Overlapping and distinct functions of SPT6, PNUTS, and PCF11 in regulating transcription termination

Abstract

The histone chaperone and transcription elongation factor SPT6 is integral to RNA polymerase II (RNAPII) activity. SPT6 also plays a crucial role in regulating transcription termination, although the mechanisms involved are largely unknown. In an attempt to identify the pathways employed by SPT6 in this regulation, we found that, while SPT6 and its partner IWS1 interact and co-localize with RNAPII, their functions diverge significantly at gene termination sites. Depletion of SPT6, but not of IWS1, results in extensive readthrough transcription, indicating that SPT6 independently regulates transcription termination. Further analysis identified that the cleavage and polyadenylation factor PCF11 and the phosphatase regulatory protein PNUTS collaborate with SPT6 in this process. These findings suggest that SPT6 may facilitate transcription termination by recruiting PNUTS and PCF11 to RNAPII. Additionally, SPT6 and PNUTS jointly restrict promoter upstream transcripts (PROMPTs), whereas PCF11 presence is essential for their accumulation in the absence of SPT6 at hundreds of genes. Thus, SPT6, PCF11, and PNUTS have both distinct and overlapping functions in transcription termination. Our data highlight the pivotal role of SPT6 in ensuring proper transcription termination at the 5' and 3'-ends of genes.

Graphical Abstract

Figure 1.

SPT6 And IWS1 chromatin binding and transcriptional roles. (A) A schematic representation of SPT6 N-terminal (1–283 amino acids), core (284–1287 amino acids), and C-terminal (1288–1726 amino acids) regions. The illustration denotes N to C terminal protein domains. (B) Overexpressed Flag-tagged SPT6 from HEK293T total protein extracts was immunoprecipitated using an anti-Flag antibody and analyzed by western blotting using indicated antibodies. (C) SPT6 and IWS1 mean ChIP-seq signal intensities plotted at RefSeq-annotated genes. The solid line represents the mean. The dark area represents the standard error and light area represents the 95% confidence interval. (D) Heatmap representing the Pearson correlation coefficients calculated for IWS1, SPT6, and RNAPII merged ChIP-seq signals, on RefSeq-annotated genes (±1 kb). (E) SPT6/RNAPII and IWS1/RNAPII mean ChIP-seq signal ratios calculated on RefSeq-annotated genes in HeLa cells. The solid line represents the mean. The dark area represents the standard error and light area represents the 95% confidence interval. (F) Total protein extracts from HeLa cells were separated using a 5%–60% glycerol gradient. A total of 12 fractions were recovered after ultracentrifugation and analyzed by western blotting. (G) Western blot showing SPT6 and IWS1 depletions upon siRNA transfection in RNA-seq experiments. Lower panel shows protein quantification relative to GAPDH and to the siCT condition (n = 3). (H) SPT6 and IWS1 target genes in HeLa cells. Positive targets correspond to genes having a |FC| > |±1.5| and a padj < 0.05. (I) Density plot showing the gene count of SPT6 and IWS1 readthrough targets, highlighting the distribution of their log₂ fold-change (FC) values. (J) SEC22B gene locus featuring RNAPII, SPT6, and IWS1 ChIP-seq profiles, as well as RNA-seq profiles upon depletion of SPT6 and IWS1 (negative strand). The arrow highlights readthrough transcription (K) Western blot showing SPT6 and IWS1 depletion upon siRNA transfection (see the “Materials and methods” section for details). (L) SEC22B readthrough and mRNA levels were assayed by RT-qPCR in three independent experiments. Values were normalized to the siCT condition arbitrarily set to 1. (M) SEC22B readthrough levels were assayed by nuclear run-on experiments. The input represents total nuclear RNAs. The siSPT6 (-BrdU) condition is used to control the specificity of the anti-BrdU immunoprecipitation. Values were normalized to the siCT condition arbitrarily set to 1, and to the KDSR and 18S housekeeping genes.

Figure 2.

Evaluating functional collaboration between SPT6 and multiple termination pathways. (A) Amplicons designed to evaluate readthrough transcription at various genes. The last exon amplicon measures full-length gene transcription. (B) Heatmap displaying the relative RNA levels assayed by RT-qPCR in five independent experiments using the amplicons designed in panel (A) at the SEC22B, ZNF180, CHST3, and FLNB genes. Values were normalized to the siCT condition arbitrarily set to 1. Statistical significance of readthrough transcripts was assessed using an unpaired t-test between each condition and the siCT, with P-values indicated as follows: .1234 (ns, nonsignificant), .0332 (*), .0021 (**), .0002 (***), <.0001 (****).

Figure 3.

SPT6 interacts with IWS1, PNUTS, and PCF11. (A) Upper panel: Diagram depicting PNUTS domains: the TND (1–147 amino acids), the PP1-binding domain (348–418 amino acids) and the RNA-binding domain (674–750 amino acids). Lower panel: Diagram depicting PCF11 domains: the CTD-interacting domain (CID) (14–142 amino acids), and the RNA-binding zinc fingers (between 1343 and 1478 amino acids). The illustration denotes N-to-C terminal protein domains. (B) Nuclear protein extracts from WT or IWS1-knockout (IWS1-/-) J-Lat A1 cells were used to immunoprecipitate endogenous SPT6. Co-immunoprecipitation of IWS1, PCF11 and PNUTS was assessed (n = 3). (C) Overexpressed Flag-tagged WT PNUTS (WT) or Flag-tagged PNUTS–W401A (W401A), a mutant for PP1 interaction, were immunoprecipitated using an anti-Flag antibody from HEK293T total protein extracts. SPT6 co-immunoprecipitation was examined for both WT and mutant PNUTS (n = 3). (D) HeLa cells infected with lentiviruses containing either a control shRNA (shCT) or an shRNA targeting PNUTS (shPNUTS). Cells nuclear fractions were used to immunoprecipitate endogenous SPT6 (n = 2). “Cyto” stands for Cytoplasmic fraction, and “Nuc” stands for Nuclear fraction used as input.

