[CANCER RESEARCH 58. 1469-1477, April 1. 1998]

Mice Transgenic for Human Carcinoembryonic Antigen as aModel for Immunotherapy1

Patrick Clarke, Jeffrey Mann, Jean F. Simpson, Karen Rickard-Dickson, and F. James Primus2

Divisions of Immunology [P. C., K. R-D.. F. J. P.] and Biology ¡J.M.¡,Beckman Research Institute of the City of Hope, and Division of Pathnli>/>\ ¡J.F. S.¡. dry iif Hope

National Medical Center. Duarte, California 91010

ABSTRACT

Mice transgenic for the human Carcinoembryonic antigen (CEA) genewere prepared for use as a preclinical model for immunotherapy. A32.6-kb fragment containing the complete human CEA gene and flanking

sequences was isolated from a genomic cosmid clone and used to producetransgenic C57BL/6 mice. A homozygous line was established that wasdesignated C57BL/6J-TgN(CE4Ge)18FJP. Southern blot analysis showed

that this line contained intact copies of the cosmid clone, with approximately 19 integrated copies at one chromosomal location. A mouse-humanchimeric anti-CEA monoclonal antibody was used to examine CEA ex

pression by immunohistochemical staining of frozen tissue sections. In thececum and colon, approximately 20% of the luminal epithelial cells hadstrong cytoplasmic staining, whereas occasional glands showed intensestaining. CEA was also expressed in gastric foveolar cells, whereas smallintestine villi had only a few (<1%) positive cells. CEA was not found byimmunohistochemistry in other tissues of the digestive tract, nor was itfound in a wide range of other tissues or organs. Concordance in resultswas obtained between immunohistochemistry and analysis of tissue extracts by enzyme immunoassay. The lone exception was the testis, whichwas positive only by enzyme immunoassay. Expression of human CEAwas not observed in tissues derived from nontransgenic mice. The fecalcontent of CEA in transgenic mice was approximately 100-fold less thanthat observed for humans. Circulating CEA was not detected. A CEA-transfected syngeneic murine colon carcinoma cell line, MC-38, was pre

pared that had stable expression of CEA in vitro and in vivo. The molecular size of CEA produced by CEA-transfected MC-38 cells and by the

colon of transgenic mice was similar to that obtained with CEA purifiedfrom human colon tumors. Anti-CEA antibody appeared in nontransgenicbut not transgenic mice bearing transfected MC-38 tumors. These find

ings demonstrate that CEA distribution and its properties in tissues ofmice transgenic for the human CEA gene are similar to that observed inhuman tissues. As in humans, immune responsiveness to CEA, as reflectedby antibody formation, was not detectable in transgenic mice bearingCEA-positive tumors. Thus, CEA transgenic mice may serve as a useful

model for studying the efficacy and safety of various immunotherapystrategies directed at this tumor self-antigen.

INTRODUCTION

Self-antigens expressed on tumor cells have been implicated astargets for human B- and T-cell responses against autologous cancer(1-5). Several of these antigens are under clinical study, in an effortto specifically boost antitumor reactions. CEA3 is abundantly ex

pressed on a variety of human carcinomas, although its presence innormal tissues is mainly limited to the large intestine (6, 7). In theimmediate years after CEA was discovered, numerous reports appeared concerning CEA's immunogenicity or lack thereof in humans

(8-11). Anti-CEA activity was even found in sera of pregnant women

Received 8/19/97; accepted 2/2/98.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 with18 U.S.C. Section 1734 solely to indicate this fact.

' This work was supported by NIH Grants CA58327 and CA70320. Additional support

was provided by NIH Cancer Core Grant CA33572 to the City of Hope.2 To whom requests for reprints should be addressed, at Vanderbilt University Medical

Center, Department of Pathology, C-3321 MCN, Nashville, TN 37232.3 The abbreviations used are: CEA, Carcinoembryonic antigen; FISH, fluorescence in

situ hybridization; MAb, monoclonal antibody; RT, room temperature; TBS. 0.1 MTris-HCl (pH 7.5)-0.9% NaCl; PI-PLC. phosphatidylinositol phospholipase C.

and was thought to arise from sensitization by antigen expressed bythe fetus (12). However, the conclusion from these early studies wasthat humans were tolerant to CEA, and when reactivity was detected,this appeared to be associated with such factors as antibodies reactingwith blood group determinants on CEA, contaminants in the CEApreparations, and nonspecific reactions. Thus, the immune repertoirefor CEA in humans appears truncated because the liberation of antigen during malignant spread does not cause an appreciable change inresponsiveness.

The issue of immune responsiveness to CEA has recently beenrevisited. In vitro immunization of peripheral blood lymphocytes fromhealthy individuals with an anti-CEA idiotype antibody was found toinduce the appearance of anti-CEA antibodies (13). Similar results

were obtained when human tumor infiltrating lymphocytes or peripheral blood T cells from colorectal cancer patients were exposed to theanti-idiotype antibody. Human anti-CEA antibodies have also been

derived from hybridomas that used draining lymph node cells fromcolorectal cancer patients as fusion partners (14).

Three recent clinical trials in cancer patients were directed atinducing anti-CEA responses. Foon et al. (15) found that treatmentwith an anti-CEA idiotype antibody produced both high titer antibodies against CEA and anti-idiotype-specific lymphoproliferative re

sponses. Two other studies have used a recombinant CEA vacciniavaccine. In one, HLA-restricted anti-CEA CTLs were identified fol

lowing repeated restimulation of peripheral blood lymphocytes withpeptide-pulsed autologous antigen-presenting cells (16). In the other,

peripheral blood lymphocytes from two patients showed a significantlymphoproliferative response to CEA (17). Thus, mounting evidencesuggests that CEA, as a self-antigen, may be a target for eliciting

autologous antitumor reactivity.A preclinical animal model is desirable for the study of vaccines

designed to elicit CEA reactivity, as well as to identify any adverseautoimmune effects that may accompany the latter. However, no trueCEA homologue has been identified in mice or rats to date ( 18). Asan alternative, we have prepared mice that are transgenic for thehuman CEA gene, and we describe herein their preparation, the tissuedistribution properties of CEA in these mice, and the development ofa syngeneic CEA-expressing tumor system.

MATERIALS AND METHODS

CEA Genomic DNA Isolation. CEA genomic DNA was isolated from thecosmid cosCEAl (19) provided by Dr. John Shively (City of Hope NationalMedical Center). cosCEAl was digested with Aatll (New England Biolabs,Beverly, MA), and the 32.6-kb fragment containing the entire CEA genomic

insert along with the flanking cosmid sequences was isolated by sucrosegradient centrifugation, as described (20). Isolated DNA was exchangedagainst 10 mM Tris-0.1 mM EDTA (pH 7.4), using a Centricon 10 filter

(Amicon Inc.. Beverly, MA). The DNA was concentrated to 2.5 j^g/ml andused directly for microinjection into the male pronucleus of C57BL/6J zygotes(The Jackson Laboratory, Bar Harbor, ME), as described (21). All animalstudies were carried out under approval of the Institutional Research AnimalCare Committee.

