Research Article

Overcoming the cytoplasmic retention of GDOWN1 modulates global transcription and facilitates stress adaptation

School of Life Sciences, Lanzhou University, China
Cuiying Honors College, Lanzhou University, China
Key Laboratory of Cell Activities and Stress Adaptations, Ministry of Education, Lanzhou University, China

Dec 7, 2022

https://doi.org/10.7554/eLife.79116

Version of Record

Accepted for publication after peer review and revision.

Download
Cite
Share
CommentOpen annotations (there are currently 0 annotations on this page).

420 views

Version of Record published: December 7, 2022 (This version)
Accepted: November 21, 2022
Received: March 31, 2022
Preprint posted: March 30, 2022 (Go to version)

1. Of interest
Cohesin-independent STAG proteins interact with RNA and R-loops and promote complex loading

Hayley Porter, Yang Li ... Suzana Hadjur

Research Article Updated Jun 2, 2023
Further reading

Article
Figures and data
Abstract
Editor's evaluation
Introduction
Results
Discussion
Materials and methods
Data availability
References
Decision letter
Author response
Article and author information
Metrics

Abstract

Dynamic regulation of transcription is crucial for the cellular responses to various environmental or developmental cues. Gdown1 is a ubiquitously expressed, RNA polymerase II (Pol II) interacting protein, essential for the embryonic development of metazoan. It tightly binds Pol II in vitro and competitively blocks the binding of TFIIF and possibly other transcriptional regulatory factors, yet its cellular functions and regulatory circuits remain unclear. Here, we show that human GDOWN1 strictly localizes in the cytoplasm of various types of somatic cells and exhibits a potent resistance to the imposed driving force for its nuclear localization. Combined with the genetic and microscope-based approaches, two types of the functionally coupled and evolutionally conserved localization regulatory motifs are identified, including the CRM1-dependent nucleus export signal (NES) and a novel Cytoplasmic Anchoring Signal (CAS) that mediates its retention outside of the nuclear pore complexes (NPC). Mutagenesis of CAS alleviates GDOWN1’s cytoplasmic retention, thus unlocks its nucleocytoplasmic shuttling properties, and the increased nuclear import and accumulation of GDOWN1 results in a drastic reduction of both Pol II and its associated global transcription levels. Importantly, the nuclear translocation of GDOWN1 occurs in response to the oxidative stresses, and the ablation of GDOWN1 significantly weakens the cellular tolerance. Collectively, our work uncovers the molecular basis of GDOWN1’s subcellular localization and a novel cellular strategy of modulating global transcription and stress-adaptation via controlling the nuclear translocation of GDOWN1.

Editor's evaluation

This important study identifies two distinct nuclear export elements and a strong cytoplasmic anchoring sequence in the GDOWN1 transcription factor that restricts its nuclear import and its ability to inhibit RNA polymerase II transcription. The study shows how this mechanism is modulated in stress conditions that promote GDOWN1 nuclear localization as part of a protective response. This study presents compelling evidence for the role of GDOWN1 in transcriptional regulation and should be of wide general interest.

https://doi.org/10.7554/eLife.79116.sa0

Introduction

In eukaryotes, RNA Polymerase II (Pol II) catalyzes the RNA synthesis of the clear majority of protein coding genes and a great number of non-coding genes in eukaryotic genomes (Haberle and Stark, 2018; Osman and Cramer, 2020). The Pol II-associated transcription machinery is composed of the 12-subunit Pol II and a collection of dynamically bound and delicately coordinated factors, including general transcription factors (TFIID, TFIIA, TFIIB, TFIIF, TFIIE, and TFIIH)(Cramer, 2019; Fischer et al., 2019), Pol II processivity-controlling factors (such as the writer, reader, and eraser factors for modifying and recognizing the carboxyl terminal domain [CTD] of RBP1 (Hsin and Manley, 2012; Jeronimo et al., 2016; Sansó and Fisher, 2013; Yurko and Manley, 2018), the positive or negative elongation factors and the termination factors) (Core and Adelman, 2019; Jonkers and Lis, 2015; Proudfoot, 2016; Zhou et al., 2012), and the co-transcriptional RNA processing and modifying factors (such as the capping enzymes, splicing machinery, RNA modification enzymes, and RNA cleavage factors) (Kachaev et al., 2020; Kilchert and Vasiljeva, 2013; Neugebauer, 2019; Noe Gonzalez et al., 2021; Schier and Taatjes, 2020; Sun et al., 2020). Many of these Pol II-binding and regulatory factors are not only essential for facilitating the production and processing of transcripts but also play critical roles in dynamic integrating the intracellular and extracellular information and adjusting Pol II’s target specificity and enzymatic activities in real time to maintain cell homeostasis (Lynch et al., 2018; McNamara et al., 2016; Muniz et al., 2021; Schier and Taatjes, 2020).

Gdown1 was initially identified as a protein copurified with Pol II in calf thymus and porcine liver, and designated as the 13th subunit of Pol II due to its high-affinity interaction with Pol II (Hu et al., 2006). Data from in vitro transcription assays along with EMSA and chemical crosslinking-based structural analyses have demonstrated that Gdown1 interacts with multiple Pol II subunits via its various regions and strongly inhibits the binding and/or functions of a series of transcription regulatory factors, including the factors required for transcription initiation, such as TFIIF (Jishage et al., 2012; Jishage et al., 2018), the factors involved in the productive elongation such as RTF1/PAF1C (Ball et al., 2022), and the transcription termination factor such as TTF2 (Cheng et al., 2012).

Although the biochemical properties support Gdown1’s potential in regulating Pol II transcription, it is largely unknown how exactly its regulatory activities are executed under the physiological circumstances. Knockout (KO) of Gdown1 in flies and mice caused embryonic lethality, and moreover, the attempt of establishing a Gdown1^-/- mouse embryonic stem cell line failed, pointing out its essential roles during the embryonic development (Jishage et al., 2020; Jishage and Roeder, 2020; Jishage et al., 2018). Interestingly, Gdown1 has been reported as a nucleocytoplasmic shuttling protein in flies. It colocalizes with Pol II in the nuclei at the transcriptionally silent syncytial blastoderm stage and moves to the cytoplasm at the later blastoderm stage when the global transcription is activated, and meanwhile, Gdown1 is found to be retained in the nuclei of the transcriptionally silent pole cells (Jishage and Roeder, 2020). Although not being proven yet, these findings imply that controlling the nucleocytoplasmic shuttling of Gdown1 may be tightly associated with tuning the transcriptional regulatory activities. At the beginning of the embryonic development, the nuclear Gdown1 may serve as a global transcriptional inhibitor, and once the embryos get adequately prepared, exclusion of Gdown1 from the nucleus provides an effective way to promote zygotic genome activation (ZGA). Further studies are certainly required to explore the functional and regulatory mechanisms behind and find out whether the above phenomena revealed in flies are similarly applied in higher animals or in other situations.

Evidence is piling up that Gdown1 also plays critical roles in somatic cells. It is expressed throughout the whole life cycle of flies (Jishage et al., 2018) and ubiquitously present across various types of mouse tissues (data not shown). Mice with Gdown1 specifically knocked out in liver were found viable and relatively normal, yet tended to trigger the quiescent hepatocytes to re-enter cell cycle in the absence of hepatic injury, highlighting its important roles in maintaining the homeostasis of hepatocytes (Jishage et al., 2020). Further ChIP-Seq analyses revealed that Gdown1 bound with the elongating Pol II at many genes in liver (Jishage et al., 2020). However, the RNA-Seq results unexpectedly revealed that the expression levels of those direct targets of Gdown1 were dramatically reduced upon its ablation, suggesting a positive effect of Gdown1 on transcription (Jishage et al., 2020). Thus, it is necessary to clarify the underlying reasons behind the apparently opposite transcriptional regulatory effects of Gdown1 observed in somatic cells and in the defined in vitro transcription assays.

To further explore GDOWN1’s functions in somatic cells, we started out by examining its subcellular localization in many cultured human cell lines and confirmed that it was predominantly localized in the cytoplasm. Based on the known functions, it’s reasonable to presume that human GDOWN1 is a nucleocytoplasmic shuttling protein. However, our data demonstrated that GDOWN1 was subjected to a very tight restriction for its nuclear import under the regular cell culture conditions. Based on these findings, we established various mutagenesis-based screening assays of GDOWN1 and identified multiple intrinsic localization-regulatory signals and their working mechanisms. In addition, manipulation of GDOWN1’s nuclear translocation caused a significant reduction of both the protein levels of Pol II and the global transcription. Its massive and constant nuclear accumulation resulted in a severe growth inhibition and even triggered cell death. Moreover, we provided the experimental evidence that the nuclear import of GDOWN1 was naturally induced upon certain cellular stresses and its genetic ablation resulted in a decreased cell viability under stresses. Overall, our data revealed a novel function of GDOWN1 in facilitating stress adaptation via modulating transcriptional homeostasis, and this protective strategy was achieved by its stress-responsive nuclear translocation.

Results

Gdown1 is primarily a cytoplasmic localized protein in mammalian somatic cells

We started out to detect the subcellular localization of GDOWN1 in the cultured human cell lines by ectopically expressing GDOWN1 fused with a fluorescent tag at its N- or C-terminus or simply with a Flag tag. Consistent with the previous observation in adult flies (Jishage et al., 2018) and the recent report in human somatic cells (Ball et al., 2022), the localization signals of GDOWN1 were exclusively present in the cytoplasm of HeLa cells, regardless of the position or size of the fused tags (Figure 1A). To explore GDOWN1’s functions in the nucleus, two nuclear localization signals (NLS) were fused to GDOWN1 at each end, which were known to be efficient for driving the 160 kDa SpCas9 protein into the nucleus in a commonly used CRISPR-vector pX459 (Ran et al., 2013). Unexpectedly, the addition of two NLS did not affect GDOWN1’s cytoplasmic localization (Figure 1A). We then detected the nucleocytoplasmic distribution of the endogenous GDOWN1 in the fractionated cell lysates from various human and mouse cell lines by Western Blotting (WB) using two KO-verified GDOWN1 antibodies (Figure 1—figure supplement 1A). The results clearly indicated that the endogenous GDOWN1 was predominantly located in the cytoplasmic fractions in all the human and mouse somatic cell lines tested and only a small fraction was seen in the nuclear extract of the mouse embryonic stem cell line E14TG2a (Figure 1B, Figure 1—figure supplement 1B). These data indicate that GDOWN1 is a strictly cytoplasmic protein in various human and mouse somatic cells.

Figure 1 with 1 supplement see all

Download asset Open asset

Detection of the subcellular localization of GDOWN1 or BiFC signal between GDOWN1 and some transcription related factors.

(A) The ectopically expressed human GDOWN1 in HeLa cells was stringently localized in the cytoplasm. Human GDOWN1 proteins fused with indicated tags, including a fluorescent tag at either terminus, a Flag tag alone or together with two NLS motifs, were ectopically expressed in HeLa cells and the subcellular localization was detected by directly monitoring the fluorescent signal or by immunofluorescence assays (IF) using an anti-FLAG antibody. (B) The endogenous human or mouse GDOWN1 was stringently located in the cytoplasm. Each indicated cell line was fractionated to separate the cytosol from the nuclei, and the cytoplasmic fraction, C,the nuclear fraction, N, and the whole cell lysate, T (total), were further detected by WB. α-TUBULIN and FBL (a nucleolus protein) were used as markers of the cytoplasmic and nuclear fractions, respectively. (C) GDOWN1 remained in the cytoplasm upon LMB treatment. The indicated cell lines were subjected to either mock or LMB treatment (detailed below) before further fractionation and WB analyses. (D) BiFC analyses of the protein-protein interactions between GDOWN1 and its potential binding partners. (E) BiFC analyses of the protein-protein interactions between NEFL-E•NEFL-A, SPT4•SPT5, GDOWN1•GDOWN1 or 3xNLS-GDOWN1•3xNLS-GDOWN1. Proteins of interest were cloned and fused with either VN (the N-terminus of Venus) or YC (the C-terminus of YFP), and each of the indicated pair of plasmids was co-transfected into HeLa cells, and the confocal microscopy images were acquired 24 hr post transfection. The LMB treatment was carried out at 20 nM final concentration for 6 hr and the mock treatment was done with an equal volume of ethanol in parallel. Nuclear DNA was stained by Hoechst 33342. scale bars—30 μm. Without further labeled with details, the GDOWN1 antibody used in WB assays was the one generated from rabbits.

Figure 1—source data 1 Raw data of WB for Figure 1B and C.: https://cdn.elifesciences.org/articles/79116/elife-79116-fig1-data1-v1.zip
Download elife-79116-fig1-data1-v1.zip

Combining GDOWN1’s potential nuclear functions and our observation of its cytoplasmic localization, it is reasonable to hypothesize that GDOWN1 is a nucleocytoplasmic shuttling protein that transiently functions in the nucleus in a tightly controlled mode. Most of the nucleocytoplasmic shuttling proteins contain a nuclear export signal (NES) and a typical NES contains a hydrophobic leucine-rich motif recognized by the ubiquitous transporter chromosome maintenance protein 1, CRM1 (la Cour et al., 2004; Xu et al., 2010). To test the possibility of GDOWN1 being a CRM1 cargo, HeLa and SW620 cells were treated with leptomycin B (LMB), a potent CRM1 inhibitor for efficiently blocking CRM1-NES interaction (Kudo et al., 1999; Kudo et al., 1998).

Using two KO-verified GDOWN1 antibodies, our WB data unambiguously demonstrated that GDOWN1 didn’t accumulate in the nucleus upon LMB treatment (Figure 1C, Figure 1—figure supplement 1C). The resistance to LMB treatment implies that either GDOWN1 does not have any NES, or this treatment alone is not sufficient to achieve the nuclear accumulation of GDOWN1.

On the other hand, we employed BiFC assays to detect the interactions between GDOWN1 and its potential nuclear binding partners in live cells. An efficient interaction between the proteins of interest drives the formation of the fluorescence complementation, achieved via the covalent interactions between the two truncated parts of a fluorescent protein (Hu et al., 2005). Therefore, BiFC signals are irreversible once generated, making this assay beneficial for capturing the transient protein-protein interactions. To support the specificity of BiFC assays, a collection of the negative controls was provided, including three cytoplasmic proteins, VDAC1 (a mitochondrial protein), GALNT2 (a Golgi protein) and PDIA3 (an endoplasmic reticulum protein) (Figure 1—figure supplement 1D). Then, a series of the transcription associated proteins were tested in HeLa cells, including a Pol II subunit (RPB5) and the RPB1-CTD binding factors (RPRD1A, RPRD1B), the Mediator components (MED1, MED26), and the transcription elongation factors (SPT4 and SPT5 in DSIF complex, NELF-E in NELF complex). The results indicated that GDOWN1 interacted to all the above factors tested except for the two Mediator components, well supporting its known characteristics as a co-purified factor of Pol II and the potential functions involved in transcriptional regulation (Figure 1D). However, all these BiFC signals were shown in the cytoplasm, yet in the parallel tests, the interaction signals between NELF-E•NELF-A, SPT4•SPT5 pairs were both exclusively present in the nucleus as expected (Figure 1E). Meanwhile, our BiFC assays detected the self-interaction of GDOWN1 in the cytoplasm, suggesting that GDOWN1 may form homodimers or oligomers in cells. It was reported that the transcription regulator RYBP contained three potent and functionally independent NLS motifs (Tan et al., 2017), and when attached to GDOWN1, the BiFC signal of the 3xNLS-GDOWN1 dimers mainly remained in the cytoplasm (Figure 1E). These results support GDOWN1’s potential functions in transcriptional regulation, while the stringent cytoplasmic localization of the BiFC signals indicates that GDOWN1 is restricted from entering the nucleus under the regular cell culture conditions. Thus, our data further confirm that the nuclear entry of GDOWN1 is subjected to a tight regulation. Apparently, alleviation of this restriction will be a prerequisite for permitting GDOWN1 to execute its transcriptional regulatory functions in the nucleus.

Human GDOWN1’s cytoplasmic localization is determined by two distinct types of localization regulatory signals

Next, we constructed a series of GDOWN1 mutants to screen for localization regulatory signals by monitoring the changes in the subcellular localization by itself or together with other proteins. Based on the information of the predicted secondary structure (Figure 2—figure supplement 1A), GDOWN1 was truncated into three parts at its structurally flexible coiled regions (the N-terminus, namely m1; the middle part, m2; the C-terminus, m3), fused with a Flag-VN tag in a BiFC vector or with an intact Venus to monitor their dynamic localization in the presence or absence of LMB treatment (Figure 2A and B). Consistent to the above cell fractionation results, both the ectopically expressed full length GDOWN1, and the BiFC signal between the full length of GDOWN1 and NELF-E remained in the cytoplasm in the presence of LMB (Figure 2B, Figure 2—figure supplement 1B). m1 was partially located in the nucleus (Figure 2B, left panel; Figure 2—figure supplement 1B), and the BiFC signal of m1•NELF-E was completely nuclear localized (Figure 2B, right panel). The other two GDOWN1 fragments, m2 and m3, remained the cytoplasmic localization on their own or together with NELF-E, which was the same as the full-length protein. Interestingly, both m2 alone and m2•NELF-E BiFC signals were translocated into the nucleus upon the LMB treatment, while either m3 alone or the m3•NELF-E signal did not respond to LMB at all (Figure 2B, Figure 2—figure supplement 1B). The quantitative and statistical analyses of these confocal images confirmed that the change in the nucleocytoplasmic distribution of m2 upon the LMB treatment was significant (Figure 2B, Figure 2—figure supplement 1B). These results clearly indicate that the middle part of GDOWN1 contains an NES motif. Given that the translocation of GDOWN1 into the nucleus may not be an autonomous and efficient process, we reasoned that monitoring the nuclear accumulation of the BiFC signals between GDOWN1 and its nuclear binding partners (such as NELF-E) would have the advantage for better mining and demonstrating the nuclear translocation potential of GDOWN1. Thus, the above BiFC system was further employed for screening the putative localization regulatory motif(s). The conserved sequence of a classical NES for CRM1 recognition was known as Ψ-(x)_1-3-Ψ-(x)_1-3-Ψ-(x)_1-3-Ψ (Ψ stands for L, I, V, M, or F, x can be any amino acid) (la Cour et al., 2004; Xu et al., 2012). We generated a ‘m1 +m2’ mutant, namely m4, and found it responded to LMB (Figure 2B). We then tested a series of GDOWN1 truncation mutants on the basis of m4 to identify the functional NES motif (Figure 2—figure supplement 1C, m5-m7), and further confirmed that a putative NES motif located between amino acids 191–201 was responsible for the LMB responsiveness. Mutation of the four hydrophobic amino acids within this region completely abolished the NES activity (Figure 2B, m4 and m4*). By co-immunoprecipitation and BiFC assays, we confirmed that GDOWN1 interacted to CRM1/RAN, the core components of the protein nuclear export machinery (Figure 2C). These results prove that GDOWN1 indeed contains a classical CRM1-dependent NES motif, and meanwhile suggest that the C-terminus of GDOWN1 contains a regulatory motif responsible for the observed resistance activity of the full-length GDOWN1 to LMB treatment.

Figure 2 with 1 supplement see all

Download asset Open asset

Identification of the Nuclear Export Signal (NES) motifs in GDOWN1.

(A) A diagram of human GDOWN1 and its mutants used in the IF or BiFC-based motif screening analyses. The mutants whose names are marked in red are the ones translocated into the nucleus in response to LMB treatment. The sequences of the identified NES motifs are shown in yellow circles and the positions are labled on each side, and the core amino acids selected for mutagenesis are highlighted in red. (B) Identification of the NES motifs in GDOWN1 via IF or BiFC-based screening analyses. Left panel: HeLa cells were transiently transfected with a plasmid carrying Flag- WT or mutant GDOWN1 as indicated, and further subjected to either mock or LMB treatment, the subcellular localization was detected by IF using a Flag antibody; Right panel: HeLa cells were transiently transfected with two BiFC plasmids, YC-NELF-E and VN-WT or mutant GDOWN1 as indicated (VN—the N-terminus of Venus; YC—the C-terminus of YFP), and further subjected to either mock or LMB treatment before signal detection by a confocal microscope. The nucleocytoplasmic distribution of the fluorescent signals was quantified using ImageJ and shown at the bottom. The P values were calculated via a t-test using the built-in tools in Graphpad Prism8, n≥2, significant: P<0.05. (C) Detection of the interaction between GDOWN1 and CRM1 or RAN by IP-WB or BiFC assays. Left panel: HeLa cells stably expressed GDOWN1-Venus-Flag were employed for IP experiment using a Flag antibody or IgG and further detected by WB with the indicated antibodies; Right panel: BiFC analyses of GDOWN1•CRM1/RAN interactions. HeLa cells were transfected with YC-GDOWN1 and VN-CRM1 or RAN. The LMB treatment was carried out at a final concentration of 20 nM for 6 hr and the mock treatment was done with an equal volume of ethanol in parallel. The nuclear DNA was stained with Hoechst 33342. scale bars—30 μm.

Figure 2—source data 1 Raw data used for the statistical analyses presented in Figure 2B.: https://cdn.elifesciences.org/articles/79116/elife-79116-fig2-data1-v1.zip
Download elife-79116-fig2-data1-v1.zip
Figure 2—source data 2 Raw data of WB for Figure 2C.: https://cdn.elifesciences.org/articles/79116/elife-79116-fig2-data2-v1.zip
Download elife-79116-fig2-data2-v1.zip

When m1 and m3 parts were combined to generate m8, it was not subjected to LMB-dependent nuclear accumulation. However, when the very end of the C-terminus was chopped off, it became LMB-responsive, which led to the identification of the second NES (Figure 3A and B, m8 and m9, Figure 3—figure supplement 1A). Further mutant screening identified the second NES located between amino acids 332–340 (Figure 3A and B, Figure 3—figure supplement 1B–D, m9*, m12, m13). Taken together, we confirm that GDOWN1 is a CRM1 cargo containing two classical CRM1-responsive NES motifs.

Figure 3 with 2 supplements see all

Download asset Open asset

Identification and mechanistic analyses of the Cytoplasmic Anchoring Signal (CAS) motif in GDOWN1.

(A) A diagram of human GDOWN1 and its mutants used in the BiFC-based motif screening analyses. The mutants whose names are marked in red are the ones translocated into the nucleus in response to LMB treatment. The sequences of the identified NES or CAS motif are shown in yellow circles and the core amino acids selected for mutagenesis are highlighted in red. (B) Identification of the second NES and the CAS motif in GDOWN1 via BiFC-based screening analyses. The experiments were carried out in the same way as described in Figure 2B. (C) The enrichment of GDOWN1 at the nuclear pore region was regulated by the CAS motif. HeLa cells stably expressing the wild type GDOWN1 (WT-Venus) or the *CAS* mutant (*mCAS*-Venus) were used for detection. The nuclear membrane was approximately represented via IF using an antibody against the nuclear lamina (α-LAMIN-A/C). Confocal Images were collected and further zoomed in for 3 folds to show more details of the nuclear membranes. (D) BiFC analyses of the interactions between GDOWN1 and some subunits of NPC in HeLa cells. Upper panel: a simplified diagram of an NPC; lower panel: BiFC results between GDOWN1 (or its *CAS* mutant) and the indicated NPC components. (E) Detection of the interaction between GDOWN1 and NUP214 by IP-WB. Parental Hela cells or HeLa cells stably expressed GDOWN1(WT or *mCAS*)-Venus-Flag were employed in IP experiment using a Flag antibody and further detected by WB with indicated antibodies. The LMB treatment was carried out as previously described. The nuclear DNA was stained with Hoechst 33342. The scale bars represent 30 μm except for the ones in C represent 15 μm.

Figure 3—source data 1 Raw data of WB for Figure 3E.: https://cdn.elifesciences.org/articles/79116/elife-79116-fig3-data1-v1.zip
Download elife-79116-fig3-data1-v1.zip

The distinct responsiveness of the m8 and m9 parts of GDOWN1 to LMB treatment clearly indicated that the C-terminus of GDOWN1 contained a CRM1-independent, cytoplasmic localization signal. The key amino acids were then examined in the BiFC reporter system by screening a series of C-terminal truncation or deletion mutants (Figure 3—figure supplement 1C, D, m14-m16). It turned out that deletion of the amino acids 352–361 abolished this cytoplasmic localization regulatory activity and switched GDOWN1 into a LMB-responsive manner (Figure 3A, B, m10, Figure 3—figure supplement 1A and B; Figure 1, m10). After testing a series of combinations of point mutations, we found that mutations of the three arginines (R352, R354, and R357) were efficient to abolish the above cytoplasmic localization activity of GDOWN1 in the presence of LMB (Figure 3A, B, m11, Figure 3—figure supplement 1B–D, m11, m17, m18). Due to its potent cytoplasmic retention activity, we named this region (352–357 aa) Cytoplasmic Anchoring Signal, CAS.

To further elucidate the working mechanism of the CAS motif, we generated a pair of stable HeLa cell lines that inducibly expressed either the wild type GDOWN1 (WT-Venus) or its CAS mutant (mCAS-Venus) (Figure 3—figure supplement 2A). In these stable cell lines, the dynamic localizations of GDOWN1 were monitored and the consistent results were obtained (Figure 3—figure supplement 2B). Interestingly, the confocal microscopy images demonstrated that the wild type GDOWN1 accumulated around the nuclear membrane, as if these molecules attempted to burst through this last defense line into the nucleus, while the CAS mutant lost this ‘ring-form’ accumulation at the nuclear periphery, and became widely scattered all over the cytoplasm (Figure 3C). We hypothesized that the Venus signal enriched around the nuclear periphery might be an indicator of GDOWN1’s association with the Nuclear Pore Complex (NPC). Due to the complicated composition of NPC, we detected the interaction of GDOWN1 to several representative NPC components via BiFC assays. RAE1 and NUP50 are two NPC components typically assembled within the cytoplasmic filaments and the nuclear baskets, respectively. BiFC results indicated that the wild type GDOWN1 strongly interacted to RAE1 at the cytoplasmic side of the nuclear membrane and in the cytoplasm while this interaction was drastically weakened in the CAS mutant (Figure 3D), suggesting that the CAS motif was involved in the GDOWN1•NPC interaction. More interestingly, the BiFC signal of the wild type GDOWN1 and NUP50 was very weak and randomly distributed throughout the cytoplasm, but when the CAS motif was mutated, this interaction signal was specifically translocated into the nucleus, especially at the inner face of the nuclear membrane where NUP50 naturally located (Figure 3D). The IP results also demonstrated that the wild type GDOWN1 interacted with the cytoplasmic NPC component, NUP214, while the CAS mutant lost this interaction (Figure 3E). Overall, these results demonstrated that GDOWN1 specifically interacted to the cytoplasmic NPC components, while the CAS mutant reduced this binding affinity and simultaneously enhanced the interaction of GDOWN1 to the nuclear NPC components. Due to the irreversible nature of BiFC signal, the nuclear signal of mCAS-GDOWN1•NUP50 interaction was a clear indication of a successful capture of this GDOWN1 mutant in the nucleus, while its wild type counterpart was restricted in the cytoplasm. The above data highlight the crucial role of the CAS motif on locking GDOWN1 in the cytoplasm, presumably through anchoring GDOWN1 to the cytoplasmic components of NPC, and imply that any cellular strategy of preventing CAS function will potentially switch GDOWN1 from a stringent cytoplasmic localized protein into a nucleocytoplasmic-shuttling protein.

The NES and CAS motifs in Gdown1 are functionally interconnected and both conserved during evolution

Based on the structural prediction of GDOWN1, its CAS motif is located within the disordered region near the carboxyl-terminus, which makes it difficult to obtain reliable structural information to predict the potential CAS-NES interaction (Figure 4—figure supplement 1A). Indeed, a previous report, carrying out chemical crosslinking with mass spectrometry readout (CX-MS) to analyze Gdown1-Pol II interaction, did not provide any information about the CAS region (Jishage et al., 2018). To clarify the functional relationship between the CAS and NES motifs, we transiently expressed GDOWN1-Venus or its localization motif mutants that carried combinations of the mutated key amino acids identified above, and tested their subcellular localization and LMB responsiveness (Figure 4A). When both NES2 and CAS were mutated to allow NES1 alone to function, the mutant GDOWN1 performed as a typical CRM1-cargo, and on the other hand, the NES1 mutant maintained the same cytoplasmic localization and LMB resistance activity as the wild type GDOWN1 (Figure 4A, a-c). Thus, NES1 was a functional NES motif working independently to NES2. Similarly, we proved that NES2 was a functional NES as well, although somewhat weaker than NES1 (Figure 4A, d and b). The subcellular distribution of GDOWN1 in these images was quantified and plotted in Figure 4B. It turned out NES1 better responded to CRM1 than NES2. The NES2 mutant regularly remained in the cytoplasm while it did not resist to LMB treatment as well as the wild type, suggesting that the cytoplasmic localization activity of CAS might be partially interfered (Figure 4A, e). Double mutations in both NES motifs made GDOWN1 distributed in both the cytoplasm and the nucleus, and did not respond further to LMB, which proved that the entire GDOWN1 contained two NES motifs, and again mutations in NES2 partially abolished the CAS activity (Figure 4A, f). Compared with the wild type, the CAS mutant responded well to the LMB treatment (Figure 4A, g). Further mutating the CAS motif on top of NES1 or NES2 increased the portion of the nuclear GDOWN1 upon LMB treatment (Figure 4A, d and c; b and e), suggesting that CAS is the main motif to keep GDOWN1 in the cytoplasm via a CRM1-independent manner. Consistently, in the stable cell lines, we found that the NES2 mutant lost the perinuclear staining, which was very similar to the phenotype of the CAS mutant (Figure 4C). The above data from the intrinsic motif analyses demonstrate that each one of the two NES motifs of GDOWN1 acts as an independent CRM1-regulated motif and the function of CAS motif partially depends on the existence of NES2. Taken together, GDOWN1 is identified as a nucleocytoplasmic shuttling protein subjected to both CRM1-dependent and CRM-independent regulation and the two layers of regulation are functionally coupled.

Figure 4 with 1 supplement see all

Download asset Open asset

The working mechanisms and conservation of the binary localization regulatory apparatus in Gdown1.

(A) Dissection of the functional independence and interplay among CAS and NES motifs. The wild type GDOWN1 or the indicated *CAS* or *NES* mutants carrying point mutations were fused with Venus and ectopically expressed in HeLa cells. The cells were subjected to mock or LMB treatment the same as described in Figure 1. The schematic diagram of each mutant is shown on the top side of the corresponding representative confocal microscopy images. (B) The quantitative and statistical analyses of A. The nucleocytoplasmic distribution of the fluorescent signals for all the mutants shown in A was quantified using ImageJ and shown on the top panel. For the statistical analyses, the P values about the distribution changes in response to LMB for each mutant and about the differences between each pair of the samples were calculated via t-test using the built-in tools in Graphpad Prism8, n=4. The P values that are smaller than 0.05 (significant) were highlighted in red. (C) Confocal images demonstrating the subcellular localization of GDOWN1 in the indicated stable cell lines upon Dox induction for 2 days. (D) The function of the NES and CAS motifs was very conservative from zebrafish and *drosophila* to mammals. Upper panel: the sequence alignment of the putative NES2-CAS regions of Gdown1 proteins from the indicated species (*Homo sapiens*, NP_056347.1, *Mus musculus*, NP_848717.1, *Danio rerio*, NP_001333109.1, *Drosophila melanogaster*, NP_650794.1). ‘*’—identical in all species analyzed; ‘:’—highly conserved; ‘.’—moderately conserved. Lower panel: the dynamic subcellular localization of the wild type or *CAS* mutants of zebrafish (zGdown1) and fly (fGdown1) was detected by IF experiments. The plasmids expressing the indicated proteins were transfected into HeLa cells and the LMB treatment was carried out as previously described. The nucleocytoplasmic distribution of the fluorescent signals was quantified using ImageJ and shown on the bottom. scale bars—30 μm.

Figure 4—source data 1 Raw data used for the statistical analyses presented in Figure 4B.: https://cdn.elifesciences.org/articles/79116/elife-79116-fig4-data1-v1.zip
Download elife-79116-fig4-data1-v1.zip

Since the nucleocytoplasmic shuttling effect of Gdown1 was reported in drosophila, we evaluated the conservation of its localization regulatory mechanisms across species. The Clustal Omega analyses were carried out to compare the Gdown1 sequences from various representative species (fly, zebrafish, mouse, and human). The NES motifs are modestly conserved across these species with the key hydrophobic amino acids roughly present in fly and zebrafish Gdown1 proteins (Figure 4D, top, Figure 4—figure supplement 1B). In terms of the CAS motifs, there is no difference between mouse and human, while there is only one or two key arginines remained present in the putative CAS motifs of zebrafish and fly Gdown1 proteins, respectively. When ectopically expressed in HeLa cells, fly and zebrafish Gdown1 proteins also located stringently in the cytoplasm and resisted to LMB treatment as same as their human counterpart (Figure 4D, Figure 4—figure supplement 1C). When the conserved amino acids in the putative CAS motifs of fly and zebrafish Gdown1 proteins were mutated, these mutants became partially nucleus localized upon LMB treatment (Figure 4D, Figure 4—figure supplement 1C), indicating that fly and zebrafish Gdown1 also contained functional NES and CAS motifs. In addition, the results from BiFC analyses demonstrated that fly and zebrafish Gdown1 proteins were able to interact to human NELF-E in the cytoplasm, indicating that these orthologs in lower animals were structurally conservative to human GDOWN1 (Figure 4—figure supplement 1D). Different from the mammalian counterpart, the BiFC signals between fly or zebrafish Gdown1 and NELF-E were partially translocated into nucleus in the presence of LMB, and when CAS regions were mutated, these BiFC signals were completely present in the nucleus, indicating that the regulatory effect of CAS in fly and zebrafish Gdown1 was present but not as potent as that in human (Figure 4—figure supplement 1D). Furthermore, we found that Drosophila CAS mutant appears to accumulate more in the nucleus than the mutated zebrafish protein in the presence of LMB (Figure 4D, Figure 4—figure supplement 1C and D). We surmised that the phenomenon is because the CAS motif in fly Gdown1 is more conserved than which in zebrafish. The above results demonstrate that both the CRM1-dependent and CRM1-independent regulatory mechanisms of Gdown1 are well conserved across the various species from flies to human, while during evolution, the cytoplasmic anchoring effect of the CAS motif seems to have been enhanced to strengthen the regulation of Gdown1’s subcellular localization.

Nuclear-accumulated GDOWN1 reduces the total Pol II and the global transcription levels, and prevents cell growth

The great effort devoted by the cells to prevent Gdown1 from entering the nucleus strongly implies that it is essential to stringently control the nuclear activities of Gdown1. To help explore the outcome of Gdown1’s nuclear accumulation in somatic cells, we generated a nuclear localized, full-length human GDOWN1 mutant by mutating all the ten key amino acids identified in the three motifs of NES and CAS (highlighted in red in Figures 2A and 3A, simply named the 10M mutant). The wild type GDOWN1 or the GDOWN1(10M) was fused with Venus, and further with or without an NLS, in a commercial pTripZ vector to achieve the doxycycline (Dox)-inducible expression. The four corresponding stable HeLa cell lines were generated. The main experimental procedures are demonstrated in a diagram of Figure 5A. The confocal images after a one-day induction were shown and the 10M mutant was evenly distributed in cells and further addition of an NLS motif switched GDOWN1 into a complete nuclear localized protein (NLS-10M) (Figure 5B, i). These stable cell lines were generated by collecting the pool of cells survived from the puromycin selection, which turned out to be heterogenous that both Venus⁺ and Venus^- cells were present upon Dox induction. The benefit of using such heterogenous cell pools instead of the single clones hereby was that the co-cultured Venus^- cells (mainly expressing the very low levels of GDOWN1-Venus) could serve as the internal negative controls for a parallel comparison. When Dox was continuously supplemented in the culture medium, the fluorescence intensity in the Venus⁺ cells and their ratio to the whole population reached nearly maximum around day 3, and remained stable hereafter in the cell lines expressing the wild type GDOWN1 (Figure 5B, ii). However, these values were significantly reduced in the cell lines expressing the nuclear localized GDOWN1(10M) mutants, especially in the NLS-GDOWN1(10M)-Venus cells which almost completely lost the Venus signal on day 9 after the initial Dox addition, suggesting that the accumulation of GDOWN1 in the nucleus was unfavorable for the cell growth (Figure 5B, ii). To dissect the underlined reasons for the signal loss, we comprehensively evaluated the growth status of the cells upon expressing either the cytoplasmic or the nuclear localized GDOWN1. The cells on day 0 (Dox addition), day 3 (Dox withdrawal), and day 9 were analyzed by FACS. The results demonstrated that the cells expressing the cytoplasmic GDOWN1 only showed the basal levels of cell death (indicated by the DAPI⁺ subgroup), while the cells expressing the nuclear GDOWN1(10M) showed a drastic reduction of expression (indicated by the decreased FITC values), and simultaneously those Venus⁺ cells mainly contributed to the significant increased death rate at the later time point (Figure 5B, iii). The expression levels of the wild type GDOWN1 and its mutants in these cell lines were tested by WB, and the expression level of the NLS-GDOWN1(10M)-Venus was comparable to that of the endogenous GDOWN1 (Figure 5—figure supplement 1A). In addition, the results from the cell counting and the real-time cell analysis assays (RTCA) also demonstrated that the cells expressing the 10M mutant had severe defects in their growth and proliferation (Figure 5B, iv, Figure 5—figure supplement 1B). The above data indicate that the translocation of GDOWN1 into the nucleus inhibits cell growth, and its continuous and massive accumulation eventually causes cell death.

Figure 5 with 1 supplement see all

Download asset Open asset

Massive accumulation of GDOWN1 in the nucleus slows down cell growth and may trigger cell death.

HeLa cells stably and inducibly expressing GDOWN1- or NLS^sv40-GDOWN1-Venus-Flag, either wild type or the *10M* mutants were used for detection. ‘i’ stands for inducible and Dox was used as the inducer. (A) The experimental scheme of the comprehensive analyses of the GDOWN1 expressing cell lines (B) The schematic diagrams of the GDOWN1 variants are shown on the top panel. (i) Confocal images are presented to show the subcellular localization of the indicated cells upon Dox induction for 1 day. The nuclear DNA was stained with Hoechst 33342. scale bars—30 μm (ii) The changes of the fluorescence intensity and the ratio of Venus⁺ cells were monitored upon the induction of GDOWN1 or its mutants. Images were acquired by Cytation 5 and data were further analyzed by Gen5. (iii) The cell death and the changes in the fluorescence intensity were detected via flow cytometry. Cells were induced by Dox for 3 days to reach the maximum expression and continuously cultured for 6 days in the absence of Dox. Then, the cells were subjected to a quick DAPI staining, followed by the flow cytometry analyses. The mean values of the FITC signal (indicating the expression levels of GDOWN1-Venus proteins) and of the DAPI signals were labeled on each graph. Meanwhile, cells were counted on days 0, 3, 6, and 9, and the growth curves were plotted and shown at the bottom. (iv) Two methods were employed to generate the cell growth curves, including the direct cell counting by a cell counter at the indicated time points (the top panels，n≥3) or detection of the cell growth status by a live cell analyzer (the bottom panels). The same amounts of the cells from each cell line were re-plated in an E-plate 16 (Agilent) after a 3 day Dox induction, and subjected to the RTCA analyses in the presence or absence of Dox (0.25 μg/mL) for another 2.5 days. The real time cell index parameters were recorded and plotted by RTCA.

Figure 5—source data 1 Raw data used for the line chart presented in Figure 5B (ii).: https://cdn.elifesciences.org/articles/79116/elife-79116-fig5-data1-v1.zip
Download elife-79116-fig5-data1-v1.zip
Figure 5—source data 2 Raw data of the cell counting presented in Figure 5B (iv).: https://cdn.elifesciences.org/articles/79116/elife-79116-fig5-data2-v1.zip
Download elife-79116-fig5-data2-v1.zip
Figure 5—source data 3 Raw data used for the line chart presented in Figure 5B (iv).: https://cdn.elifesciences.org/articles/79116/elife-79116-fig5-data3-v1.zip
Download elife-79116-fig5-data3-v1.zip

It was known from the in vitro transcription assays that Gdown1 negatively regulated Pol II transcription via competing TFIIF from binding to Pol II (Cheng et al., 2012; Jishage et al., 2012). Therefore, we reasoned the cell death effects seen here might be resulted from the GDOWN1-mediated transcriptional defects. Next, EU incorporation assays were carried out in the above four cell lines expressing WT or the 10M mutants of GDOWN1 to monitor the live transcription. Based on the literature report, the EU signals captured following a 20 minutes pulse labeling were correlated to the overall transcription, with the majority of the signals (~80%) contributed by the Pol II transcription (Jackson et al., 2000). The EU signals were pseudo-colorized based on the acquired intensity. It turned out that the expression of the wild type cytoplasmic GDOWN1 did not cause any obvious change of transcription while in the cell lines expressing the nuclear GDOWN1, the EU incorporation in the Venus⁺ cells was significantly decreased compared with the Venus^- cells, indicating that GDOWN1’s abundance in the nucleus was negatively correlated with the level of global transcription (Figure 6A and Figure 6—figure supplement 1A).

Figure 6 with 1 supplement see all

Download asset Open asset

The Nuclear GDOWN1 represses the global transcription.

All the experiments shown in this figure were carried out after four days of Dox induction. (A) The massive accumulation of GDOWN1 in the nucleus caused the global transcriptional repression detected by the EU labeling assays. Confocal images were acquired, and the EU-Apollo signals were color-coded by ImageJ as indicated by the calibration bar shown at the bottom, based on the obtained signal intensity (the original images are shown in Figure 6—figure supplement 1). The averaged EU signal per cell of the Venus⁺ cells (green, V⁺) or of the Venus^- cells (gray, V^-) was shown in the graph at the lower right corner for each of the indicated cell line. The P values were calculated via a t-test using the built-in tools in Graphpad Prism8. n=4. (B) WB analyses of the total Pol II in the GDOWN1 expressing cell lines. The cell fractionation and the following WB analyses were carried out in the same way as previously described in Figure 1B. Differently, the signals of Histone H3 were served as a nuclear protein control and also for data normalization. The RPB1 level in the whole cell lysate relative to that of H3 were calculated and shown on the right. (C) The nuclear GDOWN1 reduces the levels of both the un-phosphorylated and the S5P forms of Pol II. IF experiments were carried out to detect the changes in the indicated RPB1 levels in the four indicated cell lines. Confocal images were acquired and some representative Venus⁺ cells were pointed out with yellow arrows. The nuclear DNA was stained with Hoechst 33342. scale bars—30 μm.

Figure 6—source data 1 Raw data used for the statistical analyses presented in Figure 6A.: https://cdn.elifesciences.org/articles/79116/elife-79116-fig6-data1-v1.zip
Download elife-79116-fig6-data1-v1.zip
Figure 6—source data 2 Raw data of WB for Figure 6B.: https://cdn.elifesciences.org/articles/79116/elife-79116-fig6-data2-v1.zip
Download elife-79116-fig6-data2-v1.zip
Figure 6—source data 3 Raw data used for the statistical analyses presented in Figure 6B.: https://cdn.elifesciences.org/articles/79116/elife-79116-fig6-data3-v1.zip
Download elife-79116-fig6-data3-v1.zip

Next, we performed the cell fractionation and WB assays to check the possible changes of Pol II, and the total level of Pol II was detected using an antibody specifically recognizing the N-terminus of RPB1. Since the global transcription was dramatically repressed upon the massive accumulation of GDOWN1, we reasoned that the remained proteins were determined by their half-lives. Histone proteins were experimentally proven to be very stable (Savas et al., 2012; Toyama et al., 2013), so we employed histone H3 as the internal control for data normalization. The results indicated that the total Pol II was clearly reduced upon the nuclear accumulation of GDOWN1 (Figure 6B). The 8WG16 antibody was previously confirmed to preferentially recognize the unphosphorylated RBP1 via the in vitro kinase assays (Cheng and Price, 2007). Thus, using this antibody, we found that the levels of the unphosphorylated Pol II were significantly reduced as well (Figure 6—figure supplement 1B). The LMB-treated cells were used as a positive control in which the cytoplasmic RPB1 was known to be increased upon the treatment (Forget et al., 2010; Forget et al., 2013). In both cases, the nucleocytoplasmic ratio of the total and the unphosphorylated forms of Pol II seemed to be unaffected.

To further dissect the detailed status of the affected Pol II, the IF assays were carried out using several Pol II antibodies with distinct specificities. Again, the 8WG16 antibody was applied to monitor the signals of the unphosphorylated Pol II, and the antibodies specifically recognizing the CTD-phosphorylated form at either the Ser5 positions (S5P) or the Ser2 positions (S2P) were employed to detect the Pol II subgroup engaged in either transcriptional initiation or productive elongation, respectively. Compared with the parental HeLa cells or the stable cell lines expressing the wild type GDOWN1, the signals of the unphosphorylated and the phosphorylated forms of Pol II were all dramatically reduced in the cell lines expressing the nuclear localized GDOWN1 (Figure 6C, Figure 6—figure supplement 1C). The image quantification analyses indicated that the S5P signals were decreased more severely than those of the S2P, suggesting that the nuclear GDOWN1 may inhibit transcription initiation more effectively than transcription elongation (Figure 6—figure supplement 1D). Taken together, these data demonstrate that the massive accumulation of the nuclear GDOWN1 results in the significant loss of Pol II and the global transcriptional repression.

GDOWN1 shuttles into the nucleus in response to certain stresses and helps strengthen cellular adaptability

Next, we tested various types of reagents to search for any potential exocellular stimuli capable of triggering the nuclear translocation of the endogenous GDOWN1. No obvious change of GDOWN1’s subcellular localization was observed when cells were treated with the transcriptional inhibitors DRB or Madrasin, the translational inhibitor CHX, or the inhibitors for DNA topoisomerases such as CPT or Doxorubicin (Figure 7—figure supplement 1A). Interestingly, we found the treatment of sodium arsenite (NaAsO₂) reproducibly caused the nuclear translocation of GDOWN1. NaAsO₂-induced nuclear translocation of GDOWN1 occurred in a dose-dependent manner and reversed upon the drug removal (Figure 7A). The exposure to inorganic arsenite was known to induce a global transcriptional repression (Nelson et al., 2009; Rea et al., 2003), and eventually to accumulate severe cellular toxicity, resulting in growth inhibition, DNA damage, reactive oxygen species (ROS) production, apoptosis, and/or autophagy (Tam et al., 2020). When cells were treated with 0.5 mM NaAsO₂ for 30 minutes (the prevalently used condition in the literature), nearly all cells generated stress granules (SGs) no matter GDOWN1 was competent or knocked out, indicated by the IF signals of G3BP1, a typical SG marker (Figure 7—figure supplement 1B). However, under a milder condition (0.1 mM of NaAsO₂, for 6 hr), the GDOWN1 KO cells also generated a great number of SGs, while SGs were only detected in a very small fraction in the control cells or the cells ectopically expressing the exogenous GDOWN1-Venus (Figure 7B). The EU staining results indicated that even under this milder NaAsO₂ treatment condition, the global transcription was already significantly downregulated (Figure 7C). Furthermore, the viability of the GDOWN1 KO cells was significantly less than the GDOWN1 competent counterparts (Figure 7D), indicating that loss of GDOWN1 made the cells hypersensitive to the cell toxicity induced by the low dose of NaAsO₂ treatment. Taken together, our data demonstrate that the nucleocytoplasmic localization of the native GDOWN1 is switchable in response to NaAsO₂-induced cellular stress and potentially to other types of unidentified cellular stimuli. Our data strongly suggest that GDOWN1-mediated transcriptional control contributes to the cellular sensitivity and adaptation to those stresses.

Figure 7 with 1 supplement see all

Download asset Open asset

The expression levels of GDOWN1 correlate to the cellular sensitivity to NaAsO₂ treatment.

(A) Upon NaAsO₂ treatment, a portion of the cellular GDOWN1 was subjected to a reversible translocation into the nucleus. HeLa cells were mock treated or treated with NaAsO₂ as indicated. In some samples, the cell culture medium was refreshed after the treatment to remove NaAsO₂, and the cells were further cultured for another 24 hr before harvest. The cell fractionation and the following WB analyses were carried in the same way as previously described in Figure 1B. (B) GDOWN1 affected the formation of SGs after NaAsO₂ treatment. HeLa cells with *GDOWN1 KO* (sg#7, sg#8) or the negative control (sg#NC), and the cells stably and inducibly expressing iGDOWN1-Venus-Flag (OE) were employed. Each of the indicated cell lines was subjected with NaAsO₂ treatment at 0.1 mM for 6 hr, and the SGs were detected by IF assays using an antibody against G3BP1. The nuclear DNA was stained by Hoechst 33342. scale bars—30 μm. Left: the representative confocal images; Right: the collection of the SG parameters measured and calculated by Gen5, based on the images acquired by Cytation 5. (C) The total transcription level in HeLa cells was repressed upon NaAsO₂ treatment. HeLa cells being treated with 0.1 mM NaAsO₂ or mock treated were subjected to the EU-Apollo labeling assays. (D) Knockout of GDOWN1 made the cells more sensitive to NaAsO₂ stimulation. The relative cell viability of the indicated cell lines in the presence of 0.1 mM NaAsO₂ was monitored and calculated by Cytation 5. (E) A model summarizing the working and regulatory mechanisms in GDOWN1 (described in the main text).

Figure 7—source data 1 Raw data of WB for Figure 7A.: https://cdn.elifesciences.org/articles/79116/elife-79116-fig7-data1-v1.zip
Download elife-79116-fig7-data1-v1.zip
Figure 7—source data 2 Raw data used for the table presented in Figure 7B.: https://cdn.elifesciences.org/articles/79116/elife-79116-fig7-data2-v1.zip
Download elife-79116-fig7-data2-v1.zip
Figure 7—source data 3 Raw data used for the line chart presented in Figure 7D.: https://cdn.elifesciences.org/articles/79116/elife-79116-fig7-data3-v1.zip
Download elife-79116-fig7-data3-v1.zip

Discussion

The appropriate subcellular localization of a protein determines its potential accessibility for certain cellular processes, therefore, serves as the fundamental premise for executing functions. This study is mainly focused on the exploration of human GDOWN1’s subcellular localization and the associated functional and regulatory mechanisms in the somatic cells. Our results confirmed the cytoplasmic localization of Gdown1 in the cultured mammalian somatic cell lines. To demonstrate the nucleocytoplasmic shuttling properties of GDOWN1, we treated HeLa and other types of cells with a specific inhibitor of the nuclear exportin protein CRM1, LMB, with the expectation to observe its nuclear accumulation upon the treatment. Strikingly, it turned out that for all the cell lines tested, GDOWN1 remained its cytoplasmic localization in the presence of LMB, confirmed by both biochemical fractionation and the cell imaging-based assays. Furthermore, the artificial addition of NLS motifs to GDOWN1 did not efficiently promote its nuclear translocation either. Thus, we conclude that under the conventional cell culture conditions, GDOWN1 is strictly locked in the cytoplasm rather than dynamically shuttling between the cytoplasm and the nucleus (Figure 7D), which makes GDOWN1 remarkably different from the typical nucleocytoplasmic shuttling proteins.

Our systematic dissection of the intrinsic localization regulatory motif(s) in GDOWN1 via the detailed mutant analyses let us identify a binary localization regulatory system composed of the functionally coupled NES and CAS motifs. This delicate orchestration between CAS and NES controls the nucleocytoplasmic distribution of GDOWN1, guaranteeing the appropriate input of GDOWN1 in transcriptional regulation. The facts that both NES and CAS motifs are conservative and the CAS activity seems to be strengthened from lower to higher animals further highlight the essential role of this regulatory apparatus/mechanism in controlling Gdown1’s subcellular localization and functions.

In terms of the working mechanisms of the CAS motif, at least it is partially attributed to its participation of anchoring GDOWN1 to the cytoplasmic filament subcomplex of the NPC. NPCs are composed of ~32 conserved nucleoporin proteins. Besides their central role as nucleocytoplasmic conduits, recent studies have revealed that Nups play an important role in the maintenance of cellular homeostasis through their participation in many cellular activities such as chromatin organization, transcription regulation, DNA damage repair, genome stabilization, and cell cycle control, etc. (Raices and D’Angelo, 2022). Therefore, our results support the potential involvement of NPCs in recruitment of GDOWN1 to the nuclear periphery and the resultant cytoplasmic retention, suggesting that the nuclear periphery might be the main workplace for GDOWN1 to execute its cytoplasmic functions. When CAS is fully functional, it sufficiently locks GDOWN1 in the cytoplasm so that the function of NES becomes a backup, which explains the phenomenon that GDOWN1 is insensitive to LMB treatment under this circumstance. Thus, our data suggest that removing or at least alleviating the constraint of CAS would be a prerequisite for licensing GDOWN1’s nuclear translocation and the following transcription regulatory activities. Besides the NPC-anchoring activity, other working mechanisms of the CAS-directed cytoplasmic retention remain to be explored. In addition, the controlling mechanisms for switching off the CAS activity remain unclear. Based on our findings, one reasonable hypothesis is that post translational modifications of the core arginines within CAS or possibly other amino acids nearby might facilitate this switch via causing a conformational change or affecting the interactions of GDOWN1 to its regulatory factors (illustrated in Figure 7D), which is similar to the reported cases in the literature (Ashida et al., 2022; Navarro-Lérida et al., 2021).

Our data demonstrate that mutation of the CAS motif immediately switches GDOWN1 into an LMB-sensitive nucleocytoplasmic shuttling protein, and its nuclear abundance is determined by the dynamic balance between its functionally associated binding partners (such as Pol II) and the CRM1/RAN-mediated nuclear export machinery. This partial translocation of GDOWN1 leads to tremendous changes inside of the nucleus, including the reduction of Pol II and the global transcriptional decrease. The less Pol II, the less active transcription there is, and vice versa, and this mutual feedback causes the drastic decline of the cellular transcription levels. In the previously published in vitro biochemical data, we and others found that GDOWN1 strongly inhibited both the transcriptional initiation (Jishage et al., 2012) and elongation (Cheng et al., 2012). In this study, our data indicated that the nuclear GDOWN1 tended to affect the S5P-Pol II more than the S2P subgroup, suggesting that GDOWN1 may have a stronger or a more rapid effect on the transcription initiation than elongation. It’s also worth noting that the effects of transcription inhibition and pol II defect may be attributed to the overexpression and massive nuclear accumulation of the GDOWN1(10M) mutant, although its expression level in our stable cell line was only about one-fold higher compared with the endogenous GDOWN1 (Figure 5—figure supplement 1A). Certainly, further studies are required to figure out the working mechanisms of the nuclear GDOWN1. Moreover, it was suggested that GDOWN1 was involved in the cytoplasmic assembly of Pol II as well (Ball et al., 2022; Forget et al., 2010; Forget et al., 2013). Therefore, the nuclear translocation and accumulation of GDOWN1 may also lead to the reduced efficiency of both Pol II assembly and the following nuclear import, which contributes the observed global transcriptional repression.

Recently it was reported that GDOWN1 played a role in facilitating global transcriptional shut down during mitosis and the genetic ablation of GDOWN1 exhibited mitotic defects (Ball et al., 2022), which is consistent with GDOWN1’s stringent localization in the cytoplasm during the interphase. Our discovery of GDOWN1’s nuclear translocation upon cellular stresses further expands the context in which GDOWN1 plays an essential role in the global transcriptional repression. The cells without GDOWN1 are much more sensitive to the cellular stresses, emphasizing that GDOWN1 is a crucial factor in maintaining cellular homeostasis. Further studies are needed to explore GDOWN1’s functions in the cytoplasm and to identify more cellular situations that may trigger its nuclear translocation. Overall, this work uncovered GDOWN1’s new functions and switchable localization in mammalian somatic cells and shed a light on a new connection between the global transcriptional regulation and the cellular stress adaptation.

Materials and methods

Key resources table

Reagent type (species) or resource	Designation	Source or reference	Identifiers	Additional information
Gene (human)	POLR2M	EMBL database	ENST00000299638.8	RNA polymerase II subunit M (GDOWN1)
Gene (human)	POLR2E	EMBL database	ENST00000615234.5	RNA polymerase II subunit E (RPB5)
Gene (human)	NELFA	EMBL database	ENST00000382882.9	negative elongation factor complex member A. Provided by Dr. Ruichuan Chen
Gene (human)	NELFE	EMBL database	ENST00000375429.8	negative elongation factor complex member E. Provided by Dr. Ruichuan Chen
Gene (human)	SUPT4H1	EMBL database	ENST00000225504.8	SPT4 homolog, DSIF elongation factor subunit.
Gene (human)	SUPT5H1	EMBL database	ENST00000599117.5	SPT5 homolog, DSIF elongation factor subunit
Gene (human)	MED1	EMBL database	ENST00000300651.11	mediator complex subunit 1. Provided by Dr. Ruichuan Chen
Gene (human)	MED26	EMBL database	ENST00000263390.8	mediator complex subunit 26. Provided by Dr. Ruichuan Chen
Gene (human)	RPRD1A	EMBL database	ENST00000399022.9	regulation of nuclear pre-mRNA domain containing 1 A
Gene (human)	RPRD1B	EMBL database	ENST00000373433.9	regulation of nuclear pre-mRNA domain containing 1B
Gene (human)	VDAC1	EMBL database	ENST00000265333.8	Voltage-dependent anion-selective channel protein 1
Gene (human)	GALNT2	EMBL database	ENST00000366672.5	Polypeptide N-acetyl-galactosaminyl transferase 2
Gene (human)	PDIA3	EMBL database	ENST00000300289.10	Protein disulfide-isomerase A3
Gene (fly)	Gdown1	NCBI database	NM_142537.2	Gdown1 of Drosophila melanogaster. cDNA provided from Mr. Bingtao Niu
Gene (zebrafish)	Gdown1 (Polr2m)	NCBI database	NM_001346180.1	Gdown1 of Danio rerio. cDNA provided from Dr. Yingmei Zhang
Cell line (Homo-sapiens)	HeLa	National Collection of Authenticated Cell Cultures	TCHu187	Authenticated by STR profiling; free from mycoplasma and other microorganisms
Cell line (Homo-sapiens)	HEK293T	National Collection of Authenticated Cell Cultures	GNHu17	Authenticated by STR profiling; free from mycoplasma and other microorganisms
Cell line (Homo-sapiens)	GES-1	Provided by Dr. Kesheng Li, Gansu Provincial Academic Institute for Medical Research, China		Cell line maintained in Kesheng Li lab; Authenticated by STR profiling; free from mycoplasma and other microorganisms
Cell line (Homo-sapiens)	MKN45			Cell line maintained in Kesheng Li lab; Authenticated by STR profiling; free from mycoplasma and other microorganisms
Cell line (Homo-sapiens)	SW620	National Collection of Authenticated Cell Cultures	TCHu101	Authenticated by STR profiling; free from mycoplasma and other microorganisms
Cell line (M. musculus)	NIH3T3	National Collection of Authenticated Cell Cultures	GNM 6	Authenticated by STR profiling; free from mycoplasma and other microorganisms
Cell line (M. musculus)	E14TG2a	Provided by Dr. Qintong Li, Sichuan university, China.		Originally purchased from ATCC, further adapted to be feeder-free. Authenticated by STR profiling; free from mycoplasma and other microorganisms
Recombinant DNA reagent	pBiFC (VN- or YC-) (plasmid)	Provided by Dr. Tom Kerppola, University of Michigan Medical School, USA		Plasmids used in BiFC assay
Recombinant DNA reagent	pTripZ (plasmid)	Addgene	#127696	Lentiviral vector for inducible expression in mammalian cells
Recombinant DNA reagent	pMD2.G (plasmid)	Addgene	#12259	Lentivirus packaging vector
Recombinant DNA reagent	psPAX2 (plasmid)	Addgene	#12260	Lentivirus packaging vector
Recombinant DNA reagent	pcDNA3.1(+) (plasmid)	Addgene	#78110	Gene expression vector
Recombinant DNA reagent	pX459 (plasmid)	Addgene	#118632	CRISPR-Cas9 vector
Antibody	anti-GDOWN1 (Rabbit polyclonal)	In this study, generated in Biodragon, Suzhou, China		Immunogen: human GDOWN1 (251–368 aa); WB: 1:1000 Preferably used in this study without further indication.
Antibody	anti-GDOWN1 (Sheep polyclonal)	Provided by Dr. David Price, The University of Iowa, USA		Immunogen: human GDOWN1 (full length); WB: 1:1000
Antibody	anti-α-TUBULIN (Mouse monoclonal)	Biodragon, Suzhou, China	Cat# B1052	WB: 1:10000
Antibody	anti-FBL/Fibrillarin (Rabbit monoclonal)	Abclonal, Wuhan, China	Cat# A0850	Nucleoli marker WB: 1:10000
Antibody	anti-CRM1/XPO1 (Rabbit polyclonal)	Abclonal, Wuhan, China	Cat# A0299	WB: 1:1000
Antibody	anti-RAN (Rabbit polyclonal)	Abclonal, Wuhan, China	Cat# A0976	WB: 1:1000
Antibody	anti-RPB1-total (Rabbit monoclonal)	CST, Massachusetts, USA	Cat# 14958 S Clone D8L4Y	Immunogen: a synthetic peptide corresponding to the residues surrounding N613 of the human RPB1; WB: 1:1000
Antibody	anti-RPB1- unphosphorylated (Mouse monoclonal)	Abcam, Boston, USA	Cat# AB817 Clone 8WG16	Immunogen: the purified wheat germ Pol II; IF: 1:200; WB: 1:1000
Antibody	anti-RPB1-Ser5-Phos (Mouse monoclonal)	BioLegend, California, USA	Cat# 904001 Clone CTD4H8	Immunogen: a peptide containing 10 repeats of the synthetic peptide YSPTSPS with S5 positions phosphorylated; IF: 1:1000
Antibody	anti-RPB1-Ser2-Phos (Rabbit polyclonal)	Abcam, Boston, USA	Cat# AB5095	Immunogen: a peptide of the CTD repeats of YSPTSPS from S. cerevisiae Pol II with S2 positions phosphorylated; IF: 1:200
Antibody	anti-H3 (Mouse monoclonal)	Biodragon, Suzhou, China	Cat# B1055 Clone 1G1	WB: 1:500000
Antibody	anti-G3BP1 (Rabbit monoclonal)	Abclonal, Wuhan, China	Cat# A3968	IF: 1:500
Antibody	anti-Flag (Mouse monoclonal)	Abmart, Shanghai, China	Cat# M20008 Clone 3B9	WB: 1:2000 IF: 1:300 IP: 1:500
Antibody	HRP-conjugated goat-anti-rabbit IgG (Goat polyclonal)	Biodragon, Suzhou, China	Cat# BF03008	WB: 1:10000
Antibody	HRP-conjugated goat-anti-mouse IgG (Goat polyclonal)	Biodragon, Suzhou, China	Cat# BF03001	WB: 1:10000
Antibody	HRP-conjugated goat-anti-sheep IgG (Goat polyclonal)	Biodragon, Suzhou, China	Cat# BF03025	WB: 1:10000
Antibody	Goat-anti-mouse IgG/Alexa Fluor 594 (Goat polyclonal)	Abcam, Boston, USA	Cat# AB150116	IF: 1:200
Antibody	Goat-anti-Rabbit IgG/Alexa Fluor 594 (Goat polyclonal)	Abcam, Boston, USA	Cat# AB150080	IF: 1:200
Chemical compound, drug	Leptomycin B (LMB)	Beyotime, Shanghai, China	Cat# S1726-10
Chemical compound, drug	Doxycycline (Dox)	Biogems, California, USA	Cat# 2431450
Chemical compound, drug	Puromycin	InvivoGen, USA	Cat# ant-pr-1
Chemical compound, drug	NaAsO₂	INNOCHEM, Beijing, China	Cat# A25410
Chemical compound, drug	Hoechst 33342	Solarbio Life Sciences, Beijing, China	Cat# C0031
Chemical compound, drug	Camptothecin (CPT)	Selleck, Houston, USA	Cat# S1288
Chemical compound, drug	Doxorubicin hydrochloride	Sangon Biotech, Shanghai, China	Cat# A603456
Chemical compound, drug	Cycloheximide (CHX)	MedChemExpress, New Jersey, USA	Cat# HY-12320
	Madrasin		Cat# HY-100236
	Tubercidin		Cat# HY-100126
Chemical compound, drug	5, 6-dichloro-1-β-D-ribofuranosylbenzimidazole (DRB)	Sigma, USA	Cat# D1916
Chemical compound, drug	Propidium iodide (PI)	Solarbio Life Sciences, Beijing, China	Cat# C0080
Commercial assay or kit	Cell-Light EU Apollo643 RNA Imaging Kit	RIBOBIO, Guangzhou, China	Cat# C10316-2
Software, algorithm	ImageJ	NIH		Image analysis
Software, algorithm	GraphPad Prism 8.0.2	GraphPad Software		Data analysis
Software, algorithm	Gen5	Cytation 5		Data acquiring and analysis
Software, algorithm	RTCA Software Lite	RTCA		Data acquiring and analysis
Software, algorithm	NIS-ELEMENTS C	Nikon confocal microscope		Data acquiring and analysis
Software, algorithm	BD software	BD LSRFortessa		Data acquiring
Software, algorithm	FlowJQ v10 software	https://www.bdbiosciences.com/zh-cn/products/software/flowjo-v10-software		Data analysis
Software, algorithm	ChopChop	http://chopchop.cbu.uib.no/		sgRNA design
Software, algorithm	AlphaFold Protein Structure Database	https://alphafold.ebi.ac.uk/		Structural prediction of GDOWN1
Software, algorithm	PONDR	http://www.pondr.com/		Prediction of the natural disordered regions of human GDOWN1
Software, algorithm	The CUCKOO Workgroup	http://msp.biocuckoo.org/online.php		Prediction of arginine methylation of human GDOWN1
Other	Exfect Transfection Reagent	Vazyme, Nanjing, China	Cat# T101-02	This reagent was used in the transfection experiments throughout this article.

Cell culture, transfection, and drug treatment

HeLa cells and all the other cell lines except for E14Tg2a were cultured in Dulbecco’s Modified Eagle’s Media (12800–017, Gibco, New York, USA) supplemented with 10% Newborn Calf Serum (04-102-1A, Biological Industries, Haemek, Israel) and pen/strep. The mouse embryonic stem cell line, E14Tg2a, (gift of Dr. Qintong Li in Sichuan University) was cultured in Dulbecco’s Modified Eagle’s Media supplemented with 15% Fetal Bovine Serum (900–108, Gemini Bio-products, California, USA), 1 x non-essential amino acids (11140–035, Gibco, New York, USA), 200 mM L-glutamine, 0.1 mM β-mercaptoethanol, 10³ U/mL leukemia inhibitory factor (LIF, purified in lab), and pen/strep. All the dishes or coverslips used for culturing E14Tg2a cells were pretreated with 0.5% gelatin. All cells were maintained at 37°C, 90% humidity, and 5% CO2. Plasmid transfections were carried out using the Xfect Transfection Reagent according to the manufacturer’s protocol. 0.25 μg plasmid was used for transfecting one well of cells in a 24-well cell culture dish and normally confocal microscopy images were taken at 24 hr post transfection.

For samples treated with LMB, 20 nM final concentration of LMB was added to the culture medium at 18 hr post transfection and incubated for 6 hr before data collection (or mock treated with an equal volume of ethanol). For NaA_SO₂, DRB, CHX, Madrasin, Tubercidin, CPT, and Doxorubicin treatment, the drug was added to the complete medium at the indicated final concentration and incubated with cells for the indicated timing. Cells were washes for three times with PBS to remove the drug before further operations were pursued.

Construction of plasmids and stable cell lines

Plasmid table

Names of Plasmids	Mutant type and position(s) of mutagenesis(if applicable)	Corresponding figure No.
pCDNA3.1-GOWN1-Venus		1 A, 4 A
pBiFC-Flag-cherry-GDOWN1		1A
pBiFC-Flag-GDOWN1		1A
pBiFC-Flag-NLS-GDOWN1-NLS		1A
pX459-sg#1 for GDOWN1		1-S1A, S1B
pX459-sg#7 for GDOWN1		1-S1A, S1B, 7B, 7D, 7-S1B
pX459-sg#8 for GDOWN1		1B, 1-S1A, 7B, 7D, 7-S1B
pX459-sg#9 for GDOWN1		1-S1A, S1B
pX459-sg#10 for GDOWN1		1-S1A, S1B, 7D
pBiFC-Flag-VN-GDOWN1		1D, 1E, 1-S1D, 3D
pBiFC-Flag-YC-RPB5		1D
pBiFC-Flag-YC-SPT4		1D, 1E
pBiFC-Flag-YC-RPRD1A		1D
pBiFC-Flag-YC-MED1		1D
pBiFC-Flag-YC-NELE-E		1D, 1E, 2B, 2-S1B, 3B, 3-S1D, 4-S1D
pBiFC-Flag-YC-SPT5		1D
pBiFC-Flag-YC-RPRD1B		1D
pBiFC-Flag-YC-MED26		1D
pBiFC-Flag-VN-NELF-A		1E
pBiFC-Flag-VN-SPT5		1E
pBiFC-Flag-YC-GDOWN1		1D, 1E, 2 C
pBiFC-Flag-YC-3x NLS^RYBP-GDOWN1		1E
pBiFC-Flag-VN-3x NLS^RYBP-GDOWN1		1E
pBiFC-Flag-YC-VDAC1		1-S1D
pBiFC-Flag-YC-GALNT2		1-S1D
pBiFC-Flag-YC-PDIA3		1-S1D
pBiFC-Flag-VN-GDOWN1(m1)	A truncated mutant (1–110 aa)	2B
pBiFC-Flag-VN-GDOWN1(m2)	A truncated mutant (111–250 aa)	2B
pBiFC-Flag-VN-GDOWN1(m3)	A truncated mutant (251–368 aa)	2B
pBiFC-Flag-VN-GDOWN1(m4)	A truncated mutant (1–250 aa)	2B
pBiFC-Flag-VN-GDOWN1(m4*)	A truncated mutant with the NES1 motif mutated [1–250 aa (L191, F192, I193, I199 to Gs)]	2B
pBiFC-Flag-VN-CRM1		2C
pBiFC-Flag-VN-RAN		2C
pTripZ-GDOWN1-Venus-Flag		2-S1A, 3 C, 3E, 3-S2A, 5, 5-S1
pTripZ-GDOWN1(m1)-Venus-Flag	A truncated mutant (1–110 aa)	2-S1A
pTripZ-GDOWN1(m2)-Venus-Flag	A truncated mutant (111–250 aa)	2-S1A
pTripZ-GDOWN1(m3)-Venus-Flag	A truncated mutant (251–368 aa)	2-S1A
pBiFC-Flag-VN-GDOWN1(m5)	A truncated mutant (1–145 aa)	2-S1B
pBiFC-Flag-VN-GDOWN1(m6)	A truncated mutant (1–180 aa)	2-S1B
pBiFC-Flag-VN-GDOWN1(m7)	A truncated mutant (1–215 aa)	2-S1B
pBiFC-Flag-VN-GDOWN1(m8)	A truncated mutant (1–110+251-368 aa)	3B, 3-S1A
pBiFC-Flag-VN-GDOWN1(m9)	A truncated mutant (1–110+251-340 aa)	3B, 3-S1A
pBiFC-Flag-VN-GDOWN1(m9*)	A truncated mutant with the NES2 motif mutated [1–110+251-340 aa (L332, L336, I338 to Gs)]	3B, 3-S1A
pBiFC-Flag-VN-GDOWN1(m10)	A deletion mutant (delete 352–361 aa)	3B, 3-S1A
pBiFC-Flag-VN-GDOWN1(m11)	A three-point mutant in CAS motif (R352, R354, R357 to As)	3B, 3D, 3-S1A
pTripZ-GDOWN1(m11) -Venus-Flag	A three-point mutant in CAS motif (R352, R354, R357 to As)	3 C, 3E, 3-S2A
pBiFC-Flag-YC-RAE1		3D
pBiFC-Flag-YC-NUP50		3D
pBiFC-Flag-VN-GDOWN1(m12)	A truncated mutant (1–110+251-280 aa)	3-S1D
pBiFC-Flag-VN-GDOWN1(m13)	A truncated mutant (1–110+251-310 aa)	3-S1D
pBiFC-Flag-VN-GDOWN1(m14)	A deletion mutant (delete 366–368 aa)	3-S1D
pBiFC-Flag-VN-GDOWN1(m15)	A deletion mutant (delete 361–368 aa)	3-S1D
pBiFC-Flag-VN-GDOWN1(m16)	A deletion mutant (delete 356–368 aa)	3-S1D
pBiFC-Flag-VN-GDOWN1(m17)	A three-point mutant (E355, V356, D358 to As)	3-S1D
pBiFC-Flag-VN-GDOWN1(m18)	A four-point mutant (D358, E359, D360 D361 to As)	3-S1D
pCDNA3.1-GDOWN1(mNES2 +CAS)-Venus	A six-point mutant in NES2 and CAS motifs (L332, L336, I338 to Gs and R352, R354, R357 to As)	4A
pCDNA3.1-GDOWN1(mNES1)-Venus	A four-point mutant in NES1 motif (L191, F192, I193, I199 to Gs)	4A
pCDNA3.1-GDOWN1(mNES1 +CAS)-Venus	A seven-point mutant in NES1 and CAS motifs (L191, F192, I193, I199 to Gs and R352, R354, R357 to As)	4A
pCDNA3.1-GDOWN1(mNES2)-Venus	A three-point mutant in NES2 motif (L332, L336, I338 to Gs)	4A
pCDNA3.1-GDOWN1(mNES1 +2)-Venus	A seven-point mutant in NES1 and NES2 motifs (L191, F192, I193, I199, L332, L336, I338 to Gs)	4A
pCDNA3.1-GDOWN1(mCAS)-Venus	A three-point mutant in CAS motif (R352, R354, R357 to As)	4A
pTripZ-GDOWN1(mNES2)-Venus-Flag	A three-point mutant in NES2 motif (L332, L336, I338 to Gs)	4C
pBiFC-Flag-VN-fGdown1	The wild type Gdown1 of Drosophila melanogaster	4D, 4-S1D
pBiFC-Flag-VN-fGdown1(R325, 328 A)	A double-point mutant in CAS motif (R325, R328A)	4D, 4-S1D
pBiFC-Flag-VN-zGdown1	The wild type Gdown1 of Danio rerio	4D, 4-S1D
pBiFC-Flag-VN-zGdown1(R328A, H331A)	A double-point mutant in CAS motif (R328A, H331A)	4D, 4-S1D
pTripZ-GDOWN1(10M)-Venus-Flag	A ten-point mutant in NES1, NES2 and CAS motifs (L191, F192, I193, I199, L332, L336, I338 to Gs and R352, R354, R357 to As)	5, 5-S1
pTripZ-NLS^SV40-GDOWN1-Venus-Flag		5, 5-S1
pTripZ-NLS^SV40-GDOWN1(10M)-Venus-Flag	A ten-point mutant in NES1, NES2 and CAS motifs (L191, F192, I193, I199, L332, L336, I338 to Gs and R352, R354, R357 to As)	5, 5-S1

The pBiFC-Flag-VN (aa 1–172 of Venus) or pBiFC-Flag-YC (aa 173–238 of YFP) plasmids (gifts from Dr. Tom Kerppola, University of Michigan) were used as the parental vectors for generating all the indicated BiFC plasmids. The coding sequences of human MED1, MED26, and SPT5 genes were PCR amplified from the plasmids (gifts from Dr. Ruichuan Chen, Xiamen University, Lu et al., 2016). Human GDOWN1 (also namely POLR2M，NM_015532.5) and other genes were all amplified by RT-PCR used cDNA samples generated from HeLa cells as the templates. The Gdown1 genes in Danio rerio (NM_001346180.1) and Drosophila melanogaster (NM_142537.2) were cloned from the cDNA samples generated directly from animal lysates. Total RNA was extracted by the MolPure Cell/Tissue Total RNA Kit (19221ES50, YEASEN, Shanghai, China) and the cDNA was synthesized using the 1st Strand cDNA Synthesis SuperMix (11141ES60, YEASEN, Shanghai, China). The purified RT-PCR products were double digested by BamHI (R3136S, NEB, Massachusetts, USA) and XbaI (R0145S, NEB, Massachusetts, USA) and then ligated into pBiFC-Flag-VN or -YC vectors by T4 DNA ligase, or when these two restriction enzymes had cut sites within the cDNA sequences, the PCR products were assembled into pBiFC-Flag-VN or -YC vectors via homologous recombination using the ClonExpressII One Step Cloning Kit (C112-02, Vazyme, Nanjing, China). The two NLS motifs in a plasmid named Flag-NLS-GDOWN1-NLS in Figure 1A were adopted from the pX459 plasmid originally constructed from Dr. Feng Zhang’s lab in MIT (Ran et al., 2013). The three NLS motifs in the VN/YC-3xNLS-GDOWN1 plasmids shown in Figure 1E were cloned from the CDS sequences of human RYBP gene corresponding to the amino acids 1–94 (Tan et al., 2017). The truncated fragments of human GDOWN1 were amplified using the full-length CDS as a template and further used to construct the pBiFC-based GDOWN1 mutants. Point mutations were introduced by designing the long PCR primers containing the designated mutated sequences and then amplified the fragments in the regular PCR or bridging PCR reactions as needed. The information on the amino acids for mutagenesis was shown in Figures 2A and 3A. The above pBiFC-based plasmids series were applied in both BiFC assays (directly monitoring the BiFC signals) or in IF assays (detection via a Flag antibody) as indicated in the figure legends. For generating the GDOWN1-Venus plasmid series, pcDNA3.1(+) was used as a parental vector. The full-length, wild type GDOWN1 was amplified from the above pBiFC vector and ligated into the pcDNA3.1-Venus plasmid (previously constructed in lab). The NES and/or CAS mutant fragments were PCR amplified from the above pBiFC plasmids expressing the corresponding mutant GDOWN1 and further amplified by the bridging PCR and then assembled into the pcDNA3.1-Venus plasmid.

GDOWN1 KO HeLa cells were generated via the CRISPR-Cas9 technology. The sgRNAs were selected according to the information provided by ChopChop. The targeting sequences of sgRNAs are listed in the table down below. The pX459-sgGDOWN1-#1/#7/#8/#9/#10 plasmids were constructed and transfected into HeLa cells and the cells were selected with 0.5 μg/mL puromycin starting from 48 hr post transfection. After 5 days of selection, the survived cells were pooled and further verified by sequencing and WB. For cells transfected with pX459-sgGDOWN1-#1, the pooled cells after puromycin selection were re-plated in a p100 cell culture dish at a density of 2000 cells per dish. After 15 days of culture, the single colonies were picked to a 96-well plate and further expanded. Genomic DNA was isolated and PCR amplified using the primers shown in the table for the verification. The PCR products were gel purified and sent for sequencing (Tsingke Biotechnology, Beijing, China).

sgRNA#	Targeting sequences of sgRNA	PCR verification primers (Forward; Reverse)
1	GCGGGAAATGTTGAAGCGCC	GCATGAATGCTCACACAAGG; CGAATGTGACTGAGTCAAAGT
7	ACGAGTAAGCTGGGGTCCCG	GCATGAATGCTCACACAAGG; CGAATGTGACTGAGTCAAAGT
8	GTTACAGAGGATCACCATTG	CAGAATTCTGACCCGATAC; CTTCCCACCTCAGCCTCCTGAG
9	AACTTGACAGGCCTTTCCAG	CAGAATTCTGACCCGATAC; CTTCCCACCTCAGCCTCCTGAG
10	TTGATGACATCACAGCAGCT	GGAGGAGAATTAATTGCTAAG; GCAGTTCTAGCAACTTTGTG

For generating the HeLa cell lines stably expressing GDOWN1, pTripZ, the lentiviral expression vector, was used as a parental vector and the fragment, Venus-Flag, was initially inserted into the pTripZ empty vector to replace the original shRNA expression cassette. The wild type or mutant GDOWN1 fragments were PCR amplified from the constructed pcDNA3.1-based plasmids (for the ones with the NLS addition, the sequences of NLS^SV40 were attached to the N-terminus of the corresponding primers) and were further inserted in between the TRE-CMV promoter and the Venus gene to obtain iGDOWN1-Venus-Flag plasmids (‘i’ stands for inducible). For the viral packaging procedure, HEK293T cells cultured in a 6 well plate were transfected with 1 μg pMD2G, 2 μg pAX2, and 3 μg pTripZ-GDOWN1-Venus (WT or mutant) and medium was refreshed at 6 hr post transfection. The viral stock was harvested after 72 hr and further infected HeLa cells for 12 hr. The cells were recovered for one day and further subjected for the puromycin selection (0.5 μg/mL) for 14 days. The surviving cells were pooled and the inducible expression of the GDOWN1-Venus-Flag proteins was verified by WB with a Flag antibody after adding 2.5 μg/mL of Dox for 12 hr.

BiFC assays

For BiFC assays, HeLa cells were grown on coverslips in 24-well cell culture dishes and 0.25 μg of each pBiFC plasmid (VN- or YC-) was used for co-transfection per well. At 24 hr post transfection, the cells were fixed with 4% formaldehyde for 20 minutes at the room temperature, washed with PBS, stained with 1 μg/mL of Hoechst 33342 for 10 minutes, washed with PBS, and finally visualized in PBS.

Immunofluorescence and the data analyses

For immunofluorescence assays, the cells were grown on coverslips, fixed with 4% formaldehyde for 20 minutes at the room temperature, washed three times with PBS, dehydrated with 90% methanol at –20°C for 30 minutes, permeabilized with 0.5% Triton X-100 at the room temperature for 20 minutes, washed three times with PBS, incubated with 5% BSA for 1 hr at the room temperature. The cells were then incubated with a Flag antibody (1:200 diluted in TBST) for 12 hr at 4°C. After being washed for three times with TBST, the cells were subjected for a secondary antibody incubation for 1 hr at the room temperature. The cells were further stained with 1 μg/mL of Hoechst 33342 for 10 minutes, washed with PBS, and visualized in PBS. For data analysis in Figure 6—figure supplement 1D, the confocal images were further processed by Gen5. The cellular analysis function of Gen5 was used to calculate the Integral intensity of the Venus signal per cell and the corresponding area and these values were used to define the Venus⁺ cells from the Venus^- cells. Then the mean signals of the indicated RPB1 form per nucleus (the unphosphorylated, S2P, or S5P) for either the Venus⁺ or the Venus^- cell population were calculated separately and further processed using Graphpad Prism8.

For the statistical analyses of the stress granules, the cells were grown on coverslips in 24-well cell culture dishes (the inducible cell lines were pre-induced with Dox for one day) and then either mock treated or treated with 0.1 mM of NaAsO₂ for 6 hr. Then the cells were washed with PBS for three times and incubated with the G3BP1 antibody in the IF assays as described above. For a large-scale data quantification, the images were also acquired using Cytation 5 (Agilent) using the same set of coverslips and at least four regions of interest (ROIs) were randomly selected, and all of the images were further analyzed using the Gen5 software. The cell count (indicated by the Hoechst 33342 signals), G3BP1 spot number per cell, and the mean fluorescence intensity of each G3BP1 spot (within each ROI) were detected and calculated using the built-in tools in the Gen5 software (automatic cell count, spots count, and the subpopulation analysis). The cell containing more than 1 spot was defined as an SG⁺ cell.

% o f t h e S G^{+} c e l l s = \frac{T h e n u m b e r o f t h e S G^{+} c e l l s}{T h e n u m b e r o f t h e t o t a l c e l l s} \times 100 %

Co-immunoprecipitation

The cells stably expressed GDOWN1-Venus-Flag in a p100 dish were lysed with 500 μL of the lysis buffer (20 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.5 mM EDTA, 1% NP40, 1% Triton X-100, 50 mM NaF, 1 mM Na₃VO₄ [activated], 0.1 mM PMSF, Protease Inhibitor Cocktail [B14012, Bimake, Shanghai, China]) and incubated for 30 minutes at 4°C on a rotator. The lysate was incubated with 25 μL of the anti-Flag magnetic beads (B26101, Bimake, Shanghai, China) for 12 hr at 4°C on a rotator. The beads were washed five times with lysis buffer, and then resuspended with a 5 x loading dye (250 mM Tris-HCl, pH 6.8, 10% SDS, 50% glycerinum, 5% β-mercaptoethanol, 0.1% bromophenol blue) and further pursued for SDS-PAGE analyses.

Confocal microscopy and the nucleocytoplasmic distribution analyses

The confocal images were obtained using a 100 x oil objective (N.A. 1.45) on a Nikon A1R+ Ti2-E laser scanning microscope, equipped with a GaAsP Multi Detector Unit. Images were acquired from at least four randomly selected fields using the NIS-ELEMENTS C software. For data quantification of GDOWN1’s nucleocytoplasmic distribution, each set of the obtained images was input into the ImageJ software. The intensity of the Venus signal detected from each cell was obtained and considered as the total signal. The signal of Hoechst 33342 was utilized to define the region of each nucleus, and the Venus signal within the border of the nucleus was recorded as the nuclear signal in the corresponding cell. The cytoplasmic signal can be calculated by subtracting the nuclear signal from the total and the nucleocytoplasmic distribution can be calculated and plotted. For the statistical analyses, the calculated values from each field were averaged and the standard errors (SE) among the four fields were calculated and plotted. The P values were calculated via a t-test using the built-in tools in Graphpad Prism8.

EU-Apollo assay and the data analyses

For the EU-Apollo assays, either the parental HeLa cells or the derived stable cell lines were grown on coverslips in 48-well cell culture dishes, and 0.25 μg/mL of Dox was used for the protein induction. 250 μM of EU was added into the culture medium 20 minutes before cell harvest. All the procedures were carried out following the manufacturer’s instructions of the Cell-Light EU Apollo 643 RNA Imaging Kit. The confocal images acquired from at least four fields per treatment were used for data analyzed. Based on the fluorescence intensity of the Venus signals, the cells were separated into either Venus⁺ or Venus^- group. The fluorescence intensity of the EU-Apollo signal was measured cell by cell with ImageJ, and the averaged EU-Apollo signal for each group of the cells per field was calculated and plotted in bar graphs. For the statistical analyses, the averaged values and SEs of the EU-Apollo signals among the four fields were calculated and further processed to obtain the p values via a t-test using the built-in tools in Graphpad Prism8.

Cell fractionation and the quantitative analyses of WB data

About 2x10^6 cells were used for each cell fractionation assay. Freshly harvested cell pellet was resuspended with 100 μL of the cytoplasmic extraction buffer (20 mM Hepes, 1 mM EDTA, 10 mM KCl, 2 mM MgCl₂, 0.1% NP-40, 1 mM DTT, 0.1 mM PMSF, Protease Inhibitor Cocktail), and incubated at 4°C for 30 mintes. The cell lysate was centrifuged at 1500 rpm for 3 minutes at 4°C, and the supernatant was saved as the cytoplasmic fraction. The remained nuclei pellet was washed for three times and further resuspended with 100 μL of the cytoplasmic extraction buffer. The resuspended nuclei samples were used as the nuclear fraction (containing both the soluble nucleoplasm and the insoluble chromatin). 5 x protein loading dye was added into the above cytoplasmic (C) and nuclear (N) fractions to generate 1 x samples for SDS-PAGE and WB analyses.

For data quantification, ImageJ was employed to acquire the IntDen value (integral optical density) of each band in the obtained WB images.

R e l a t i v e g r a y v a l u e o f t h e R P B 1 (n o r m a l i z e d w i t h H 3) = \frac{I n t D e n (R P B 1 i n t h e w h o l e c e l l l y s a t e)}{I n t D e n (H 3 i n t h e w h o l e c e l l l y s a t e)}

Live cell analyses and data analyses

For live cell analyses, the cells were plated in a 48-well cell culture dish 24 hr before the treatment. For the results shown in Figure 7D, the cells were incubated with the complete medium supplemented with 0.2 mM NaAsO₂, 0.1 μg/mL Hoechst 33342, and 1 μg/mL PI, then immediately analyzed using Cytation 5. Four ROIs from each well were randomly selected and images were acquired using Gen5 software. The count of total cells (using the Hoechst 33342 signal as an indicator) and the dead cells (using the PI signal as an indicator) in each ROI were calculated using the built-in tools (automatic cell count and subpopulation analysis) in the Gen5 software. The calculated values were further processed using Graphpad Prism8.

R e l a t i v e c e l l v i a b i l i t y (%) = \frac{T h e n u m b e r o f t h e t o t a l c e l l s - T h e n u m b e r o f t h e d e a d c e l l s}{T h e n u m b e r o f t h e t o t a l c e l l s} \times 100 %

For the results shown in Figure 5B (ii), cells were grown in 6-well cell culture dishes and induced by 0.25 μg/mL of Dox for 0, 3, 6, 9 days, respectively. Upon harvest, cells were fixed with 4% formaldehyde for 20 minutes at the room temperature, washed with PBS, and visualized in PBS. Images of at least four randomly selected ROIs from each well were acquired using the Gen5 software. For data analyses, the number of the total cells and the Venus⁺ cells, and the fluorescence intensity of the Venus⁺ cells were calculated using the built-in tools in Gen5. The obtained data were further processed in Graphpad Prism8 to export figures.

R a t i o o f t h e V e n u s^{+} c e l l s = \frac{T h e n u m b e r o f t h e V e n u s^{+} c e l l s}{T h e n u m b e r o f t h e t o t a l c e l l s}

For live cell analyses, using RTCA in Figure 5B (iv), the cells were seeded in an E-plate 16 PET (Agilent), and cultured in the xCELLigence RTCA S16 (Agilent) inside of a cell culture incubator. The cell index values and the slope of their changes were acquired by the RTCA Software Lite along with the cell growth. The obtained values were further processed in Graphpad Prism8 for data export.

Cell counting and flow cytometry analyses (FACS)

For the results shown in Figure 5B (iv), 6×10^4 cells of each cell line were plated on day 0 and cultured with the complete medium. Two experiments were performed at the same time, with the cells either being seeded in 12-well plates for cell counting, or in 6 well plates for FACS. For generating the growth curve, the cells were incubated with 0.25 μg/mL of Dox for 9 days, and the equal number (6×10^4) of cells from each cell line were re-plated into a fresh cell culture dish every three days (on day 3 and 6). The number of cells was counted on days 3, 6, and 9 using a cell counter (Bodboge, Guangzhou, China), and the averaged value was calculated based on at least three independent readings. The finalized cell counts were calculated based on the averaged values and the distinct dilution coefficient upon each passage.

For FACS analyses, the cells were harvested on days 0, 3, and 9 after Dox induction. To gather the dead cells, the culture medium was centrifuged at 1000 ×g for 5 minutes and the cells at the bottom of the tubes were collected. Then the adherent cells in the plates were trypsinized (0.25% trypsin) and collected via a centrifugation at 800 xg for 5 minutes. Both the attached cells and the dead cells originally from the same sample were combined and resuspend with 500 μL of PBS, and further incubated with DAPI (at a final concentration of 5 μg/mL) at dark for 15 minutes. Then, the cells were immediately handled in the Flow Cytometer (LSRFortessa, BD, New Jersey, USA) for the fluorescence detection. For each sample, a minimum of 10,000 cells were analyzed with the FlowJo 7.6 software. A control sample generated by mixing 2/3 of live cells with 1/3 of the formaldehyde fixed dead cells (indicated as DAPI⁺ cells) was utilized to facilitate the accurate setting of the gating parameters. The data were exported using the FlowJo 7.6 software.

Data availability

All data generated or analyzed during this study are included in the manuscript and supporting file; Source Data files have been provided for Figures whenever is necessary.

References

1. Ashida S
2. Morita R
3. Shigeta Y
4. Harada R
(2022) Phosphorylation in the accessory domain of yeast histone chaperone protein 1 exposes the nuclear export signal sequence
Proteins 90:317–321.
https://doi.org/10.1002/prot.26240
- PubMed
- Google Scholar
1. Ball CB
2. Parida M
3. Santana JF
4. Spector BM
5. Suarez GA
6. Price DH
(2022) Nuclear export restricts Gdown1 to a mitotic function
Nucleic Acids Research 50:1908–1926.
https://doi.org/10.1093/nar/gkac015
- PubMed
- Google Scholar
1. Cheng B.
2. Price DH
(2007) Properties of RNA polymerase II elongation complexes before and after the P-TEFb-mediated transition into productive elongation
The Journal of Biological Chemistry 282:21901–21912.
https://doi.org/10.1074/jbc.M702936200
- PubMed
- Google Scholar
1. Cheng Bo
2. Li T
3. Rahl PB
4. Adamson TE
5. Loudas NB
6. Guo J
7. Varzavand K
8. Cooper JJ
9. Hu X
10. Gnatt A
11. Young RA
12. Price DH
(2012) Functional association of Gdown1 with RNA polymerase II poised on human genes
Molecular Cell 45:38–50.
https://doi.org/10.1016/j.molcel.2011.10.022
- PubMed
- Google Scholar
1. Core L
2. Adelman K
(2019) Promoter-Proximal pausing of RNA polymerase II: a nexus of gene regulation
Genes & Development 33:960–982.
https://doi.org/10.1101/gad.325142.119
- PubMed
- Google Scholar
1. Cramer P
(2019) Organization and regulation of gene transcription
Nature 573:45–54.
https://doi.org/10.1038/s41586-019-1517-4
- PubMed
- Google Scholar
1. Fischer V
2. Schumacher K
3. Tora L
4. Devys D
(2019) Global role for coactivator complexes in RNA polymerase II transcription
Transcription 10:29–36.
https://doi.org/10.1080/21541264.2018.1521214
- PubMed
- Google Scholar
(2010) The protein interaction network of the human transcription machinery reveals a role for the conserved GTPase RPAP4/GPN1 and microtubule assembly in nuclear import and biogenesis of RNA polymerase II
Molecular & Cellular Proteomics 9:2827–2839.
https://doi.org/10.1074/mcp.M110.003616
- PubMed
- Google Scholar
(2013) Nuclear import of RNA polymerase II is coupled with nucleocytoplasmic shuttling of the RNA polymerase II-associated protein 2
Nucleic Acids Research 41:6881–6891.
https://doi.org/10.1093/nar/gkt455
- PubMed
- Google Scholar
1. Haberle V
2. Stark A
(2018) Eukaryotic core promoters and the functional basis of transcription initiation
Nature Reviews. Molecular Cell Biology 19:621–637.
https://doi.org/10.1038/s41580-018-0028-8
- PubMed
- Google Scholar
1. Hsin JP
2. Manley JL
(2012) The RNA polymerase II CTD coordinates transcription and RNA processing
Genes & Development 26:2119–2137.
https://doi.org/10.1101/gad.200303.112
- PubMed
- Google Scholar
(2005) Visualization of protein interactions in living cells using bimolecular fluorescence complementation (bifc) analysis
Current Protocols in Protein Science 19:ps1910s41.
https://doi.org/10.1002/0471140864.ps1910s41
- PubMed
- Google Scholar
1. Hu X
2. Malik S
3. Negroiu CC
4. Hubbard K
5. Velalar CN
6. Hampton B
7. Grosu D
8. Catalano J
9. Roeder RG
10. Gnatt A
(2006) A mediator-responsive form of metazoan RNA polymerase II
PNAS 103:9506–9511.
https://doi.org/10.1073/pnas.0603702103
- PubMed
- Google Scholar
(2000) The balance sheet for transcription: an analysis of nuclear RNA metabolism in mammalian cells
FASEB Journal 14:242–254.
https://doi.org/10.1096/fasebj.14.2.242
- PubMed
- Google Scholar
(2016) The RNA polymerase II CTD: the increasing complexity of a low-complexity protein domain
Journal of Molecular Biology 428:2607–2622.
https://doi.org/10.1016/j.jmb.2016.02.006
- PubMed
- Google Scholar
1. Jishage M
2. Malik S
3. Wagner U
4. Uberheide B
5. Ishihama Y
6. Hu X
7. Chait BT
8. Gnatt A
9. Ren B
10. Roeder RG
(2012) Transcriptional regulation by Pol II (G) involving mediator and competitive interactions of Gdown1 and TFIIF with pol II
Molecular Cell 45:51–63.
https://doi.org/10.1016/j.molcel.2011.12.014
- PubMed
- Google Scholar
1. Jishage M
2. Yu X
3. Shi Y
4. Ganesan SJ
5. Chen WY
6. Sali A
7. Chait BT
8. Asturias FJ
9. Roeder RG
(2018) Architecture of Pol II (G) and molecular mechanism of transcription regulation by Gdown1
Nature Structural & Molecular Biology 25:859–867.
https://doi.org/10.1038/s41594-018-0118-5
- PubMed
- Google Scholar
1. Jishage M
2. Ito K
3. Chu CS
4. Wang X
5. Yamaji M
6. Roeder RG
(2020) Transcriptional down-regulation of metabolic genes by Gdown1 ablation induces quiescent cell re-entry into the cell cycle
Genes & Development 34:767–784.
https://doi.org/10.1101/gad.337683.120
- PubMed
- Google Scholar
1. Jishage M
2. Roeder RG
(2020) Regulation of hepatocyte cell cycle re-entry by RNA polymerase II-associated gdown1
Cell Cycle 19:3222–3230.
https://doi.org/10.1080/15384101.2020.1843776
- PubMed
- Google Scholar
1. Jonkers I
2. Lis JT
(2015) Getting up to speed with transcription elongation by RNA polymerase II
Nature Reviews. Molecular Cell Biology 16:167–177.
https://doi.org/10.1038/nrm3953
- PubMed
- Google Scholar
(2020) Interplay of mRNA capping and transcription machineries
Bioscience Reports 40:BSR20192825.
https://doi.org/10.1042/BSR20192825
- PubMed
- Google Scholar
1. Kilchert C
2. Vasiljeva L
(2013) Mrna quality control goes transcriptional
Biochemical Society Transactions 41:1666–1672.
https://doi.org/10.1042/BST20130202
- PubMed
- Google Scholar
1. Kudo N
2. Wolff B
3. Sekimoto T
4. