HUMAN GENE THERAPY 13:1383–1390 (July 20, 2002)© Mary Ann Liebert, Inc.

Brief Report

lacZ Transgenic Rats Tolerant for b-Galactosidase: Recipientsfor Gene Transfer Studies Using lacZ as a Reporter Gene

SÉVERINE MÉNORET,1 DOMINIQUE AUBERT,2 LAURENT TESSON,1 CÉCILE BRAUDEAU,1

VIRGINIE PICHARD,2 NICOLAS FERRY,2 and IGNACIO ANEGON1

ABSTRACT

Gene transfer of reporter genes may trigger immune responses against the heterologous protein resulting inshortening of gene expression and inflammation. We generated transgenic rats expressing the lacZ gene un-der the control of the human immunodeficiency virus type 1 (HIV-1) long-terminal repeat (LTR) (HIV-lacZ)to obtain rats with undetectable transgene expression using histologic methods, thus avoiding interferencewith b-galactosidase (b-gal) expression from gene transfer, and displaying immune tolerance toward b-gal.LacZ transgenic mice with tolerance toward b-gal have already been used for gene transfer but rats consti-tute unique animal models with several advantages compared to mice. Two transgenic lines displayed low lev-els of b-gal mRNA in most organs tested, as detected only by reverse transcription-polymerase chain reac-tion (RT-PCR). The protein was undetectable by immunohistology and was only detected in the thymus andspleen using a sensitive enzyme-linked immunosorbent assay (ELISA). HIV-lacZ transgenic rats displayed im-mune tolerance to b-gal because immunization with b-gal resulted in markedly lower cellular and antibodyresponses compared to wild-type controls, whereas immunization with a nonrelated antigen, keyhole limpethemocyanin (KLH), resulted in comparable immune responses. The usefulness of this model in gene transferwas tested using a retroviral vector, which does not elicit destructive immune responses against transducedcells. Retroviral-mediated nlslacZ gene transfer in the liver resulted in nuclear b-gal expression for longerthan 12 months in HIV-lacZ transgenic rats, whereas wild-type controls showed nuclear b-gal expression forless than 1 month. After gene transfer of nlslacZ to the liver, antibodies, cytotoxic T lymphocytes (CTLs), andproliferation against b-gal were detected in wild-type controls but not in HIV-lacZ transgenic rats. In con-clusion, HIV-lacZ transgenic rats displaying low b-gal expression and immune tolerance toward b-gal are auseful tool to analyze the spatial and temporal expression of the b-gal protein in gene transfer experimentsusing lacZ as a reporter gene.

1383

INTRODUCTION

THE AMELIORATION OF GENE THERAPY demands accurate mea-surement of gene transfer efficiency, inflammation, and du-

ration of gene expression for each vector type. Gene transfer ofreporter genes is used to this end. However, most of these mark-ers are heterologous for the recipient and often result in im-mune responses against the foreign proteins, such as b-galac-tosidase (b-gal), resulting in local inflammation and eliminationof transduced cells (Yang et al., 1994, 1995 Juillard et al., 1995;Brubaker et al., 1996; Sarukhan et al., 2001a).

The ROSA26 lacZ transgenic mice (Zambrowicz et al.,1997) (tolerant toward b-gal), has been used to analyze b-galexpression after lacZ gene transfer in some models (Yang etal., 1995; Vassalli et al., 1999) but high basal transgene-derivedb-gal in some organs hampers its use.

The use of rats offers several advantages compared to mice,and in many cases, rats represent an alternative and/or com-plementary model to mice. Because of its bigger size, the rat ismore accessible than the mouse to microsurgery, multiple tis-sue and organ sampling, injection of substances into the brainfollowed by precise neuroanatomic localization (in mice there

1INSERM U437 and ITERT, 2INSERM ERM 0105 Centre Hospitalier Universitaire de Nantes, 44093, Nantes, France.

is lack of resolution through diffusion seen in a smaller brain),and analysis of organ function ex vivo (e.g., heart perfusion).Furthermore, a large body of physiologic data and disease mod-els (Gill et al., 1989) as well as transgenic technology (Char-reau et al., 1996, 1999) are available in the rat. Rats transgenicfor reporter genes and with immune tolerance toward the trans-gene could represent a useful tool for the gene transfer field.

The goal of this study was to generate rats transgenic forlacZ with low b-gal expression that would allow easy detec-tion of lacZ gene transfer experiments, and displaying immunetolerance toward b-gal, which would eliminate immune re-sponses against b-gal. Transgenic rats were obtained using aDNA construct in which lacZ was under the transcriptionalcontrol of the HIV-1 LTR, which has been shown to be ex-pressed at low levels in lymphoid tissues and Langerhans cellsfrom transgenic mice (Leonard et al., 1989; Cavard et al.,1990). These rats were analyzed for transgene-derived b-galexpression, for tolerance toward b-gal after immunization withEscherichia coli b-gal or retrovirus-mediated gene transfer oflacZ, as well as for duration of retrovirus-mediated expressionof b-gal.

