, Zymed) was added at 1/500dilution (50 @.dJwell),and the plates were incubated at 37°Cfor 1 h. The plateswere washed four times with 200 @lof buffer per well. Plates were developedby the addition ofSO p1/wellof O-phenylene diamine (1 mg/ml; Sigma). Plateswere allowed to develop for 15 mm, and absorbance was measured at A@5,,,@in a plate reader (V,,...,@;Molecular Devices Co., Menlo Park, CA).

Quantitative ELISA. All ELISA reagents were supplied by Zymed unless

noted otherwise. Microtiter plates (96-well) were coated with 50 @lof purifiedIgGi, IgG2a, IgG2b, IgG3, or 1gM murine myeloma proteins (2 @g/ml-l ng/ml

in PBS, pH 7.4) at 37°C for 2 h. Plates were incubated overnight with 200 @l

ofblocking solution (5% dry milk in PBS) at 4°C.PBS with 1% Tween 20 wasused for all scm dilutions, antibody dilutions, and washes. Diluted rem (50 p.1)were added to the appropriate wells. Following four washes, 50 p.1of rabbitantimouse isotype specific antibody at 1/300 dilution was incubated for 45 mmat 37°C.Following four washes, 50 p.1of goat antirabbit alkaline phosphataseconjugated antibody at 1/1000 dilution was incubated for 45 mm at 37°C.Following four washes, 50 p.1ofsubstrate buffer(lO% diethanolamine, pH 9.8)containing p-nitrophenyl phosphate (1 mg/ml) was added for color develop

ment. The reaction was stopped with the addition of 50 p.10.5 MNaOH, andcolor intensity was quantified at A,,@5nm Myeloma protein standards were usedto generate isotype-specific antibody binding curves. Within the linear range,the myeloma standards were used to determine the concentration of peptidespecific antibody bound in diluted sera samples.

RESULTS

Production of MUC1.Tg Mice. Three mice out of a total of 104mice screened were found positive by Southern blot analysis for theMUCJ transgene. Of these founders, two lines transmitted the transgene to their progeny. Both lines were fertile and healthy, and one line(79.24) was selected for further studies. For immunohistochemicalanalysis, the MUCJ transgene was moved onto the mouse Muddeficient background (30), and the pattern of expression was determined using an antibody to the MUC1 cytoplasmic tail (27). Forinvestigation of tumor growth and humoral immunity in MIJC1.Tgmice, transgenic animals with a mouse Mud-expressing backgroundwere used.

Immunohistochemical staining revealed that the level and pattern ofMUC1 expression in cells and organs of the transgenic mice wassimilar to that seen in humans, including the lung, mammary gland,and pancreas (Fig. 1) and the kidney (distal convoluted tubules and

316

on June 5, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Table1 TissueexpressionofMUCI in transgenicmiceTissuesfrom MUC1.tg mice were stained with CFI antibody or peptide-blockedCl]antibody.Positive

tissues NegativetissuesGallbladder

+a Brain—KidneyColon—Collecting

ducts + Heart—Distalconvoluted tubules + Liver-Lung

Muscle—Acini- Skin-Bronchi+ Small intestine—Type

II pneumocyte + Spleen—Mammarygland+Pancreas

+Stomach(Bninner's glands)+Salivary

glandAcini-Collecting

ducts+Uterus(virgin)+a

@ tissues that stained positively for MUC1 expression; —, tissues thatstainednegatively

for MUC1 expression.

TOLERANCE IN MUC1 TRANSGENIC MICE

:.:.:‘@•@•:‘@@

Pancreas@@ . . @.@ .

. . ,

. 4- 1.1

S

PeptideBlocked

@@ Tl,,_:

‘—S@@

@ ,‘@ .@147@‘•“@@4I'°•% @..@.: , -:@@-@@ ,@‘

:d1.@;.@@@ •@:@ ;:k “..-@- ..@—, .... ‘ c .. -@- @(.

@@@ @T;

Lung

. , , @....

@.. p

‘ *4# 4. @••

.. ;i@

@.‘:‘@

S •@

.1@4

.5 .. - ,..

