Although tightly linked, the TCR α and δ genes are expressed specifically in T lymphocytes, whereas the Dad1 gene is ubiquitously expressed. Between TCR α and Dad1 are eight DNase I hypersensitive sites (HS). HS1 colocalizes with the TCR α enhancer (Eα) and is T cell-specific; HS2, -3, -4, -5, and -6 map downstream of HS1 and are tissue-nonspecific. The region spanning HS2–6 was reported to display chromatin-opening activity and to confer copy number-dependent and integration site-independent transgene expression in transgenic mice. Here, we demonstrate that HS2–6 also displays enhancer-blocking activity, as it can block an enhancer from activating a promoter when located between the two in a chromatin-integrated context, and can do so without repressing either the enhancer or the promoter. Multiple enhancer-blocking elements are arrayed across HS2–6. We show that HS2–6 by itself does not activate transcription in chromatin context, but can synergize with an enhancer when located upstream of an enhancer and promoter. We propose that HS2–6 primarily functions as an insulator or boundary element that may be critical for the autonomous regulation of the TCR α and Dad1 genes.

The TCR αδ locus is a complex genetic region that spans more than one megabase in both mouse and man (1, 2). Contained within this locus are the variable (V), diversity (D), joining (J), and constant (C) gene segments that encode two different TCR chains, δ and α. A large number of Vα gene segments as well as several Vδ gene segments are arrayed across the 5′ portion of the locus. A large number of Jα gene segments, as well as Cα, is found at the extreme 3′ end. Nested between the V gene segments and Jα gene segments lie the Dδ, Jδ, and Cδ gene segments. TCR δ and TCR α gene segments are assembled by the process of V(D)J recombination during the differentiation of T lymphocytes in the thymus. VδDδJδ rearrangement occurs with retention of the Vα, Jα, and Cα gene segments and can lead to the production of a functional TCR δ polypeptide, whereas VαJα rearrangement occurs with deletion of the Dδ, Jδ, and Cδ gene segments and can lead to the production of a functional TCR α polypeptide.

Although the TCR δ and TCR α gene segments are tightly linked and are both rearranged and expressed specifically in T lymphocytes, they are clearly under distinct regulatory control. TCR δ gene rearrangement and expression initiates in the double negative population of thymocytes and at day 14 of murine embryogenesis, whereas TCR α gene rearrangement and expression initiates in the subsequent double-positive population of thymocytes and at day 16 of murine embryogenesis (3, 4, 5, 6). Interestingly, recent studies have identified an unrelated and distinctly regulated antiapoptosis gene, Dad1, to be closely linked to the TCR αδ locus, only 12 kb 3′ of Cα (7). The Dad1 gene is expressed in all tissues examined and at least as early as day 7 of murine embryogenesis.

The cis-acting elements that control Dad1 expression remain to be identified. However, much is known about the cis-acting elements that control TCR δ and TCR α gene rearrangement and expression. Of particular importance are the TCR δ enhancer (Eδ)4, situated in the intron between the Jδ3 and Cδ gene segments (8, 9, 10), and the TCR α enhancer (Eα), situated immediately 3′ of Cα (11, 12). Both enhancers have been shown to direct the developmental stage-specific activation of V(D)J recombination in transgenic V(D)J recombination substrates in vivo, with Eδ activating V(D)J recombination at the double-negative stage, and Eα activating V(D)J recombination at the double-positive stage (13, 14, 15). Further, Eα has been shown to be critical for V(D)J recombination and expression of the endogenous TCR α gene (16). Thus Eα and Eδ appear to confer region-specific developmental control to the process of V(D)J recombination within the endogenous TCR αδ locus. The activation of V(D)J recombination by these enhancers is thought to occur as a consequence of their ability to modify local chromatin structure to allow access to the V(D)J recombinase (17, 18).

The tight linkage of the TCR δ, TCR α, and Dad1 genes raises the question of how their distinct regulation is maintained. One potential mechanism is suggested by our recent studies identifying an enhancer-blocking element, blocking element αδ-1 (BEAD-1), in a 2-kb region between the Cδ and Jα gene segments (19). We showed that BEAD-1 could block the TCR δ enhancer from activating a nearby promoter when situated between the two in a chromatin-integrated construct in stably transfected cells. We proposed that BEAD-1 might function in vivo to prevent Eδ from prematurely activating the Jα gene segments for V(D)J recombination during the double-negative stage of T cell development, thereby helping to maintain independent regulation of TCR δ and TCR α. BEAD-1 has properties that are similar to those of a class of regulatory elements, initially described in Drosophila, that are known as boundary elements or insulators (20, 21, 22). These elements include scs and scs′, which flank the Drosophila 87A7 hsp70 locus (23, 24), su(Hw) protein binding sites in the Drosophila gypsy transposon (25, 26), 5′ hypersensitive site (HS) 4 in the chicken β-globin locus (27, 28), and Fab-7 in the Drosophila bithorax complex (29, 30, 31, 32). Like BEAD-1, these elements can block an enhancer from activating a promoter when positioned between the two. Furthermore, they can insulate a transgene from position effects. Boundary elements are thought to play a critical role in dividing chromatin into independently regulated units, or domains, thereby preventing interlocus regulation. The Fab-7 element in the Drosophila bithorax complex has been explicitly shown to fulfill these expectations in vivo, as it is required for normal parasegment-specific gene expression and parasegment differentiation during Drosophila development (29, 30).