MATERIALS AND METHODS

Generation of HIV-lacZ transgenic rats

Transgenic rats were produced following previously pub-lished procedures (Charreau et al., 1996, 1999). Embryos weremicroinjected with a DNA construct (2 mg/ml), previously usedto generate transgenic mice (Cavard et al., 1990) (kindly pro-vided by P. Briand, Institut Cochin, Paris, France). This con-struct contained the U3R sequence of the HIV-1 LTR upstreamof the cytoplasmic form of the lacZ gene and a SV40 polyadeny-lation signal. Tail DNA from the offspring was analyzed bypolymerase chain reaction (PCR) and Southern blot to identifytransgenic founders carrying lacZ sequences.

Analysis of b-gal expression

For histochemical analysis, tissue biopsies were embeddedin optimal cutting temperature compound (Miles Laboratories,Elkhardt, IN) and snap-frozen in liquid nitrogen precooledmethyl-butane. Cryostat sections (12 mm) were fixed and in-cubated with 5-bromo-4-chloro-3-indolyl-b-D-galactopyra-noside (X-Gal, Interchim, France) as previously described indetail (David et al., 1998). For immunohistologic analysis, tis-sue biopsies were fixed in formalin and embedded in paraffin.Tissue sections (5 mm) were deparaffinized and incubated (18hr, 4°C) with a polyclonal rabbit anti-b-gal antibody (Chemi-con, Temecula, CA) diluted in phosphate-buffered saline (PBS)containing Tween (0.1% v/v), then incubated with biotinylatedgoat anti-mouse immunoglobulin (Jackson Laboratories, WestGrove, PA), horseradish peroxidase-conjugated streptavidin,and amino ethyl carbazol substrate (Vector Laboratories,Burlingame, CA). The presence of b-gal was also analyzed in200 ml of tissue homogenates (containing 250 mg of protein)using a sensitive (detection limit 40 pg/ml) enzyme-linked im-munosorbent assay (ELISA) kit according to the manufacturer’sinstructions (Boehringer Mannheim, Mannheim, Germany). Fordetection of b-gal transcripts, RNA was isolated and reverse-

transcribed as previously described (David et al., 1998). Primersequences for b-gal were as follows: 59-AGTTCAGATGT-GCGGCGAGTT-39 (sense) and 59-TTCATTCCCCAGCGAC-CAGAT-39 (antisense). PCR was performed with 35 cycles ofamplification at 94°C for 30 sec, 61°C for 30 sec, and 72°C for30 sec. The mRNA of hypoxanthine phosphoribosyltransferasewas amplified as an internal control (David et al., 1998). Am-plified sequences were blotted and hybridized with the respec-tive probes labeled with radioactive phosphorus (32P).

Immunizations

One hundred micrograms of b-gal (Boehringer Mannheim)or keyhole limpet hemocyanin (KLH; Sigma, St. Louis, MO)emulsified in 200 ml of complete Freund’s adjuvant (CFA) wereinjected into the footpad. Sera and draining lymph nodes wereharvested 10 days later.

ELISA for detection of anti-b-gal and anti-KLH IgG antibodies

Detection of anti-b-gal and anti-KLH immunoglobulin G(IgG) antibodies was performed by ELISA as previously de-scribed (Guillot et al., 2000), using recombinant b-gal (Sigma,l’isle d’Abeau, France). Results were expressed as the recipro-cal serum dilution providing 50% of the maximal optical den-sity.

Proliferation assays against b-gal and anti-KLH

Lymph node cells and splenocytes were cultured in 96-wellplates (Nunc, Roskilde, Denmark) (105 cells per well) in trip-licate cultures in RPMI culture medium (Guillot et al., 2000)in the presence of different concentrations of recombinant b-gal or KLH and concanavalin A (ConA) (12.5 mg/ml) for 3days. Proliferation was evaluated by 3H-thymidine incorpora-tion during the final 8 hr of culture. For groups of animals, re-sults were expressed as the mean proliferation index (prolifer-ation against b-gal or KLH/proliferation in the presence ofculture medium) 6 standard error of the mean (SEM). For in-dividual animals, results were expressed as the mean D countsper minute (Dcpm) (proliferation against b-gal or KLH 2 pro-liferation in the presence of culture medium) 6 standard devi-ation [SD] of triplicate wells.