@. .p

, - . .. S

Ij@@:'@ .:‘@,@.,@@ ‘‘;;a'%,@ . d'. .., . . -@ ::-@ .@@@ @. . .‘:.@ •i:

, @:çp@ . . .@ - . .@@ f\.@ ‘1.@l@@@ ‘@ ,@

@34 ‘@@i@ bc::@ @:@@‘:@@@ , ,‘‘,°4@,i,,,etFig.

1.MUC1expressionintissuesofthe MUC1.Tgmouse.TissueswerestainedwithCT1antiserum,reactivewiththeMUCIcytoplasmictall(A.C,andE)todemonstrateMUCIexpressionor with peptide-blocked CF1 (B, D, and F) as a negative control. Specific staining was observed on the lumenal surface of bronchial epithelium in the lungs (A), theaciniin

the pancreas (C), and the lumenal surface of mammary epithelial cells (E). Staining was not observed in peptide-blocked control sections (B, D, and F).

collecting ducts), gall bladder, salivary gland, stomach, and uterus(datanotshown).A similarpatternof stainingwasobservedusingthemurine monoclonal antibody HMFG-2, although background staining

was higher. The MUCJ transgene was not expressed in tissues that did

not normally express Muc 1, such as colon, heart, liver, muscle, andspleen (Table 1).

Transfection of the B16 Murine Melanoma Cell Line. Bl6 is awell-characterized murine melanoma cell line derived from theC57B1/6 mouse strain (24). C57B1/6 mice developed progressivelygrowing tumors after s.c. injection with Bl6 cells (24). An epitopetagged MUCI (MUC1F) cDNA was subcloned into the BamHI site ofthe expression vector pHf3APr-1-neo, which has previously been usedfor establishing stable expression of MUC1 in a number of human androdent cell lines (26), and transfected into the B 16 cell line. Severalclonal lines expressing MUC1F were derived. The immunoblot shownin Fig. 2 demonstrates MUC1F expression in three B16 clones,designated Bl6.2, Bl6.5, and B16.9. Flow cytometric analysis wasperformed to determine MUC1F expression on cells of the B 16.2,B16.5, and B16.9 transfectant lines. B16.2, B16.5, and Bl6.9 linesexpressed detectable levels of MUC1F on 80—92%of cells (data notshown). Control Bl6 transfectants were produced by transfecting Bl6cells with the pH@APr-1-neo vector alone.

MUC1.Tg Mice Develop Progressive MUC1-positive Tumors;MUC1-negative Variants Develop In wt Mice. To determinewhether endogenous MUC1 expression would affect the growth rateof MUC1-expressing tumors, groups of seven wt and MUCI.Tg mice

317

aMUCI

Mam.Gland

on June 5, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TOLERANCE IN MUCI TRANSGENIC MICE

shown). Interestingly, substantial differences in humoral responseswere observed when mice were immunized with the 105-amino acidpeptide comprising 5.25 MUC1 TR (105-mer). wt mice producedstrong antibody responses to the MUC1 TR in 1gM and all IgGisotypes (Fig. 6). In contrast, MUC1.Tg mice exhibited a strong 1gMresponse but only weak IgG subtype responses (Fig. 6). The differences in humoral responses of the wt and MUC1 .Tg mice to thel05-mer peptide were highly significant (P < 0.05).

DISCUSSION

The mechanisms that regulate immunological tolerance to tumorantigens are formidable obstacles that prevent effective tumor immunotherapy. An inbred C57B1/6 MUC1 .Tg mouse has been developedto facilitate study of the parameters that regulate immunologicaltolerance to the tumor-associated antigen MUC1. This experimentalmodel will enable us to study the effect of endogenous expression ofthe MUCJ gene on the ability of mice to produce protective immuneresponses to tumors, and it represents an improved model system forevaluating the efficacy of anti-MUC1 vaccine formulations in vivowithin the context of existing tolerance mechanisms.

Immunohistochemical analysis demonstrated that the 10-kb SacIIgenomic DNA fragment resulted in reliable expression of the MUCJgene in tissues that normally express MUC 1; no expression was seenin ectopic sites. This finding is consistent with previous results usingthis construct in outbred mice, in which the correct temporal andspatial pattern of expression were conferred by the transgene (23).