Recently, a series of eight DNase I HS were mapped between the TCR α and Dad1 genes (7, 33). One of these, HS1, is T cell-specific and colocalizes with the previously defined Eα. The region encompassing HS2, -3, -4, -5, and -6 (hereafter denoted HS2–6) was originally proposed to represent a locus control region (LCR), or an essential component of an LCR, that functioned to control V(D)J recombination and gene expression throughout the entire TCR αδ locus. In conjunction with Eα, HS7, and HS8, HS2–6 was found to confer high-level, integration site-independent, and copy number-dependent TCR α gene expression in the thymus of transgenic mice, whereas Eα, HS7, and HS8 alone could not (33). HS2–6 was reported to have chromatin-opening activity, as it could activate an erythroid-specific β-globin reporter in a tissue-nonspecific manner in transgenic mice (34). However, elimination of HS2–6 from the endogenous TCR αδ locus had no measurable effect on either TCR α or TCR δ rearrangement and transcription in thymocytes and had only a relatively small effect on the relative proportions of thymocyte subsets that could not be readily attributed to a specific molecular defect (7). Hence, the true regulatory function of HS2–6 remains unclear.

LCRs are typically composed of a series of tissue-specific DNase HS that function in a dominant fashion to open chromatin and activate gene expression. By comparison, boundary elements are relatively neutral; although they can interfere with enhancer-promoter communication in a position-dependent manner, they typically lack an intrinsic and dominant ability to activate gene expression (20, 21, 35). Either LCR or boundary activity can be invoked to explain the ability of HS2–6 to confer copy number-dependent and integration site-independent expression onto a linked transgene. However, a boundary function for HS2–6 might best explain the ability of the TCR α and Dad1 genes to maintain dramatically different expression patterns, despite their very close linkage.

In this report, we show that HS2–6 can block an enhancer from activating a promoter when located between the two in a chromatin-integrated substrate. HS2–6, by itself, failed to dominantly activate transcription from a linked promoter. However, it synergized with an enhancer to activate the promoter when located upstream of the two elements. Although position-dependent enhancer blocking and enhancer synergy may both play a role within the endogenous TCR αδ locus, we propose that the primary function of HS2–6 is that of an insulator/boundary element that helps to maintain independent regulation of the TCR α and Dad1 genes.

The constructs P-Neo, E-P-Neo-scs′, and E-2.7-P-Neo-scs′ were described previously (19). In these constructs, E, P, Neo, and scs′ represent the 380-bp Eδ, the 1.6-kb Vδ1 promoter, the bacterial neomycin resistance gene, and the Drosophila hsp70 scs′ boundary element, respectively (Fig. 1,A). 2.7 represents two copies of the 1.35-kb phage φX HaeIII DNA fragment in tandem. Plasmid pLCR8.0 carries an 8-kb DNA fragment of the TCR αδ locus isolated by Dr. P. Lauzurica from a strain 129 mouse genomic DNA library (kindly provided by Dr. T. Tedder, Duke University, Durham, NC). The 8.0-kb RSa8.0 and 5.5-kb RSm5.5 DNA fragments were released from this plasmid by digestion with EcoRI plus SalI and EcoRI plus SmaI, respectively. These fragments, as well as the 2.3- and 2.8-kb HindIII-HindIII (HH2.3 and HH2.8) and 2.4-kb HindIII-BglII (HB2.4) fragments (Fig. 1 B), were inserted into the SalI site between Eδ and the Vδ1 promoter of plasmid E-P-Neo-scs′ by blunt-end ligation, to generate plasmids E-RSa8.0-P-Neo-scs′, E-RSm5.5-P-Neo-scs′, E-HH2.3-P-Neo-scs′, E-HH2.8-P-Neo-scs′, and E-HB2.4-P-Neo-scs′. RSm5.5-P-Neo-scs′ was generated by deletion of Eδ from E-RSm5.5-P-Neo-scs′ using NotI and ClaI digestion followed by religation of blunt ends. HH2.3, HH2.8, and HB2.4 were also introduced by blunt-end ligation into the XbaI site upstream of Eδ in E-P-Neo-scs′, generating plasmids HH2.3-E-P-Neo-scs′, HH2.8-E-P-Neo-scs′, and HB2.4-E-P-Neo-scs′, respectively. Note that all constructs are named in a descriptive fashion that identifies each of the elements, in order, within the plasmid. Plasmids were purified by two CsCl density gradient centrifugation steps and were linearized by NotI or SacII digestion. Following three phenol and two chloroform extractions, linearized plasmids were ethanol precipitated and resuspended in 10 mM Tris (pH 8.0) and 1 mM EDTA.