PCR Analysis. The initial screening of founder mice for the CEA trans-

gene was carried out by PCR analysis on mouse tail DNA (22). Tail samples(1-3 mm) were digested overnight at 55°Cwith proteinase K (0.7 mg/ml) in

1469

Research. on December 18, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

CEA TRANSOENIC MICE

0.1 M Tris (pH 8.0), containing 0.01 M EDTA, 0.5% Nonidet P-40, and 0.5%

Tween 80. Digests were extracted with phenol and precipitated with ethanolprior to PCR. PCR used 50 ng of tail sample DNA and AmpliTaq DNApolymerase (Perkin-Elmer, Branchburg, NJ) in standard reactions using 1.5mM MgCU. The sense strand primer (TCCCCTGGCAGAGGCTCCTGCT-CAC) and antisense strand primer (GCTAGGAGAGAGCAGAGAG-CATCGG) corresponded to the CEA genomic exon 1 (nucleotides 1806-1830) and CEA intron 1 (nucleotides 2725-2701), respectively (19). PCRs

using these primers generated a fragment of 920 bp. The size of the CEAcDNA copy in CEA-transfected cells was determined by PCR using as prim

ers: sense strand, CAGATGGTGCATCCCCTGGCAGAGG; and antisensestrand, CCTTAATTAATGCTATATCAGAGCAACCCCAACC. Oligonu-

cleotides were synthesized by the City of Hope DNA synthesis facility usingan Applied Biosystems 394 DNA RNA Synthesizer.

Southern Analysis. Transgenic mouse genomic DNA for Southern analysis was isolated from liver samples as described (20). Normal C57BL mousegenomic DNA was purchased from The Jackson Laboratory. DNAs wererestricted with Sad and then fractionated on 0.8% agarose gels using aTris-acetate buffering system. Samples were transferred to Zeta-probe nylonfilters (Bio-Rad, Hercules, CA). The filters were probed using the CEA Aatllgenomic fragment described above, labeled with 12P (ICN, Irvine, CA) by the

random primer method (20). Hybridizations were carried out in 1 mM EDTA,0.25 M Na2HPO4 (pH 7.2), and 7% SDS at 65°Cfor 24 h. Filters were washedtwice in 1 mM EDTA, 0.04 M Na,HPO4 (pH 7.2), and 5% SDS at 65°Cfor l h

and then twice in 1 mM EDTA, 0.04 M Na2HPO4 (pH 7.2), and 1% SDS at65°Cfor 1 h. Filters were then autoradiographed.

To assess the CEA transgene copy number. Southern blots were run inparallel with C57BL mouse DNA spiked with known amounts of cosCEAlDNA. For this experiment, a reference standard was created by adding serialdilutions of cosCEAl DNA to normal C57BL mouse DNA. This standardDNA was then restricted with Sad. Southern analysis was then carried out asdescribed above. Autoradiographs were scanned using a Bio-Rad GS-670

imaging densitometer. CEA transgene copy number was determined by com

paring the intensities of bands appearing in the transgenic mouse lanes withthose observed in the cosCEA containing standard lanes. Copy number wascalculated using 3.3 x IO9 bp per haploid mouse genome as a reference (23).

FISH. The purified 32.6-kb fragment of cosCEAl was labeled with digoxi-genin-11-dUTP using a nick translation kit according to the manufacturer's

instructions (Boehringer Mannheim. Mannheim, Germany). Mononuclear cellswere isolated from mouse blood, fixed in acetone-methanol, and dropped onto

glass slides (24). The slides were treated with 0.25 mg/ml proteinase K in 2XSSC at 45°C,followed by incubation for 1 h at 37°Cin 2x SCC and then

2-min dehydrations in graded alcohols. The slides were treated with 70%formamide-2x SCC at 70°Cfor 2 min, dehydrated with graded alcohols, andthen hybridized with 15 ng of denatured (5 min at 70°C)probe in Hybrisol VI(Oncor, Gaithersburg, MD) for 18 h at 37°C.Detection and amplification werecarried out using FITC-digoxigenin according to the manufacturer's instruc

tions (Oncor). After counterstaining with propidium iodide-antifade (Oncor),

the slides were viewed through a Nikon Labophot II fluorescence microscopeequipped with a triple-band pass filter.

CEA Measurement by ELISA. A two-site MAb ELISA was used for the

detection and quantitation of CEA in samples. The T84.1 MAb (500 ng/well),which cross-reacts with CEA-related antigens (25, 26), was coated overnight at4°Cin 0.1 M sodium bicarbonate (pH 9.6) onto polystyrene microtiter plates.

Plates were then blocked for I h at RT with 0.5% BSA in coating buffer.Following incubation at 37°Cfor 1 h with 100 /j.1of test sample or standard,

plates were washed with PBS containing 0.05% Tween 20 and then exposed tobiotinylated T84.66 MAb. The CEA-specific T84.66 MAb (25, 26) wasbiotinylated with NHS-LC-Biotin (Pierce, Rockford, IL) at a 200:1 mole inputratio of biotin to antibody. After an additional 1 h incubation at 37°C,the plates

were washed, reacted with streptavidin-alkaline phosphatase conjugate (Jackson ImmunoResearch, West Grove, PA) for 1 h at 37°C,and then exposed to

disodium p-nitrophenyl phosphate (Sigma Chemical Co.. St. Louis, MO) in

10% (v/v) diethanolamine. The plates were read at 405 nm using a Bio-Rad

Microplate reader. Standard curves were generated with CEA purified fromhuman colonie adenocarcinomas (provided by Dr. John Shively) and calibratedin a commercial CEA assay (Hoffman-LaRoche, Nutley. NJ).

Supernatants from fecal and tissue extracts were used for CEA detectionand/or quantitation. Feces was dispersed in PBS containing 1.0% Triton X-100

(1:10, w/v), sonicated for 30 s, and centrifuged at 13,000 X g for 30 min.Tissue samples were homogenized in PBS containing 0.5% Nonidet P-40-2

mM phenylmethylsulfonyl fluoride (1:10, w/v) using a Polytron homogenizer(Brinkman, Westbury, NY). Following sonication for 30 s, the tissue homo-

genates were centrifuged as described above. Protein in fecal and tissueextracts was determined with bicinchoninic acid reagent (Pierce) using BSA asa standard. For detection of circulating CEA, 100 /j.1 of neat serum sampleswere assayed.

Immunohistochemistry. Acetone-fixed, 5-f^m cryostat tissue sections

were used to localize CEA in tissues following procedures described elsewhere(27). Sections were reacted with chimeric mouse-human T84.66 MAb (28),followed by biotin-labeled goat antihuman IgG (Vector Laboratories, Burl-

ingame, CA). Sites of antigen localization were disclosed using peroxidase-streptavidin conjugate (Vector) and diaminobenzidine. The primary anti-CEA

MAb was omitted on control sections.Western Blotting. Colons from nontransgenic and transgenic mice were

cleared of fecal material and then homogenized at 4°Cin 0.01 M Tris-HCl

buffer (pH 8.0), containing 0.5% Triton X-100, 1 mM phenylmethylsulfonyl

fluoride, 5 mM iodoacetamide, and 0.025% NaN,. Following centrifugaron at13,000 x g for 30 min at 4°C,the supernatant was precleared by incubation for

1 h at 4°Cwith a 1:1 (v/v) slurry of Sepharose 4B-normal mouse IgG adsorban!

(Pharmacia, Piscataway. NJ). The gel was removed from the latter mixture bycentrifugation at 200 X g for 5 min, and the precleared supernatant was mixedwith 1:1 slurry of Sepharose 4B-T84.66 immunoadsorbant. The immunoad-

sorbants were prepared by coupling 5 mg of normal mouse IgG or T84.66MAb per ml of cyanogen bromide-activated Sepharose 4B (Pharmacia). Incubation of the supernatant-immunoadsorbant mixture was carried out overnightat 4°Cwith continuous agitation. The latter mixture (100 /xl) was poured into

a small pipette column, the unbound fraction was eluted, and then boundmaterial was removed with 300 p.1 of SDS-PAGE sample buffer lacking

2-mercaptoethanol.