318

M2

lrt O@ C@1 @1) O@@,o @1á@@ ‘@o @.o @‘c ‘@o— — — — — — — —

@ m@@@@@

were injected s.c. with 2 X iO@MUC1F-expressing Bl6.9 cells (Fig.3A). MUC1.Tg mice developed progressively growing tumors with anaverage volume of 1.31 ±1.05 cm3 at day 26 postinjection. Conversely, tumor growth in wt mice was significantly delayed(P < 0.05); only one of seven wt mice had a tumor at day 26postinjection. By day 34 postinjection, five of seven wt mice developed tumors with an average volume of 0.1 1 ±0.06 cm3. Similarresults were obtained with Bl6.2 cells (data not shown). Groups of 7MUC1.Tg and wt mice were injected s.c. with control B16 transfectants, and both strains developed tumors at approximately the samerate (Fig. 3B). Similar results were observed when wt and MUC1.Tgmice were injected with 4 X iO@cells per mouse.

Expression of MUC1 in tumors from MUCJF-transfected Bl6 cellswas compared in wt and MUC 1.Tg mice. Tissue and tumor sampleswere obtained for immunohistochemical analysis 4.5—6weeks afterinjection. Tumors from MUC1.Tg mice (n = 10) were heterogeneously positive for MUC1 expression by immunohistochemistry(Fig. 4A). In contrast, tumors from wt mice (n = 12) were consistentlyMUC1-negative as determined by immunohistochemical analysis(Fig. 4C). These results suggested that MUC1-expressing tumor cellswere eliminated in the wt mice and that MUC1 negative variants werenot eliminated and eventually progressed in wt mice.

Humoral Immunity in MUC1.Tg Mice. Sera from wt and transgenic mice challenged with MUC1-expressing tumors were tested forantibodies that bind to a 40-residue synthetic peptide, correspondingto two copies of the TR (Fig. 5). Five of seven wt mice developedstrong MUC1-specific 1gM and IgG1 responses, whereas none of theMUC1.Tg mice developed an 1gM response, and only one of theMUC1.Tg mice developed an IgGl antibody response. Similarly,three of seven wt mice produced a strong IgG3 antibody response,whereas zero of seven MUC1 .Tg produced this response. MUC1-reactive IgG2a or IgG2b antibodies were not detected in any mice(data not shown). Antibody responses to control antigens (KeyholeLimpet Hemocyanin and other peptides) were similar among thecontrol and MUC1.Tg animals (data not shown).

To investigate the immune response of MUC1 .Tg mice to selfpeptides derived from the TR region of the MUC1 protein, wt andMUC1 .Tg mice were immunized with peptides corresponding to 1.2copies (24 amino acids) or 5.25 copies ( 105 amino acids) of theMUC1 TR. Immunization of mice with a synthetic peptide corresponding to 1.2 copies of the MUC1 TR (TAPPA) resulted in a weak1gM and no IgG antibody response in both wt and Tg mice (data not

@@p• iØ@ 250 kDa

@4—205 kDa

Fig. 2. MUCIF expression in three Bl6 melanoma clones, Bl6.2, Bl6.5, and B16.9.Equal quantities of cell lysates from parental cells and MUC1F transfectants wereseparated on 6% denaturing polyacrylamide gels, transferred to polyvinylidene difluoridemembranes, and tested for reactivity with MUC1F (M2 antibody) or an isotype controlantibody. The transfectants expressed large quantities of MUCIF. as demonstrated by M2reactivity with an antigen of greater than Mr 200,000. Specificity was confirmed by thelack of reactivity in isotype control lanes.

30* 34*

Days

A

flimor volume(cmi)

B

Thmor volume(cin@)

Days

Fig. 3. Rates of tumor growth in transgenic and wt mice. wt (EJ) and MUCI.Tg (0)C57BI/6 mice were injected with 2 X l0@MUC1-expressing Bl6 murine melanoma cells(A) and 2 X l0@°Bl6 control transfectants (B). a, statistically significant difference intumor volume (P < 0.05) as determined by the two-sample t test. The results arerepresentative of six independent experiments. Bars, SEM.