FIGURE 1.

Test plasmids and fragments. A, The backbones of several test plasmids containing combinations of E (Eδ), P (Vδ1 promoter), Neo (bacterial neomycin resistance gene), and scs′ (Drosophila blocking element) are shown. Positions where test fragments are introduced are marked by triangles. B, Schematic map of the TCR αδ locus and the Dad1 genes, with test fragments indicated. Vertical arrows indicate DNase I HS. The HS2–6 region is drawn to scale based on Diaz et al. (7) and Hong et al. (33) and additional mapping. Restriction enzymes are: R, EcoRI; H, HindIII; E, EcoRV; X, XbaI; Bm, BamHI; B, BglII; and S, SmaI. The SalI site at the 3′ end of of Rsa8.0 is vector-derived. BE1 is the previously defined BEAD-1 enhancer-blocking element (19). Dad1 and the single Vδ segment located downstream of Cδ are transcribed from right to left (7). All other elements are transcribed from left to right. The direction of transcription through Jα-Cα and Dad1 is indicated by horizontal arrows.

FIGURE 1.

Test plasmids and fragments. A, The backbones of several test plasmids containing combinations of E (Eδ), P (Vδ1 promoter), Neo (bacterial neomycin resistance gene), and scs′ (Drosophila blocking element) are shown. Positions where test fragments are introduced are marked by triangles. B, Schematic map of the TCR αδ locus and the Dad1 genes, with test fragments indicated. Vertical arrows indicate DNase I HS. The HS2–6 region is drawn to scale based on Diaz et al. (7) and Hong et al. (33) and additional mapping. Restriction enzymes are: R, EcoRI; H, HindIII; E, EcoRV; X, XbaI; Bm, BamHI; B, BglII; and S, SmaI. The SalI site at the 3′ end of of Rsa8.0 is vector-derived. BE1 is the previously defined BEAD-1 enhancer-blocking element (19). Dad1 and the single Vδ segment located downstream of Cδ are transcribed from right to left (7). All other elements are transcribed from left to right. The direction of transcription through Jα-Cα and Dad1 is indicated by horizontal arrows.

Close modal

Each linearized test DNA construct was transfected into the human T cell leukemia Jurkat by electroporation in triplicate, and transfectants were plated in 30 ml 0.33% agar (Sigma, St. Louis, MO) in RPMI 1640 supplemented with 10% FBS and 1000 μg/ml active G418 (Life Technologies, Gaithersburg, MD), as described previously (19). G418-resistant colonies were counted 3–4 wk after plating and selection.

Jurkat cells were cotransfected with linearized test construct and linearized pTK-hyg (19) at a molar ratio of 6:1 by electroporation. Individual hygromycin B-resistant clones were generated by limiting dilution cloning 24 h after transfection as described previously (19). Test construct integration and copy number were determined by slot blot analysis of genomic DNA isolated from individual clones.

We previously established a stable transfection/soft agar colony forming assay in human Jurkat T cells to test enhancer and enhancer-blocking activities in a chromatin context (19). The basic reporter plasmid includes a bacterial neomycin resistance gene whose expression is driven by the human Vδ1 promoter (P-Neo) (Fig. 1 A). Eδ is inserted upstream of the Vδ1 promoter (E-P-Neo) to measure enhancer activity, and test fragments can be inserted between Eδ and the Vδ1 promoter to measure enhancer-blocking activity. The Drosophila boundary element scs′, which was previously demonstrated to be functional as an enhancer-blocking element in Jurkat cells, is inserted 3′ of neo (E-P-Neo-scs′) to prevent activation of the promoter by downstream copies of the enhancer in tandemly arrayed multicopy integrants.

Using this assay, we asked whether there is enhancer-blocking activity within HS2–6 of the murine TCR αδ locus. To this end, 8.0-kb (RSa8.0) and 5.5-kb (RSm5.5) DNA fragments (Fig. 1,B), both of which contain HS2–6, were individually cloned between Eδ and the Vδ1 promoter of construct E-P-Neo-scs′. Insertion of either RSa8.0 (E-RSa8.0-P-Neo-scs′) or RSm5.5 (E-RSm5.5-P-Neo-scs′) between Eδ and Vδ1 promoter completely blocked the ability of Eδ to activate the Vδ1 promoter, suggesting the existence of enhancer-blocking elements in HS2–6. Enhancer blocking was independent of the orientation of HS2–6 (Fig. 2 A).

FIGURE 2.