The bound fractions from the T84.66 immunoadsorbant were applied to a7.5% SDS-polyacrylamide gel after samples were heated at 100°Cfor 5 min.

After electrophoresis, transfer to nitrocellulose (Hybond; Amersham, Arlington Heights, IL) was carried out at 100 mA/gel for 50 min using a semidrytransfer cell (Bio-Rad) and 0.048 MTris, 0.039 glycine, and 20% methanol (pH9.2) buffer. After transfer, the nitrocellulose was blocked overnight at 4°Cwith

5% nonfat dry milk in TBS. The transfer paper was reacted with T84.66 MAb(0.5 /Ag/ml in 3% BSA-TBS) for 2 h at RT with agitation, followed by washing

in TBS containing 0.1% Tween 20. Peroxidase coupled goat antimouse IgG

(Pierce) in 3% BSA, TBS, and Tween 20 was allowed to react for 1 h at RT,washed, and then developed for 5 min with Pierce SuperSignal Substrateaccording to the manufacturer's instructions. The transfer paper was then

exposed to Biomax X-ray film (Eastman Kodak Company, Rochester, NY).

Transfections. The chemically induced colon adenocarcinoma cell line,MC-38, was obtained from Dr. Steven A. Rosenberg (National Cancer Institute, Bethesda. MD) and was maintained in DMEM high-glucose medium

supplemented with nonessential amino acids. 4 mM glutamine, 2 mM sodiumpyruvate, 100 units/ml penicillin, 100 ng/ml streptomycin (all obtained fromIrvine Scientific, Santa Ana. CA), and 10% fetal bovine serum (Gemini,Carlsbad, CA). MC-38 cells were transfected with Lipofectin (Life Technol

ogies, Inc., Gaithersburg, MD) and the CEA cDNA-containing plasmidpH/3Apr-l-neo provided by Dr. John Shively. Transfected cells were selected

with 500 ¿ig/mlG418 (Life Technologies), and CEA-expressing clones were

identified first by ELISA, followed by flow cytometry using T84.66 MAb forCEA detection. Flow cytometry analysis was carried out in the City of HopeFlow Cytometry and Sorting facility using a MoFlo (Cytomation, Ft. Collins,CO).

For PI-PLC digestion, cells were suspended in RPMI 1640 (Irvine Scien

tific) containing 0.1% BSA, 0.02 M EDTA, and 0.5 units/ml PI-PLC. ThePI-PLC obtained from Bacillus ¡hiiringiensis was provided by Dr. Donald

Diamond (Department of Bone Marrow Transplantation, City of Hope National Medical Center). Cells were incubated for 45 min at 37°Cwith agitation,

washed 3 times, and analyzed by flow cytometry.Growth of CEA-transfected MC-38 Cells in Vivo. CEA-expressing

MC-38 cells were released from culture flasks with trypsin-EDTA (IrvineScientific), washed twice in serum-free RPMI 1640, and 2.5 X IO5 cells were

implanted s.c. Tumor growth was monitored by caliber measurements in two

1470

Research. on December 18, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

CEA TRANSOENIC MICE

perpendicular dimensions, and tumor volume was determined according to theformula: kb 123

Tumor volume =(Short dimension)2 X (long dimension)

The expression of CEA on transfected MC-38 cells growing in vivo

was determined by flow cytometry. Tumors were excised and minced,and single-cell suspensions were prepared by digestion of fragments

at RT for 4 h with a mixture of 0.1 mg/ml hyaluronidase type IV, 1mg/ml collagenase type IV, and 0.1 mg/ml DNase Type IV (Sigma) inRPMI 1640 (29). Cells were placed in cell culture and propagated inthe absence of G418 for 8-13 days prior to analysis by flow cytom

etry.The presence of anti-CEA antibody in the serum of tumor-bearing

mice was measured by ELISA. Polystyrene microtiter wells werecoated with 200 ng of purified CEA in 0. l M sodium bicarbonate (pH9.6) overnight at 4°C. After blocking with 0.5% BSA, doubling

dilutions of serum, beginning at 1:100, were added to the wells, andthe plates were incubated at 37°Cfor 1 h. The plates were washed in

PBS containing 0.05% Tween 20 and then reacted with alkalinephosphatase-labeled goat antimouse IgG+IgM (Jackson ImmunoRe-search) for l h at 37°C. The plates were washed, reacted with

chromogen, and read in a microplate reader.

RESULTS

Establishment of CEA Transgenic Mouse. A 32.6-kb Aatll re

striction fragment that contains the entire human CEA genomic regionwas used to generate transgenic mice. This DNA fragment has beenshown to contain the regulatory sequences required for tissue-specific

CEA expression (19). Of 19 mice born, 4 were found to be positive forthe CEA gene, as determined by PCR analysis of DNA derived fromtail samples (data not shown). To determine whether any of thepositive mice expressed CEA, we took advantage of the observationthat CEA can be detected in human feces (30). Extracts of mouse fecalpellets were assayed for CEA in a two-site MAb ELISA. Three of the

four founder mice were found to contain immunoreactive CEA intheir feces. The A405nm for the three CEA-expressing mice rangedbetween 0.8 and 1.3, as compared to a range of 0.05-0.11 for fecesobtained from the nontransgenic littermates, as well as control non-transgenic mice. Of the three CEA-expressing founder mice, success

ful breeding was only obtained with one, and this mouse line has beendesignated C57BL/6J-TgN(C£4Ge)18FJP or line 18, as it will bereferred to hereafter. Using the 32.6-kb CEA genomic fragment forFISH analysis of interphase nuclei from peripheral blood, homozy-

gous mice were identified and bred. FISH also demonstrated integration of the CEA gene into one chromosomal site (data not shown).

Southern Blot and Copy Number Determination. Fig. 1 showsthe results of Southern analysis using transgenic and normal mouseDNAs. The Aatll CEA genomic fragment used to generate the transgenic mice contains the 13.0-, 9.0-, 4.8-, and 2.6-kb fragments. All of

the latter fragments were also contained within the transgenic DNAdigest, indicating that this line harbors an intact CEA gene. ThecosCEAl digestion gives rise to an additional 4.4-kb band, corresponding to CEA-cosmid junction fragments. The two minor bands at

3.7 and 5.1 kb obtained with the transgenic DNA digest are consistentwith head-to-tail or head-to-head junction fragments created by mul

tiple inserts. Determination of copy number revealed that line 18contains 19.2 ±2.9 (n = 3) copies of the CEA gene.

Tissue Distribution of CEA Expression. Immunoperoxidasestaining was carried out on acetone-fixed, frozen tissue sections usinga mouse/human chimeric anti-CEA MAb, identical to the antibody

3.7-

2.6- «•>*>

Fig. 1. Southern blot analysis of DNA from CEA transgenic mice. Genomic or plasmidDNAs were digested with Sad and then fractionated on a 0.8% agarose gel. Lane I,cosCEAl; Lane 2. transgenic DNA; Lane 3, normal C57BL DNA.

used by Eades-Perner et al. (31 ) for demonstration of CEA in their

CEA transgenic mice. With the latter MAb, it was possible to use anantihuman secondary antibody to avoid nonspecific background staining arising from endogenous mouse immunoglobulin in tissue sections that would result from the application of antimouse secondaryantibodies. In the cecum and large intestine, 10-20% of the luminal

epithelial cells had strong cytoplasmic and apical staining, whereasoccasional glands showed intense staining. An area from a transgeniccolon is depicted in Fig. 2 and shows a region in which the majorityof epithelial cells are CEA positive. By contrast, the colon from anontransgenic mouse did not display any positive staining (Fig. 2).The pattern of CEA localization in the transgenic colon is very similarto CEA staining that we have previously reported for morphologicallynormal human colon (32). Besides the cecum and colon, CEA stainingwas also observed in gastric foveolar cells and in a sporadic epithelialcell in the proximal portion of the small intestine. All other tissues(tongue, esophagus, middle and distal portion of the small intestine,liver, spleen, kidney, bladder, lung, thyroid, salivary gland, skin,testis, ovary, uterus, and cervix) were negative for CEA expression, asmeasured by immunohistochemistry.