Isotype control

on June 5, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TOLERANCE IN MUCI TRANSGENIC MICE

f@;.;J@i:@ #..%#3@@ @@a,@ @PI@$@ P-.' @‘q

@ La@ •;@?s4al@c3 brk44@%,@L@@ S

@@@@@ ‘S;:@@@

@@ t@a@:: ‘@4r@ r!@t@A@%i

- -. 4.b@@ .‘-:;@@@ —p..

@4―f,i.:4@ 5,....:@P'.@@ ,S C $@, j@ #;i. •@h,4,: br@@@

@ @.I :.. @-@‘@@;: @.

.; . ,.‘@â€@- -â€tt;,@ 5@ ‘@‘@‘:@@ ..@ ‘@‘•@‘1@.@‘@:; J

:-@@@ @, @‘:@. ..@@ £@@@@@@@ .. - @-

.‘v..@@ ;q' ...i;@aè' @% .@ @•;‘@.@ .) .:@ @‘@ _ ‘@I@

—.@ @, @@,7L-@ .@ •@ , @‘L@i'@@ ‘:¼1―@@. .. L' @‘ ., *e @.

4.. ,@ . .@@ ‘,@ ‘.@@@@@@@ â€,@@@

@ :@ ‘@@ ; p :@@ .@:T: ,@

@!@@@ ,@., .@ ‘r .@ @:‘°@@ . ;@‘C@L@ ;@ $

@@@ .@ t#,'. —. .

@ ;ø :@ e'.@ .;

@ *,@@@ @@*c'@-@••@:a.@@@@@ .b_t.•.,.@ ?@

@ ;@, ‘* ;.@@ #- C@

‘-@ c@@ p@@ @, @1@t1@ &@L'@@@'@@ ‘@@@ .@.v @::@@

I . @,4,.@@ *@. @.

. -@ 4:@@ . & , . p@ @,@@ ‘@;

I@'1 l@@ •@ @:@ :@ @@:;1Rd:4t@@;\t::.i$4o -

* ç@1@ 4 .t,1 40; @Q eP;.IId@;U@@@ ..@ ..‘ ‘S @I@4@. ft# i . 0 “@ -a b@@ 0 .@ .,@@@@@ % LF V P •â€0 ‘@ @†‘: ..@@ •. fl@@

4@ *4@ )@ __l@.@44@ ?@ •@ , 4@ , @2'@6'@@ 1:@a@

@@ c':@@@ @@&‘.:@4@; r@@ :1..@@@

a @H*@ . 4S@@@ ‘ •• • .,‘‘‘@

:;A@ ;@@;4@@@ ,@ .

_@ft @4 @,‘.,@‘@;:@@@@ (:.;:tQ @‘31w@:.@@ . ;@%.@ s:@;4:@., @‘t .@ F@,'41.@t@ 6:@q@Id.1@*s@@

.@@@@ S •:: .@@ ‘:@ dç @@@:2' •(I :@@‘i4@@

Fig. 4. Immunohistochemical staining of MUCI-transfected Bl6 melanoma tumors from wt and MVCI.Tg mice. Serial sections of tumor tissue were stained for expression ofMUCIF with HMFG2 antibody, reactive with the MUC1 TR, or with an isotype control antibody as a negative controL Tumors from MUC1.Tg mice(A; n = 10) were heterogeneouslypositive for MUC1F expression by immunohistochemistry. In contrast, tumors from wt mice (C; n = 12) were consistently MUCIF negative as determined by immunohistochemicalanalysis. Specificity of the antibody staining was confirmed by the lack of staining in tumors from either group when stained with an isotype control antibody (B and D).

