Deletions of SUS1, SAC3, and THP1 lead to a loss of association of GAL-GFP-RZ and GAL-GFP-pA mRNA dots with their respective genes. (A) GAL-GFP-RZ mRNA or (B) GAL-GFP-pA mRNA is visualized by FISH (red); respective reporter loci are visualized using the TetR-GFP/TetO₄₄₈ system (green); and nuclear periphery is visualized using Nup49-GFP (green). Images are taken at the 30-min time point after the transcriptional shutoff. Scale bar, 2 μm. (C) relative kinetics of the GAL-GFP-RZ and GAL-GFP-pA mRNA dots' separation from respective genes after transcriptional shutoff in sus1Δ cells. (Very few GAL-GFP-pA mRNA dots remain at 60 min in sus1Δ cells; dot/locus separation in WT never exceeded 5%.) Error bars represent standard deviation.

A model that integrates the multiple interactions between the activated GAL genes (green), non-nascent mRNP pools (red), and nuclear pores (black), emphasizing the dual role of the Sac3-Thp1-Sus1-Cdc31 complex (blue) at the site of transcription as well as at the nuclear periphery. (A) In WT cells, perinuclear repositioning of the GAL-GFP-pA locus occurs via a combined action of multiple mechanisms including transcriptional activator (A)-dependent recruitment as well as transcription (T)-, chromatin (C)-, and mRNP (M)-mediated capture and retention. (B) Upon shutoff of transcription, persistent chromatin- and mRNP-mediated retention prevents the immediate departure of the GAL-GFP-pA locus from the nuclear rim. (C) The abnormal and large GAL-GFP-RZ mRNP pool produced when mRNA 3′-end processing is bypassed by RZ makes additional contacts at the nuclear pore, which are not Sac3-Thp1-Sus1-Cdc31 complex-dependent, and hence is strongly retained at the nuclear rim. (D) Upon transcriptional shutoff, the synergy of these additional contacts with the NPC, acting together with the Sac3-Thp1-Sus1-Cdc31 complex-dependent and chromatin-dependent contacts, allows the GAL-GFP-RZ mRNP to persist at the nuclear periphery longer than GAL-GFP-pA. (E) Even in the absence of the Sac3-Thp1-Sus1-Cdc31 complex at the NPC, the synergy of the activator-mediated contacts, ongoing transcription, chromatin-mediated and the Sac3-Thp1-Sus1-Cdc31 complex-independent interactions of the exaggerated GAL-GFP-RZ mRNP with the NPC support a modest degree of peripheral tethering. (F) Upon transcriptional shutoff, this modest tethering fails immediately, while the aberrant mRNP remodeling (illustrated by the change in shape of the mRNP pool) and/or compromised export of the GAL-GFP-RZ mRNP in the absence of Sac3-Thp1-Sus1-Cdc31 complex causes its separation from the gene.

Altered nuclear-retained (A,B,E,F) TDH-GFP-RZ and (C,D) GAL-GFP-RZ mRNA dot morphology in sus1Δ, sac3Δ, and thp1Δ cells. (A,B,E–H), fluorescent in situ hybridization (FISH) for the TDH-GFP-RZ mRNA in (A) wild-type (WT), (B) sus1Δ, (E) gcn5Δ, (F) spt20Δ, (G) sac3Δ, and (H) thp1Δ cells. (Left panels) FISH signal (red); (middle panels) DAPI staining for total DNA (blue); (right panels) overlay. Scale bar, 2 μm. (C,D) FISH for the GAL-GFP-RZ mRNA with simultaneous visualization of the TetO₄₄₈ array integrated at the BMH1 locus on Chromosome V, <5kb away, using TetR-GFP in (C) sus1Δ cells and (D) sus1Δ cells expressing an ectopically integrated WT copy of SUS1 as well as a Nup49-GFP (the presence of Nup49-GFP has no effect on the dot morphology phenotype). (Left panels) FISH signal (red); (middle panels) FISH signal (red) overlaid on TetR-GFP signal (green); (right panels) triple overlay of FISH signal, TetR-GFP signal, and DAPI staining for the total DNA (blue). (I) In contrast to the GAL-GFP-RZ mRNA dot (C,D), no enlargement of the GAL-GFP-pA mRNA dot is observed in sus1Δ cells.

Deleting SUS1, SAC3, and THP1 has distinct transcription-independent effects on retention of the activated GAL-GFP-pA and GAL-GFP-RZ genes at the nuclear periphery. (A) Intranuclear positioning of the GAL-GFP-pA gene is indistinguishable in sus1Δ, sac3Δ, and thp1Δ cells under activating and repressing conditions. The proportion of cells with (left panel) peripherally and (right panel) internally positioned loci is plotted for galactose-grown (Gal) and glucose-grown (Glu) cells. The intranuclear position of the locus was revealed using the TetR-GFP/TetO₄₄₈ system. The nuclear periphery was visualized using Nup49-GFP fusion coexpressed in the same strain. Subperipheral fraction remains unchanged at ∼20% (data not shown). Error bars represent standard deviation. (B) Dissociation of the GAL-GFP-RZ gene locus from the nuclear periphery in sus1Δ, sac3Δ, and thp1Δ cells is dramatically accelerated and parallels the kinetics of the RNAP II runoff from the gene. The proportion of cells with a peripherally positioned GAL-GFP-RZ gene locus after 0, 30, or 60 min of transcriptional shutoff or in steady-state glucose conditions is plotted. For comparison, RNAP II occupancy after transcriptional shutoff as well as in steady-state glucose conditions, measured by ChIP, is shown as vertical bars. Error bars represent standard deviation. (C) RNAP II occupancy of the GAL-GFP-pA and GAL-GFP-RZ reporter constructs in sus1Δ cells is affected to the same extent. The RNAP II levels are normalized to the RNAP II occupancy on the endogenous PGK1 gene, which is unaffected by the loss of SUS1. Error bars represent standard deviation. (D) Northern blot quantification of the steady-state GAL-GFP-pA and GAL-GFP-RZ mRNA levels using a GFP-specific probe in WT and sus1Δ cells grown on galactose and glucose. (SCR1) Loading control used for normalization.