1471

Research. on December 18, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

CEA TRANSGENIC MICE

:& »"V-'

-negative for CEA staining (data not shown), b,section obtained from nontransgenic mouse andstained with anti-CEA MAb as in a. Sections werelightly counterslained with hematoxylin. Finalmagnification. X300.

•¿�V>v5P 4'-

The expression of CEA was also measured by ELISA in extracts ofthe same tissues examined by immunohistochemistry. Tissues fromthree males and one female were used for this analysis. In thegastrointestinal tract, the highest level of CEA content was found inthe colon (25.9 ± 7.5 ng/mg protein), followed by stomach(12.2 ±5.9) and cecum (5.5 ±1.6). CEA was also found in the testis(9.1 ±2.8), which represented the only discrepancy between immunohistochemistry and ELISA results. CEA expression was not observed in the same tissues derived from nontransgenic mice whenexamined by ELISA.

The CEA content of feces and serum was compared between line18 mice and normal humans in ELISA. To enhance extraction of CEAfrom fecal samples, both detergent (1% Triton X-100) and sonication

were used. Although CEA was present in transgenic mouse feces, itslevel was approximately 100-fold less than that measured in feces

from normal humans (Table 1). CEA was not detected in the serum oftransgenic mice.

Molecular Size of CEA Expressed in Colonie Epithelial Cells ofLine 18 Transgenic Mice. Supernatants from homogenates of colonsobtained from either transgenic or nontransgenic mice were passedover an affinity gel containing T84.66 anti-CEA MAb after a pre-

Table I CEA conteníof feces and serum obtained from CEA transiente miceFecal samples were resuspended in PBS containing 1% Triton X-100 and then were

sonicated followed by centrifugation at 10,000 x #. Protein and CEA content of fecalsupernatants were quantitated in bicinchoninic acid assays and ELISA assays, respectively. The presence of circulating CEA was analyzed by ELISA using 100-/il serumsamples from mice.

Source Feces CEA (ng/mg of totalprotein)MouseTransgenic

NontransgenicHumanNormal11.7±4.0(n

= 4)01570±683(n

= 3)Serum

(ng/ml)<2.5

<2.5<2.5"

1Literature value (53).

1472

Research. on December 18, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

CEA TRANSOENIC MICE

A B12345 1234

kOa

I^^B

—¿�2095bp201—¿�

117—78—48.6—

33—24—

Fig. 3. Characterization of CEA and/or CEA gene in colons of transgenic mice orCEA-transfected MC-38 cells. A, analysis by Western blotting using chemiluminescenceof CEA in colon of CEA transgenic mice and CEA-expressing MC-38 cells. Colonextracts were adsorbed to an immunoadsorbant containing T84.66 anti-CEA MAb. andthen applied nonreduced to a 7.5% acrylamide gel for SDS-PAGE. Lane 1, CEA purified

from human colonie adenocarcinomas (40 ng); Lane 2, immunoadsorbant bound fractionfrom colon of nontransgenic mouse {20-/J.1 sample); Lime 3. immunoadsorbant boundfraction from colon of CEA transgenic mouse (20-^1 sample); Lane 4, extract of CEA-transfected MC-38 cells, clone C 15-4.3; Lane 5, extract of parental MC-38 cells. Blotswere stained with T84.66 MAb. ß.analysis by PCR of CEA DNA in clone C 15-4.3 cells.Lane I, length standard (mixture of A////idIII and <f>xl74Hae\\\Y. Lane 2. PCR product ofplasmid containing CEA; Lane 3. PCR product of parental MC-38 cells; Lane 4. PCRproduct of CEA-transfected MC-38 cells.

clearing step with a mouse IgG adsorban!. The gel was eluted withSDS-PAGE sample buffer, and the eluate was analyzed by Western

blotting. The same volume of material bound from either nontransgenic or transgenic extracts was applied to the SDS-polyacrylamide

gel. As shown in Fig. 3/4, a major band in the same size range as thatof CEA purified from human colonie tumors was observed with colonextracts from transgenic mice. A second intense band of slowermigration was also observed in transgenic colon extracts that mostlikely represented aggregates arising from sample processing (33).Similar band patterns were observed in reduced and nonreduced gels.Bands in the same size range as those obtained with transgenic colonextracts were also observed with extracts from colons of nontransgenic mice (Fig. 3A). However, the bands appearing in colons fromthe latter mice were much weaker than those observed with colonsfrom transgenic mice. We have repeatedly observed this backgroundstaining with nontransgenic colon extracts from different mice, but itis always much weaker than that obtained with colons from differenttransgenic mice. Furthermore, exposure of blots to an irrelevantmouse MAb instead of the anti-CEA MAb also resulted in similar

weak bands for colon extracts from both transgenic and nontransgenicmice (data not shown). Although the basis for this nonspecific reactivity is unknown, these results show that the CEA molecule producedby the transgenic mice is of the same molecular size range as CEAderived from human colonie adenocarcinomas.

Establishment of Syngeneic Colon Tumor Cell Line ExpressingCEA. Transfection of MC-38 colon carcinoma cells with the CEAcDNA-containing plasmid, pH/3Apr-l-neo, resulted in the establishment of a clone, C15-4.3, that has maintained CEA expression for

over one year in continuous culture in the absence of drug selection.The flow cytometry profile for C15-4.3 cells stained with the anti-

CEA MAb T84.66 is depicted in Fig. 4, and it shows that over 90%of the cells are expressing CEA on their surface. CEA expressed onthe surface of human colon carcinoma cells is linked to the cell

surface by a phosphatidylinositol glycan linkage (34). To determinewhether the CEA expressed on the C 15-4.3 cell surface had a similarlinkage, the cells were treated with PI-PLC. which cleaves phosphatidylinositol glycan-linked proteins from the cell surface (35). Themajority of C 15-4.3 cells became CEA negative following treatmentwith PI-PLC (Fig. 4), demonstrating that C 15-4.3 cells have CEA

attached to their cell surface via a phosphatidylinositol glycan linkage.Quantitation of the amount of CEA produced by tissue cultured

C 15-4.3 cells was carried out by ELISA. C 15-4.3 cells secreted27.3 ±1.0 ng of CEA per IO6 cells every 24 h, whereas CEA was not

detected in supernatants derived from the parental MC-38 cells. Undersimilar conditions, the human colonie adenocarcinoma cell line, LS-

174T, secreted 71.4 ±8.1 ng of CEA.The molecular size of the CEA produced by C15-4.3 cells was

analyzed by Western blotting. Cultured cells (0.75-1.6 X IO6 cells)

were washed with PBS, resuspended in SDS-PAGE sample buffer,and heated at 100°Cfor 5 min prior to electrophoresis. As shown in

Fig. 3A. two bands were observed with C15-4.3 cell extracts, one

major band migrating the same as the reference purified human CEAand a second weaker band appearing at M, —¿�50,000.The fastermigrating band was absent in other C15-4.3 cell extracts (data not

shown), suggesting that it may have represented a breakdown product.Cell extracts from the parental MC-38 cells were consistently nega

tive.Analysis by PCR of C 15-4.3 DNA for the size of the integrated

CEA cDNA insert showed a single band at the predicted size of 2095bp (Fig. 3ß),demonstrating that an intact copy of the CEA cDNA waspresent in these cells.

Growth of C 15-4.3 Cells in Line 18 Mice. The growth of C 15-4.3cells was evaluated in both transgenic and nontransgenic mice following s.c. implantation. Despite the expression of the foreign CEAmolecule, the C15-4.3 cells grew in nontransgenic mice (Fig. 5).However, these CEA-expressing tumor cells grew at a faster rate in

transgenic mice; all of these mice consistently developed progressively growing tumors. One nontransgenic mouse had a tumor thatgrew transiently and regressed by day 14. On day 28, the sizes of thetumors growing in transgenic and nontransgenic mice were 418 ±91and 179 ±47 mm3, respectively (P = 0.03). The latter experiment

was repeated, and a similar difference in growth rate of C 15-4.3 cells

was observed between transgenic and nontransgenic mice. To determine whether the C 15-4.3 cells maintained expression of CEA whilegrowing in transgenic mice, tumors were removed from 3 mice 31-43days after implantation. The tumors were enzyme-digested, and the

Iz"ou•¿�

o>DC

Log Fluorescence Intensity

Fig. 4. Expression of CEA on CEA-transfected MC-38 colon carcinoma cells, asmeasured by flow cytometry. Clone C15-4.3 was stained with T84.66 anti-CEA MAb( ) or control IgG ( ). Cells were also treated with PI-PLC prior to .staining withT84.66 (shaded area). Bound antibody was detected with fluorescein-labeled antimouse

IgG.

1473

Research. on December 18, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

CEA TRANSGENIC MICE

ìscoko

700-

600-

500-

400-

300-

200-

100-

Transgenic

Nontransgenic

10 15 20 25 30 35

DAY POST TUMOR IMPLANTATIONFig. 5. Growth of CEA-transfecled MC-38 cells in transgenic mice. Clone C15-4.3

cells (2.5 X 10s) were implanted s.c. into transgenic (D; n = 5) or nontransgenic (O;

n = 6) mice.

released cells were placed in tissue culture using medium lackingG418. Flow cytometry was then carried out on the cultured cells 3-8days later. All three cultures stained positive for CEA (94-99%), apercentage similar to that obtained with C 15-4.3 cells maintained in

continuous culture (data not shown). However, the density of CEAexpression appeared to be reduced because the mean fluorescenceintensity for the in vivo passaged cells was 63-73% of that observed

for the cells grown in continuous culture. The reduction in the intensity of CEA expression may result from a loss of a subpopulation ofhighly positive cells in vivo, a slow adaptation of the latter cells togrowth in vitro because the assay was carried out 3-8 days after

removal of the tumors from the animals or a reduction in expressionlevel by the general cell population. Nonetheless, these results demonstrate that propagation of the C15-4.3 cells in vivo did not result in

a complete or rapid loss of CEA expression.Anti-CEA Antibody in Transgenic and Nontransgenic Mice

with Progressively Growing C15-4.3 Tumors. When tumor-bearing

mice from the previous experiment required euthanasia, blood wascollected to test by ELISA for the presence of antibody against CEA.The day after implantation for removal of a blood sample rangedbetween 28 and 36 and 28 and 59 days for transgenic and nontransgenic mice, respectively. The shorter residence time of tumors intransgenic mice resulted from their faster growth rate in these mice.However, there was not a significant difference in the mean tumorsizes between transgenic (1030 ± 615 mm3) or nontransgenic(1189 ±653 mm3) mice when the tumors were removed from the

animals. With nontransgenic mice, five of six mice developed anti-

CEA antibody, with four showing elevated titers (Fig. 6). By contrast,CEA antibody was not detected in the sera obtained from tumor-

bearing transgenic mice at the lowest dilution analyzed (1:100).Lower dilutions of sera were not examined because of increasedbackground that can appear when high concentrations of sera are usedin ELISA. Because the tumor sizes were very similar between transgenic and nontransgenic mice when antibody assays were carried out,it is unlikely that the tumors of transgenic mice were an antibody sinkand prevented its detection in these animals. Furthermore, becausecirculating antigen was not detected in both transgenic and nontransgenic mice (data not shown), it is also unlikely that immune complexes prevented the detection of anti-CEA antibody in transgenic

DISCUSSION

A major limitation in immunotherapy studies of human tumorantigens is the general lack of appropriate preclinical models. Clinicalstudies can be difficult to implement, particularly when a clear understanding of the potential efficacy, limitations, and safety of animmunotherapeutic strategy is not available from relevant animalinvestigations. Immune-mediated therapies directed at tumor self-

antigens such as CEA also deal with a balance between desirableantitumor responses and unwanted autoimmune reactions. However,mice carrying a transgene for a human tumor self-antigen may provide

a more acceptable experimental model in which knowledge about theabove issues can be enhanced prior to initiating clinical trials. Withrespect to CEA, two laboratories have previously described the preparation of CEA transgenic mice, differing chiefly in the type of CEADNA used (31,36).

Hasegawa et al. (36) were the first to create a CEA transgenicmouse line by using a full-length CEA cDNA fragment containing the

SV40 early promoter. Northern and Southern blotting revealed expression of CEA in all tissues examined, including the brain, thymus,lung, spleen, liver, kidney, and colon. The ectopie expression of CEAby the brain, thymus, spleen, liver, and kidney was not unexpectedbecause of the incorporation of the SV40 promoter. However, somecellular specificity of expression was evident because only the epithelial cells in the lung and colon were positive by immunohistochem-

ical staining.

2.5-

1.5-

ino*

qO

A. Nontransgenic Mice

0.5-

100 1000 10000 100000

2.5-

2-

1.5-

1-

0.5-

B. Transgenic

10 100 1000 10000

RECIPROCAL OF DILUTIONFig. 6. Anti-CEA antibody in mice bearing CEA-lransfected MC-38 tumors (C15-4.3

clone). Serum obtained from nontransgenic (H = 6; A) or transgenic (n = 5; B) mice wasassayed for the presence of anti-CEA antibody by ELISA beginning at l:IOO dilution.D (A), reaction produced by nontransgenic C57BL/6 mouse serum.