It was hypothesized that the endogenous expression of MUC1would result in immunological tolerance to MUC1. MUC1.Tg micechallenged with MUC1-expressing B16 melanoma cells developedprogressively growing MUC1-positive tumors. In contrast, wt micechallenged with MUC1-expressing B16 cells failed to produceMUC1-positive tumor cells; only MUC1-negative variants were detected. One interpretation of these results is that the wt mice develo_ a MUC1-specific immune response that effectively eliminatedthe MUC1-positive tumor cells. This interpretation is consistent withprevious studies that demonstrated immunological rejection ofMUC1-transfected tumor cells in wt mice (12—14).The developmentof MUC1-negative tumor variants observed in the wt mice may beexplained by the presence of 10—18%of cells that were MUC1-negative; these tumor cells were not eliminated by the anti-MUC1response. The MUC1-negative variants probably appeared in theMUC1-transfected B16 cell lines (although they were cloned following transfection) as result of down-regulation or inactivation of cxpression of the MUC1 transgene or promoter. The outgrowth of

MUC1-negative tumor variants in the wt mice underscores a fundamental problem in the development of tumor antigen-specific vaccines. Tumor populations display antigemc heterogeneity and retainthe capacity to alter antigen expression by different mechanisms. Theobserved loss of MUC1 expression reinforces the widely held beliefthat effective tumor vaccines will need to include multiple antigeniccomponents to completely eradicate the entire population of tumorcells.

Evidence that MUC1.Tg mice are tolerant to MUC1 is provided bythe finding that sera from MUC1.Tg mice challenged with MUC1-expressing tumors (with one exception) contained neither 1gM norIgG MUC1-reactive antibodies. In contrast, wt mice challenged withMUC1-expressing tumors produced approximately 10-30 p@g/mlofMUC1 TR-specific 1gM and IgO antibodies. Further evidence ofimmune tolerance is provided by the results of immunizing theMUC1.Tg animals with MUCI synthetic peptides. wt mice immunized with the 105-mer developed strong 1gM and IgG immuneresponses. In contrast, MUC1.Tg mice developed strong 1gM antibody responses but only weak IgG antibody responses. These resultsindicate that in MUC1.Tg mice, the B-cell compartment does notundergo robust antibody class switching when challenged with the

MUC1 TR 105-mer, which suggests that the T-cell compartment isaffected by the endogenous expression of MUC1. Immunologicaltolerance to transgene products has been shown in several othertransgenic model systems (31, 32). For example, transgenic miceexpressing the soluble antigen hen egg lysozyme (HEL) showedimmunological tolerance at the level of T cells; low-level expressionof HEL resulted in T-cell tolerance, whereas the induction of B-celltolerance required high levels of HEL expression (31). Studies that

evaluate the status of the MUC1.Tg T-cell compartment are in pro

gross.An alternateexplanation for a lack of detectable circulating anti

bodies in the MUC1.Tg mice is that MUC1-specific antibodies maycomplex with endogenously expressed MUC1. The likelihood of this

319

on June 5, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TOLERANCEIN MUC1TRANSOENICMICE

is decreased by the fact that MUC1 is expressed at the lumenal orapical membrane of epithelial cells, a site of limited antibody access.Within the limits of immunoperoxidase detection, the MUC1.Tganimals showed no evidence of antibody deposition in the kidneys orother organ sites. Moreover, no inflammatory cell infiltrates wereobserved in any organs, except at sites of tumor growth. Recentstudies suggest that the inability of B cells to undergo class switchingcan be overcome by immunizing mice with the TAPPA peptideconjugated to a very strong carrier protein (33). Using this immunogen, both wt and MUC1.Tg mice could be shown to produce strong1gM and IgG antibody responses to the TAPPA portion of the molecule (33). The observation that sera obtained from the MUC1.Tgmice immunized with TR-KLH produce detectable 1gM and IgGMUC1-specific antibodies provides evidence that circulating MUC1-reactive antibodies in MUC1.Tg mice can be detected by conventional

solid-phase ELISA techniques.

