Repression of RNA polymerase II elongation in vivo is critically dependent on the C-terminus of Spt5

Abstract

The stalling of RNA polymerase II (RNAPII) at the promoters of many genes, including developmental regulators, stress-responsive genes, and HIVLTR, suggests transcription elongation as a critical regulatory step in addition to initiation. Spt5, the large subunit of the DRB sensitivity-inducing factor (DSIF), represses or activates RNAPII elongation in vitro. How RNAPII elongation is repressed in vivo is not well understood. Here we report that CTR1 and CTR2CT, the two repeat-containing regions constituting the C-terminus of Spt5, play a redundant role in repressing RNAPII elongation in vivo. First, mis-expression of Spt5 lacking CTR1 or CTR2CT is inconsequential, but mis-expression of Spt5 lacking the entire C-terminus (termed NSpt5) dominantly impairs embryogenesis in zebrafish. Second, NSpt5 de-represses the transcription of hsp70-4 in zebrafish embryos and HIVLTR in cultured human cells, which are repressed at the RNAPII elongation step under non-inducible conditions. Third, NSpt5 directly associates with hsp70-4 chromatin in vivo and increases the occupancy of RNAPII, positive transcription elongation factor b (P-TEFb), histone H3 Lys 4 trimethylation (H3K4Me3), and surprisingly, the negative elongation factor A (NELF-A) at the locus, indicating a direct action of NSpt5 on the elongation repressed locus. Together, these results reveal a dominant activity of NSpt5 to de-repress RNAPII elongation, and suggest that the C-terminus of Spt5 is critical for repressing RNAPII elongation in vivo.

Figure 1. Injection of F-Nspt5 RNA into WT dominantly impairs embryonic development in zebrafish.

(A) The functional domains of Spt5 based on previous in vitro analysis , . (B–E) Morphological phenotypes of WT or fog^s30 embryos (B), WT or fog^s30 embryos injected with F-spt5 RNA (C), F-spt5ΔCTR1 RNA (D), or F-Nspt5 RNA (E).

Figure 2. NSpt5 de-represses hsp70-4 expression in the absence of heat shock.

(A–C) Embryonic morphology or GFP fluorescence of hsp70-GFP transgenic embryos. hsp70-GFP transgenic WT injected with F-spt5 RNA (A), hsp70-GFP transgenic fog^s30 mutant (B), and hsp70-GFP transgenic WT injected with F-Nspt5 RNA (C). (D–E) Confocal images of FLAG- and GFP- double immuno-labeled embryos, injected with F-spt5 RNA (D), or with F-Nspt5 RNA (E). (F) GFP fluorescence in hsp70-GFP transgenic embryos injected with F-spt5 RNA (left) or F-Nspt5 RNA (right) and subjected to heat shock for one hour. (G–H) Quantitative RT-PCR analysis shows de-repression of hsp70-4 expression in 6 hpf Nspt5-expressing embryos (G), and no significant difference of hsp70-4 expression between F-Nspt5-expressing, F-Spt5-expressing, and control embryos upon heat shock (H). (I) F-NSpt5 increases transcription from the HIVLTR. CAT activity of Hela cells that express RD and F-Spt5 (lanes 1 and 3), or F-NSpt5 (lanes 2 and 4), in the absence (lanes 1 and 2) or presence of Tat (lanes 3 and 4). Results are presented in arbitrary units. Error bars represent S.E.M. from three independent experiments.

Figure 3. NSpt5 interferes with the repressive but not the stimulatory activity of endogenous DSIF in a dominant manner in vitro.

(A) The elongation stimulation activity of DSIF was assayed using pSLG402 as a template, which generates short (promoter-proximal) and long (promoter-distal, dependent on the elongation stimulatory activity of DSIF) RNase T1-resistant products under the control of the adenovirus major-late promoter. Transcription initiation/elongation was allowed to proceed for the indicated times. Time-dependent increase of distal transcripts is observed in the control (first three lanes), while NSpt5 slightly enhanced transcription at 2X or 4X concentration but not at 8X concentration. (B) pTF3-6C2AT, which generates a 380-nt RNase T1-resistant product under the control of the adenovirus E4 promoter, was used as a template, and transcription was allowed to proceed for 10 minutes. This product is sensitive to the elongation repressive activity of DSIF (in the presence of the P-TEFb inhibitor DRB)(first two lanes). NSpt5 inhibits the repression activity of endogenous DSIF at 8X concentration.

Figure 4. RNAPII elongates on the HIVLTR and hsp70-4 chromatin in the presence of F-NSpt5.

(A) ChIP and qRT-PCR were performed with RNAPII antibodies and indicated primers (arrows) in Hela cells. Data are presented as percent of input material immunoprecipitated with specific antibodies over those with the IgG control. Error bars represent S.E.M. of triplicate measurements from three independent experiments. (B) A scheme of the in vivo ChIP analysis in zebrafish. (C) The gene structure of hsp70-4, highlighting the location of the hsp70-4 primer sets that are used during real-time PCR amplification of the immuoprecipitated material. The long arrow indicates the start of transcription. (D) The relative enrichment of RNAPII on hsp70-4 with RNAPII antibody over the IgG control under heat shock condition. (E) The relative enrichment of RNAPII on hsp70-4 in NSpt5 expressing embryos. Error bars represent S.E.M. of duplicate measurements from two independent experiments.

Figure 5. NSpt5 directly interacts with the hsp70-4 chromatin in vivo.

Charts show the percent of input material immunoprecipitated in different regions of hsp70-4 chromatin. The relative enrichment of ChIP and qRT-PCR values obtained with Flag antibody over the IgG control. Error bars represent S.E.M. of duplicate measurements from two independent experiments.

Figure 6. ChIP at the hsp70-4 chromatin in Nspt5 expressing zebrafish embryos.

(A to D) Charts show the percent of input material immunoprecipitated in different regions of hsp70-4 chromatin. The relative enrichment of ChIP and qRT-PCR values obtained with CDK9 antibody over the IgG control (A), H3K4Me3 antibody over the IgG control (B), or H3K79Me2 antibody over the IgG control (C), NELF-A antibody over the rabbit serum control (D). Error bars represent S.E.M. of duplicate measurements from two independent experiments.

Figure 7. A model depicts the role of Spt5 in regulating RNAPII elongation in vivo.

(Top) Stalled RNAPII complex on a gene that is subjected to elongation regulation, in an un-induced state. (Bottom left) RNAPII complex with the incorporation of NSpt5. (Bottom right) RNAPII complex in an induced state.