1474

Research. on December 18, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

CEA TRANSGENIC MICE

Schrewe et al. (19) cloned the complete gene for CEA, includingflanking regulatory elements, and showed that this genomic cloneconveyed cell-type specific expression of CEA. Furthermore, trans-fection of this cosmid clone into CHO cells resulted in CEA expression, demonstrating that the factors responsible for activation of theCEA gene were conserved between humans and rodents (37). Thisprompted Eades-Perner et al. (31), followed by ourselves in this study,to prepare transgenic mice with genomic CEA in hopes that thetissue-specific expression observed in humans would be also reflectedin transgenic mice. The former laboratory established four independent lines having a C57BL/6 X CB6 background, all of which contained intact copies of the cosmid insert. A restricted pattern of CEAexpression was observed in all four lines, as determined by Northernblot and immunohistochemical analysis. However, ectopie appearanceof CEA with respect to its expression in adult humans (6) wasobserved in the esophagus, small intestine, trachea, and lung of alltransgenic lines. In the line 18 mouse described here, CEA was onlydetected by immunohistochemistry in the large intestine, cecum, andstomach, more closely approximating the adult human pattern. Because the same MAb and similar tissue processing procedures wereused in our studies as those of Eades-Perner et al. (31), differences inlevels of gene expression most likely account for the discordance intissue expression between the mice derived in the two studies. Thelevels of CEA in the colons of transgenic mice produced by Eades-Perner et al. (31) were over 50- and 10- fold higher than that of theline 18 mouse and humans, respectively. Similarly, fecal content ofCEA in the former mice was 2-4 times higher than that of humanfeces, whereas line 18 mouse feces contained 100-fold less CEA. Thehigh level of CEA production was not related to copy number amongthe four lines studied by Eades-Perner et al. (31), nor was it related tothe line 18 mouse because the latter had twice the number of copiesof the mouse line with the highest copy number in their study. Theformer transgenic lines also had elevated levels (14-30 ng/ml) ofcirculating CEA, whereas it was below detection limits in the serumfrom line 18 mice, as it is in normal human serum. Although differences between transgenic mice and humans in posttranslational regulatory mechanisms or turnover rates may explain the lack of correlation in CEA expression with copy number or CEA mRNA levels(31), our findings with the line 18 mouse suggest that the level of CEAtranscription may also be important and is perhaps influenced by siteof insertion. Nonetheless, the pattern of CEA expression of the line 18mouse confirms previous findings (31) that the regulatory mechanisms contributing to tissue-specific CEA expression in humans areconserved in mice.

The CEA produced by the colon of line 18 mice was similar in sizeto that of CEA purified from human colonie adenocarcinomas, demonstrating that the CEA molecule was processed normally. However,it is likely that differences exist in glycosylation of CEA in mice ascompared to humans (38). The CEA produced by the colonie epithelium of line 18 mice was predominantly localized to the cells in theapical regions of the crypts, similar to the pattern observed in humancolons (32). In contrast, Eades-Perner et al. (31) found that CEA wasexpressed throughout the lumen of the colonie crypts in their fourtransgenic lines, again demonstrating heightened CEA expression inthese mice.

Depending upon the intended study, the high level of CEA expression by the four lines produced by Eades-Perner et al (31) may be lessdesirable than a transgenic mouse with lower antigen expression. Asa potential model for antibody targeting studies, high circulatinglevels of CEA could result in the formation of complexes with aninjected antibody, as has been observed clinically with radioantibodies(39). This could compromise the use of the mouse model becauseserum CEA levels are rare in early-stage colorectal cancer, although

95% of colorectal cancers express CEA (40). Aberrant expression ofantigen in ectopie sites could also enhance inappropriate localizationof antibody to these locations. Furthermore, as discussed below,vaccine development may be more arduous because of difficulties inbreaking tolerance to CEA and may not truly mimic requirements tobreach the tolerant state in humans.

Besides potential limitations caused by elevated levels of CEAexpressed by the transgenic lines generated by Eades-Perner et al.(31), our interest in producing another CEA transgenic mouse linewas motivated by the desire to have a syngeneic model for tumorimmunotherapy. For a tumor system, a syngeneic, transplantablecolon carcinoma tumor, MC-38 (41), was transfected with humanCEA cDNA. One clone (C 15-4.3) contained a full-length copy ofthe CEA cDNA, expressed a high level of CEA on its surface, andsecreted antigen at a rate that was about one-third the level ofantigen produced by the LS-174T human colon adenocarcinomacell line. This clone also expressed antigen that was similar inmolecular size to that of CEA purified from human colonie tumorsand was attached to the cell surface via a phosphatidylinositolglycan linkage, which is typical for CEA. The transfected MC-38cells produced progressively growing tumors following implantation into transgenic mice. Importantly, expression of CEA similarto that of cells under continuous cell culture was maintainedthroughout tumor growth in vivo.

Other groups have previously described the production of CEA-positive rodent colon carcinomas (42, 43). The CEA-transfectedMC-38 cell line, MC-38-cea-2 (43), has been used in severalexperimental immunotherapy studies (44-48). However, aboutone-half of the CEA gene was missing in these cells, and this cellline was apparently selected based on its superior growth properties in normal mice, as compared to transfectants harboring thecomplete CEA gene (49). An underlying anti-CEA immune response appears to be the reason for the retarded growth of the lattercells in normal mice (49). Likewise, we found that the C15-4.3cells did not grew as well in nontransgenic as did transgenic mice.Pèlegrinet al. (42) described the growth of several CEA transfected rat colon carcinoma lines. They reported that, in normal rats,the CEA-positive cell lines grew at a slower rate than did non-transfected cells, and some of the highest CEA expressers regressed. Thus, although other approaches exist for the preparationof tumor models in transgenic mice, such as direct chemicalinduction or mating with mice with a high tumor incidence, the useof CEA-transfected tumor cells, as shown here, appears to providea suitable alternative. Unlike the use of transfectants bearing aforeign antigen in normal mice, transfectants expressing authenticantigen grow well in transgenic mice.

The growth in vivo of CEA-transfected MC-38 cells was accompanied by the appearance of antibody in nontransgenic but not transgenic mice. Despite the presence of a humoral response, the C 15-4.3produced progressively growing tumors in nontransgenic mice, albeitat a slower rate than in transgenic animals. The role that anti-CEAantibody has in inhibiting tumor growth, as well as contributions ofcell-mediated components to an antitumor response in the formermice, are not known. More importantly, the failure of transgenic miceto develop anti-CEA antibody under these conditions shows that theline 18 mice are tolerant to CEA. Ongoing studies are currentlyevaluating the extent to which tolerance to CEA in the latter mice canbe assigned to the B- and/or T-helper cell compartments. Adelstein etal. (50), in experiments carried out in hen egg lysozyme transgenicmice, showed that T cells were tolerant, irrespective of the level oflysozyme expressed. B-cell tolerance did not occur when the antigenconcentration in the serum was below 0.1 nM.On the basis of theseobservations, it appears likely that B-cell tolerance to CEA in line 18

1475

Research. on December 18, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

CEA TRANSGENIC MICE

mice is absent or weak and certainly less than that of the linesdeveloped by Eades-Perner et al. (31). Because the latter mouse linesexpress CEA to much higher levels than that found in either line 18mice or in humans, it may be more difficult to break tolerance to CEAin these mice.

We expect that the line 18 CEA transgenic mice will prove to bea valuable model for antigen-specific vaccine studies. Like humans, these mice are tolerant to CEA, although the characteristicsof this tolerance has yet to be defined in both humans and line 18mice. Nonetheless, the tolerant state of the latter mice presents amore demanding challenge than do nontransgenic animals forexamining the ability of vaccines to elicit anti-CEA responses thatmay also be therapeutically effective. We have shown here thatsyngeneic transplantable tumor transfectants expressing high levels of authentic CEA can be prepared and that these grow well intransgenic mice. Thus, these CEA transfectants provide a means tobegin testing the antitumor potency of vaccines, whereas othertumor models are under development. Because line 18 mice alsoexpress antigen in normal tissue locations, as is found in humans,it is possible to search for the appearance and properties of autoimmune reactions against these tissues once tolerance to CEA iscircumvented. Finally, the line 18 transgenic mouse model provides an alternative for antibody targeting studies, one in whichexperiments can be carried out in an immunocompetent host. Theanalysis of immunoconjugates in which their activities at leastpartly depend on an intact immune system would be enhanced intransgenic mice (51, 52).