I-

I

14000@

12000•

10000@

0:@ 8000

I-

6000•

4000@

200O@

I100

7525

20

15

10

5

- gM ‘ IgGi ‘ lgG2a ‘ lgG2b ‘ lgG3

Fig. 6. Humoral anti-MUC1 immune responses in MUCI.Tg and wt mice immunizedwith a 105-residue MUC1 TR peptide. One week following the third injection withMUCI-l05-merpeptide,seraofwt mice(n = 5)exhibitedhigh1gMandIgGanti-MUCI105-mar titers. In contrast, MUC1.Tg mice (n = 5) developed high 1gM titers but onlyweak IgG titers 1 week following the third injection of MUC1-lO5-mer. IgG responseswere significantly different between wt and MUC1.Tg mice (P < 0.05) as detennined bythe two-sample : test. Titer is expressed as the reciprocal of the dilution of sera thatresulted in A@5@ readings that were 2.1 times greater than those observed in nonimmunized mice. *, statistically significant difference in antibody titer (P < 0.05) asdetermined by the two-sample : test.

A

U

WT MUC1 Tg

U

In summary, we have produced MUC1.Tg mice that express thehuman MUCJ transgene in a pattern consistent with that observed inhumans. The findings that MUC1.Tg mice are unable to reject tumorsthat express the MUC1 protein and that B lymphocytes from thesemice fail to class switch in response to immunization with MUC1peptides provide indirect evidence that MUC1.Tg mice are tolerant tostimulation by MUC1 at the level of the T cell. Thus, the MUC1.Tgmice provide a useful model for investigating immunity and mechanisms of central and peripheral tolerance to the tumor antigen MUC1.This model also provides opportunities to evaluate anti-MUC1 vaccine formulations on the background of existing tolerogenic mechanisms and to evaluate the autoimmune consequences of producinganti-MUC1 immune responses.

ACKNOWLEDGMENTS

We acknowledgethe expertise of SureshSavarirayanin embryo injectionand implantation, Antia Jennings for expert tissue processing and sectioning,and Amy Weaver for statistical assistance. We thank Dr. Olia Finn for thegenerous gift of the MUC1—l05-merpeptide. We acknowledge the graphicalexpertise of Julie Jenson and Marvin Ruona.

WT MUC1 Tg

U

A

IgGi(@ig/mlof sore)

B25

20

IgG3@(j.tgfml of sore)

10

5,

0'

C

1gM(pg/mI of sore)

Fig. 5. Humoral immune responses of MUC1.Tg and wt mice injected with MUC1-expressing Bl6 melanoma cells. Sees from MUC1.Tg and wt mice primed with 2 X 10@MUC1-expressing Bl6 murine melanoma cells were analyzed at 26-34 days postinjectionfor levelsof MUC1reactiveantibodiesbyELISA.wi micegeneratedstronganti-MUCI,IgGI (5 of 7; A), Ig03 (3 of 7; B), and 1gM (5 of 7; C) antibody responses. In contrast,MUC1.Tgmicefailedtodemonstrate1gMorIgG3antibodyresponsestoMUC1,andonlyone mousedemonstratedan lgOl response.The resultswere representativeof fourindependentexperiments.

320

MUCI.Tg

LI “@‘@Type

111

REFERENCES

1. Gendler, S. J., Lancaster, C. A., Taylor-Papadimitriou, J., Duhig, T., Peat, N.,Burchell. J., Pemberton, L., Lalani, E. N., and Wilson, D. Molecular cloning andexpression of human tumor-associated polymorphic epithelial mucin. 3. Biol. Chem.,265:15286—15293,1990.

2. Gendler, S. 3., Taylor-Papadimitriou, 3., Duhig, T., ROthbard, J., and Burchdll, 3. Ahighlyimmunogenicregionof a humanpolymorphicepithelialmucinexpressedbycarcinomas is made up of tandem repeats. J. Biol. Chem., 263: 12820—12823,1988.

3. Lan, M. S., Batra, S. K., Qi, W-N., Metzgar, R. S., and Hollingswonh, M. A. Cloningand sequencing of a human pancreatic apomucin cDNA. J. Biol. Chem., 265:15294—15299,1990.

4. Gendler, S. J., and Spicer, A. P. Epithelial mucin genes. Annu. Rev. Physiol., 57:607—634,1995.

5. Barnd, D. L., Lan, M. S., Metzgar, R. S.. and Finn, 0. J. Specific, major histocom

on June 5, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TOLERANCE IN MUC1 TRANSOENIC MICE

patibility complex-unrestricted recognition of turner-associated mucins by humancytotoxic T cells. Proc. Nail. Aced. Sci. USA, 86: 7159—7163, 1989.

