of August 16, 2017.This information is current as

Differentiation Thymocyte+CD8+ to CD4−CD8−During CD4 Gene Locus βChromatin Changes at the TCR-

Biochemical and Functional Analyses of

and Jianzhu ChenSamit Chattopadhyay, Charles E. Whitehurst, Frieder Schwenk

http://www.jimmunol.org/content/160/3/12561998; 160:1256-1267; ;J Immunol 

Referenceshttp://www.jimmunol.org/content/160/3/1256.full#ref-list-1

, 43 of which you can access for free at: cites 101 articlesThis article

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists All rights reserved.Copyright © 1998 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on August 16, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on August 16, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

Biochemical and Functional Analyses of Chromatin Changes atthe TCR-b Gene Locus During CD42CD82 to CD41CD81

Thymocyte Differentiation1

Samit Chattopadhyay,2* Charles E. Whitehurst,2* Frieder Schwenk,† and Jianzhu Chen3*

Allelic exclusion is the process wherein lymphocytes express Ag receptors from only one of two possible alleles, and is effectedthrough a feedback inhibition of further rearrangement of the second allele. The feedback signal is thought to cause chromatinchanges that block accessibility of the second allele to the recombinase. To identify the putative chromatin changes associated withallelic exclusion, we assayed for DNase I hypersensitivity, DNA methylation, and transcription in 100 kb of the TCR-b locus.Contrary to current models, we identified chromatin changes indicative of an active and accessible locus associated with theoccurrence of allelic exclusion. Of 11 DNase I hypersensitive sites identified, 3 were induced during CD42CD82 to CD41CD81

thymocyte differentiation, and demethylation and increased germline transcription of the locus were evident. We further examinedthe role of the most prominently induced site near the TCR-b enhancer (Eb) in allelic exclusion by targeted mutagenesis. Twoother sites were also examined in New Zealand White (NZW) mice that have a natural deletion in the TCR-b locus. TCR-b generecombination and allelic exclusion were normal in both mutant mice, negating dominant roles for the three hypersensitive sitesin the control of allelic exclusion. The data suggest that alternativecis-regulatory elements, perhaps contained in the Eb enhancerand/or in the upstream Vb region, are involved in the control of TCR-b allelic exclusion. The Journal of Immunology,1998, 160:1256–1267.

I mmunoglobulins and TCRs are heterodimers, consisting ofheavy and light chains ora- andb-chains, respectively. TheAg-binding domains of Ig and TCR chains are encoded by

variable gene segments, V, D, and J, for IgH and TCRb-chains,and V and J for IgL and TCRa-chains. Assembly of the variablegene segments through V(D)J recombination is required before theAg receptor genes can be expressed (1–3). Because vertebrates arediploid and one-third of V(D)J recombination events result in theexpression of functional protein chains (1, 4, 5), one-fifth of lym-phocytes are expected to express any given Ag receptor chain fromboth alleles. However, the majority of lymphocytes express Agreceptors from only one of the two alleles. This phenomenon isknown as allelic exclusion (reviewed in Refs. 6 and 7). Allelicexclusion was first observed in rabbits using allotypic markers forIg heavy chains. Individual B cells were found to express Ig heavychain from only one of the two alleles (8, 9). Similar studies withhuman and mouse B cells yielded the same results. Using Vb-specific Abs, 99% of T cells were also found to express TCRb-chain from only one of the two alleles (10, 11). Compared withthe IgH, IgL, and TCR-b genes, allelic exclusion at the TCR-agene locus is less stringent (12, 13).

Over the years, many studies were performed in an attempt toelucidate the mechanisms that underlie the control of allelic ex-clusion. The presence of DJ rearrangements at the second excludedIgH locus in mature B cells and at the TCR-b locus in mature Tcells strongly suggests that allelic exclusion is achieved through amechanism regulating V(D)J recombination (5, 14). According tocurrent understanding, each differentiating lymphocyte has twochances of assembling a productive VDJ rearrangement. BecauseDJ rearrangements occur on both alleles, the regulated step is at theV to DJ rearrangement step. If the first VDJ rearrangement is pro-ductive, then V to DJ joining on the other allele is inhibited. If theinitial VDJ rearrangement is nonproductive, then V to DJ rear-rangement can proceed on the second allele. Feedback inhibitionof further rearrangement by a prior productive recombinationevent has been demonstrated by studies with transgenic and gene-targeted mice. Expression of the membrane-bound, but not thesecretedm heavy chain inhibits VH to DJH rearrangements at theendogenous IgH loci (15–18). Targeted deletion ofmm exons alsoabolishes IgH gene allelic exclusion (19). Similarly, expression ofa TCR-b transgene blocks rearrangements of the endogenous lociat the DbJb stage (20, 21). Complete rearrangement and expres-sion of the endogenous alleles occur only when the TCR-b trans-gene is deleted (22). In pro-B cells, them heavy chains form pre-Bcell receptors by associating with surrogate light chains,l5 andVpreB, and Iga/Igb heterodimers (pre-BCR)4 (23, 24). Signalsinitiated from the pre-BCR mediate not only IgH gene allelic ex-clusion, but also pro-B to pre-B cell differentiation and expansionof pre-B cells (6, 7, 25, 26). In an analogous manner, TCR-bchains associate with the pre-TCRa-chains (pTa) and CD3 com-ponents to form pre-TCRs (pre-TCR) (27, 28), which signals for

*Center for Cancer Research and Department of Biology, Massachusetts Institute ofTechnology, Cambridge, MA 02139; and†Institute for Genetics, University of Co-logne, Cologne, Germany

Received for publication August 5, 1997. Accepted for publication October 21, 1997.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby markedadvertisementin accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.1 This work is supported in part by AI40146 and Arthritis Foundation (to J.C.), Can-cer Center Core Grant CA1405, Training Grant AI07463, and Cancer Research In-stitute Fellowship (to C.E.W.).2 The first two authors contributed equally to the work.3 Address correspondence and reprint requests to Dr. Jianzhu Chen, Center for CancerResearch, Massachusetts Institute of Technology, E17-128, 40 Ames Street, Cam-bridge, MA 02139. E-mail address: jchen@mit.edu

4 Abbreviations used in this paper: BCR, B cell receptor; DN, double negative; DP,double positive; Eb, T cell receptor-b gene enhancer; ES, embryonic stem; HS,DNase I hypersensitive site; neo, neomycin-resistance gene; NZB, New ZealandBlack; NZW, New Zealand White; RAG, recombination activating gene; RSB, re-ticulocyte standard buffer; SP, single positive.

Copyright © 1998 by The American Association of Immunologists 0022-1767/98/$02.00

by guest on August 16, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

TCR-b gene allelic exclusion as well as DN to DP thymocytedifferentiation and the expansion of DP thymocytes (29–31). Thus,allelic exclusion at IgH and TCR-b loci is mediated through afeedback inhibition of V to DJ rearrangement, occurring duringpro-B to pre-B cell and DN and DP thymocyte differentiation,respectively.

The feedback signal initiated from pre-BCR and pre-TCR couldeffect allelic exclusion by regulating recombinase activity as wellas modulating substrate gene segment accessibility. It is knownthat pro-B to pre-B cell and DN to DP thymocyte differentiation isaccompanied by a transient down-regulation of RAG-1 andRAG-2 gene transcription (32, 33). Furthermore, RAG-2 protein isdegraded before cells enter the S phase, as occurring in rapid pro-liferating cells during pre-B cell and DP thymocyte expansion(34–36). Although the down-regulation of RAG expression fol-lowing the feedback signaling has been postulated to prevent fur-ther V to DJ rearrangement on the second allele, thereby enablingthe establishment of allelic exclusion (37), coupling of V(D)J re-combination to the cell cycle is not essential for allelic exclusion(36, 38). RAG genes are re-expressed in pre-B cells and DP thy-mocytes for rearranging IgL and TCR-a genes, respectively (39–41). To prevent V to DJ rearrangement on the second IgH andTCR-b allele in pre-B cells and DP thymocytes, respectively, anadditional mechanism is required to specifically maintain their al-lelic exclusion. Because V(D)J recombinations at all Ig and TCRloci are mediated by conserved recombination signal sequences (1,3), it seems unlikely that the control is achieved solely by modi-fying the recombinase so that it no longer utilizes IgH or TCR-balleles as substrates. Rather, it seems more likely that allelic ex-clusion is maintained by controlling the accessibility of the in-volved gene segments to the recombinase (6, 7). In this model,signals from pre-BCR and pre-TCR induce the differentiation andproliferation of pre-B cells and DP thymocytes as well as stimulatechromatin structural remodeling at the IgH and TCR-b loci, re-sulting in inaccessibility to the recombinase. Supporting thismodel, exogenously added RAG-1 and RAG-2 core proteins canintroduce dsDNA breaks at the recombination signal sequencesflanking DJH and VH gene segments in nuclei of pro-B cells, butnot pre-B cells (42). However, chromatin changes associated withallelic exclusion, their molecular nature, and the underlyingcis-regulatory elements involved still need to be defined to prove thismodel.

Changes in DNase I hypersensitive sites, DNA methylation, andtranscription have traditionally been used as indicators of changesin chromatin structure (43–48). V(D)J recombination activity isassociated with increased DNase I hypersensitivity, hypomethyla-tion, and transcription. Before recombination, IgH andk locusbecomes hypersensitive to DNase I treatment, hypomethylated,and actively transcribed (49–52). Hypomethylation of recombina-tion substrates also promotes their rearrangements in both celllines and transgenic mice (53, 54). In addition, transcription en-hancers such as Em, Ek, Eb, and Ea are all known to promoterecombination of their respective genomic loci and minilocus re-combination substrates, although their mechanism of action re-mains elusive (41, 55–64). According to current understanding,allelic exclusion should be associated with a decrease in HS, hy-permethylation, and diminished transcription at the IgH andTCR-b locus, all indicators of an inaccessible locus.

To characterize the chromatin changes associated with allelicexclusion at the DP stage of thymocyte development and to iden-tify the cis-regulatory elements involved, we assayed for DNase Ihypersensitive sites, DNA methylation, and transcription in DNand DP thymocytes in a 100-kb region of the TCR-b locus starting20 kb upstream of Db1 and ending 50 kb downstream of Vb14.

Contrasting the current understanding of the control of V(D)J re-combination, our data suggest that DNase I hypersensitivity, DNAmethylation, and transcription in general may not be reliable mo-lecular indicators of inaccessibility to the recombinase during al-lelic exclusion. In addition, we demonstrate that three of theDNase I hypersensitive sites identified are not critical for TCR-bgene allelic exclusion. These findings suggest that alternativecis-regulatory elements within the assayed region such as Eb enhancerand/or in the upstream Vb region are likely to play a dominant rolein allelic exclusion control.

