Chromatin remodeling directly activates V(D)J recombination

Abstract

V(D)J recombination substrate choice is regulated to ensure that the appropriate gene segments are rearranged during lymphocyte development. It has been proposed that regulation of substrate usage is determined by changes in accessibility of the DNA targets. We show that Rag-mediated recombination of an episomal substrate in cells is affected by its packaging into chromatin. Chromatinized substrates were inefficiently rearranged, and methylation further reduced recombination. Disruption of nucleosomes by using butyrate on methylated substrates was sufficient to activate recombination, and dexamethasone could activate recombination in the absence of detectable transcription. Therefore, chromatin structure, and its manipulation by altering nucleosome positioning, can directly affect recombination efficiencies.

Figure 1

Recombination efficiency is affected by the chromatin structure of the substrate. (A) Recombination substrate pSC1 contained a 12 RSS (black triangle) embedded in the MMTV LTR. The 23 RSS (white triangle) is downstream of the luciferase reporter gene. The Epstein–Barr origin of replication (Orip) is included to allow plasmid maintenance. After V(D)J recombination, signal-joint products could be amplified by PCR (arrows) and detected in a Southern blot probed with an oligonucleotide that spanned the newly formed signal joint (rectangle). The other product of the recombination reaction (the coding joint) is shown but not assayed. (B) Drug treatment does not increase the efficiency of recombination of transiently introduced substrates. pSC1 (lanes 1–4) or methylated pSC1, pSC1 ME (lanes 6–9) were cotransfected with Rag-1 and Rag-2 expression vectors (RAG1/2) into 293 cells, and the drugs butyrate (But) or dexamethasone (Dex) were added as indicated. Assays were performed as described in A. Fourfold less pSC1 DNA than in lanes 1–4 was transfected to generate lane 5. Relative activities were determined by using the titration series in which transient pSC1 plus RAG1/2 was defined as 1. (C) Stably maintained episomes are inhibited by ≈100-fold. Tenfold dilutions of transiently transfected pSC1 and RAG1/2 extract into empty extract generates a titration curve (lanes 1–4). Stably maintained pSC1 cells transfected with RAG1/2 (lane 6) or without RAG1/2 (lane 5) were amplified in parallel. (D) Stably maintained substrates are refractory to recombination, and methylation further decreases the efficiency of recombination. Cell lines stably maintaining pSC1 (lanes 1–4) or methylated pSC1, pSC1 ME (lanes 7–10) were transfected with RAG1/2 vectors and were treated with drugs as indicated. Note: the absolute amount of recombination in D, lane 2 is ≈100-fold less than in B, lane 2. Relative activities were calculated as in B except stable pSC1 plus RAG1/2 was defined as 1. (E) The levels of Rag proteins are unaffected by drug treatments. Expression vectors for Rag-1 and Rag-2 were cotransfected into 293 cells and were treated with dexamethasone or butyrate as indicated. Bands labeled control hybridized nonspecifically with the immunodetection reagents used.

Figure 2

The chromatin structure of a recombination substrate directly affects the accessibility of the recombinase for its DNA substrate. (A) Schematic diagram of the LM-PCR assay. DNA samples were ligated to blunt linkers and were subjected to PCR by using a linker primer and a second primer internal to the Rag-dependent double-strand break (arrows) at the 12 RSS (black triangle). The amplified products were subjected to Southern blot hybridization and were probed with an oligonucleotide internal to the amplified product (rectangle). (B) Measurement of double-strand break formation. Cell lines stably maintaining pSC1 or pSC1 ME were cotransfected with Rag expression vectors and were treated with drugs as indicated (lanes 1–8). Titration analysis was performed in parallel to verify the linearity of the assay (lanes 9–12). All samples were generated from the same experiment and gel. Assays were performed as described in A.

Figure 3

Stably maintained pSC1 adopts a nucleosomal structure that can be disrupted by butyrate treatment. Nucleosome positioning on transiently (lanes 1–6) or stably maintained (lanes 7–12) pSC1, isolated either from untreated or butyrate-treated cells, was determined by treating purified nuclei with micrococcal nuclease (MNase, as indicated), and digested DNA was detected by Southern blot. Molecular weight markers are indicated on the left, and the nucleosomal ladder is indicated on the right.

Figure 4

Transcriptional activation does not accurately reflect accessibility of a recombination substrate. (A) Transcription of the MMTV LTR is increased by butyrate treatment when the promoter is stably maintained. Cells harboring transiently unmethylated (white), stably unmethylated (gray), or stably methylated (black) pSC1 were assayed for luciferase activity after drug treatment as indicated. The fold activation was calculated by setting the untreated samples to one, and the numerical values are listed above the bars. The average of three experiments is shown. The absolute level of luciferase activity of the untreated stable pSC1 was 1,062 ± 490; stable pSC1 ME was 323 ± 78; and that with no extract was 237 ± 76. (B) Western blot analysis reveals that glucocorticoid receptor (GR) is expressed in 293E cells as compared with U2OS cells that do not express glucocorticoid receptor and HeLa cells known to express glucocorticoid receptor (42, 43). (C) E1a can block glucocorticoid-dependent transactivation of the MMTV LTR. U2OS cells were transfected with glucocorticoid receptor in the presence (black) or absence (white) of E1a and dexamethasone (Dex) and were monitored for luciferase activity.

Figure 5

Model of accessibility. (A) Transiently introduced recombination substrate not packaged into nucleosomes is recombinationally and transcriptionally active. (B) Stably maintained, nucleosomal targets are refractory to recombination and transcription. (C) Steroid-dependent glucocorticoid receptor binding to the MMTV LTR disrupts nucleosome activating recombination but not transcription. (D) Disruption of the nucleosomes allows assembly of transcription factors and coactivators, allowing transcriptional transactivation and presumably recombination. E1a can block the transition from C to D by preventing p300 assembly on the promoter.