6. Jerome,K.R.,Domenech,N.,andFinn,0. J. Tumor-specificcytotoxicT cellclonesfrompatientswithbreastandpancreaticadenocarcinomarecognizeEBV-immortalized B cells transfected with polymorphic epithelial mucin complementary DNA. J.Immunol., 151: 1654—1662,1993.

7. Kotera, Y., Fontenot, J. D., Pecher, G., Metzgar, V., and Finn, 0. J. Humoralimmunity against a tandem repeat epitope of human MUC1 in sera from breast,pancreas and colon cancer patients. Cancer Res., 54: 2856—2860,1994.

8. Taylor-Papadimitriou, J. Report on the first international workshop on carcinomaassociated mucins. hit. 3. Cancer, 49: 1-5, 1991.

9. Xing, P. X., Prenzoska,J., Quelch, K., and McKenzie, I. F. Second generationanti-MUC1 peptide monoclonal antibodies. Cancer Res., 52: 2310-2317, 1992.

10. Thang,S., Graeber,L A., Helling, V., Ragupathi,V.. AdIUri,S.. Lloyd, K. 0., andLivingston, P. 0. Augmenting the immunogenicity of synthetic MUCI peptidevaccines in mice. Cancer Rca., 56: 3315—3319,1996.

11. Spicer, A. P., Parry, G., Patton, S., and Gendler, S. J. Molecular cloning and analysisofthemousehomologueof thetumor-associatedmucin,MUCI,revealsconservationof potentialO-glycosylationsites,transmembrane,and cytoplasmicdomainsand aloss of minisatellite-like polymorphism. 3. Biol. Chem., 266: 15099-15109, 1991.

12. Apostolopoulos,V.,Xing,P.X.,andMcKenzie,I. F.C. Murineimmuneresponsetocells transfected with human MUCI: immunization with cellular and syntheticantigens. Cancer Rca., 54: 5186-5193, 1994.

13. Graham,R.A.,Burchell,J. M.,Beverley,P.,andTaylor-Papadimitriou,J. Intramuscular immunizationwith MUC1eDNA can protectC57 mice challengedwithMUCI-expressing syngeneic mouse tumour cells. hit. J. Cancer, 65: 664-670, 1996.

14. Lalani, E. N., Berdichevsky,F., Boshell, V., Shearer,M., Wilson, D., Status, H.,Gendler, S. I., and Taylor-Papadimitriou, 3. Expression of the gene coding for ahuman mucin in mouse mammaiy tumor cells can affect their tumorigenicity. 3. Biol.Chem., 266: 15420—15426.1991.

15. Disis,M.L.,andCheever,M.A. Oncogenicproteinsas tumorantigens.Curr.Opin.Immunol., 8: 637—642,1996.

16. Robbins,P. F., andKawakami,Y. Humantumorantigensrecognizedby Tcells. Curr.Ophs. Immunol., 8: 628—636,1996.

17. Jones, L A., Chin, L T., Longo, D. L, and KrUisbeek, A. D. Peripheral clonalelimination of functional T cells. Science (Washington DC), 250: 1726—1729,1990.

18. Kappler,J. W., Roehm,N., andMarrack,P. 1 cell toleranceby clonal eliminationinthe thymus.Cell,49: 273—280,1987.

19. Mondino,A., Khonits,A.,andJenkins,M. K.Theanatomyof T-cellactivationandtolerance. Proc. NatI. Aced. Sci. USA, 93: 2245—2252,1996.

20. Goodnow,C. C., Crosbie,J., Adelstein,S., Lavoie,T. B., Smith-Gill,S. J., Brink,R. A., Pritchard-Btiscoe,H.. Wotherspoon,J. S., Loblay, R. H., Raphael,K.. Trent,R. J., and Basten, A. Altered immunoglobulinexpression and functionalsilencing

of self-reactiveB lymphocytesin transgenicmice.Nature(Lond.),334: 676—682,1988.