Materials and MethodsMice

RAG-1- and RAG-2-deficient mice were obtained from Drs. SusumuTonegawa (Massachusetts Institute of Technology, MA) and Fred Alt(Harvard Medical School, MA), respectively (65, 66). Activatedlck trans-genic mice were from Dr. Roger Perlmutter (University of Washington,WA) (67). Two different TCR-b transgenic mouse strains, one from Dr.Mark Davis (Stanford University School of Medicine, CA) and one fromDr. Eugenia Spanopoulou (Mount Sinai School of Medicine, NY), wereused in the present studies. The former was constructed from a genomicTCR-b transgene, and the latter from a cDNA TCR-b transgene (68, 69).The two strains of TCR-b transgenic mice are essentially the same in ourassays, except that Jb2-Cb2 intronic probes did not hybridize to the cDNATCR-b transgene, and thus simplified Southern blot hybridization patternsin this assay. The same pattern of DNase I hypersensitive sites was detectedin DN thymocytes from either RAG-1- or RAG-2-deficient mice (data notshown). Similarly, the same pattern of DNase I hypersensitive sites wasdetected in DP thymocytes from anti-CD3e Ab-treated RAG-1- or RAG-2-deficient mice (data not shown), using a protocol developed by Shinkaiand Alt (70), and later experiments were conducted with RAG-2-deficientmice only. TCR-b and lck transgenes were also on the RAG-2-deficientbackground. Cre transgenic deleter mouse was from Klaus Rajewsky (Uni-versity of Cologne, Germany) (71). NZW mice were from The JacksonLaboratory (Bar Harbor, ME). The genomic TCR-b transgene was alsointroduced into the NZW mice and mice harboring a targeted deletion ofHS1 (see below). Mice were maintained under specific pathogen-free con-dition in the animal facilities at Massachusetts Institute of Technology(Cambridge, MA).

Probes and genomic clones

Two cosmid clones containing Db1, Jb1, Cb1, Db2, Jb2, Cb2, Vb14, and30 kb downstream of Vb14 were kindly provided by Dr. Marie Malissen(Centre d’Immunologie, INSERM-CNRS de Marseille-Luminy, France)(72). A genomic clone extending 20 kb further downstream of the existingclones was identified by PCR from a Stratagene (La Jolla, CA) AKR mouseSuperCos library and by filter hybridization using probe H (Table I and Fig.2A). Probes isolated from the cosmid clones are summarized in Table I andin Figure 2A. The probes used to assay for homologous recombination in

Table I. Summary of the probes used to identify DNase I hypersensitivesites and their respective hybridizing fragments

Probes Hybridizing Fragments

Name Size (kb) Digest Size (kb) Digest

A 0.6 HindIII-NcoI 9.0 BglII-HpaIB 0.9 ClaI-EcoRI 3.3 EcoRVC 1.9 NcoI-HindIII 6.0 HpaID 1.2 BamHI-SpeI 4.0 EcoRVE 1.6 HindIII-BgII 6.7 BamHI-BgIIF 1.2 BgII-BgIII 6.0 BgIIG 1.1 BgIII-EcoRI 9.3 BgIIH 0.8 PstI-XbaI 9.0 EcoRII 0.8 ApaI-EcoRI 10.5 HpaIJ 0.7 ApaI-EcoRI 16.0 ApaIJ 0.7 ApaI-EcoRI 10.0 HpaIK 0.9 EcoRV-PstI 9.4 EcoRIK 0.9 EcoRV-PstI 6.0 BamHIL 2.4 HindIII-HindIII 10.0 EcoRVM 1.3 EcoRI-EcoRV 6.0 EcoRI

1257The Journal of Immunology

by guest on August 16, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

ES cells and for allelic exclusion in thymocytes are as follows: probe 1,0.6-kb HindIII-PstI fragment; probe 2, 0.9-kbEcoRV-PstI fragment; andprobe 3 is the same as probe H. Probes used for Northern hybridizationwere provided by Dr. Susumu Tonegawa and are as follows: Cb2, a 430-bpcDNA fragment; Vb8, a 190-bpEcoRI-PstI fragment; and Vb14, a 668-bpAccI genomic fragment.

Isolation of thymocytes

Single cell suspensions of thymocytes were prepared for the isolation ofDNA and RNA. Because RAG-deficient thymus contains only a few mil-lion cells compared with a hundred million cells in normal thymus (TableII), larger numbers of thymus were pooled to obtain enough cells for DNAand RNA isolation. Special caution was taken to remove thymic stromalcells by filtering through a nylon mesh. Thymocyte preparations were as-sayed for purity by FACS staining for CD4 and CD8. In general, thymo-cytes contained at least 95% DN cells from RAG mice and 95% DP cellsfrom TCR-b/RAG, lck/RAG, and CD3/RAG mice. Thymocytes of TCR-bmice contained 80% DP cells.

DNase I treatment

Nuclei from DN and DP thymocytes were prepared according to Forresteret al. (73). Briefly, 13 108 cells were washed in cold PBS, centrifuged at1000 rpm for 5 min, and resuspended into 10 ml reticulocyte standardbuffer (RSB) containing 10 mM Tris-Cl, pH 8, 10 mM NaCl, and 10 mMMgCl2. After slowly adding an equal volume of cold RSB containing 0.2%Nonidet P-40, the mixture was kept on ice for 10 min and centrifuged at900 rpm for 5 min, and the resulting nuclei were resuspended in 2 ml ofRSB. Two-hundred-microliter aliquots of nuclei were then treated with arange of DNase I (0.125–15mg/ml) at 37°C, and the digestion was stoppedby adding 200ml of 23 lysis buffer containing 1.2 M NaCl, 20 mM Tris-Cl, pH 8, 10 mM EDTA, and 1% SDS. The lysates were treated withproteinase K overnight at 55°C, and DNA isolated by phenol-chloroformextraction.

Southern and Northern hybridization

For Southern analysis, 10 to 15mg of DNA from DN and DP thymocyteswere digested with specific enzymes and were fractionated on a 0.9% aga-rose gel. After denaturation and neutralization, DNA was transferred toZ-probe filters and hybridized with specific probes labeled with[a-32P]dCTP. For Northern analysis, 10 or 20mg of RNA from DN and DPthymocytes were fractionated on a 1% formamide-agarose gel. RNA wastransferred onto Z-probe filters and hybridized with specific probes labeledwith [a-32P]dCTP. Filters were washed twice for 30 min in 23 SSC and0.1% SDS at 65°C. Hybridization signals were detected by phosphor im-aging and autoradiography.

Targeted mutagenesis of HS1

Mice harboring a targeted deletion of HS1 were constructed by targetedmutation in embryonic stem (ES) cells and then deriving a mouse strainfrom the mutant ES cells. The targeting vector was constructed by replac-ing a 780-bpXcmI-BglII fragment containing HS1 with a phosphoglyceratekinase promoter-driven neomycin-resistance (neo) gene flanked by loxP

sites (Fig. 5). The targeting vector contained 5.9-kb and 8.2-kb homolo-gous sequences at the 59 and 39 ends, respectively, and the thymidinekinase gene was inserted just outside of the 39 homologous sequence. Thetargeting vector was transfected into J1 ES cells, and doubly resistantclones were picked from three independent transfections. Homologous re-combinants were identified byEcoRI digestion of DNA and hybridizationwith a 0.6-kbHindIII-PstI fragment that does not hybridize to randomlyintegrated constructs. Among 258 clones analyzed, 31 yielded the expectedhybridizing fragments at 10 and 7.4 kb from the normal and the targetedalleles, respectively. The correct targeting of the 31 positive ES cell cloneswas confirmed by Southern blotting using the deleted 780-bp fragment asa probe, and then a 0.9-kbBglII-NcoI fragment containing the Eb as aprobe. ES cells from 5 of the 31 clones were injected into C57BL/6 blas-tocysts. Chimeras with more than 95% agouti coat color were generatedfrom all five clones and were bred with cre transgenic deleter mice forgermline transmission and deletion of the introduced neo gene at the sametime (71). Heterozygous mutant mice were interbred to obtain homozygousmutant mice, and the genomic TCR-b transgene was bred onto the mutantbackground. Data shown are from littermates and/or age-matched mice.

ResultsExperimental design

To identify chromatin changes associated with allelic exclusion,we chose the TCR-b locus for study because it fulfilled two re-quirements (Fig. 1). First, the DNA region in which we sought tocharacterize differences in DNase I hypersensitive sites, DNAmethylation, and transcription could be narrowed down to a man-ageable size. The IgH locus, which spans several megabases (74),is too large for such a study. In contrast, the entire TCR-b locus iswithin 600 kb (75, 76). Although it is still not feasible to analyzethe entire TCR-b locus, we postulated that the chromatin changesassociated with TCR-b allelic exclusion would most likely occurin a 100-kb region, starting 20 kb upstream of Db1 and ending 50kb downstream of Vb14 (Fig. 1, seeDiscussion). Second, DN andDP thymocytes representing the two developmental stages beforeand after allelic exclusion were readily available (77). Althoughthe normal thymocyte population is a mixture containing approx-imately 5% DN, 80% DP, and 15% CD4 or CD8 single-positive(SP) thymocytes, a relatively pure population of DN and DP thy-mocytes can be obtained easily from several mutant/transgenicmouse strains (Table II). In RAG-2-deficient mice (referred to asRAG mice), thymocytes are all DN due to a block in V(D)J re-combination (66). These DN thymocytes are arrested at theCD442CD251 stage at which TCR-b gene recombination nor-mally occurs (78, 79). Thus, the status of DNase I hypersensitivity,

FIGURE 1. Genomic organization of the TCR-b locus. Vertical lines and boxes indicate Vb, Db, Jb, and Cb gene segments, and horizontal arrowsindicate transcriptional orientations of the respective gene segments. A trypsinogen gene (Try) is located 20 kb upstream of Db1.