Enhancer-blocking activity in the HS2–6 region as measured by colony formation. Constructs were transfected in triplicate into Jurkat cells and colony number was determined following growth in soft agar medium containing G418. Results are presented as mean ± SD, with the colony number for E-P-Neo-scs′ or E-2.7-P-Neo-scs′ normalized to 100. E is Eδ, P is the Vδ1 promoter, Rsa8.0, Rsm5.5, HH2.3, HH2.8, and HB2.4 are test fragments shown in Fig. 1. 2.7 is two of the 1.35-kb phage φX HaeIII fragment in tandem. A, Enhancer blocking by RSa8.0 and RSm5.5. The absolute number of colonies for E-P-Neo-scs′ was 585. The data shown are representative of two independent experiments. B, Multiple enhancer-blocking elements in HS2–6. The absolute number of colonies for E-2.7-P-Neo-scs′ was 131. The data shown are representative of four independent experiments. C, Enhancer blocking by components of HS2–6 is distinguishable from silencing. The absolute number of colonies for E-2.7-P-Neo-scs′ was 101. The data shown are representative of three independent experiments.

FIGURE 2.

Enhancer-blocking activity in the HS2–6 region as measured by colony formation. Constructs were transfected in triplicate into Jurkat cells and colony number was determined following growth in soft agar medium containing G418. Results are presented as mean ± SD, with the colony number for E-P-Neo-scs′ or E-2.7-P-Neo-scs′ normalized to 100. E is Eδ, P is the Vδ1 promoter, Rsa8.0, Rsm5.5, HH2.3, HH2.8, and HB2.4 are test fragments shown in Fig. 1. 2.7 is two of the 1.35-kb phage φX HaeIII fragment in tandem. A, Enhancer blocking by RSa8.0 and RSm5.5. The absolute number of colonies for E-P-Neo-scs′ was 585. The data shown are representative of two independent experiments. B, Multiple enhancer-blocking elements in HS2–6. The absolute number of colonies for E-2.7-P-Neo-scs′ was 131. The data shown are representative of four independent experiments. C, Enhancer blocking by components of HS2–6 is distinguishable from silencing. The absolute number of colonies for E-2.7-P-Neo-scs′ was 101. The data shown are representative of three independent experiments.

Close modal

The above result could reflect inhibition of neo expression due to the increased distance between Eδ and the Vδ1 promoter or inhibition of transfection efficiency due to increased plasmid size. To eliminate these possibilities and to better localize enhancer-blocking activity, a 2.3-kb HindIII fragment (HH2.3) that includes HS2 and HS3, a 2.8-kb HindIII fragment (HH2.8) that includes HS4, HS5, and HS6, and a 2.4-kb HindIII-BglII fragment (HB2.4) that contains HS4 and HS5 (Fig. 1), were individually cloned into E-P-Neo-scs′. Compared with a 2.7-kb DNA fragment (two tandem copies of the phage φX 1.35 kb HaeIII fragment; E-2.7-P-Neo-scs′) previously shown to have no enhancer-blocking activity (19), HH2.3 (E-HH2.3-P-Neo-scs′) reduced the colony number, but not to the level of P-Neo, indicating partial blockade of Eδ (Fig. 2B). By contrast, HH2.8 (E-HH2.8-P-Neo-scs′) was a highly effective enhancer blocker. Deletion of the HS6 region from HH2.8 (E-HB2.4-P-Neo-scs′) resulted in partial loss of enhancer blocking, indicating that the HS6 region carries a component of the enhancer-blocking activity. Hence, at least three discrete segments of HS2–6 (the HS2–3 segment, the HS4–5 segment, and the HS6 segment) contribute to enhancer-blocking activity in the colony assay. We note that the colony number for E-HH2.8-P-Neo-scs′ was actually slightly below the level for P-Neo in this experiment; the smaller size of P-Neo might have resulted in an elevated transfection efficiency relative to the other constructs. This effect does not compromise our conclusions about enhancer-blocking relative to E-2.7-P-Neo-scs′, as all other constructs are of similar size.

The enhancer-blocking activity of HH2.8 was further confirmed by a cotransfection/limiting dilution cloning assay (19). In this assay, E-HH2.8-P-Neo-scs′ and E-2.7-P-Neo-scs′ were each cotransfected into Jurkat cells with a hygromycin resistance gene. Ten hygromycin B resistant clones containing E-HH2.8-P-Neo-scs′ and seven hygromycin B resistant clones containing E-2.7-P-Neo-scs′ were tested for Neo gene expression by Northern blot analysis. Levels of neo transcripts in the E-HH2.8-P-Neo-scs′ clones were found to be significantly reduced in comparison to E-2.7-P-Neo-scs′ (Fig. 3,A). After correction for neo gene copy number and RNA loading, insertion of HH2.8 between Eδ and the Vδ1 promoter resulted in, on average, a 93% decrease of neo gene expression (Fig. 3 B).

FIGURE 3.