ACKNOWLEDGMENTS

We thank Dr. John Shively for his helpful discussions and for reviewing thismanuscript and Jennifer L. Ho for her assistance in identifying homozygousmice by FISH.

REFERENCES

1. Boon, T., Cerottini, J-C., Van den Eynde. B., van der Bruggen, P., and Van Pel, A.Tumor antigens recognized by T lymphocytes. Annu. Rev. Immunol., 12: 337-365,1994.

2. Finn, O. Tumor-rejection antigens recognized by T lymphocytes. Curr. Opin. Immunol., 5: 701-708, 1993.

3. Houghton, A. N. Cancer antigens: immune recognition of self and altered self. J. Exp.Med., 180: 1-4, 1994.

4. Livingston, P. O. Construction of cancer vaccines with carbohydrate and protein(peptide) tumor antigens. Curr. Opin. Immunol., 4: 624-629, 1992.

5. Mitchell, M. S., Harel, W., Kempf, R. A., Hu. E., Kan-Mitchell, J., Boswell, W. D.,Dean, G., and Stevenson, L. Active-specific immunotherapy for melanoma. J. Clin.Oncol., 8: 856-869, 1990.

6. Nap, M., Mollgard, K., Burtin, P., and Fleuren, G. J. Immunohistochemistry ofcarcino-embryonic antigen in the embryo, fetus, and adult. Tumor Biol., 9: 145-153,

1988.7. Wittekind, C. Carcinoembryonic antigen family members as diagnostic tools in

immunohistopathology. Tumor Biol., 16: 42-47, 1995.8. Matzku. S., Zöller,M., Schulz, U., Dietze, W., Barth, H. O., Saeger, H-D., and Price,

M. R. Lack of correlation between Carcinoembryonic antigen content of tumorextracts and leukocyte migration reactivity of tumor patients. J. Nati. Cancer Inst.(Bethesda), 64: 1345-1348, 1980.

9. Müller,M., Irmscher, J., Heidi, G., Grossmann, H., Wagner, H., Kotzsch, M-, andZotter, S. Sensitization of human lymphocytes to Carcinoembryonic antigen (CEA):neutralization experiments with anti-CEA sera. Exp. Pathol., 21: 5-9, 1982.

10. Hill, R., Khoo, S. K., Daunter, B., Silburn, P. A., and Mackay, E. V. Immuno-globulins reactive to Carcinoembryonic antigen and their relationship to theantigen in malignant ascitic fluid of ovarian carcinoma. Int. J. Cancer, 30:587-592, 1982.

11. Lo Gerfo, P., Herter, F. P., and Bennet, S. J. Absence of circulating antibodies toCarcinoembryonic antigen in patients with gastrointestinal malignancies. Int. J. Cancer, 9: 344-348, 1972.

12. Gold, P. Circulating antibodies against Carcinoembryonic antigens of the humandigestive system. Cancer (Phila.), 20: 1663-1667, 1967.

13. Durrani, L. G., Dentón, G. W. L., Jacobs, E., Mee, M., Moss, R., Austin, E. B.,Baldwin, R. W., Hardcastle, J. D., and Robins, R. A. An idiotypic replica ofCarcinoembryonic antigen inducing cellular and humoral responses directed againsthuman colorectal tumours. Int. J. Cancer, 50: 811-816, 1992.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

Koda, J., Glass, M. C., and Chang, H. R. Generation of human monoclonal antibodiesagainst colon cancer. Arch. Surg. 725: 1591-1597, 1990.Foon, K. A., Chakraborty, M., John, W. J., Sherratt, A., Köhler,H., and Bhattacharya-

Chatterjee, M. Immune response to the Carcinoembryonic antigen in patients treatedwith an anti-idiotype antibody vaccine. J. Clin. Invest., 96: 334-342, 1995.

Tsang, K. Y., Zaremba, S., Nieroda, C. A., Zhu, M. Z., Hamilton, J. M., and Schlom,J. Generation of human cytotoxic T cells specific for human Carcinoembryonicantigen epitopes from patients immunized with recombinant vaccinia-CEA vaccine.J. Nati. Cancer Inst. (Bethesda), 87: 982-990, 1995.

Conry, R. M., Saleh, M. N., Schlom, J., and LoBuglio, A. F. Breaking tolerance toCarcinoembryonic antigen with a recombinant vaccinia virus vaccine in man. Proc.Am. Assoc. Cancer Res., 36: 492, 1995.McCuaig, K., Rosenberg, M., Nedellec, P., Turbide, C., and Beauchemin, N. Expression of the BGP gene and characterization of mouse colon biliary glycoproteinisoforms. Gene, 127: 173-183, 1993.

Schrewe, H., Thompson, J., Bona, M., Hefta, L. J. F., Maruya, A., Hassauer, M.,Shively, J. E., von Kleist, S., and Zimmermann, W. Cloning of the complete gene forCarcinoembryonic antigen: analysis of its promoter indicates a region conveying celltype-specific expression, Mol. Cell. Biol., 10: 2738-2748, 1990.

Sambrook, J., Fritsch, E. F., and Maniatis, T. (eds.). Molecular Cloning: A LaboratoryManual, Ed. 2, pp. 9.16-9.19. Cold Spring Harbor, NY: Cold Spring Harbor Labo

ratory, 1989.Mann, J. R., and McMahon, A. P. Factors influencing production frequency oftransgenic mice. Methods Enzymol. 225: 771-781, 1993.

Erlich, H. A. PCR Technology: Principles and Applications for DNA Amplification,p. 246. New York: Stockton Press, 1989.Miklos, G. L. G., and Rubin, G. M. The role of the genome project in determininggene function: insights from model organisms. Cell, 56: 521-529, 1996.Dim-link. J. E., Kelley, K., and Boyle, A. L. Fluorescence in situ hybridization of

interphase nuclei isolated from whole blood of transgenic mice. Biotechnology, 17:954-959, 1994.

Esteban, J. M., Paxton, R., Mehta, P., Battifora, H., and Shively, J. E. Sensitivity andspecificity of Gold types 1 to 5 anti-carcinoembryonic antigen monoclonal antibodies:

immunohistologic characterization in colorectal cancer and normal tissues. Hum.Pathol., 24: 322-328, 1993.Nap, M., Hammarström, M-L., Börmer,O., Hammarström, S., Wagener, C., Handt,S., Schreyer, M., Mach, J-P., Buchegger, F., von Kleist, S., Grunert, F., Sequin, P.,

Fuks, A., Holm, R., and Lamerz, R. Specificity and affinity of monoclonal antibodiesagainst Carcinoembryonic antigen. Cancer Res., 52: 2329-2339, 1992.

Simpson, J. F., Primus, F. J., and Schlom, J. Complementation of expression ofCarcinoembryonic antigen and tumor associated glycoprotein (TAG-72) in humancolon adenocarcinomas. Int. J. Biol. Mark., 6: 83-90, 1991.Neumaier, M., Shively, L., Chen, F-S., Gaida, F-J., Ilgen, C., Paxton, R. J., Shively,

J. E., and Riggs, A. D. Cloning of the genes for T84.66, an antibody that has a highspecificity and affinity for Carcinoembryonic antigen, and expression of chimerichuman/mouse T84.66 genes in myeloma and Chinese hamster ovary cells. CancerRes., 50: 2128-2134, 1990.

Rosenberg, S. A., Mulé,J. J., Spiess, P. J., Reichert, C. M., and Schwarz, S. L.Regression of established pulmonary métastasesand subcutaneous tumor mediated bythe systemic administration of high-dose recombinant interieukin 2. J. Exp. Med.,161: 1169-1188, 1985.

