Regulation of Antisense Transcription by NuA4 Histone Acetyltransferase and Other Chromatin Regulatory Factors

Abstract

NuA4 histone lysine (K) acetyltransferase (KAT) promotes transcriptional initiation of TATA-binding protein (TBP)-associated factor (TAF)-dependent ribosomal protein genes. TAFs have also been recently found to enhance antisense transcription from the 3' end of the GAL10 coding sequence. However, it remains unknown whether, like sense transcription of the ribosomal protein genes, TAF-dependent antisense transcription of GAL10 also requires NuA4 KAT. Here, we show that NuA4 KAT associates with the GAL10 antisense transcription initiation site at the 3' end of the coding sequence. Such association of NuA4 KAT depends on the Reb1p-binding site that recruits Reb1p activator to the GAL10 antisense transcription initiation site. Targeted recruitment of NuA4 KAT to the GAL10 antisense transcription initiation site promotes GAL10 antisense transcription. Like NuA4 KAT, histone H3 K4/36 methyltransferases and histone H2B ubiquitin conjugase facilitate GAL10 antisense transcription, while the Swi/Snf and SAGA chromatin remodeling/modification factors are dispensable for antisense, but not sense, transcription of GAL10. Taken together, our results demonstrate for the first time the roles of NuA4 KAT and other chromatin regulatory factors in controlling antisense transcription, thus illuminating chromatin regulation of antisense transcription.

FIG 1

NuA4 KAT promotes GAL10 antisense transcription. (A) Schematic diagram showing the experimental strategy for analysis of the GAL10 antisense transcript. The P1 primer targeted toward the 5′ end of the GAL10 antisense transcript was extended by AMV reverse transcriptase-based reverse transcription at 42°C, and subsequently the extended primer was amplified by primer pairs targeted to the coding regions, M and N, of GAL10 and GAL7, respectively. The numbers are presented with respect to the position of the translational stop codon (TGA) of GAL10. Regions P, M, O, and N represent GAL1 core, GAL10 5′ end, GAL10 3′ end, and GAL7 5′ end (or GAL7 open reading frame [ORF]), respectively, in the ChIP assay. (B and C) Analysis of the GAL10 antisense transcript in the esa1-ts mutant and wild-type (WT) strains in dextrose-containing growth medium. The RNA level in the wild-type strain was set to 100, and the relative RNA level in the mutant strain was plotted in panel C. (D) Sense transcription analysis of GAL10, RPS5, and ACT1 in the esa1-ts and wild-type strains in galactose-containing growth medium by RT-PCR.

FIG 2

Analysis of NuA4 KAT recruitment to the 3′ end of the GAL10 coding sequence. (A and B) Esa1p is associated with the 3′ end of the GAL10 coding sequence in dextrose-containing growth medium. Immunoprecipitation (IP) was carried out using an anti-Myc antibody against Myc-tagged Esa1p. Immunoprecipitated DNA was analyzed by PCR using the primer pairs encompassing the 3′ end of the GAL10 coding sequence (region O in Fig. 1A) and a region with 18S rDNA. The ratio of the immunoprecipitate over the input in the autoradiogram (termed a ChIP signal) was measured. The maximum ChIP signal was set to 100, and other ChIP signals relative to the maximum ChIP signal (represented as relative ChIP signal or relative occupancy) were plotted. (C, D, and E) ChIP analysis of Myc-tagged Esa1p at the 3′ and 5′ ends (regions O and M, respectively, in Fig. 1A) of the GAL10 coding sequence, the GAL1 core promoter (region P in Fig. 1A), the RPS5 UAS, and an inactive region within chromosome V (Chr.-V). (F) ChIP analysis of Myc-tagged Esa1p at the 3′ end of the GAL10 coding sequence and RPS5 UAS in the presence and absence of a Reb1p-binding site at the 3′ end of the GAL10 coding sequence. (G) ChIP analysis of Myc-tagged Eaf5p at the 3′ and 5′ ends of the GAL10 coding sequence and Chr.-V in dextrose-containing growth medium. (H) ChIP analysis of Myc-tagged Eaf5p at the 3′ end of the GAL10 coding sequence in the wild-type and Δeaf1 strains.