Enhancer blocking by HS4–6 (HH2.8) as measured by Northern blot analysis of neo gene expression in stably transfected cell clones. A, Constructs were cotransfected into Jurkat cells along with pTK-hyg, and total RNA isolated from individual hygromycin B resistant clones was analyzed on a Northern blot that was serially hybridized with 32P-labeled neo and GAPDH probes. Neo gene copy number in each clone was determined by slot blot analysis of genomic DNA using the same probes. Northern blot and slot blot hybridization signals were quantified using a Phosphorimager (Molecular Dynamics, Sunnyvale, CA). Note that the control RNA samples from E-2.7-P-Neo-scs′-transfected cells are identical to those analyzed previously (19). B, The neo gene transcript level for each clone was determined on a per copy basis. The average level of neo transcripts in E-2.7-P-Neo-scs′ and E-HH2.8-P-Neo-scs′ was 27.28 ± 16.12 and 2.02 ± 1.58 arbitrary units per transgene copy, respectively.

FIGURE 3.

Enhancer blocking by HS4–6 (HH2.8) as measured by Northern blot analysis of neo gene expression in stably transfected cell clones. A, Constructs were cotransfected into Jurkat cells along with pTK-hyg, and total RNA isolated from individual hygromycin B resistant clones was analyzed on a Northern blot that was serially hybridized with 32P-labeled neo and GAPDH probes. Neo gene copy number in each clone was determined by slot blot analysis of genomic DNA using the same probes. Northern blot and slot blot hybridization signals were quantified using a Phosphorimager (Molecular Dynamics, Sunnyvale, CA). Note that the control RNA samples from E-2.7-P-Neo-scs′-transfected cells are identical to those analyzed previously (19). B, The neo gene transcript level for each clone was determined on a per copy basis. The average level of neo transcripts in E-2.7-P-Neo-scs′ and E-HH2.8-P-Neo-scs′ was 27.28 ± 16.12 and 2.02 ± 1.58 arbitrary units per transgene copy, respectively.

Close modal

Enhancer-blocking activity should be evident only when an element is positioned between an enhancer and a promoter, whereas silencing activity should be evident regardless of position. The inhibition of gene expression by HS2–6 represents enhancer blocking rather than silencing, since insertion of HS2–6 DNA fragments HH2.3, HH2.8, and HB2.4 upstream of Eδ did not inhibit Eδ from activating neo gene expression (Fig. 2 C). In contrast, insertion of these elements upstream of Eδ resulted in 3- to 4-fold increases in colony number. This result indicates that HS2–6 can either inhibit or augment neo gene expression in a fashion that is dependent on its position relative to the enhancer and promoter. The increase in expression in this experiment could be due to an insulating activity that may protect the reporter from negative position effects, or could result from an intrinsic activating property of HS2–6, as has been reported previously (33, 34). An increase in expression is not intrinsically associated with the positioning of enhancer-blocking elements in an upstream position, as no such effect was observed in a previous study (19).

LCRs activate gene expression in a dominant manner through the cooperation between multiple cis-elements. Typically, an LCR contains a classical enhancer that can activate transcription in transient transfection assays, as well as chromatin-opening elements that activate gene expression in a chromatin context (20, 35). In the β-globin LCR, HS2 is an example of the former, while HS3 and HS4 are examples of the latter (36, 37). It has been reported that HS2–6 between TCR α and Dad1 genes does not activate transcription in a transient transfection assay (33) but does display chromatin-opening activity (34). Therefore, we asked whether HS2–6 can activate a chromatin-integrated Vδ1 promoter in the absence of an enhancer. HS2–6 fragment RSm5.5 was cloned upstream of the Vδ1 promoter of P-Neo-scs′ in both orientations. In the soft agar colony forming assay, inclusion of HS2–6 in either orientation upstream of the Vδ1 promoter (RSm5.5-P-Neo-scs′) failed to increase colony number over that obtained with P-Neo (Fig. 4,A). This result was confirmed by the cotransfection/limiting dilution assay. In all eight hygromycin-resistant Jurkat cell clones containing RSm5.5-P-Neo-scs′, neo gene expression was either undetectable or at much lower levels than in clones containing E-2.7-P-Neo-scs′ (Fig. 4,B). From these experiments, we conclude that HS2–6 by itself is unable to significantly activate transcription from the Vδ1 promoter. Thus, the increase in neo gene expression when HS2–6 is introduced upstream of both Eδ and the Vδ1 promoter (Fig. 2 C) is most likely dependent on the presence of a linked enhancer (Eδ). From these transfection experiments, we conclude that HS2–6 possesses an enhancer-blocking activity that is manifest when it is positioned between an enhancer and promoter. Further, we conclude that although HS2–6 is unable to activate a promoter by itself, it can synergize with an enhancer to up-regulate gene expression when positioned upstream of an enhancer and promoter.

FIGURE 4.