CTL assay against b-gal–expressing cells

Splenocytes were harvested more than 30 days after retro-virus-mediated lacZ gene transfer and cultured for 5 days in thepresence of recombinant vaccinia-lacZ viruses (kindly providedby Dr. Drillien, Strasbourg, France) at a multiplicity of infec-tion (MOI) of 10. Skin fibroblasts from newborn syngeneic ratseither nontransduced or transduced with supernatants contain-ing retroviruses coding for lacZ fused to a nuclear nls local-ization signal (retro-nlslacZ) or green flurorescent protein(GFP) were used as target cells. Target fibroblasts were labeledwith 51Cr and seeded in 96-well, V-bottom plates. Stimulatedeffector splenocytes were added to the target cells at variouseffector-to-target cell ratios. Spontaneous release (no effectors)and maximum release (10% sodium dodecyl sulfate [SDS]added) were included. The plates were incubated (6 hr, 37°C)and 50 ml of supernatant was removed and radioactivity deter-

MÉNORET ET AL.1384

mined. The percentage of specific 51Cr release was calculatedas follows: (experimental release-spontaneous release)/(maxi-mum release-spontaneous release) 3 100. Spontaneous releasewas lower than 10%. Results were expressed as mean 6 SDpercentage of specific lysis for each group of animals at dif-ferent target/effector ratios performed in triplicate.

Retrovirus production and gene transfer in vivo

Retro-nlslacZ were produced using the the TELCeB6 AF7cell line (Cosset et al., 1995). Viral titers were 108 transducingparticles per milliliter. Two thirds partial hepatectomy was per-formed as previously described (Ferry et al., 1991) and retro-viral delivery was performed 24 hr after partial hepatectomy byintravenous injection of 2.5 ml of medium containing retro-viruses and 8 mg/ml of polybrene (Sigma).

Statistics

Statistical significance was evaluated using a one-way anal-ysis of variance test and a p , 0.05 was considered as signifi-cant.

RESULTS AND DISCUSSION

Generation of rats transgenic for lacZ

The HIV-lacZ construct was microinjected into rat eggs and146 of these embryos were transferred into foster mothers. Tworats among 33 pups were identified by Southern blot and PCRanalysis as carrying the lacZ gene. Southern blot analysis of theprogeny of both transgenic founders showed transgene trans-mission in both lines and the presence of 50 and 10 copies inlines 12 and 18, respectively (data not shown).

HIV-lacZ transgenic rats express ubiquitously lowlevels of b-gal

In transgenic mice using the HIV-LTR, transgene expressionhas been previously detected in eye tissues and hair follicles ofembryos and in lymphoid tissues and skin Langerhans cells ofadult mice (Leonard et al., 1989; Cavard et al., 1990). In ratembryos of both HIV-lacZ transgenic lines, b-gal expressionwas detected in eye tissues by X-gal staining (data not shown).In adult rats of both lines, analysis of b-gal mRNA levelsshowed low levels of expression in the thymus, lymph nodes,spleen, skin, liver, heart, lung, and brain but none in the kid-ney, as detected by reverse transcription-polymerase chain re-action (RT-PCR) only after blotting and hybridization of theamplified sequences and not by northern blot analysis (Table1). Analysis of b-gal expression by ELISA showed detectablelevels of the protein in the thymus and spleen but not in theother organs of transgenic rats (Table 1). Analysis of b-gal ex-pression by X-Gal and immunohistologic staining of the lym-phoid organs, skin, liver, heart, lung, and kidney showed no la-beling in any of these tissues (Table 1). Nontransgeniclittermates used as negative controls in all experiments for anal-ysis of b-gal expression at the mRNA or protein level did notdisplay positive signals.

Thus, HIV-lacZ transgenic rats displayed undetectable lev-els of b-gal protein as assessed with the usual histology tech-

niques used to detect b-gal after gene transfer. Low levels ofb-gal protein level were detected in the thymus and spleen sug-gesting that these animals could potentially display b-gal–spe-cific immune tolerance by central and peripheral tolerancemechanisms.

HIV-lacZ transgenic rats display specific immunetolerance for b-gal

Antigen-specific immune tolerance is defined as diminishedor absent immune responses against a given antigen with nor-mal immune responses against unrelated antigens being re-tained. To test whether HIV-lacZ transgenic rats displayed b-gal–specific immune tolerance, we immunized adult transgenicrats from line 12 or nontransgenic control animals via the foot-pad with b-gal or the unrelated KLH antigen using the strongcomplete Freund’s adjuvant and analyzed their respective im-mune responses 10 days later. Lymphocytes of footpad drain-ing lymph nodes from HIV-lacZ transgenic animals showedmarkedly lower proliferation against b-gal in vitro comparedto those from immunized nontransgenic littermates but slightlyhigher than those of nonimmunized controls (Fig. 1A). In con-trast, lymph node cells from HIV-lacZ transgenic and non-transgenic rats showed the same mean proliferation index us-ing the nonspecific mitogen ConA (58.8 6 10.7, n 5 11 vs.46.4 6 9.4, n 5 12, respectively). Addition of interleukin-2(IL-2) increased proliferation of lymphocytes from HIV-lacZtransgenic lymph nodes, albeit at much lower levels comparedto those of lymphocytes from nontransgenic rats (Fig. 1A, in-set). This result suggests that anergy, a peripheral tolerancemechanism, may play a role in tolerance of HIV-lacZ trans-genic animals. Nevertheless, we cannot exclude the possibilitythat tolerance is caused by deletion in the thymus of high avid-ity anti-b-gal T cells and that the remaining low-avidity T cellsrequire IL-2 to proliferate in response to b-gal.

After footpad injection, HIV-lacZ transgenic rats showed sig-

lacZ TRANSGENIC RATS TOLERANT FOR b-gal 1385

TABLE 1. ANALYSIS OF b-GAL EXPRESSION IN ORGANS FROM

HIV-LTR TRANSGENIC ADULT RATS

Organs mRNA ELISA Immunohistology

Thymus 1 300 6 100 —Lymph nodes 1 ,40 —Spleen 1 586 6 289 —Skin 1 ,40 —Liver 1 ,40 —Heart 1 ,40 —Lung 1 ,40 —Brain 1 ,40 —Kidney 2 ,40 —

Detection of b-gal mRNA was performed by RT-PCR fol-lowed by blotting and hybridization with a radioactive probe(n 5 2–3). Detection of b-gal by ELISA (n 5 3) is expressedas picograms of b-gal/mg of total proteins. Immunohistologywas performed on paraffin sections using an anti-b-gal anti-body (n 5 2–3). Organs from nontransgenic rats were includedin each experiment and were consistently negative for the threeassays. Organs injected with an adenovirus coding for lacZ wereused as positive controls in every immunohistology experiment.

nificantly reduced levels of anti-b-gal IgG antibodies comparedto those of nontransgenic littermates and higher levels thanthose of nonimmunized controls (Fig. 1B). Intraperitoneal im-munization with soluble b-gal, a weaker immunizing protocolcompared with the use of complete Freund’s adjuvant, also re-sulted in generation of anti-b-gal antibodies by HIV-lacZ trans-genic rats, albeit at much lower levels than in nontransgeniccontrols (data not shown). It can be concluded that HIV-lacZtransgenic animals displayed drastically reduced immune re-sponse against exogenous b-gal, irrespectively of the immu-nization protocol. To assess whether the reduction of anti-b-galimmune responses in HIV-lacZ transgenic rats was due to TH1to TH2 deviation, we analysed the isotype of IgG anti-b-gal an-tibodies. In the rat, IgG2b predominates in TH1 responses,whereas IgG1 and IgG2a predominate in TH2 responses (Gra-cie and Bradley et al., 1996; David et al., 1998). In HIV-lacZtransgenic rats immunized with b-gal (n 5 10), IgG1, IgG2a,and IgG2b anti-b-gal antibodies showed a 2.7-, 3.5-, and 3.7-fold decrease, respectively, compared to the control group (n 5

12) (data not shown). Because TH1- and TH2-dependent IgGisotypes showed a comparable decrease, it can be concludedthat TH1 to TH2 immune deviation does not play a role in tol-erance toward b-gal in HIV-lacZ transgenic rats.

When KLH was used as an immunogen, lymph node cellsfrom HIV-lacZ transenic and nontransgenic rats draining the

footpad injected with KLH both showed comparably high pro-liferative responses whereas cells from nonimmunized controlsshowed no proliferation (Fig. 1C). Similarly, anti-KLH IgG an-tibodies were comparable in HIV-lacZ transgenic and non-transgenic rats immunized with KLH (Fig. 1D). Injection of b-gal into one footpad and of KLH in the contralateral footpad ofthe same transgenic (n 5 2) and nontransgenic (n 5 2) rats alsorevealed significantly decreased proliferative and antibody re-sponses toward b-gal and comparable immune responses toKLH in transgenic versus nontransgenic animals (data notshown). Collectively, these results demonstrate specific toler-ance toward b-gal in HIV-lacZ transgenic rats.

Persistent b-gal expression in HIV-lacZ transgenicrats after retrovirus-mediated gene transfer of lacZ

To explore the usefulness of HIV-lacZ transgenic rats in genetransfer protocols, we performed gene transfer of lacZ usingretroviral vectors. Retroviruses can elicit primary immune re-sponses but do not result in destruction of transduced cells be-cause of absence of viral genes to be expressed by transducedcells. Retroviral-mediated gene transfer has been shown in cer-tain cases to induce an immune response directed against thetransgene product (Song et al., 1997; Izembart et al., 1999;Aubert et al., 2001). Gene transfer of lacZ into the liver of rats

MÉNORET ET AL.1386

FIG. 1. Analysis of immune responses after immunization with b-gal or KLH. HIV-lacZ transgenic rats and nontransgenic ratswere immunized with 100 mg of either b-gal or KLH emulsified in CFA. Ten days later, draining lymph nodes and sera wereharvested. A: Proliferation of lymph node cells from transgenic (squares, n 5 11) nontransgenic (circles, n 5 12), or naïve (di-amonds, n 5 4) animals against the indicated concentrations of b-gal. *p , 0.0002 vs. non-Tg. Inset: experiments performed inthe presence of IL-2 (50 U/ml). B: Anti-b-gal IgG antibodies were analyzed in serially diluted serum from transgenic (n 5 11),nontransgenic (n 5 12), and naïve (n 5 4) animals by ELISA. *p , 0.0002 vs. non-Tg. C: Proliferation of lymph node cellsdraining the KLH injected footpad from transgenic (squares, n 5 6) non-transgenic (circles, n 5 6), and naïve (diamonds, n 52) animals against the indicated concentrations of KLH. D: Anti-KLH IgG antibodies were detected in serially diluted serumfrom transgenic (n 5 6), nontransgenic (n 5 6) or naïve (n 5 2) animals by ELISA. Proliferation results are expressed as meanproliferation index 6 SEM. ELISA results are expressed as the mean reciprocal serum dilution providing 50% of the maximaloptical density 6 SEM.

was performed using high doses of retro-nlslacZ. Delivery ofretro-nlslacZ into nontransgenic rats resulted in approximately10% of hepatocytes expressing b-gal at days 7 and 14 but noneat day 30 after gene transfer (Fig. 2A, 2C, and 2E). Livers oftransgenic rats showed no detectable b-gal expression in the ab-sence of gene transfer (Fig. 2B, inset) and persistent expressionat days 7, 30, and 260 after transduction with retro-nlslacZ (Fig.2B, 2D and 2F), suggesting that immune tolerance toward b-galallowed a prolonged expression after gene transfer.

Absence of anti-b-gal immune responses in HIV-lacZtransgenic rats after retrovirus-mediated gene transferof nlslacZ

To confirm that the persistent expression of b-gal in HIV-lacZ transgenic rats was associated with immune tolerance to-ward b-gal we analyzed the levels of antibodies, proliferation,and CTL responses against b-gal in transgenic and nontrans-genic rats after nlslacZ gene transfer.

After gene transfer of nlslacZ, anti-b-gal IgG antibodies wereundetectable in HIV-lacZ transgenic rats, as for controls thatdid not receive retro-nlslacZ, whereas nontransgenic ratsshowed high titers of these antibodies (Fig. 3A).

As for nontransduced controls, splenocytes from HIV-lacZtransgenic rats showed low levels of proliferation but only inthe presence of high concentrations of b-gal with no prolifera-tion at lower levels (Fig. 3B). In contrast, nontransgenic ratsdisplayed higher proliferative responses at high and low con-centrations of b-gal (Fig. 3B). Proliferative responses to ConA

were comparable in HIV-lacZ transgenic and nontransgenic rats(Dcpm for one animal of each group; 54,530 6 811 vs.63,080 6 215, respectively).

After gene transfer of nlslacZ, anti-b-gal CTL activity wasundetectable in splenocytes from HIV-lacZ transgenic rats, asfor nontransduced controls, whereas significant CTL activitywas detectable in splenocytes from nontransgenic rats (Fig. 3C).Nonspecific CTL activity against both nontransduced or GFP-transduced target cells was below 10 % in both nontransgenicand transgenic splenocytes (data not shown).

Thus, indefinite b-gal expression by hepatocytes in HIV-lacZtransgenic rats is because of tolerance toward b-gal and the useof retroviruses as gene transfer vectors. Tolerance toward b-galwas included epitopes presented by major histocompatibilitycomplex (MHC) class II antigens, because anti-b-gal CD41 cellproliferation and B cell responses were suppressed, and also byMHC class I antigens, because anti-b-gal CD81 CTL responseswere also absent. Anti-b-gal immune responses in nontrans-genic rats may have resulted from either cross-priming of CD41

T cells and CTLs by b-gal–derived peptides presented by MHCclass II and I molecules on antigen presenting cells (APCs) orby direct transduction of APCs, as previously shown for retro-viruses (Song et al., 1997).

Importantly, anti-b-gal immune responses were absent afterlacZ gene transfer (Fig. 3) but only partially inhibited after in-jection of b-gal (Fig. 1A and 1B), as previously described inlacZ transgenic mice after injection with b-gal (Salmon et al.,1998; Theopold and Köhler et al., 1990). This difference couldbe because of the use of an E. coli-derived nonglycosylated

lacZ TRANSGENIC RATS TOLERANT FOR b-gal 1387

FIG. 2. Analysis of b-gal expression in hepatocytes of transgenic and nontransgenic rats after retrovirus-mediated gene trans-fer of lacZ. HIV-lacZ transgenic rats and nontransgenic rats were partially hepatectomized, retro-nlslacZ was injected intra-venously 24 hr later and liver were harvested at the indicated time points for analysis of b-gal expression using anti-b-gal anti-bodies. A, C, and E: Hepatocytes from nontransgenic rats display nuclear expression of b-gal at day 7 and 14 but not at day 30after gene transfer. B, D, and F: Hepatocytes from HIV-lacZ transgenic rats display persistent nuclear expression of b-gal at day7, 30, and 260 after gene transfer. The inset in (D) shows that b-gal–positive hepatocytes in HIV-lacZ transgenic rats, which didnot receive retro-nlslacZ, were undetectable. Results are representative of 10 nontransgenic and 10 transgenic animals. Originalmagnification, 403.

form of b-gal for immunization, whereas the b-gal expressedby the cells after gene transfer is glycosylated. Alternatively,the presence of the b-gal protein in an extracellular form afterfootpad or systemic injection may be more immunogenic thanintracellular b-gal expression after gene transfer of nlslacZ andtherefore resulting in partial release of peripheral tolerancemechanisms. In confirmation of this idea, higher titers of anti-b-gal IgG antibodies were observed in nontransgenic animalsinjected with b-gal versus those that received retro-lacZ(30,750 6 5,221 vs. 4,983 6 1,833; from Figs. 1 and 3, re-spectively). Increased immune responses have already been de-scribed after gene transfer of secreted or membrane versus in-tracellular forms of b-gal because of more efficient antigenpresentation (Sarukhan et al., 2001b).

b-gal offers several convenient features as a reporter gene.Precise cellular identification within tissues can be assessed intissue sections by histochemical or immunohistochemical tech-niques. Furthermore, as is the case for GFP but not for other re-

porter genes, b-gal–expressing cells can easily be stained andsorted alive using fluorescein di-b-galactopyranoside (Nolan etal., 1988). There is minimal endogenous galactosidase activitywhereas endogenous phosphatase alkaline activity (PAP) (Mac-Gregor et al., 1995) and autofluorescence can interfere with thedetection of PAP and GFP, respectively. Finally, in contrast tocertain transgenic models overexpressing GFP (Huang et al.,2000; Martinez-Serrano et al., 2000), even high levels of b-galexpression early in embryonic life has not been associated withtoxicity (Zambrowicz et al., 1997). Rats transgenic for lacZ havebeen previously described (Charreau et al., 1996; Taniyama etal., 2001) but the use of promoters with nonlymphoid tissue-re-stricted specificity or even of immunoprivileged organs (such asthe eye) and the lack of reported studies on immune tolerancetoward the transgene product do not enable their use as tolerantrecipients in gene transfer experiments. Rats transgenic for GFP(Hakamata et al., 2001) or PAP (Kisseberth et al., 1999) havebeen recently described and, although not reported, these ani-

MÉNORET ET AL.1388

FIG. 3. Analysis of anti-b-gal immune responses in transgenic and nontransgenic rats after retrovirus-mediated gene transferof lacZ. HIV-lacZ transgenic rats and nontransgenic rats were partially hepatectomized, retro-nlslacZ was injected intravenously24 hr later and sera as well as spleens were harvested 21 days after gene transfer. A: Anti-b-gal IgG antibodies were analyzedin the sera of nontransgenic (n 5 9), transgenic (n 5 8), or naïve (n 5 4) animals. *p 5 0.0002. B: Splenocytes from transgenic(squares), nontransgenic (circles), or naïve (diamonds) rats were cultured for 3 days in the presence of b-gal at the indicated con-centrations. Each curve corresponds to a single animal assayed in the same experiment and these results are representative of 5other experiments with 11 transgenic and 10 nontransgenic animals. Results are expressed as mean proliferation index 6 SD. C:CTL activity of splenocytes from immunized transgenic (squares, n 5 4) nontransgenic (circles, n 5 4), or naïve (diamonds, n 52) rats was assayed against 51Cr-labeled syngeneic fibroblasts expressing lacZ. *p , 0.003.

mals could be tolerant for the transgenic proteins since trans-gene expression was high in immune cells. Nevertheless, thesetransgenes were expressed ubiquitously and at high levels mak-ing them unsuitable as recipients of gene transfer experimentsbecause of high endogenous reporter gene expression.

In conclusion, HIV-lacZ transgenic rats displayed unde-tectable levels of b-gal as evaluated by histologic techniquesand were tolerant toward b-gal. These features enable their useas recipients in gene transfer and transplantation experimentsusing vectors or cells expressing b-gal as a reporter gene.

ACKNOWLEDGMENTS

This work was financed in part by the Fondation Transvieand EEC grant Biomed2 BMM-CT98-3277. We are grateful toDr. Jean-Gérard Guillet and Pedro Lowenstein for criticallyreading the manuscript, to researchers who contributed withreagents, and to the Gene Vector Production Network from theAssociation Française contre les Myopathies of the Centre Hos-pitalier Universitaire de Nantes for producing recombinantviruses.

REFERENCES

AUBERT, D., MÉNORET, S., CALISE, D., IZEMBART, A., CHIARI,E., PODEVIN, G., PICHARD, V., TESSON, L., MOULLIER, P.,ANEGON, I., and FERRY, N. (2002) Highly efficient retroviral me-diated hepatic gene transfer is blunted by an immune response againstthe transgene product. Mol. Ther. 5, 388–396.

BRUBAKER, J., THOMPSON, C., MORRISON, L., KNIPE, D.,SIBER, G., and FINBERG R. (1996) Th1-associated immune re-sponses to b-galactosidase expressed by a replication-defective her-pes simplex virus. J. Immunol. 157, 1598–1604.

CAVARD, C., ZIDER, A., VERNET, M., BENNOUN, M.,SARAGOSTI, S., GRIMBER, G., and BRIAND, P. (1990) In vivoactivation by ultraviolet rays of the human immunodeficiency virustype I long terminal repeat. J. Clin. Invest. 86, 1369–1374.

CHARREAU, B., MENORET, S., TESSON, L., AZIMZADEH, A.,AUDET, M., WOLF, P., MARQUET, R., VERBAKEL, C., IJZER-MANS, J., COWAN, P., PEARSE, M., D’APICE, A., SOULILLOU,J.P., and ANEGON, I. (1999) Protection against hyperacutexenograft rejection of transgenic rat hearts expressing human decayaccelerating factor (DAF) transplanted into primates. Mol. Med. 5,617–630.

CHARREAU, B., TESSON, L., SOULILLOU, J.P., POURCEL, C.,and ANEGON, I. (1996). Transgenic rats: Technical aspects andmodels. Transgenic Res. 5, 223–234.

COSSET, F.L., TAKEUCHI, Y., BATTINI, J.L., WEISS, R.A., andCOLLINS, M.K.L. (1995). High-titer packaging cells producing re-combinant retroviruses resistant to human serum. J. Virol. 69,7430–7436.

DAVID, A., COUPEL-CLAUCE, H., CHETRITT, J., TESSON, L.,CASSARD, A., SOULILLOU, J.-P., and ANEGON, I. (1998). Anti-adenovirus immune responses in rats are enhanced by interleukin-4but not interleukin-10 produced by recombinant adenovirus. Hum.Gene Ther. 9, 1755–1768.

FERRY, N., DUPLESSIS, O., HOUSSIN, D., DANOS, O., andHEARD, J.-M. (1991). Retroviral-mediated gene transfer into hepa-tocytes in vivo. Proc. Natl. Acad. Sci. U.S.A. 88, 8377–8381.

GILL, T., SMITH, G.J., WISSLER, R.W., and KUNZ, H.W. (1989).The rat as an experimental animal. Science 245, 269–276.

GRACIE, J.A., and BRADLEY, J.A. (1996). Interleukin-12 induces in-terferon-gamma-dependent switching of IgG alloantibody subclass.Eur. J. Immunol. 26, 1217–1221.

GUILLOT, C., MATHIEU, P., COATHALEM, H., LE MAUFF, B.,CASTRO, M.G., TESSON, L., USAL, C., LAUMONIER, T.,BROUARD, S., SOULILLOU, J.P., LOWENSTEIN, P.R., CU-TURI, M.C., and ANEGON, I. (2000). Tolerance to cardiac allo-grafts via local and systemic mechanisms after adenovirus-mediatedCTLA4Ig expression. J. Immunol. 164, 5258–5268.

HAKAMATA, Y., TAHARA, K., UCHIDA, H., SAKUMA, Y.,NAKAMURA, M., KUME, A., MURAKAMI, T., TAKAHASHI,M., TAKAHASHI, R., HIRABAYASHI, M., UEDA, M., MIYOSHI,I., KASAI, N., and KOBAYASHI, E. (2001) Green fluorescent protein-transgenic rat: A tool for organ transplantation research.Biochem. Biophys. Res. Commun. 286, 779–785.

HUANG, W.-Y., ARAMBURU, J., DOUGLAS, P., and IZUMO, S.(2000). Transgenic expression of green fluorescence protein cancause dilated cardiomyopathy. Nat. Med. 6, 482–483.

IZEMBART, A., AGUADO, E., GAUTHIER, O., AUBERT, D.,MOULLIER, P., and FERRY, N. (1999) In vivo retrovirus-mediatedgene transfer to the liver of dogs results in transient expression andinduction of a cytotoxic immune response. Hum. Gene Ther. 10,2917–2925.

JUILLARD, V., VILLEFROY, P., GODFRIN, D., PAVIRANI, A.,VENET, A., and GUILLET, J.-G. (1995) Long-term humoral andcellular immunity unduced by a single immunization with replica-tion-defective adenovirus recombinant vector. Eur. J. Immunol. 25,3467–3473.