21. Nemazee, D. A., and Burki, K. Clonal deletion of B lymphocytes in a transgenicmouse bearing anti-MHC class I antibody genes. Nature (Loud.), 337: 562-566,1989.

22. Yule, 1. D., Basten. A., and Allen, P. M. Hen egg-white lysozyme-specific T cellselicited in hen egg-white lysozyme-transgenic mice retain an imprint of self-tolerance. J. Immunol., 151: 3057—3069,1993.

23. Peat, N., Gendler, S. i., Lalani, E-N., Duhig, T., and Taylor-Papadimitriou, J.Tissue-specificexpressionof a humanpolymorphicepithelialmucin(MUC1)intransgenic mice. Cancer Res., 52: 1954-1960. 1992.

24. Talmadge,J. E.,Talmadge,C.B..Thar,B.,McEwen,R.,Meeker,A.K.,andTribble,H. in vivo immunologicselection of class I majorhistocompatibilitycomplex genedeletion variants from the B16-BL6 melanoma. J. NatI. Cancer Inst., 78: 1215—1221,1987.

25. Gunning, P., Leavitt, J., Muscat, G., Ng, S-Y., and Kedes, L. A human @-actinexpressionvectorsystemdirectshigh-levelaccumulationof antisensetranacripis.Proc. NatI. Acad. Sci. USA, 84: 4831-4835, 1987.

26. Burdick, M. D., Harris, A., Reid, C. 3., Iwamura, T., and Hollingsworth, M. A.Oligosaccharides expressed on MUC1 produced by pancreatic and colon tumor celllines. 3. Biol. Chem., 272: 24198-24202, 1997.

27. Pemberton, L, Taylor-Papadimitriou, J., and Gendler, S. J. Antibodies to the cytoplasmic domain of the MUC1 mucin show conservation throughout mammals.Biochem.Biophys.Res.Commun.,185: 167—175,1992.

28. Fontenot, J. D., Tjandra, N., Bu, D., Ho, C., Montelaro, R. C., and Finn, 0. J.Biophysical characterization ofone-, two-, and three-tandem repeats ofhuman mucin(MUC1) protein core. Cancer Res., 53: 5386—5394,1993.

29. Fontenot,J. D.,Finn,0. J.,Dales,N.,Andrews,P.C.,andMontelaro,R.C.Synthesisof largemulti-determinateimmunogensusingthe poly-proline13-turnhelixmotif.Pept. Rca., 6: 330—336, 1993.

30. Spicer, A. P., Rowse, G. J., Lidner, T. K., and Gendler, S. J. Delayed mammazytumorprogression in Mud null mice. J. Biol. Chem., 270: 30093—30101, 1995.

31. Adelstein,S., Pritchard-Btiscoe,H., Anderson,T. A., Crosbie,3., Gammon,G.Loblay, R. H., Basten, A., and Goodnow, C. C. Induction of self-tolerance in T cellsbutnotB cellsof transgenicmiceexpressinglittleselfantigen.Science(WashingtonDC), 251: 1223—1225,1991.

32. Milich, D. R., Jones, J. E., Hughes, J. L., Maruyama, T., Price, J., Melhado, I., andJirik, F. Extrathymic expression ofthe intracellular hepatitis B core antigen results inT cell tolerancein transgenicmice.J. Immunol.,152:455—466,1994.

33. Rowse, G. J., Tempero, R., Hollingsworth, M. A., and Gendler, S. J. MUCI peptideimmunogens produce different antibody responses in native and MUCI transgemcmice. in: J. Taylor Papadimitriou (ed), 4th International Workshop on CarcinomaAssociated Mucins, p. 83. Cambridge, UK: imperial Cancer Research Fund, 1996.

321

on June 5, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1998;58:315-321. Cancer Res   Gerald J. Rowse, Richard M. Tempero, Michelle L. VanLith, et al.   Murine ModelTolerance and Immunity to MUC1 in a Human MUC1 Transgenic

  Updated version

  http://cancerres.aacrjournals.org/content/58/2/315

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/58/2/315To request permission to re-use all or part of this article, use this link

on June 5, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from