Table II. List of transgenic/knockout mice and their thymocyte phenotypes and numbers

Mouse Description

Thymocyte

Phenotype Number

RAG RAG-2-deficient mice DN 106

CD3/RAG RAG-2-deficient mice treated with anti-CD3e Abs DP 108

lck/RAG RAG-2-deficient mice complemented with an activated form oflck transgene DP 108

TCR-b/RAG RAG-2-deficient mice complemented with a TCR-b transgene DP 108

TCR-b TCR-b transgenic mice on normal (RAG1) background DP (80%) 108

1258 CHROMATIN CHANGES AND TCR-b GENE ALLELIC EXCLUSION

by guest on August 16, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

DNA methylation, and transcription in the region most likely rep-resents the chromatin structure that is accessible to the V(D)J re-combinase. On the other hand, introduction of a functionally as-sembled TCR-b transgene or an activated form of protein tyrosinekinase lck transgene onto the RAG-2-deficient background (re-ferred to as TCR-b/RAG andlck/RAG mice, respectively), or in-jection of anti-CD3e Abs into RAG-2-deficient mice (referred to asCD3/RAG mice) restores the differentiation of DP thymocytes tonormal levels (Table II) (70, 80–82). Thymocytes in TCR-b/RAG, lck/RAG, and CD3/RAG mice do not undergo further dif-ferentiation into SP T cells due to their inability to rearrange theTCR-a locus and express a functional TCR, and thus are more than95% DP. As a result of signaling through the pre-TCR pathway(28), the endogenous TCR-b loci in these DP thymocytes arelikely to have undergone chromatin changes associated with allelicexclusion, even though they are still in the germline configuration.Nevertheless, to more closely mimic the normal situation, we alsoanalyzed DNase I hypersensitive sites in thymocytes (80% DP)from TCR-b transgenic mice (RAG1) (referred to as TCR-b mice)in which the endogenous TCR-b loci undergo DbJb rearrange-ment (Table II) (20, 21). If the same pattern of DNase I hypersen-sitive sites, methylation, and transcription is observed in DP thy-mocytes from TCR-b, TCR-b/RAG, lck/RAG, and CD3/RAG

mice, these changes most likely represent the chromatin structurethat is inaccessible to the V(D)J recombinase.

Analysis for DNase I hypersensitive sites

To begin our investigation, DNase I hypersensitive sites at theTCR-b locus were assayed in DN and DP thymocytes. Thymo-cytes were harvested from all five mutant/transgenic mice (TableII). Aliquots were analyzed by flow cytometry for CD4 and CD8expression. Thymocytes were 98% DN from RAG mice; 97% DPfrom TCR-b/RAG, lck/RAG, and CD3/RAG mice; and 3% DN,80% DP, and 17% SP from TCR-b mice (data not shown). Nucleiwere prepared from the remaining thymocytes and treated withdifferent amounts of DNase I (seeMaterials and Methods). DNAisolated was assayed for DNase I hypersensitive sites by Southernblotting using specific probes isolated from the 100-kb region. Fig-ure 2Ashows the strategy of the analysis and summarizes probesused and their hybridizing fragments, and the latter are describedin more detail in Table I. Southern blotting using probes B, C, E,F, and K detected DNase I hypersensitive sites. As an example, theidentification of hypersensitive site 1 to 4 with probe E is detailedbelow. Southern blotting using probes A, D, G, H, I, J, L, and Mdid not detect any hypersensitive sites, and these assays will not bedescribed further.

FIGURE 2. Analysis of DNase I hypersensitive sites at the TCR-b locus in DN and DP thymocytes.A, Summary of DNase I hypersensitive sites (HS)identified in the 100-kb region of the TCR-b locus. Probes used in the study are labeled A to M, and their corresponding hybridizing fragments are shownas the lines below (see Table I). Arrows indicate the positions of the DNase I hypersensitive sites 1 through 11. Major and minor hypersensitive sites areindicated by larger and smaller arrows, respectively. Open circles are sites that did not change between DN and DP thymocytes, and filled circles are sitesthat were induced in DP thymocytes.B, Autoradiograms of Southern hybridization experiments showing HS1 through HS4. DNA was isolated from DNaseI-treated nuclei of DN (RAG) and DP thymocytes (CD3/RAG,lck/RAG, TCR-b/RAG, and TCR-b), digested withBamHI plusBglI, and hybridized withprobe E. In all five panels, DNA in thefirst lanewas isolated from nuclei that were not treated with DNase I. For DNase I treatment, the lowest concentrationwas 0.125mg/ml. DNase I concentration was doubled for each successive treatment. For example, DNase I concentrations used to treat nuclei of DNthymocytes were 0.125, 0.25, 0.5, and 1mg/ml. DNase I concentrations used to treat nuclei of DP thymocytes fromlck/RAG mice were 0.125, 0.25, 0.5,1, and 2mg/ml.

1259The Journal of Immunology

by guest on August 16, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

DNA was digested byBamHI plusBglI, and the blots werehybridized with probe E, a 1.6-kbHindIII-BglI fragment from 59of the Vb14 gene segment (Table I and Fig. 2A). In addition to theexpected 6.7-kb fragment, with increased DNase I concentrations,four smaller hybridizing fragments representing HS1, HS2, HS3,and HS4 were detected at 4.4, 3, 1.2, and 0.8 kb, respectively (Fig.2B). The minor fragments (HS3 and HS4) were not hybridizationto sequences between HS1 and HS2, because probe E does notextend to the region (Fig. 2A). Furthermore, they were detected byprobe E whenNcoI that cuts in between HS2 and HS3 was used todigest DNAs (data not shown). HS2 and HS3 were detected in allfive DNA samples, and therefore are present in both DN and DPthymocytes. In contrast, HS1 was only faintly detectable in DNAfrom DN thymocytes and was strongly induced in DNAs from DPthymocytes. Similarly, HS4 appeared to be present only in DPthymocytes from CD3/RAG,lck/RAG, TCR-b/RAG, and TCR-bmice (Fig. 2B).

Based on the fragment sizes, HS1 corresponds to a previouslyidentified DNase I hypersensitive site found in pre-T cell lines, andHS2 corresponds to Eb (83–86). This was further confirmed bydigesting DNA with the same enzymes and hybridization with thesame probes as in previous studies (data not shown) (85). In ad-dition, DNA was digested withBamHI plus one of the followingenzymes,BglI, EcoRI, NcoI, PstI, XbaI, or HpaI, which cut atdifferent positions successively downstream of HS1. Hybridizationof the filter with probe D from theBamHI (59) end of the fragmentgave rise to different sizes of the endogenous fragments (data notshown). Comparing the positions of HS1 and HS2 with the en-dogenous fragments enabled a more precise determination of thelocations of HS1 and HS2 and showed that HS1 and HS2 colo-calize with the previously identified hypersensitive sites.

In summary, we have assayed for DNase I hypersensitive sitesin a 100-kb region, starting 20 kb upstream of Db1 and ending 50kb downstream of Vb14, and have found a total of 11 DNase Ihypersensitive sites (Fig. 2). All of the identified sites are locatedwithin a 30-kb region, from 3 kb upstream of Db1 to 3 kb down-stream of Vb14. Eight of the sites (HS2, HS3, HS5–8, HS10, andHS11) are present in both DN and DP thymocytes; three (HS1,HS4, and HS9) are strongly induced in DP thymocytes. These data

show that in DP thymocytes, the region of the TCR-b locus that wehave analyzed remains accessible, and in some regions there areincreases in accessibility to factor-binding activities.

Analysis for DNA methylation

To further characterize chromatin changes at the TCR-b locus,DNA methylation status was assayed in DN and DP thymocytes.Within the 100-kb region in which DNase I hypersensitive siteswere assayed, 46 kb containing all of the identified DNase I hy-persensitive sites have been sequenced (L. Rowen and L. Hood,personal communication). Within this 46-kb region, there are 23MspI/HpaII sites and 19HhaI sites, most of them clustered in twosubregions, one surrounding Jb2 and the other downstream of theVb14 gene segment (Fig. 3C). The methylation status of thesesites in DN and DP thymocytes was determined by Southern blot-ting using methylation-sensitive enzymes.

For example, DNA from DN and DP thymocytes was digestedwith EcoRI plusMspI,EcoRI plusHpaII, or EcoRI plusHhaI, andthe blot was probed with the 5.1-kbEcoRI fragment covering theVb14 gene segment (Fig. 3B). Sequence data showed the presenceof oneMspI/HapII site and threeHhaI sites in the 5.1-kb region.EcoRI andMspI digestion yielded the two expected size frag-ments, one at 3.8 kb and the other at 1.3 kb in all of the samples(Fig. 3A,lanes 1,4,7, and10). EcoRI andHpaII digestion of DNAfrom DN thymocytes of RAG mice yielded only the 5.1-kb frag-ment (Fig. 3A,lane 2), indicating thatHpaII site is methylated inthese cells. In contrast, DNA from DP thymocytes of CD3/RAG,lck/RAG, and to a lesser extent, TCR-b/RAG mice yielded the5.1-kb as well as the 3.8- and 1.3-kb fragments (Fig. 3A, lanes 5,8, and11). Quantitation of the fragments showed that 5 to 20% ofthe DNAs are demethylated at theHpaII site in DP thymocytes.

EcoRI andHhaI digestion of DNA from DN thymocytes gen-erated 2.6- and 2.4-kb hybridizing fragments at equal intensities(Fig. 3A,lane 3). Since there are threeHhaI sites in the fragment,the only possibility for this hybridization pattern is that the firstHhaI site is unmethylated and the other two sites are completelymethylated (Fig. 3B). Similarly, this HhaI site is also unmethyl-ated in DP thymocytes from TCR-b/RAG, CD3/RAG, andlck/

FIGURE 3. Analysis of DNA methylation status at the TCR-b locus in DN and DP thymocytes.A,Autoradiogram of a representative Southern hybridization showing the methylation status of sites aroundthe Vb14 gene segment. DNA from DN (RAG) and DP (CD3/RAG,lck/RAG, TCR-b/RAG) thymocyteswere digested withEcoRI (lanes 1–12) plusMspI,HpaII, or HhaI, and hybridized with the 5.1-kbEcoRIprobe.B, Positions ofMspI (M) andHhaI (H) sites in the 5.1-kbEcoRI (RI) fragment. Possible fragmentsizes (kb) after digestion withEcoRI plusMspI (orHpaII), andEcoRI plusHhaI are indicated. Open boxrepresents Vb14 gene segment.C, Summary of DNA methylation status of TCR-b locus in DN and DPthymocytes. The positions ofMspI/HpaII andHhaI sites are indicated by vertical lines, and open circlesrepresent unmethylated sites. X represents partially methylated/demethylated sites, and filled circlesrepresent fully methylated sites.