HS2–6 cannot directly activate a promoter in stably transfected cells. A, Colony forming assay. The experiment was conducted as described in the legend to Fig. 2. The absolute number of colonies for E-P-Neo-scs′ was 216. The data shown are representative of two independent experiments. B, Northern blot analysis of neo gene expression in stable transfectants. The experiment was conducted as described in the legend to Fig. 3. Note that the control E-2.7-P-Neo-scs′ RNA samples analyzed in this experiment are identical to those analyzed in Fig. 3.

FIGURE 4.

HS2–6 cannot directly activate a promoter in stably transfected cells. A, Colony forming assay. The experiment was conducted as described in the legend to Fig. 2. The absolute number of colonies for E-P-Neo-scs′ was 216. The data shown are representative of two independent experiments. B, Northern blot analysis of neo gene expression in stable transfectants. The experiment was conducted as described in the legend to Fig. 3. Note that the control E-2.7-P-Neo-scs′ RNA samples analyzed in this experiment are identical to those analyzed in Fig. 3.

Close modal

In this report, we have shown that HS2–6 blocks an enhancer from activating a promoter when located between the two in a chromatin context and that it does so without repressing either the enhancer or promoter. Enhancer blocking by HS2–6 results from the activity of multiple enhancer-blocking elements dispersed among fragments HS2–3, HS4–5, and HS6. In addition to enhancer-blocking activity, HS2–6 also shows a synergistic activation property that is suggested by the several-fold increase of reporter expression when HS2–6 is located upstream of both the enhancer and promoter. However, synergistic activation by HS2–6 is distinguished from activities of classic enhancers. HS2–6 does not activate gene expression in transient transfection assays (33). Further, using two different assays, we demonstrate that in the absence of an enhancer, HS2–6 is incapable of activating transcription from a linked promoter in a chromatin context. Thus, HS2–6 displays two different activities that are dependent on its position. When located between an enhancer and promoter, it blocks the enhancer from activating the promoter. When located upstream of both the enhancer and promoter, it can synergize with the enhancer in gene activation. At present, it is unclear whether or not the two activities result from a single mechanism. The two activities appear tightly associated since the HS2–3, HS4–5, and HS4–6 enhancer-blocking fragments all show enhancer synergism and since deletion of the HS6 region from the HH2.8 fragment reduces both enhancer-blocking and enhancer synergism (Fig. 2 C). It is also unclear whether either of these activities are attributable to the HS per se, or to other associated elements. Further studies will be required to address whether the two activities can be dissociated from each other or from the mapped HS.

Our data are compatible with the results of recent gene-targeting studies in the endogenous TCR αδ locus. A deletion of Eα that leaves HS2–6 intact abolishes V(D)J recombination and transcription of TCR α gene (16), indicating that HS2–6 cannot provide chromatin accessibility for V(D)J recombination and transcription in the absence of Eα. In contrast, a deletion of HS2–6 that leaves Eα intact affects neither TCR α nor TCR δ gene expression (7), indicating that HS2–6 is not required for normal expression of these genes. Our data are also consistent with the observation by Ortiz et al. (34) that HS2–6 is a relatively weak activator of heterologous reporter gene expression in the thymus of transgenic mice. Considering all of these data, we propose that the primary function of HS2–6 in the endogenous locus is that of an insulator or boundary element. Such activity may be at least partially responsible for the LCR-like ability of HS2–6 to confer copy number-dependent and integration site-independent expression to an Eα-containing TCR α transgene, as reported previously (33). Within the endogenous locus, HS2–6 may prevent the as yet unidentified regulatory elements responsible for ubiquitous Dad1 expression from activating TCR α. Alternatively, it may prevent Eα from superactivating Dad1 expression in T cells. Eα is known to be a potent activator of transcription and V(D)J recombination whose influence extends at least 90 kb in the 5′ direction (16); yet, although Dad1 is positioned only 8 kb 3′ of Eα, it is striking that Dad1 expression is not significantly perturbed in double-positive thymocytes of Eα−/− mice (B. Sleckman, personal communication).

The notion that HS2–6 is required for appropriately regulated Dad1 expression in thymocytes must remain a tentative hypothesis, at least in part because the Dad1 promoter has not been defined and was not specifically tested in our assays. A role for HS2–6 in Dad1 expression could best be assessed in mice carrying a targeted deletion of HS2–6 (7). However, several properties of these previously generated mice suggest that they would be inappropriate for this analysis. First, in addition to eliminating HS2–6, the gene-targeting event disrupted the Dad1 transcription unit by eliminating a portion of the Dad1 3′ untranslated region. Second, the gene-targeting event resulted in the integration of a PGK-neo cassette, which has been documented to perturb the expression of linked genes in numerous instances (38, 39, 40). Perhaps as a consequence of the above features of the gene-targeting event rather than of HS2–6 deletion per se, no Dad1 transcripts are detectably expressed by the mutant allele, and mice homozygous for the deletion display early embryonic lethality (7). Together, these complexities argue that additional gene-targeting strategies will be required to address the role of HS2–6 in unambiguous fashion.