Figure 4.

PNUTS, PCF11, and SPT6 chromatin distribution and transcriptional effect according to gene categories. (A) PNUTS ChIP-seq peaks (n =29 328) distribution was assessed at active and inactive enhancers, repressed regions, and RefSeq-gene promoter (TSS ± 1.5 kb), body and termination (TES ± 1.5 kb) regions. (B) RNAPII displaying promoter and termination region enrichment (Category A), promoter-only enrichment (Category B) or termination-region-only enrichment (Category C). The plots represent the mean RNAPII ChIP-seq signals. (C) Promoter-pausing index defined by the ratio between RNAPII signal in a 500 bp window surrounding the TSS and RNAPII signal in a 500 bp window starting 750 bp downstream of the TSS (regions a and b in the scheme, respectively). (D) Promoter-pausing index for the three gene categories defined in panel (B). (E) Average RNA-seq signal calculated in the promoter region (TSS ± 1.5 kb) of genes belonging to Categories A, B, and C. (F) Length (in kilobases) of genes belonging to Categories A, B, and C. (G) PNUTS, PCF11, and SPT6 binding profiles depending on gene category. Mean ChIP-seq signals are plotted. (H) RNAPII binding profile upon PNUTS depletion (left panel, our data), PCF11 depletion [44] and SPT6 depletion [5], depending on the gene category. Figure 4B, G, and H: The solid line represents the mean. The dark area represents the standard error and light area represents the 95% confidence interval.

Figure 5.

Differential readthrough transcripts upon depletion of SPT6, PNUTS, and PCF11. (A) Mean RNA-seq signal intensity shown at SPT6 readthrough targets (TES + 5 kb) in the different conditions. (B) Heatmap displaying differentially expressed readthrough transcripts in HeLa cells following SPT6 and/or PNUTS depletion. Readthrough transcripts with a log₂FC >1 in the siSPT6 condition compared to the control condition were considered positive. The readthrough regions were identified as described in Supplementary Fig. S1F. The heatmap plots the average log2FC scores (siRNA/siCT) of these regions. Three distinct clusters were defined to characterize the varying responses to SPT6 and/or PNUTS depletion. The analysis was conducted using three independent biological replicates. (C) Box-plots showing the median log₂FC of readthrough transcripts relative to the control condition for the three clusters shown in panel (B). (D–F) Same as in panels (A)–(C) but for SPT6 and/or PCF11 depletion. (G) CTNND1 gene locus featuring RNA-seq data (positive strand) following the use of the indicated siRNAs. The lower panel (PAS) displays the locations of human PASs as defined by Zhang et al. [82]. The arrow highlights readthrough transcription. The scale is shown at the top right corner of the figure.

Figure 6.

Differential PROMPTs regulation upon depletion of SPT6, PNUTS, and PCF11. (A) Mean RNA-seq signal intensity shown at SPT6 PROMPT targets (TSS, -5 kb) in the different conditions. (B) Heatmap displaying differentially expressed PROMPTs in HeLa cells following SPT6 and/or PNUTS depletion. PROMPTs with a log₂FC >1 in the siSPT6 condition compared to the control condition were considered positive. PROMPT regions were identified as described in Supplementary Fig. S6A. The heatmap plots the average log₂FC scores (siRNA/siCT) of these regions. Three distinct clusters were defined to characterize the varying responses to SPT6 and/or PNUTS depletion. The analysis was conducted using three independent biological replicates. (C) Box-plots showing the median log₂FC of PROMPTs relative to the control condition for the three clusters shown in panel (B). (D) PXDN PROMPT levels were assayed by RT-qPCR in three independent experiments. Values were normalized to the siCT condition arbitrarily set to 1. (E–G) Same as in panels (A)–(C) but for SPT6 and/or PCF11 depletion. (H) Same as in panel (D) but for the GGCT gene. (I) PVT1 gene locus featuring RNA-seq data (positive strand is in positive values and negative strand in negative values) following the use of the indicated siRNAs. The lower panel (PAS) displays the locations of human PASs as defined by Zhang et al. [82]. The arrow highlights PROMPTs. The scale is shown at the top right corner of the figure.

Figure 7.

Model for the regulation of transcription termination by SPT6, PNUTS, and PCF11 at promoters and gene ends. At transcription termination sites, SPT6, PNUTS, and PCF11 prevent readthrough transcription. At gene promoters, SPT6 and PNUTS establish an initial block to limit PROMPTs. When this primary block is removed, PCF11 promotes PROMPTs, either directly or by inhibiting the recruitment of another PROMPT repressor. Created in BioRender. Bejjani, F. (2025) https://BioRender.com/f07e489.