KISSEBERTH, W.C., BRETTINGEN, N.T., LOHSE, J.K., and SAND-GREN, E.P. (1999). Ubiquitous expression of marker transgenes inmice and rats. Dev. Biol. 214, 128–138.

LEONARD, J., KHILLAN, J., GENDELMAN, H., ADACHI, A.,LORENZO, S., WESTPHAL, H., MARTIN, M., and MELTZER,M. (1989). The human immunodeficiency virus long terminal repeatis preferentially expressed in Langerhans cells in transgenic mie.AIDS Res. Hum. Retroviruses 5, 421–430.

MACGREGOR, G., ZAMBROWICZ, B., and SORIANO, P. (1995).Tissue non-specific alkaline phosphatase is expressed in both em-bryonic and extraembryonic lineages during mouse embryogenesisbut is not required for migration of primordial germ cells. Develop-ment 121, 1487–1496.

MARTINEZ-SERRANO, A., VILLA, A., NAVARRO, B., RUBIO, F.-J., and BUENO, C. (2000). Human neural progenitor cells: betterblue than green ? Nat. Med. 6, 483.

NOLAN, G., FIERING, S., NICOLAS, J.-F., and HERZENBERG, L.(1988). Fluorescence-activated cell analysis and sorting of viablemammalian cells based on b-d-galactosidase activity after transduc-tion of Escherichia coli lacZ. Proc. Natl. Acad. Sci. U.S.A. 85,2603–2607.

SALMON, A., BRUAND, C., CARDONA, A., CHANGEUX, J., andBERRIH-AKNIN, S. (1998). An acetylcholine receptor alpha sub-unit promoter confers intrathymic expression in transgenic mice. Im-plications for tolerance of a transgenic self-antigen and for autore-activity in myasthenia gravis. J. Clin. Invest. 101, 2340–2350.

SARUKHAN, A., CAMUGLI, S., GJATA, B., VON BOEHMER, H.,DANOS, O., and JOOSS K. (2001a). Successful interference withcellular immune responses to immunogenic proteins encoded by re-combinant viral vectors. J. Virol. 75, 269–277.

SARUKHAN, A., SOUDAIS, C., DANOS, O., and JOOSS, K. (2001b).Factors influencing cross-presentation of non-self antigens expressedfrom recombinant adeno-associated virus vectors. J. Mol. Ther. 3,260–270.

SONG, E.S., LEE, V., SURH, C.D., LYNN, A., BRUMM, D., JOLLY,D.J. WARNER, J.F., and CHADA, S. (1997). Antigen presentationin retroviral vector-mediated gene transfer in vivo. Proc. Natl. Acad.Sci. U.S.A. 94, 1943–1948.

lacZ TRANSGENIC RATS TOLERANT FOR b-gal 1389

TANIYAMA, Y., SATO, K., SUGAWARA, A., URUNO, A., IKEDA,Y., KUDO, M., ITO, S., and TAKEUCHI, K. (2001). Renal tubule-specific transcription and chromosomal localization of rat thiazide-sensitive Na-Cl cotransporter gene. J. Biol. Chem. 276, 26260–26268.

THEOPOLD, U., and KÖHLER, G. (1990). Partial tolerance in b-galactosidase-transgenic mice. Eur. J. Immunol. 20, 1311–1316.

VASSALLI, G., AGAH, R., QIAO, R., AGUILAR, C., and DICHEK,D. (1999). A mouse model of arterial gene transfer. Antigen specificimmunity is a minor determinant of the early loss of adenovirus-me-diated transgene expression. Circ. Res. 85, e25–e32.

YANG, Y., ERTL, H.C.J., and WILSON, J. (1994). MHC class-I re-stricted cytotoxic T lymphocytes to viral antigens destroy hepato-cytes in mice infected with E1-deleted recombinant adenovirus. Im-munity 1, 433–442.

YANG, Y., LI, Q., ERTL, H.C.J., and WILSON, J. (1995). Cellularand humoral immune responses to viral antigens create barriers tolung-directed gene therapy with recombinant adenoviruses. J. Virol.69, 2004–2015.

ZAMBROWICZ, B., IMAMOTO, A., FIERING, S., HERZENBERG,

L., KERR, W., and SORIANO, P. (1997). Disruption of overlappingtranscripts in the ROSA bgeo 26 gene trap strain leads to widespreadexpression of b-galactosidase in mouse embryos and hematopoieticcells. Proc. Natl. Acad. Sci. U.S.A. 94, 3789–3794.

Address reprint requests to:Ignacio Anegon, M.D.

INSERM U43730 Boulevard Jean Monnet

44093 NantesFrance

E-mail: ianegon@nantes.inserm.fr

Received for publication January 7, 2002; accepted after revi-sion June 20, 2002.

Published online: July 10, 2002.

MÉNORET ET AL.1390