We thank T. Tedder for providing a strain 129 mouse genomic DNA library and P. Lauzurica for providing plasmid pLCR8.0. We also thank Y. Zhuang and C. Hernandez-Munain for their helpful comments on the manuscript.

1

This work was supported by National Institutes of Health Grant GM41052. X.-P.Z. was supported in part by Public Health Service Training Grant CA09058.

4

Abbreviations used in this paper: E, enhancer; BEAD-1, blocking element αδ-1; HS, hypersensitivity site; LCR, locus control region; P, promoter; Neo, neomycin resistance gene.

1
Wang, K., J. L. Klotz, G. Kiser, G. Bristol, E. Hays, E. Lai, E. Gese, M. Kronenberg, L. Hood.
1994
. Organization of the V gene segments in mouse T-cell antigen receptor α/δ locus.
Genomics
20
:
419
2
Koop, B. F., L. Rowen, C. L. Kuo, D. Seto, J. A. Lenstra, S. Howard, W. Shan, P. Deshpande, L. Hood.
1994
. The human T-cell receptor TCRAC/TCRDC (Cα/Cδ) region: organization, sequence, and evolution of 97.6 kb of DNA.
Genomics
19
:
478
3
Raulet, D. H., R. D. Garman, H. Saito, S. Tonegawa.
1985
. Developmental regulation of T-cell receptor gene expression.
Nature
314
:
103
4
Chien, Y.-H., M. Iwashima, K. B. Kaplan, J. F. Elliot, M. M. Davis.
1987
. A new T-cell receptor gene located within the α locus and expressed early in T-cell differentiation.
Nature
327
:
677
5
Wilson, A., W. Held, H. R. MacDonald.
1994
. Two waves of recombinase gene expression in developing thymocytes.
J. Exp. Med.
179
:
1355
6
Petrie, H. T., F. Livak, D. Burtrum, S. Mazel.
1995
. T cell receptor gene recombination patterns and mechanisms: cell death, rescue and T cell production.
J. Exp. Med.
182
:
121
7
Hong, N. A., D. Cado, J. Mitchell, B. D. Ortiz, S. N. Hsieh, A. Winoto.
1997
. A targeted mutation at the T-cell receptor α/δ locus impairs T-cell development and reveals the presence of the nearby antiapoptosis gene Dad1.
Mol. Cell. Biol.
17
:
2151
8
Redondo, J. M., S. Hata, C. Brocklehurst, M. S. Krangel.
1990
. A T cell specific transcriptional enhancer within the human T cell receptor δ locus.
Science
247
:
1225
9
Bories, J. C., P. Loiseau, L. d’Auriol, C. Gontier, A. Bensussan, L. Degos, F. Sigaux.
1990
. Regulation of transcription of the human T cell antigen receptor δ chain gene: a T lineage-specific enhancer element is located in the Jδ3-Cδ intron.
J. Exp. Med.
171
:
75
10
Gill, L. L., D. Zaninetta, K. Karjalainen.
1991
. A transcriptional enhancer of the mouse T cell receptor δ gene locus.
Eur. J. Immunol.
21
:
807
11
Ho, I.-C., L.-H. Yang, G. Morle, J. M. Leiden.
1989
. A T-cell-specific transcriptional enhancer element 3′ of Cα in the human T-cell receptor α locus.
Proc. Natl. Acad. Sci. USA
86
:
6714
12
Winoto, A., D. Baltimore.
1989
. A novel, inducible and T cell-specific enhancer located at the 3′ end of the T cell receptor α locus.
EMBO J.
8
:
729
13
Lauzurica, P., M. S. Krangel.
1994
. Enhancer-dependent and -independent steps in the rearrangement of a human T cell receptor δ transgene.
J. Exp. Med.
179
:
43
14
Capone, M., F. Watrin, C. Fernex, B. Horvat, B. Krippl, L. Wu, R. Scollay, P. Ferrier.
1993
. TCRβ and TCRα gene enhancers confer tissue- and stage-specificity on V(D)J recombination events.
EMBO J.
12
:
4335
15
Lauzurica, P., M. S. Krangel.
1994
. Temporal and lineage-specific control of T cell receptor α/δ gene rearrangement by T cell receptor α and δ enhancers.
J. Exp. Med.
179
:
1913
16
Sleckman, B. P., C. G. Bardon, R. Ferrini, L. Davidson, F. W. Alt.
1997
. Function of the TCRα enhancer in αβ and γδ T cells.
Immunity
7
:
505
17
Stanhope-Baker, P., K. M. Hudson, A. L. Shaffer, A. Constantinescu, M. S. Schlissel.
1996
. Cell type-specific chromatin structure determines the targeting of V(D)J recombinase activity in vitro.
Cell
85
:
887
18
McMurry, M. T., C. Hernandez-Munain, P. Lauzurica, M. S. Krangel.
1997
. Enhancer control of local accessibility to V(D)J recombinase.