1260 CHROMATIN CHANGES AND TCR-b GENE ALLELIC EXCLUSION

by guest on August 16, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

RAG mice because the largest and dominant hybridizing frag-ments are at 2.6 and 2.5 kb (Fig. 3A, lanes 6,9, and 12). Inaddition, two smaller hybridizing fragments of 1.7 and 1.1 kb weredetected at submolar levels in DNA from DP thymocytes. (A0.6-kb fragment was also visible on the original autoradiogram.The 0.6-kb fragment was expected to hybridize least because theentire 5.1-kb fragment was used as probe.) These patterns weregenerated when the other twoHhaI sites were partially digesteddue to their partial demethylation. Quantitation of the fragmentssuggests 5% of the DNA is demethylated at the other two sites.Together, these data suggest that among the four detectable meth-ylation sites in the 5.1-kbEcoRI fragment, one is unmethylated inboth DN and DP thymocytes, and the other three sites are meth-ylated in DN thymocytes and partially demethylated in DPthymocytes.

We have determined the methylation status of most of theMspI/HpaII andHhaI sites in the 46-kb region. The results are summa-rized in Figure 3C. First, most of the sites in and around Jb2 genesegments are unmethylated in both DN and DP thymocytes. Threeof the sites in the region become completely unmethylated in DPthymocytes. Second, among the four sites immediately down-stream of the Vb14 gene segment, one site is unmethylated in bothDN and DP thymocytes, and the other three become partially dem-ethylated during DN to DP thymocyte differentiation. Third, mostsites farther downstream of the Vb14 gene segment, including thecluster of nineHhaI sites, are methylated in both DN and DPthymocytes. These findings suggest that DN to DP thymocyte dif-ferentiation is associated with demethylation in the TCR-b locus.

Analysis for germline transcription

Chromatin changes at the TCR-b locus were also assayed by com-paring germline transcripts from the locus in DN and DP thymo-cytes. Northern blot hybridization was performed with total RNA(10mg) from thymocytes of RAG, CD3/RAG,lck/RAG, and TCR-b/RAG mice. As controls, RNA from the EL4 T cell line and athymoma derived from a DN thymocyte of a p53 and RAG-1 dou-ble-knockout mouse were used. Hybridization with a 430-bp Cb2cDNA probe revealed the presence of 1- and 1.6-kb germline tran-scripts from DN thymocytes that comigrated with those from thecontrol thymoma (Fig. 4A). These germline transcripts were thesame sizes as those identified previously from the TCR-b locus inDN thymocytes from normal and mutant mice (30, 78, 87). Theunique 1.3-kb transcript present in RNA from TCR-b/RAG thy-mocytes comigrated with the mature TCR-b transcript from theEL4 T cell line and was most likely derived from the transgene inthese mice. The hybridization signal larger than 28S was not in-formative, as it may have resulted from trace genomic DNA con-tamination in the RNA samples.

A duplicate filter was hybridized with a 680-bp Vb14 probe.Two weakly hybridizing species, one at 1 kb and the other at 0.8kb, were detected in RNA from DP thymocytes of CD3/RAG,lck/RAG, and TCR-b/RAG mice, but not from DN thymocytes ofRAG mice, thymoma, or EL4 (Fig. 4B). These transcripts repre-sent germline transcription from the endogenous Vb14 gene seg-ment, as the TCR-b transgene in the TCR-b/RAG mice utilizes theVb3 gene segment. To determine the transcription status of 59 Vbgene segments in DN and DP thymocytes, the same two filterswere stripped and hybridized with a 190-bp Vb8 probe or otherVb probes, including Vb6, 7, 9, and 13. In contrast to Vb14, thelevel of Vb8 germline transcript (including Vb8.1, 8.2, and 8.3)was much higher in DN thymocytes from RAG mice than in DPthymocytes from CD3/RAG,lck/RAG, and TCR-b/RAG mice(Fig. 4C). No RNA transcripts were detected when Vb6, 7, 9, or13 probes were used individually (data not shown). Together, these

data suggest that germline transcription in Cb-Vb14 region is in-creased in DP thymocytes, indicative of increased accessibility inthis region, whereas germline transcription in the upstream Vbregion is nonexistent or decreased in DP thymocytes, suggesting aregion with decreased accessibility.

Analysis of the role of HS1 in TCR-b gene allelic exclusion

We next focused our investigation on determining whethercis-regulatory elements contained within HS1 control TCR-b geneallelic exclusion. HS1 was interesting because it was the mostprominent chromatin structural change identified in the TCR-blocus between DN and DP thymocytes, and because it was only400 bp upstream of Eb. Considering thatcis elements within Ebhave a dominant role in promoting TCR-b VDJ recombination (41,55, 63, 64), it was possible that factors binding tocis elements inHS1 might interact with and modulate those binding to Eb andeffectively suppress recombination at the entire locus, thereby me-diating allelic exclusion. To test this possibility, a 780-bpXcmI-BglII fragment containing HS1 was replaced with a floxed neogene by homologous recombination in ES cells, and chimeric micewere produced for germline transmission of the alteration (Fig. 5;Materials and Methods). To minimize effects from the introducedneo gene, it was deleted by Cre/loxP-mediated recombination bybreeding mutant mice with deleter mice, in which Cre is expressedat the two-cell stage of embryonic development (71). After Cre-mediated recombination, a single 34-nucleotide loxP site remainsthat is unlikely to influence the phenotype resulting from the HS1

FIGURE 4. Analysis of transcripts from the TCR-b locus in DN and DPthymocytes.A, Autoradiogram of a Northern hybridization with a Cb2probe to detect all Cb-containing transcripts in total RNA from DN (RAG)and DP (CD3/RAG,lck/RAG, TCR-b/RAG) thymocytes. EL4 represents amature CD4-positive thymoma, and expresses the mature 1.3-kb TCR-btranscript from a fully assembled allele as a control. Thymoma was froma p53 and RAG-1 double-mutant mouse, and expresses germline transcriptsof 1 and 1.6 kb from unrearranged alleles as a control.B, Autoradiogramof a duplicate filter hybridized with a Vb14 probe.C, Autoradiogram of thesame filter, as inA, hybridized with a Vb8 probe.D, Ethidium bromidestaining of 28S RNA on blot A to show the amount of RNA loaded in eachlane.

1261The Journal of Immunology

by guest on August 16, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

deletion. The resulting mutant allele is detected as a 7.4-kbEcoRIfragment by Southern hybridization using probe 1 (Fig. 5B). Ho-mozygous mutant mice with or without the neo gene deletedyielded normal numbers of thymocytes and peripheral T cells inspleen and lymph nodes (see below and data not shown). A com-parison of DNase I hypersensitivity of thymocyte nuclei from nor-mal and homozygous mutant mice, in which the neo gene wasdeleted, confirmed that HS1 was deleted successfully in mutantmice (Fig. 5C). HS2, HS3, and HS4 remained in the mutant micedespite the deletion of HS1, indicating that these DNase I hyper-sensitive sites form independently of HS1.

Phenotypic analyses of cells from spleen, lymph node, and thy-mus by FACS showed that T cell development and TCR-ab ex-pression in the mutant mice were indistinguishable from that ofnormal mice (Fig. 6). Additional FACS analyses of lymph node T

cells stained with a panel of Abs specific for individual Vbs re-vealed no significant difference in Vb repertoire usage betweenmutant and normal mice (data not shown). In addition, amplifica-tion of thymocyte DNA from normal and mutant mice by PCR todetect Db1 to Jb1.1 through Jb1.5, and Db1 to Jb2.1 throughJb2.5 rearrangements revealed no differences between the normaland mutant mice in D to J rearrangement (data not shown). Thus,the deletion of HS1 does not grossly affect TCR-b gene rearrange-ment, expression, and T cell development.

The effect of HS1 deletion on TCR-b gene allelic exclusion wasnext investigated by introducing a functionally assembled TCR-btransgene onto the homozygous mutant background and determin-ing the extent of allelic exclusion by two different assays. First,lymph node T cells from normal and mutant mice with and withoutthe TCR-b transgene were assayed by FACS for endogenous Vbgene expression, including Vb4, 5.1 and 2, 6, 7, 8.1 and 2, 10b, 12,13, and 14. In the presence of the TCR-b transgene, the percent-ages of T cells expressing the endogenous Vb were reduced to thesame extent in both normal and mutant mice (data not shown).Simultaneous staining for Vb3 of the TCR-b transgene and a panelof endogenous Vbs did not detect any dual TCR-b-expressing Tcells in the mutant mice (data not shown), indicating a normalTCR-b gene allelic exclusion after HS1 deletion. Second, the ex-tent of allelic exclusion was measured by Southern blot hybrid-ization for the retention of a 2.2-kbPstI-fragment located upstreamof Db1 (Fig. 7A). In normal mice, the level of hybridization to the

FIGURE 5. Deletion of HS1 from the TCR-b locus by targeted mu-tagenesis and Cre/loxP-mediated recombination.A, Schematic diagrams ofthe targeting vector and targeted alleles.B, Autoradiogram of a represen-tative Southern hybridization for genotyping mutant mice. Tail DNA wasdigested withEcoRI and hybridized with the probe 1. Wild-type allele, 10kb, and mutant allele, 7.4 kb.C, Autoradiograms of Southern hybridizationshowing the successful deletion of HS1 in mutant mice. DNA was isolatedfrom DNase I-treated thymocyte nuclei of normal (1/1) and homozygousmutant (2/2) mice, digested withBamHI plusBglI, and hybridized withprobe E (see Fig. 2).

FIGURE 6. Phenotypic analysis of the effect of HS1 deletion on T celldevelopment.A andB, Dot plots from FACS analysis of thymocytes (A)and lymph node cells (B) from normal and homozygous mutant micestained with anti-CD4 and anti-CD8 Abs. The number indicates the per-centage of cells in each quadrant.C, Dot plots of lymph node cells fromnormal and homozygous mutant mice stained with anti-TCR-ab and anti-CD3e Abs. The number indicates the percentage of cells in each quadrant.D, Histograms showing the levels of TCR-ab expression on T cells fromnormal and homozygous mutant mice (gated on R1).

1262 CHROMATIN CHANGES AND TCR-b GENE ALLELIC EXCLUSION

by guest on August 16, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

2.2-kbPstI fragment was almost undetectable in thymocyte DNA(Fig. 7B). In the presence of the TCR-b transgene, the levels ofhybridization were the same in DNA from thymocytes and kidney,indicating the occurrence of allelic exclusion. The retention of thesignal in mice expressing the TCR-b transgene is not due to dif-ferences in DNA loading, as indicated by the similar level of si-multaneous hybridization of probe 3 to a 10-kbPstI control frag-ment located downstream of Vb14 (Fig. 7, A and B). If thedeletion of HS1 abolished allelic exclusion, we expected that Vbto DbJb recombination would persist in the presence of the TCR-btransgene and result in a lower level of retention of the 2.2-kbPstIfragment located upstream of Db1. However, the level of hybrid-ization was the same in the mutant samples as in the normal mice(Fig. 7B). The same results were also obtained when DNA isolatedfrom purified lymph node T cells of mutant and normal mice wereused for the assay (data not shown). In addition, ligation-mediatedPCR measuring the levels of dsDNA break at the 59 of Db1 and atthe 39 of Vb14 gene segment in thymocytes failed to detect anyconsistent difference between the normal and homozygous mutantmice (data not shown). Therefore, these combined data show thatcis elements within HS1 are not required for TCR-b gene allelicexclusion.

Analysis of the role of HS7 and HS8 in allelic exclusion

We next considered the possibility thatcis-regulatory elementswithin HS7 and/or HS8 might play a role in controlling allelic

exclusion because these sites are maintained after Db to Jb re-combination (Fig. 2). In NZW mice, a natural deletion of 8.8 kb ofDNA in the TCR-b locus has removed Cb1, Db2, and Jb2 genesegments, as well as HS7 and HS8 (88). Therefore, to assess therequirement ofcis-regulatory elements associated with HS7 andHS8 for TCR-b gene allelic exclusion, we bred a functionalTCR-b transgene onto the NZW mice, isolated thymocyte DNA,and then measured for the extent of allelic exclusion by Southernhybridization of probe 2 to the 2.2-kbPstI fragment located up-stream of Db1. As shown in Figure 7C, Vb to DbJb rearrange-ment of TCR-b in NZW mice was inhibited effectively by expres-sion of the TCR-b transgene, as indicated by the increasedhybridization signal to the 2.2-kbPstI fragment compared withnormal NZW mice. Therefore, deletion of HS7 and HS8 does notgrossly affect TCR-b gene allelic exclusion. FACS analyses oflymph node T cells from NZW mice stained with a panel of Absto endogenous Vbs to detect dual TCR-b expression confirmedthat TCR-b allelic exclusion was intact in the NZW mice (data notshown). These data demonstrate thatcis-regulatory elementswithin HS7 and HS8 are not required for TCR-b gene allelicexclusion.

DiscussionChromatin structure at the TCR-b locus in DN and DPthymocytes

Studies to identify all potential chromatin changes of the TCR-blocus are inherently difficult due to its immense size of 600 kb(Fig. 1) (75, 76). In our study, we first considered the followingfacts about the TCR-b locus to narrow down the region of analysis.Because allelic exclusion is regulated at the V to DJ rearrangementstep, it is unlikely thatcis elements involved in maintaining allelicexclusion reside between Db1 and Jb2.7. Therefore, we focusedour effort on identifying chromatin changes occurring in a 100-kbregion from 20 kb upstream of Db1 to nearly 70 kb downstream ofJb2. This represents the most comprehensive study ever under-taken for an Ag receptor locus. Second, we did not study beyond20 kb upstream of Db1 because of the presence of a trypsinogengene that, in our thinking, may separate upstream Vbs from thedownstream gene segments and associated regulatory elements.Third, it was impractical to study chromatin changes in the entireupstream Vb region due to its immense size. In addition, previousstudies using TCR-b minilocus substrates have suggested that afunctional Vb promoter is not required for allelic exclusion andthat the signal for allelic exclusion may be mediated by sequences39 of the Jb2.2 gene segment (89). Therefore, we focused our studyon the DNA flanking the unique Vb14 gene segment, which islocated downstream of the Db, Jb, and Cb gene segments andseems to be subjected to the same allelic exclusion control as allother Vb gene segments. The Vb14 gene segment is also inter-esting because of its inverted transcriptional orientation relative toall other TCR-b gene segments and the conservation of this con-figuration in human, mouse, and chicken. Because Vb14 rear-ranges to the DbJb complex by inversional joining, this configu-ration ensures that DNA between Vb14 and the DbJb complex isretained after rearrangement, suggesting that the Jb2 to Vb14 in-tervening region may contain criticalcis-regulatory elements re-quired for TCR-b gene recombination and/or transcription.

We identified a total of 11 DNase I hypersensitive sites in the100-kb region. Three of the sites, HS1, HS4, and HS9, are uniqueor strongly induced in DP thymocytes, whereas the other eightsites identified are present in both DN and DP thymocytes (Fig. 2).Six of the sites are located between the Jb2-Vb14 interveningregion, consistent with the importance of this region for TCR-b

FIGURE 7. Analyses of the effect of HS1 deletion, and HS7 and HS8deletion on TCR-b gene allelic exclusion.A, A schematic diagram ofTCR-b locus showing the position of probes 2 and 3 and their hybridizingfragments.B, Autoradiogram of Southern hybridization for retention ofDNA upstream of Db1 in normal and homozygous mutant mice with andwithout the TCR-b transgene. DNA from kidney (K) and thymocyte (T)was digested withPstI. Blot was hybridized with probe 2 to the 2.2-kbPstIfragment upstream of Db1 and with probe 3 to a 10-kbPstI fragmentlocated 20 kb downstream of Vb14 that serves as a normalization controlfor the amount of DNA loaded per lane.C, Autoradiogram of Southernhybridization for retention of DNA upstream of Db1 in DNA from kidney(K) and thymocytes (T) of NZW mice and NZW mice with a TCR-btransgene. DNA was digested withPstI and hybridized with probes 2 and3 simultaneously, as inB.

1263The Journal of Immunology

by guest on August 16, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

gene regulation. A general correlation between transcriptional ac-tivity of the TCR-b locus and the presence of the DNase I hyper-sensitive sites is apparent. For example, HS2 maps to the previ-ously identified Eb enhancer (83, 84, 86), and is present in bothDN and DP thymocytes (Fig. 2), consistent with the enhancer be-ing active at both stages of development. Indeed, we detectedgermline transcripts (1 and 1.6 kb) containing Cb sequences inboth DN and DP thymocytes (Fig. 4), and mostMspI/HpaII andHhaI sites in the Cb region are unmethylated (Fig. 3), a stateusually associated with active transcription. In DP thymocytes,demethylation of threeMspI/HpaII sites in the Jb2 region corre-lated with increased levels of Cb-containing transcripts and theinduction of HS1 and HS9. HS9 is located immediately upstreamof Db1, and analogous to Ig and TCR-a loci (51, 52, 90), mayserve as the promoter for germline transcription through the Db,Jb, and Cb regions. Consistent with this notion, a germline tran-script initiated immediately upstream of Db1 was identified re-cently in human fetal intestine (91). HS1 is 400 bp upstream of Eband has been implicated to influence transcription, perhaps bymodulating Eb activity (see below). Another site, HS5, maps to theVb14 transcriptional promoter. Although HS5 is present in bothDN and DP thymocytes, Vb14 germline transcripts are only de-tectable in DP thymocytes (Fig. 4). Vb14 transcripts correlate withthe partial demethylation of threeMspI/HpaII sites located up-stream of the Vb14 gene segment in DP thymocytes. These ob-servations suggest that if the factors that cause HS5 are involved intranscriptional activity, their binding alone is not sufficient for theinduction of Vb14 germline transcription. Additional factors orsignals induced during the DP thymocyte stage are also most likelyrequired for the induction of Vb14 transcription. In summary, ourfindings suggest that chromatin structure in the 100-kb region ofthe TCR-b locus is dynamic and accessible in both DN and DPthymocytes, as reflected by the presence of multiple DNase I hy-persensitive sites, hypomethylation, and active transcription. Wedid not find evidence supporting a model of allelic exclusion,wherein accessibility of the TCR-b locus is decreased dramaticallyduring the DP stage of thymocyte development.

Chromatin changes and TCR-b gene allelic exclusion

Our studies reveal that none of the chromatin changes at theTCR-b locus in DN and DP thymocytes is indicative of a typicalinaccessible locus. In DP thymocytes, 3 of 11 identified DNase Ihypersensitive sites are either unique or strongly induced, 6 of 33HpaII/HhaI sites show partial to complete demethylation, and in-creased germline transcription initiated from the Db-Cb regionand Vb14 is evident. Moreover, all DNase I hypersensitive sites,hypomethylation, and transcription observed in DN thymocytesare also maintained in DP thymocytes. The observed changes mostlikely represent genuine chromatin alterations that are induced bypre-TCR signaling because the same changes are observed in DPthymocytes by either the expression of a TCR-b or an activatedlcktransgene, or by anti-CD3e treatment. Because pre-TCR also sig-nals for TCR-b gene allelic exclusion (28), the observed chromatinchanges are either coincident or associated with allelic exclusion(38). Our data show that the TCR-b locus is accessible in DPthymocytes to the various enzymes and factors that bind DNAsequences and mediate the transcription and demethylation of thelocus. However, in spite of this, the locus remains inaccessible tothe V(D)J recombinase because the TCR-b gene is excluded fromfurther rearrangement in DP thymocytes (28). This observation issignificant for several reasons. First, although chromatin changeshave been postulated to block recombinase accessibility or to re-target the recombinase for achieving allelic exclusion, the under-lying mechanism is unknown (92). Our findings have eliminated

some possibilities by showing that the mechanism does not involvethe general shutdown of the locus, which would be reflected by anabsence of transcription and active hypermethylation of the locus,and perhaps the disappearance of DNase I hypersensitive sites.Second, and contrary to our expectations, the locus becomes hy-pomethylated and more actively transcribed. This suggests thatwhile transcription and hypomethylation may be required for theinitiation of V(D)J recombination (6, 7), and may be used to mea-sure general chromatin changes and accessibility, the status oftranscription and methylation may not be indicative of the recom-bination inaccessibility associated with the excluded TCR-b allele.Third, new DNase I hypersensitive sites are induced at the ex-cluded TCR-b locus. Therefore, ifcis-regulatory elements associ-ated with these new DNase I hypersensitive sites are involved inthe control of allelic exclusion, it seems likely that the associatedfactors would function as repressors of V to DJ recombination.

HS1, HS7, and HS8 are not required for TCR-b gene allelicexclusion

HS1 is 400 bp upstream of Eb and maps to a previously identifiedDNase I hypersensitive site found in precursor T cell lines (Fig. 2)(85). Although HS1 was shown to correlate with TCR-b expres-sion in cell lines (85), findings from our more physiologic exper-imental system demonstrate that HS1 is primarily present in DPthymocytes induced in RAG-deficient mice by expression of aTCR-b or activatedlck transgene or by anti-CD3e Ab treatment.Importantly, HS1 was also detected in DP thymocytes from normalTCR-b transgenic mice, in which the endogenous TCR-b loci un-derwent DbJb rearrangement and were excluded from Vb toDbJb rearrangement. Therefore, the data provide strong supportthat HS1 is induced by pre-TCR engagement during DN to DPthymocyte differentiation and could represent a chromatin changeassociated with TCR-b gene allelic exclusion.

Based on the close proximity of HS1 to Eb, we hypothesizedthat factors binding to HS1 during the DP thymocyte stage mightsuppress Eb function specifically in recombination, but not tran-scription, and consequently would promote allelic exclusion. Totest this hypothesis, we deleted a 780-bp DNA fragment encom-passing HS1 from the endogenous TCR-b locus by targeted mu-tagenesis. We found that while HS1 was deleted by this mutation,HS2, HS3, and HS4 remained (Fig. 5), suggesting that these sitesmay function independently of HS1. Compared with normal mice,homozygous mutant mice displayed normal TCR-b gene recom-bination, expression, and allelic exclusion either with or without aTCR-b transgene (Figs. 6 and 7). Therefore, our data show thatHS1 is not required for the control of allelic exclusion. This findingsuggests the possibility that othercis-regulatory elements in thelocus may have a primary role in allelic exclusion control. If HS1does participate in the process, it may do so in cooperation withother more essential and dominant elements. Although we do notcurrently know the function of HS1, its role in transcription hasbeen suggested both by its proximity to Eb and by the presence ofseveral conserved sequence motifs homologous to known tran-scriptional elements (85). Octamer, c-myb, AP-1, and enhancercore sequences of SV40 and polyoma virus are present within twosubregions that are most conserved between human and mouse (85and 95% homology). Nevertheless, HS1 has no enhancer activityin both transient transfection assays and transgenic mice (83, 84)(Erik Selsing and Jim Miller, personal communication). Further-more, our current biochemical evidence suggests that in DP thy-mocytes, factor binding to HS1 occurs in a region different fromthat described by the above consensus motifs and may function torepress TCR-b transcription (unpublished data). Additional studies

1264 CHROMATIN CHANGES AND TCR-b GENE ALLELIC EXCLUSION

by guest on August 16, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

are required to understand the function of the nuclear factors bind-ing to HS1.

HS7 and HS8, located within the Jb2-Cb2 intron, were detectedpreviously in T cell lines, and nuclear factor-kB and additionalfactors were found to interact with specific sequences in the region(93, 94). We found that HS7 and HS8 were present in both DN andDP thymocytes, suggesting a role in transcription and/or recom-bination of the TCR-b locus at both stages of development. Inter-estingly, in mice harboring a targeted replacement of 15 kb of thelocus encompassing Jb1.3 to Cb2, recombination of the remainingDb1, Jb1, and Vb gene segments did not occur. Although therecombination defect could be caused by the removal of otherunknown elements within the region important for recombination,or due to the insertion of a neomycin gene cassette that has beenshown to perturb recombination in other loci (56, 57, 61, 62, 95),we considered the possibility that the deletion of HS7 and HS8might be responsible for the observed defect. In addition, HS7 andHS8 might also function in allelic exclusion control. NZW micehave naturally deleted 8.8-kb region encompassing Cb1, Db2, andJb2 gene segments as well as HS7 and HS8 (88). From the liter-ature, there was no evidence of a defect in the surface expressionof TCR-b on NZW T lymphocytes, suggesting that transcriptionand recombination of the locus are normal. However, whether al-lelic exclusion was normal in these mice remained to be tested,especially considering that (NZW3 NZB)F1 mice have a highincidence of autoimmune disease characterized by high levels ofautoantibodies in the circulation (96–100), and that this diseasestate might be influenced indirectly by a loss of TCR-b gene allelicexclusion control. However, our analyses from Southern blottingand FACS demonstrated no apparent defect in the control ofTCR-b gene allelic exclusion in NZW mice after introduction of aTCR-b transgene (Fig. 7 and data not shown), indicating thatcis-regulatory elements within HS7 and HS8 are not required for theprocess. Therefore, it seems that the loss of HS7 and HS8 in NZWmice is unlikely to be a contributing factor to the development ofautoimmune disease in (NZB3 NZW)F1 mice.

Cis-regulatory elements involved in TCR-b gene allelicexclusion

DNase I hypersensitive sites are generated when nucleosomes arereplaced by specifictrans-acting factors that competitively bind totheir cognate sequences. The presence of hypersensitive sites usu-ally implies the presence ofcis-regulatory elements and their as-sociated factors. Although targeted deletion of HS1 or natural de-letion of HS7 and HS8 has no apparent effect on TCR-b generecombination, expression, or allelic exclusion, it remains possiblethat in normal circumstances these sites may contribute to any ofthese processes in cooperation with other more dominant regula-tory elements such as Eb in the assayed region. The Eb enhancerwas identified initially as a transcription enhancer (83, 84); it isalso critical for the recombination of the TCR-b gene. Targeteddeletion of a 0.5-kb region containing Eb from the endogenouslocus in mice results in a complete absence of Db to Jb rearrange-ment and therefore a block ofab T cell development (63, 64).Inclusion of Eb in a minilocus recombination substrate readilypromotes Vb to DbJb rearrangement in transgenic mice (41, 58).Thus, an effective mechanism of suppressing recombination of theentire TCR-b locus may be through the specific suppression of Ebfunction in recombination. It should be emphasized that our datasuggest that such a mechanism could not operate through decreas-ing transcription, as we observed increased transcriptional activityand associated demethylation of the TCR-b gene in DP thymo-cytes under conditions of allelic exclusion. Our data are consistentwith a mechanism wherein Eb may contain independent and sep-

arablecis elements that function to regulate either transcription orrecombination/allelic exclusion of the TCR-b gene (101).

If the mechanism of TCR-b gene allelic exclusion does involvea shutdown in chromatin accessibility to the recombinase, our dataare also consistent with the possibility that the mechanism mayinvolve cis elements residing completely outside of the 100-kbregion of DNA that we have analyzed, perhaps in the upstream Vbregion. Indeed, we did observe a significant down-regulation ofVb8 germline transcription in DP thymocytes compared with DNthymocytes, suggesting decreased accessibility to the transcrip-tional machinery. In this scenario, allelic exclusion of the down-stream Vb14 gene segment may involve a different mechanismsince we observed increased germline transcription of this genesegment and demethylation near its promoter in DP thymocytes.However, it remains possible that germline transcription of Vbgene segments may not be a reliable indicator of accessibility tothe recombinase. This is consistent with previous studies utilizingminilocus recombination substrates, in which a correlation be-tween transcription and recombination has not always been found(58, 101, 102). Interestingly, Alvarez et al. showed that while de-letion of the decamer motif of a Vb gene segment abolished tran-scription to below detectable levels, the rearrangement and allelicexclusion of the Vb gene segment remained unaffected (89).Therefore, allelic exclusion of each Vb gene segment may not becontrolled individually, but by dominantcis-regulatory elementsthat control allelic exclusion of all Vb gene segments (exceptVb14). If such dominantcis elements are present, they do notreside in the 100-kb region encompassing Vb5.2 to Vb9, becausethis 100 kb is deleted in SJL mice, and these mice have no apparentdefect in TCR-b gene allelic exclusion (75, 76). Alternatively,allelic exclusion could be mediated by multiple localcis-regula-tory elements, analogous to the T earlya element for the rear-rangement of the 59 Ja gene segments (90).

AcknowledgmentsThe authors thank Drs. Lee Rowen and Leroy Hood for making murineTCR-b sequences available before publication; Dr. Marie Malissen forproviding genomic clones containing the Db, Jb, Cb, and Vb14 regions;Drs. Susumu Tonegawa, Fred Alt, Roger Perlmutter, Mark Davis, and Eu-genia Spanopoulou for providing RAG-1, RAG-2,lck, and TCR-b trans-genic mice, respectively; Drs. Ranjan Sen and Susumu Tonegawa forprobes; Drs. Hidde Ploegh, Richard Hynes, Fred Alt, and Christiaan Levelt,and members of the laboratory for helpful discussions and for critical read-ing of the manuscript; and Gail Ackermann for technical assistance.

References1. Tonegawa, S. 1983. Somatic generation of antibody diversity.Nature 302:575.2. Alt, F., T. K. Blackwell, and G. Yancopoulos. 1987. Development of the pri-

mary antibody repertoire.Science 238:1079.3. Davis, M. M., and P. J. Bjorkman. 1988. T-cell antigen receptor genes and

T-cell recognition.Nature 334:395.4. Alt, F. W., and D. Baltimore. 1982. Joining of immunoglobulin heavy chain

gene segments: implications from a chromosome with evidence of three D-JH

fusions.Proc. Natl. Acad. Sci. USA 79:4118.5. Alt, F. W., G. Yancopoulos, T. K. Blackwell, C. Wood, E. Thomas, M. Boss,

R. Coffman, N. Rosenberg, S. Tonegawa, and D. Baltimore. 1984. Orderedrearrangement of immunoglobulin heavy chain variable region segments.EMBO J. 3:1209.

6. Lansford, R., A. Okada, J. Chen, E. M. Oltz, T. K. Blackwell, F. W. Alt, andG. Rathbun. 1996. Mechanism and control of immunoglobulin gene rearrange-ment. InMolecular Immunology, 2nd Ed. B. D. Hames and D. M. Glover, eds.IRL Press, Oxford, p. 1.

7. Sleckman, B. P., J. R. Gorman, and F. W. Alt. 1996. Accessibility control ofantigen-receptor variable-region gene assembly: role ofcis-acting elements.Annu. Rev. Immunol. 14:459.

8. Pernis, B., G. Chiappino, A. S. Kelus, and P. G. H. Gell. 1965. Cellular local-ization of immunoglobulins with different allotypic specificities in rabbit lym-phoid tissues.J. Exp. Med. 122:853.

9. Cebra, J. J., J. E. Colberg, and S. Dray. 1966. Rabbit lymphoid cells differen-tiated with respect toa, g, and m heavy polypeptide chains and to allotypicmarkers Aa1 and Aa2.J. Exp. Med. 123:547.

1265The Journal of Immunology

by guest on August 16, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

10. Padovan, E., C. Giachino, M. Cella, S. Valitutti, O. Acuto, and A. Lanzabecchia.1995. Normal T lymphocytes can express two different T cell receptorb chains:implications for the mechanism of allelic exclusion.J. Exp. Med. 181:1587.

11. Davodeau, F., M. A. Peyrat, F. Romagne, A. Necker, M. M. Hallet, H. Vie, andM. Bonneville. 1995. Dual T cell receptorb chain expression on human Tlymphocytes.J. Exp. Med. 181:1391.

12. Malissen, M., J. Trucy, F. Letourner, N. Rebai, D. E. Dunn, F. W. Fitch,L. Hood, and B. Malissen. 1988. A T cell clone expresses two T cell receptora genes but uses oneab heterodimer for allorecognition and self MHC-re-stricted antigen recognition.Cell 55:49.

13. Padovan, E. G., P. Casorati, P. Dellabona, S. Meyer, M. Brockhaus, andA. Lazavecchia. 1993. Expression of two T cell receptora chains: dual receptorT cells.Science 262:422.

14. Lee, N. E., and M. M. Davis. 1988. T cell receptorb-chain genes in BW5147and other AKR tumors: deletion order of murine Vb gene segments and possible59 regulatory regions.J. Immunol. 140:1665.

15. Rusconi, S., and G. Kohler. 1985. Transmission and expression of a specific pairof rearranged immunoglobulinm andk genes in a transgenic mouse line.Nature314:330.

16. Weaver, D., F. Costantini, T. Imanishi-Kari, and D. Baltimore. 1985. A trans-genic immunoglobulin Mu gene prevents rearrangement of endogenous genes.Cell 42:117.

17. Storb, U., C. Pinkert, B. Arp, P. Engler, K. Gollahon, J. Manz, W. Brady, andR. L. Brinster. 1986. Transgenic mice with mu and kappa genes encoding an-tiphosphorylcholine antibodies.J. Exp. Med. 164:627.

18. Nussenzweig, M. C., A. C. Shaw, E. Sinn, D. B. Danner, K. L. Holmes,H. C. I. Morse, and P. Leder. 1987. Allelic exclusion in transgenic mice thatexpress the membrane form of immunoglobulinm. Science 236:816.

19. Kitamura, D., and K. Rajewsky. 1992. Targeted disruption ofm chain membraneexon causes loss of heavy-chain allelic exclusion.Nature 356:154.

20. Uemtasu, Y., S. Ryser, Z. Dembic, P. Borgulya, P. Krimpenfort, A. Berns,H. von Boehmer, and M. Steinmetz. 1988. In transgenic mice the introducedfunctional T cell receptorb gene prevents expression of endogenousb genes.Cell 52:831.

21. Krimpenfort, P., F. Ossendorp, J. Borst, C. Melief, and A. Berns. 1989. T celldepletion in transgenic mice carrying a mutant gene for TCR-b. Nature 341:742.

22. Bluthmann, H., P. Kisielow, Y. Uematsu, M. Malissen, P. Krimpenfort,A. Berns, H. von Boehmer, and M. Steinmetz. 1988. T-cell-specific deletion ofT-cell receptor transgenes allows functional rearrangement of endogenousa-andb-genes.Nature 334:156.

23. Reth, M. 1992. Antigen receptors on B lymphocytes.Annu. Rev. Immunol. 10:97.

24. Melchers, F., D. Haasner, U. Grawunder, C. Kalberer, H. Karasuyama,T. Winkler, and A. G. Rolink. 1994. Roles of IgH and L chains and of surrogateH and L chains in the development of cells of the lymphocyte lineage.Annu.Rev. Immunol. 12:209.

25. Loffert, D., A. Ehlich, W. Muller, and K. Rajewsky. 1996. Surrogate light chainexpression is required to establish immunoglobulin heavy chain allelic exclusionduring early B cell development.Immunity 4:133.

26. Gong, S., and M. C. Nussenzweig. 1996. Regulation of an early developmentalcheckpoint in the B cell pathway in Igb deficient mice.Science 272:411.

27. Groettrup, M., and H. von Boehmer. 1993. A role for a pre-T-cell receptor inT-cell development.Immunol. Today 14:610.

28. Levelt, C. N., and K. Eichmann. 1995. Receptors and signals in early thymicselection.Immunity 3:667.

29. Fehling, H. J., A. Krotkova, C. Saint-Ruf, and H. von Boehmer. 1995. Crucialrole of the pre-T-cell receptora gene in development ofab but notgd T cells.Nature 375:795.

30. Malissen, M., A. Gillet, L. Ardouin, G. Bouvier, J. Trucy, P. Ferrier, E. Vivier,and B. Malissen. 1995. Altered T cell development in mice with a targetedmutation of the CD3-e gene.EMBO J. 14:4641.

31. Xu, Y., L. Davidson, F. W. Alt, and D. Baltimore. 1996. Function of the pre-Tcell receptora chain in T cell development and the allelic exclusion at the T-cellreceptorb locus.Proc. Natl. Acad. Sci. USA 93:2169.

32. Wilson, A., W. Held, and H. R. MacDonald. 1994. Two waves of recombinasegene expression in developing thymocytes.J. Exp. Med. 179:1355.

33. Grawunder, U., T. M. J. Leu, D. G. Schatz, A. Werner, A. G. Rolink,F. Melchers, and T. H. Winkler. 1995. Down-regulation ofRAG1and RAG2gene expression in preB cells after functional immunoglobulin heavy chain re-arrangement.Immunity 3:601.

34. Lin, W.-C., and S. Desiderio. 1993. Regulation of V(D)J recombination acti-vator protein RAG-2 by phosphorylation.Science 260:953.

35. Lin, W.-C., and D. Desiderio. 1994. Cell cycle regulation of V(D)J recombi-nation-activating protein RAG-2.Proc. Natl. Acad. Sci. USA 91:2733.

36. Li, Z., D. I. Dordai, J. Lee, and S. Desiderio. 1996. A conserved degradationsignal regulates RAG-2 accumulation during cell division and links V(D)J re-combination to the cell cycle.Immunity 5:575.

37. Dudley, E. C., H. T. Petrie, L. M. Shah, M. J. Owen, and A. C. Hayday. 1994.T cell receptorb chain gene rearrangement and selection during thymocytedevelopment in adult mice.Immunity 1:83.

38. Tourigny, M. R., S. Mazel, D. B. Burtrum, and H. T. Petrie. 1997. T cellreceptor (TCR)-b gene recombination: dissociation from cell cycle regulationand developmental progression during T cell ontogeny.J. Exp. Med. 185:1549.

39. Hardy, R. R., C. E. Carmack, S. A. Shinton, J. D. Kemp, and K. Hayakawa.1991. Resolution and characterization of pro-B and pre-pro-B cell stages innormal mouse bone marrow.J. Exp. Med. 173:1213.

40. Ehlich, A., S. Schaal, H. Gu, D. Kitamura, W. Muller, and K. Rajewsky. 1993.Immunoglobulin heavy and light chain genes rearrange independently at earlystages of B cell development.Cell 72:695.

41. Capone, M., F. Watrin, C. Fernex, B. Horvat, B. Krippl, L. Wu, R. Scollay, andP. Ferrier. 1993. TCRb and TCRa gene enhancers confer tissue- and stage-specificity on V(D)J recombination events.EMBO J. 12:4335.

42. Stanhope-Baker, P., K. M. Hudson, A. L. Shaffer, A. Constantinescu, andM. S. Schlissel. 1996. Cell type-specific chromatin structure determines thetargeting of V(D)J recombinase activity in vitro.Cell 85:887.

43. Weintraub, H., and M. Groudine. 1976. Chromosomal subunits in active geneshave an altered conformation.Science 193:848.

44. Gross, D. S., and W. T. Garrard. 1988. Nuclease hypersensitive sites in chro-matin.Annu. Rev. Biochem. 57:159.

45. Cedar, H. 1988. DNA methylation and gene activity.Cell 53:3.46. Weintraub, H. 1993. Summary: genetic tinkering—local problems, local solu-

tions.Cold Spring Harbor Symp. Quant. Biol. 58:819.47. Dillon, N., and F. Grosveld. 1994. Chromatin domains as potential units of

eukaryotic gene function.Curr. Biol. 4:260.48. Martienssen, R. A., and E. J. Richards. 1995. DNA methylation in eukaryotes.

Curr. Opin. Genet. Dev. 5:234.49. Mather, E. L., and R. P. Perry. 1983. Methylation status and DNase I sensitivity

of immunoglobulin genes: changes associated with rearrangement.Proc. Natl.Acad. Sci. USA 80:4689.

50. Ford, A. M., S. M. Watt, A. J. W. Furley, H. V. Molgaard, and M. F. Greaves.1988. Cell lineage specificity of chromatin configuration around the immuno-globulin heavy chain enhancer.EMBO J. 7:2393.

51. Schlissel, M. S., and D. Baltimore. 1989. Activation of immunoglobulin kappagene rearrangement correlates with induction of germline kappa gene transcrip-tion. Cell 58:1001.

52. Alessandrini, A., and S. Desiderio. 1991. Coordination of immunoglobulin DJH

transcription and D-to-JH rearrangement by promoter-enhancer approximation.Mol. Cell. Biol. 11:2096.

53. Engler, P., D. Haasch, C. A. Pinkert, L. Doglio, M. Glymour, R. Brinster, andU. Storb. 1991. A strain-specific modifier on mouse chromosome 4 controls themethylation of independent transgene loci.Cell 65:939.

54. Hsieh, C., and M. Lieber. 1992. CpG methylated minichromosomes becomeinaccessible for V(D)J recombination after undergoing replication.EMBO J.11:315.

55. Ferrier, P., B. Krippl, T. K. Blackwell, A. J. Furley, S. Heikyung, A. Winoto,W. Cook, L. Hood, F. Costantini, and F. Alt. 1990. Separate elements control DJand VDJ rearrangement in a transgenic recombination substrate.EMBO J.9:117.

56. Serwe, M., and F. Sablitzky. 1993. V(D)J recombination in B cells is impairedbut not blocked by targeted deletion of the immunoglobulin heavy chain intronenhancer.EMBO J. 12:2321.

57. Chen, J., F. Young, A. Bottaro, V. Stewart, R. Smith, and F. Alt. 1993. Muta-tions of the intronic lgH enhancer and its flanking sequences differentially affectaccessibility of the JH locus.EMBO J. 12:4635.

58. Okada, A., M. Mendelsohn, and F. Alt. 1994. Differential activation of tran-scription versus recombination of transgenic T cell receptorb variable regiongene segments in B and T lineage cells.J. Exp. Med. 180:261.

59. Lauzurica, P., and M. S. Krangel. 1994. Temporal and lineage-specific controlof T cell receptor alpha/delta gene rearrangement by T cell receptor alpha anddelta enhancers.J. Exp. Med. 179:1913.

60. Roberts, J. L., P. Lauzurica, and M. S. Krangel. 1997. Developmental regulationof VDJ recombination by the core fragment of the T cell receptor alpha en-hancer.J. Exp. Med. 185:131.

61. Xu, Y., L. Davidson, F. W. Alt, and D. Baltimore. 1996. Deletion of the Igklight chain intronic enhancer/matrix attachment region impairs but does notabolish VkJk rearrangement.Immunity 4:377.

62. Gorman, J. R., N. van der Stoep, R. Monroe, M. Cogne, L. Davidson, andF. W. Alt. 1996. The Igk 39 enhancer influences the ration of Igk versus Igl Blymphocytes.Immunity 5:241.

63. Bories, J.-C., J. Demengeot, L. Davidson, and F. W. Alt. 1996. Gene-targeteddeletion and replacement mutations of the T-cell receptorb-chain enhancer: therole of enhancer elements in controlling V(D)J recombination accessibility.Proc. Natl. Acad. Sci. USA 93:7871.

64. Bouvier, G., F. Watrin, M. Naspetti, C. Verthuy, P. Naquet, and P. Ferrier. 1996.Deletion of the mouse T-cell receptorb gene enhancer blocksab T-cell devel-opment.Proc. Natl. Acad. Sci. USA 93:7877.

65. Mombaerts, P., J. Iacomini, R. S. Johnson, K. Kerrup, S. Tonegawa, andV. E. Papioannou. 1992. RAG-1-deficient mice have no mature B and T lym-phocytes.Cell 68:869.

66. Shinkai, Y., G. Rathbun, K. Lam, E. M. Oltz, V. Stewart, M. Mendelsohn,J. Charron, M. Datta, F. Young, A. M. Stall, and F. W. Alt. 1992. RAG-2-deficient mice lack mature lymphocytes owing to inability to initiate V(D)Jrearrangement.Cell 68:855.

67. Anderson, S. J., K. M. Abraham, T. Nakayama, A. Singer, and R. Perlmutter.1992. Inhibition of T-cell receptorb-chain gene rearrangement by overexpres-sion of the non-receptor protein tyrosine kinase p56lck. EMBO J. 11:4877.

68. Berg, L. J., B. Fazekas de St. Groth, A. M. Pullen, and M. M. Davis. 1989.Phenotypic differences betweenab versusb T-cell receptor transgenic miceundergoing negative slection.Nature 340:559.

69. Pircher, H., T. W. Mak, R. Lang, W. Ballhausen, E. Ruedi, H. Hengartner,R. M. Zinkernagel, and K. Burki. 1989. T cell tolerance to Mlsa encoded anti-gens in T cell receptor Vb8.1 chain trangenic mice.EMBO J. 8:719.

1266 CHROMATIN CHANGES AND TCR-b GENE ALLELIC EXCLUSION

by guest on August 16, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

70. Shinkai, Y., and F. W. Alt. 1994. CD3e-mediated signals rescue the develop-ment of CD41CD81 thymocytes in RAG-22/2 mice in the absence of TCRbchain expression.Int. Immunol. 6:995.

71. Schwenk, F., U. Baron, and K. Rajewsky. 1995. A cre-transgenic mouse strainfor the ubiquitous deletion of loxP-flanked gene segments including deletion ingerm cells.Nucleic Acids Res. 23/24:5080.

72. Malissen, M., C. McCoy, D. Blanc, J. Trucy, C. Devaux, A.-M. Schmitt-Ver-hulst, F. Fitch, L. Hood, and B. Malissen. 1986. Direct evidence for chromo-somal inversion during T-cell receptorb-gene rearrangements.Nature 319:28.

73. Forrester, W. C., E. Epner, M. C. Driscoll, T. Enver, M. Brice,T. Papayannopoulou, and M. Groudine. 1990. A deletion of the humanb-globinlocus activation region causes a major alteration in chromatin structure andreplication across the entireb-globin locus.Genes Dev. 4:1637.

74. Krawinkel, U., T. Christoph, and T. Blankenstein. 1989. Organization of the IgVH locus in mice and humans.Immunol. Today 10:339.

75. Lindsten, T., M. D. Lee, and M. M. Davis. 1987. Organization of the T-cellantigen-receptorb-chain locus in mice.Proc. Natl. Acad. Sci. USA 84:7639.

76. Chou, H. S., C. A. Nelson, S. A. Godambe, D. D. Chaplin, and D. Y. Loh. 1987.Germline organization of the murine T cell receptorb-chain genes.Science238:545.

77. Chen, J., Y. Shinkai, F. Young, and F. W. Alt. 1994. Probing immune functionsin RAG-deficient mice.Curr. Opin. Immunol. 6:313.

78. Godfrey, D. I., J. Kennedy, P. Mombaerts, S. Tonegawa, and A. Zlotnik. 1994.Onset of TCR-b gene rearrangement and role of TCR-b expression duringCD32CD42CD82 thymocyte differentiation.J. Immunol. 152:4783.

79. Penit, C., B. Lucas, and F. Vasseur. 1995. Cell expansion and growth arrestphases during the transition from precursor (CD4282) to immature (CD4181)thymocytes in normal and genetically modified mice.J. Immunol. 154:5103.

80. Mombaerts, P., S. J. Anderson, R. M. Perlmutter, T. W. Mak, and S. Tonegawa.1994. An activatedIck transgene promotes thymocyte development in RAG-1mutant mice.Immunity 1:261.

81. Shinkai, Y., S. Koyasu, K. Nakayama, K. M. Murphy, D. Y. Loh,E. L. Reinherz, and F. W. Alt. 1993. Restoration of T cell development inRAG-2-deficient mice by functional TCR transgenes.Science 259:822.

82. Levelt, C. N., P. Mombaerts, A. Iglesias, S. Tonegawa, and K. Eichmann. 1993.Restoration of early thymocyte differentiation in T-cell receptorb-chain-defi-cient mutant mice by transmembrane signaling through CD3e. Proc. Natl. Acad.Sci. USA 90:11401.

83. Krimpenfort, P., R. de Jong, Y. Uematsu, Z. Dmebic, S. Ryser, H. von Boehmer,M. Steinmetz, and A. Berns. 1988. Transcription of T cell receptorb-chaingenes is controlled by a downstream regulatory element.EMBO J. 7:745.

84. McDougall, S., C. L. Peterson, and K. Calame. 1988. A transcriptional enhancer39 of Cb2 in the T cell receptorb locus.Science 241:205.

85. Hashimoto, Y. 1989. T cell receptorb has two downstream DNase I hypersen-sitive regions.J. Exp. Med. 169:2097.

86. Bier, E., Y. Hashimoto, M. I. Greene, and A. M. Maxam. 1985. Active T-cellreceptor genes have intron deoxyribonuclease hypersensitive sites.Science 22:528.

87. Fagioli, M., A. Care, E. Ciccone, L. Moretta, A. Moretta, E. Meccia, U. Testa,B. Falini, F. Grignani, C. Peschle, and P. G. Pelicci. 1991. Molecular hetero-

geneity of the 1.0-kb Tb transcript in natural killer andg/d lymphocytes.Eur.J. Immunol. 21:1529.

88. Kotzin, B. L., V. L. Barr, and E. Palmer. 1985. A large deletion within the T-cellreceptor beta-chain gene complex in New Zealand white mice.Science 229:167.

89. Alvarez, J. D., S. J. Anderson, and D. Y. Loh. 1995. V(D)J recombination andallelic exclusion of a TCRb-chain minilocus occurs in the absence of a func-tional promoter.J. Immunol. 155:1191.

90. Villey, I., D. Caillol, F. Selz, P. Ferrier, and J.-P. de Villartay. 1996. Defect inrearrangement of the most 59 TCR-Ja following targeted deletion of T earlya(TEA): implications for TCRa locus accessibility.Immunity 5:331.

91. Koningsberger, J. C., A. Chott, T. Logtenberg, L. J. J. M. Wiegman,R. S. Blumberg, G. P. van Berge Henegouwen, and S. P. Balk. 1997. TCRexpression in human fetal intestine and identification of an early T cell receptorb-chain transcript.J. Immunol. 159:1775.

92. Schlissel, M. S., and P. Stanhope-Baker. 1997. Accessibility and the develop-mental regulation of V(D)J recombination.Semin. Immunol. 9:161.

93. Jamieson, C., F. Mauxion, and R. Sen. 1989. Identification of a functionalNF-kB binding site in the murine T cell receptorb2 locus.J. Exp. Med. 170:1737.

94. Mauxion, F., M. G. Pray, and R. Sen. 1990. Multiple nuclear factors interactwith sequences within the Jb-Cb intron of the murine T cell receptorb-chaingene.J. Immunol. 145:1577.

95. Takeda, S., Y. Zou, H. Bluethmann, D. Kitamura, U. Muller, and K. Rajewsky.1993. Deletion of the immunoglobulink chain intron enhancer abolishesk chaingene rearrangement incis but notl chain gene rearrangement intrans. EMBOJ. 12:2329.

96. Yanagi, Y., S. Hirose, R. Nagasawa, T. Shirai, T. W. Mak, and T. Tada. 1986.Does the deletion within T cell receptorb-chain gene of NZW mice contributeto autoimmunity in (NZB3 NZW)F1 mice?Eur. J. Immunol. 16:1179.

97. Kotzin, B. L., and E. Palmer. 1987. The contribution of NZW genes to lupus-like disease in (NZB3 NZW)F1 mice.J. Exp. Med. 165:1237.

98. Noonan, D. J., P. J. McConahey, and G. J. Cardenas. 1990. Correlations ofautoimmunity with H-2 and T cell receptorb chain genotypes in (NZB3NZW)F2 mice.Eur. J. Immunol. 20:1105.

99. Hirose, S., K. Tokushiga, K. Kinoshita, S. Nozawa, J. Saito, H. Nishimura, andT. Shirai. 1991. Contribution of the gene linked to the T cell receptorb chaincomplex of NZW mice to the autoimmunity of (NZB3 NZW)F1 mice. Eur.J. Immunol. 21:823.

100. Kono, D. H., R. W. Burlingame, D. G. Owens, A. Kuramochi, R. S. Balderas,D. Balomenos, and A. N. Theofilopoulos. 1994. Lupus susceptibility loci inNew Zealand mice.Proc. Natl. Acad. Sci. USA 91:10168.

101. Fernex, C., M. Capone, and P. Ferrier. 1995. The V(D)J recombinational andtranscriptional activities of the immunoglobulin heavy-chain intronic enhancercan be mediated through distinct protein-binding sites in a transgenic substrate.Mol. Cell. Biol. 15:3217.

102. Kallenbach, S., C. Babinet, S. Pournin, P. Cavelier, M. Goodhardt, andF. Rougeon. 1993. The intronic immunoglobulink gene enhancer acts indepen-dently on rearrangement and on transcription.Eur. J. Immunol. 23:1917.

1267The Journal of Immunology

by guest on August 16, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from