Kuroki, M., Koga, Y., and Matsuoka, Y. Purification and characterization of Carcinoembryonic antigen-related antigens in normal adult feces. Cancer Res., 41: 713-

720, 1981.Eades-Perner, A-M., van der Putten, H., Hirth, A., Thompson, J., Neumaier, M., von

Kleist, S., and Zimmermann, W. Mice transgenic for the human Carcinoembryonicantigen gene maintain its spatiotemporal expression pattern. Cancer Res., 54: 4169-4176, 1994.Primus, F. J., Clark, C. A., and Goldenberg, D. M. Immunoperoxidase localization ofCarcinoembryonic antigen in normal human intestinal mucosa. J. Nati. Cancer Inst.(Bethesda), 67: 1031-1039, 1981.Lisowska, E., Krop-Watorek, A., and Sedlaczek, P. The dimeric structure of Carcinoembryonic antigen (CEA). Biochem. Biophys. Res. Commun., 115: 206-211,1983.Hefta, S. A., Hefta, L. J., Lee, T. D., Paxton, R. J., and Shively, J. E. Carcinoembryonic antigen is anchored to membranes by covalent attachment to a glycosylphos-phatidylinositol moiety: identification of the ethanolamine linkage site. Proc. Nati.Acad. Sci. USA, 85: 4648-4652, 1988.

Cross, G. A. M. Glycolipid anchoring of plasma membrane proteins. Annu. Rev. CellBiol., 6: 1-39, 1990.

Hasegawa, T., Isobe, K.. Tsuchiya, Y., Oikawa, S., Nakazato, H., Ikezawa, H.,Nakashima, I., and Shimokata, K. Establishment and characterization of humanCarcinoembryonic antigen transgenic mice. Br. J. Cancer, 64: 710-714, 1991.Hefta, L. J. F., Schrewe, H., Thompson, J. A., Oikawa, S., Nakazato, H., and Shively,J. E. Expression of complementary DNA and genomic clones for Carcinoembryonicantigen and nonspecific cross-reacting antigen in Chinese hamster ovary and mousefibroblast cells and characterization of the membrane-expressed products. CancerRes., 50: 2397-2403, 1990.Matsui, T., Hamako, J., Kameyama, K-Z., Kurosawa, Y., Titani, K., and Mizuochi, T.The occurrence of mouse-type oligosaccharides in mouse-human chimeric immuno-globulin G. Biochem. Biophys. Res. Commun., 164: 245-250, 1989.Primus, F. J., Bennett, S. J., Kim, E., DeLand, F. H., Zahn, M. C., and Goldenberg,D. M. Circulating immune complexes in cancer patients receiving goat radiolocaliz-ing antibodies to Carcinoembryonic antigen. Cancer Res., 40: 497-501, 1980.

1476

Research. on December 18, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

CEA TRANSOENIC MICE

40. Ballesta, A. M., Molina, R., Fuella, X., Jo, J., and Giménez,N. Carcinoembryonicantigen in staging and follow-up of patients with solid tumors. Tumor Biol., 16:32-41, 1995.

41. Corbe«,T. H., Groswold, D. P., Roberts, B. J., Peckham, J. C., and Schabel, F. M.Tumor induction relationships in development of transplantable cancers of the colonin mice for chemotherapy assays, with a note on carcinogen structure. Cancer Res.,35: 2434-2439, 1975.

42. Pèlegrin, A., Terskikh, A., Hayoz, D., Chalandon, Y., Olsson, N-O., Folli, S.,Buchegger, F., Kromer, B., Schwarz, K., Martin, M., Martin, F., and Mach, J-P.

Human Carcinoembryonic antigen cDNA expressed in rat carcinoma cells can function as target antigen for tumor localization of antibodies in nude rats and and asrejection antigen in syngeneic rats. Int. J. Cancer, 52: 110-119, 1992.

43. Robbins, P. F.. Kantor, J. A., Salgaller, M., Hand, P. H., Fernsten, P. D.. and Schlom,J. Transduction and expression of the human Carcinoembryonic antigen gene in amurine colon carcinoma cell line. Cancer Res., 5/: 3657-3662, 1991.

44. Bei, R., Kantor, J., Kashmiri, S. V. S., Abrams, S., and Schlom, J. Enhanced immuneresponses and anti-tumor activity by baculovirus recombinant Carcinoembryonic

antigen (CEA) in ice primed with the recombinant vaccinia CEA. J. Immunother., 16:275-282, 1994.

45. Conry, R. M., LoBuglio, A. F., Loechel, F., Moore, S. E., Sumerel, L. A., Barlow,D. L., and Curiel, D. T. A Carcinoembryonic antigen polynucleotide vaccine has inviva antitumor activity. Gene Ther., 2: 59-65, 1995.

46. Hodge, J. W., McLaughlin, J. P., Abrams, S. I., Shupert, W. L., Schlom, J., andKantor, J. A. Admixture of a recombinant vaccinia containing the gene for thecostimulatory molecule B7 and a recombinant vaccinia virus containing a tumor-

associated antigen gene results in enhanced specific T-cell responses and antitumorimmunity. Cancer Res., 55.- 3598-3603. 1995.

47. McLaughlin. J. P.. Schlom. J.. Kantor, J. A., and Greiner, J. W. Improved immuno-therapy of a recombinant Carcinoembryonic antigen vaccinia vaccine when given incombination with interleukin 2. Cancer Res.. 56: 2361-2367, 1996.

48. Pervin, S., Chakraborty. M., Bhattacharya-Chatterjee, M., Zeytin, H., Foon. K. A.,and Chatterjee, S. K. Induction of antitumor immunity by an anti-idiotype antibodymimicking Carcinoembryonic antigen. Cancer Res., 57: 728-734, 1997.

49. Hand, P. H., Robbins, P. F.. Salgaller, M. L.. Poole, D. J., and Schlom, J. Evaluationof human carcinoembryonic-antigen (CEA)-transduced and non-transduced murinetumors as potential targets for anti-CEA therapies. Cancer Immunol. Immunother.,36: 65-75, 1993.

50. Adelstein, S., Pritchard-Briscoe. H., Anderson. T. A.. Crosbie, J., Gammon, G.,Loblay, R. H., Basten, A., and Goodnow, C. C. Induction of self-tolerance in T cells

but not B cells of transgenic mice expressing little self antigen. Science (WashingtonDC), 251: 1223-1225, 1991.

51. Clarke, P., Szalai, G., and Primus, F. J. A single gene-encoded antibody for targetingIL-2 to tumors. FASEB J., 10: AI350. 1996.

52. Bruno, R., Mach, J-P., Mani, J-C, Ychou, M., Folli, S., Artus, J-C, and Pèlegrin,

A. Cytokine targeting in tumors using a bispecific antibody directed againstCarcinoembryonic antigen and tumor necrosis factor a. Cancer Res., 56: 4758-

4765, 1996.53. Sikorska, H., Shuster, J., and Gold, P. Clinical application of Carcinoembryonic

antigen. Cancer Detect. Prev. 120: 321-355, 1988.

1477

Research. on December 18, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

1998;58:1469-1477. Cancer Res   Patrick Clarke, Jeffrey Mann, Jean F. Simpson, et al.   Model for ImmunotherapyMice Transgenic for Human Carcinoembryonic Antigen as a

  Updated version

  http://cancerres.aacrjournals.org/content/58/7/1469

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/58/7/1469To request permission to re-use all or part of this article, use this link

Research. on December 18, 2020. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from