Mol. Cell. Biol.
17
:
4553
19
Zhong, X.-P., M. S. Krangel.
1997
. An enhancer-blocking element between α and δ gene segments within the human T cell receptor α/δ locus.
Proc. Natl. Acad. Sci. USA
94
:
5219
20
Schedl, P., F. Grosveld.
1995
. Domains and boundaries. S. C. R. Elgin, ed.
Chromatin Structure and Expression
172
IRL Press, Oxford.
21
Gerasimova, T. I., V. G. Corces.
1996
. Boundary and insulator elements in chromosomes.
Curr. Opin. Genet. Dev.
6
:
185
22
Geyer, P. K..
1997
. The role of insulator elements in defining domains of gene expression.
Curr. Opin. Genet. Dev.
7
:
242
23
Kellum, R., P. Schedl.
1991
. A position-effect assay for boundaries of higher order chromosomal domains.
Cell
64
:
941
24
Kellum, R., P. Schedl.
1992
. A group of scs elements function as domain boundaries in an enhancer-blocking assay.
Mol. Cell. Biol.
12
:
2424
25
Roseman, R., V. Pirrotta, P. K. Geyer.
1993
. The su(Hw) protein insulates expression of the Drosophila melanogaster white gene from chromosomal position-effects.
EMBO J.
12
:
435
26
Geyer, P. K., V. G. Corces.
1992
. DNA position-specific repression of transcription by a Drosophila zinc finger protein.
Genes Dev.
6
:
1865
27
Chung, J. H., M. Whitely, G. Felsenfeld.
1993
. A 5′ element of the chicken β-globin domain serves as an insulator in human erythroid cells and protects against position effect in Drosophila.
Cell
74
:
505
28
Chung, J. H., A. C. Bell, G. Felsenfeld.
1997
. Characterization of the chicken β-globin insulator.
Proc. Natl. Acad. Sci. USA
94
:
575
29
Gyurkovics, H., J. Gausz, J. Kummer, F. Karch.
1990
. A new homeotic mutation in the Drosophila bithorax complex removes a boundary separating two domains of regulation.
EMBO J.
9
:
2579
30
Galloni, M., H. Gruykovics, P. Schedl, F. Karch.
1993
. The bluetail transposon: evidence for independent cis-regulatory domains and domain boundaries in the bithorax complex.
EMBO J.
12
:
1087
31
Zhou, J., S. Barolo, P. Szymanski, M. Levine.
1996
. The Fab-7 element of the bithorax complex attenuates enhancer-promoter interactions in the Drosophila embryo.
Genes Dev.
10
:
3195
32
Hagstrom, K., M. Muller, P. Schedl.
1996
. Fab-7 functions as a chromatin domain boundary to ensure proper segment specification by the Drosophila bithorax complex.
Genes Dev.
10
:
3202
33
Diaz, P., D. Cado, A. Winoto.
1994
. A locus control region in the T cell receptor α/δ locus.
Immunity
1
:
207
34
Ortiz, B., D. Cado, V. Chen, P. W. Diaz, A. Winoto.
1997
. Adjacent DNA elements dominantly restrict the ubiquitous activity of a novel chromatin-opening region to specific tissues.
EMBO J.
16
:
5037
35
Kioussis, D., R. Festenstein.
1997
. Locus control regions: overcoming heterochromatin-induced gene inactivation in mammals.
Curr. Opin. Genet. Dev.
7
:
614
36
Ellis, J., K. C. Tan-Un, A. Harper, D. Michalovich, N. Yannoutsos, S. Philipsen, F. Grosveld.
1996
. A dominant chromatin-opening activity in 5′ hypersensitive site 3 of the human β-globin locus control region.
EMBO J.
15
:
562
37
Fraser, P., S. Pruzina, M. Antoniou, F. Grosveld.
1993
. Each hypersensitive site of the human β-globin locus control region confers a different developmental pattern of expression on the globin genes.
Genes Dev.
7
:
106
38
Takeda, S., Y.-R. Zou, H. Bluethmann, D. Kitamura, U. Muller, K. Rajewsky.
1993
. Deletion of the immunoglobulin κ chain intron enhancer abolishes κ chain gene rearrangement in cis but not λ chain gene rearrangement in trans.
EMBO J.
12
:
2329
39
Serwe, M., F. Sablitzky.
1993
. V(D)J recombination in B cells is impaired but not blocked by targeted deletion of the immunoglobulin heavy chain intron enhancer.
EMBO J.
12
:
2321
40
Manis, J. P., N. van der Stoep, M. Tian, R. Ferrini, L. Davidson, A. Bottaro, F. W. Alt.
1998
. Class switching in B cells lacking 3′ immunoglobulin heavy chain enhancers.
J. Exp. Med.
188
:
1421