FIG 3

Analysis of histone H4 acetylation at the 3′ end of the GAL10 coding sequence. (A to C) Analysis of the levels of histone H4 acetylation and histone H3/H4 tetramer at the 3′ and 5′ ends of the GAL10 coding sequence and GAL7 ORF (or GAL7 5′ end) in the esa1-ts mutant and wild-type strains. Immunoprecipitation was carried out using an anti-histone H4 acetylation antibody against acetylated histone H4 or an anti-histone H3 antibody against histone H3 of the histone H3/H4 tetramer. (D) Analysis of histone H4 acetylation and histone H3 at the 5′ and 3′ ends of the GAL10 coding sequence and 5′ end of the GAL7 coding sequence. (E and F) The Reb1p-binding site regulates histone H4 acetylation at the 3′ end of the GAL10 coding sequence. (G) Analysis of Rpb1p association with GAL10 in dextrose-containing growth medium in the esa1-ts mutant and wild-type strains.

FIG 4

NuA4 KAT-dependent GAL10 antisense transcription is not altered in response to rapamycin treatment. (A) RT-PCR analysis of the GAL10 antisense transcript with (+R) or without (−R) rapamycin treatment. (B) The transcription data shown in panel A were plotted.

FIG 5

GAL10 antisense transcription is regulated by histone H3 K4 and K36 methyltransferases. (A) Analysis of histone H3 K4 trimethylation (H3-K4-Me₃) and histone H3 levels at the 5′ and 3′ ends of the GAL10 coding sequence and the 5′ end of the GAL7 coding sequence. (B) ChIP analysis of Myc-tagged Set1p at the 5′ and 3′ ends of the GAL10 coding sequence, Chr.-V, and 5′ ends of the ADH1 and GAL7 coding sequence. The fold increase of Set1p ChIP signal relative to HA is plotted. (C and D) RT-PCR analysis of GAL10 antisense RNA in the Δset1 and wild-type strains. The transcription data shown in panel C were plotted in panel D. (E and F) RT-PCR analysis of GAL10 antisense RNA in the Δset3 and wild-type strains. (G) RT-PCR analysis of GAL10 antisense RNA in the Δrpd3 and wild-type strains. (H) ChIP analysis of histone H4 acetylation and histone H3 levels at GAL10 in the wild-type and Δset1 strains. (I and J) RT-PCR analysis of GAL10 antisense RNA in the Δset2 and wild-type strains. (K and L) ChIP analysis of histone H3 K36 trimethylation (H3 K36-Me₃) and histone H3 at GAL10, GAL7, and Chr.-V in dextrose-containing growth medium. The maximum ChIP signal was set to 100, and other ChIP signals relative to the maximum ChIP signal were plotted.

FIG 6

GAL10 antisense transcription is regulated by histone H2B ubiquitin conjugase and Paf1p. (A and B) Analysis of GAL10 antisense RNA in the Δrad6 and wild-type strains. (C) ChIP analysis of Myc-tagged Rad6p at the 5′ and 3′ ends of the GAL10 coding sequence, Chr.-V, and 5′ ends of the ADH1 and GAL7 coding sequences. (D) ChIP analysis of Flag-tagged histone H2B and HA-tagged ubiquitin at the 5′ and 3′ ends of the GAL10 coding sequence in the absence of the RING domain of Bre1p (the bre1Δ500 strain without 500 amino acids at Bre1p's C terminus that contain the RING domain). Immunoprecipitation was carried out using anti-Flag and anti-HA antibodies against Flag-tagged histone H2B and HA-tagged ubiquitin as described previously (31). (E and F) ChDIP analysis for the levels of ubiquitylated histone H2B in the wild-type (YKH045) and bre1Δ500 strains expressing Flag-tagged H2B and HA-tagged ubiquitin. (G) Analysis of histone H4 acetylation and histone H3 levels at GAL10 in the wild-type and Δrad6 strains. (H) Analysis of Rpb1p association with the 3′ and 5′ ends of the GAL10 coding sequence in the wild-type and bre1Δ500 strains. The maximum ChIP signal was set to 100, and the other ChIP signals relative to maximum ChIP signal were plotted. (I and J) Analysis of GAL10 antisense RNA in the Δpaf1 and wild-type strains.

FIG 7

GAL10 antisense transcription is not regulated by SAGA and Swi/Snf. (A and B) Analysis of GAL10 antisense RNA in the Δspt20 and wild-type strains. (C and D) Analysis of GAL10 antisense RNA in the Δswi2 and wild-type strains.