Biochemical analysis of TREX complex recruitment to intronless and intron-containing yeast genes

Abstract

The TREX complex is involved in both transcription elongation and mRNA export and is recruited to nascent transcription complexes. We have examined Yra1p, Sub2p and Hpr1p recruitment to nine genes of varying lengths and transcription frequencies. All three proteins increase from the 5' to the 3' ends of the four intronless genes examined. A modified chromatin immunoprecipitation assay that includes an RNase step indicates that Sub2p is bound to nascent RNA, Yra1p is associated with both RNA and DNA, and Hpr1p is associated with DNA. Although Hpr1p is recruited similarly to both intronless and intron-containing genes, low Yra1p and Sub2p levels are present on a subset of intron-containing genes. The residual Yra1p and Sub2p recruitment is less RNA-associated, and this correlates with high levels of U1 SnRNP on these genes. These experiments support a model in which TREX is recruited via the transcription machinery and then Yra1p and Sub2p are transferred to the nascent RNA. On some intron-containing genes, retention and/or transfer of Yra1p and Sub2p to nascent RNA are inhibited.

Figure 1

RNA PolII is differentially recruited to four intronless genes. (A) The schematic diagram shows the relative positions of the primer pairs used to analyze the recruitment to the intronless genes PMA1, PGK1, ADH3 and FBA1. (B) The amount of PolII detected via ChIP using a monoclonal antibody against PolII (8WG16; white bars) or an anti-HA antibody to recognize an HA-tagged RNA PolII (gray bars) is very similar. The position of the PMA1 primer pairs is indicated in (A). (C) The amount of PolII recruited to four different intronless genes was examined using ChIPs with the monoclonal PolII antibody (8WG16). The data are presented as the fold enrichment of PolII relative to the background binding and represent the average results from greater than 10 experiments. The striped bars indicate primer pairs that are approximately 1000 bp after transcription initiation.

Figure 2

TREX recruitment to four different intronless genes. Single chromatin preparations from yeast strains expressing epitope-tagged Yra1p (KAY136), Sub2p (FSY1473) or Hpr1p (FSY1525) were immunoprecipitated using either antibodies recognizing PolII (8WG16) or the epitope tag. To determine the amount of TREX component recruited to the site of transcription per PolII/nascent RNA complex, the amount of TREX component recruited was divided by the amount of PolII recruited (see Materials and methods for details). The striped bars indicate which primer pairs are located approximately 1000 bp after transcription initiation. (A) Yra1p, (B) Sub2p and (C) Hpr1p. The data presented are the average of two independent experiments.

Figure 3

The recruitment of the MS2-HA fusion protein to the nascent transcription complex is sensitive to RNase. ChIPs were performed on yeast expressing an MS2-HA fusion protein and pGAL-MS2-GFP-pA-CEN (KAY406). (A) The diagram shows the features of the GAL1-MS2-GFP-pA construct and the position of three primer sets used in the ChIPs described in (B, C). (B) An MS2-HA fusion protein is cotranscriptionally recruited to a GFP mRNA containing two MS2 binding sites in the 5′-UTR (gray bars). The recruitment of MS2-HA is dependent on transcription; the MS2 fusion protein is not recruited when transcription from the GAL1 promoter is repressed by the addition of glucose (white bars). (C) An RNase treatment step was added to ChIPs with antibodies against either PolII or MS2-HA to determine whether RNase can distinguish between RNA- and DNA-bound proteins in a ChIP assay. The data are presented as the percentage of protein that remains after RNase treatment (see Materials and methods). Only between 20 and 40% of the MS2-HA recruited to GAL1-MS2-GFP remains after RNase treatment. The data presented are the average of two independent experiments.

Figure 4

Sub2p, Yra1p and Hpr1p are differentially associated with the nascent mRNA. ChIPs with an RNase treatment step were performed on yeast strains expressing epitope-tagged Sub2p (FSY1473), Yra1p (KAY136) and Hpr1p (FSY1525). The methods used are identical to those described in Figure 3. The PMA1 primer pairs used are described in Figure 1. The RNase sensitivity of Sub2p is almost identical to MS-HA; only 20–40% of the Sub2p remains at the site of transcription after RNase treatment (n=4). In contrast, the RNase sensitivity of Hpr1p (n=4) resembles that of PolII (n=10); it is insensitive to RNase treatment. Yra1p shows partial RNase sensitivity (n=7): it is more RNase sensitive than PolII at both the 5′ and middle primer; however this difference is statistically significant only at the 5′ end of PMA1.

Figure 5

RNA PolII recruitment reflects the transcriptional frequency of nine intronless and intron-containing genes. (A) The schematic diagram shows the relative positions of the primer pairs used to examine the recruitment of PolII and TREX components to intron-containing genes. The striped regions indicate the position of the intron. (B) ChIPs with the monoclonal PolII antibody (8WG16) indicate that different amounts of PolII are recruited to four intronless (PMA1, PGK1, ADH3 and FBA1) and five intron-containing (ASC1, ACT1, RPL28, DBP2 and SEC27) genes. The striped bars indicate the amount of PolII recruitment approximately 1000 bp after transcription initiation. The data are presented as fold PolII enrichment relative to background binding and represent the average results from greater than 10 experiments (see Materials and methods for details).

Figure 6

Yra1p and Sub2p are poorly recruited to a subset of intron-containing genes. ChIPs were performed on strains expressing epitope-tagged Yra1p (KAY136), Sub2p (FSY1473), Hpr1p (FSY1525) and the U1 SnRNP component, Prp42p (YKK25). The amount of Yra1p, Sub2p, Hpr1p and U1 snRNP recruited per PolII/nascent RNA complex is examined on five different intron-containing genes. The primer pairs used are described in Figure 5A and the striped bars (A–C) indicate primer pairs that are approximately 1000 bp after transcription initiation and the checkered bars (D) indicate the position of the intron. (A) Hpr1p, (B) Yra1p, (C) Sub2p and (D) U1 snRNP. PCR reactions for all nine genes were performed from the same ChIP samples; therefore, gene-to-gene differences in TREX and U1 snRNP recruitment cannot be attributed to experimental variation. The data presented are the average of two independent experiments.

Figure 7

Poor cotranscriptional recruitment of Yra1p and Sub2p is correlated with a decrease in RNA association. Chromatin preparations from strains expressing epitope-tagged Yra1p (KAY136) or Sub2p (FSY1473) were incubated with or without RNase. The primer pairs used in these experiments are described in Figure 5A. (A) Sub2p and (B) Yra1p. The data presented are the average of two independent experiments.

Figure 8

Removal of the intron from ASC1 restores Yra1p levels and its RNA association. ChIPs were performed using epitope-tagged Yra1p strains expressing either ASC1+intron (SALY56) or ASC1−intron (SALY59). (A) The amount of Yra1p recruited to ASC1 increases when the intron is removed; compare squares (−intron) with circles (+intron). The data are presented relative to the distance from the ATG to account for the differences in transcript length between the two ASC1 constructs. (B) Yra1p becomes more RNA associated on ASC1 when the intron is removed (compare squares (−intron) with circles (+intron).

Figure 9

Model for TREX component recruitment to intronless and intron-containing genes. (A) Our data indicate that once the TREX complex is recruited to the site of transcription, presumably via the transcription machinery, the complex dissociates. The THO complex remains associated with the DNA or transcription machinery. Sub2p is immediately and efficiently transferred to the nascent RNA. The Yra1p divides into two pools: one remains associated with the DNA and the other is transferred to the RNA. (B) On intron-containing genes, the TREX complex is recruited normally but the transfer of Sub2p and Yra1p to the RNA may be hindered by the spliceosome. As a result, less Sub2p and Yra1p are retained at the site of transcription (S=Sub2p; Y=Yra1p).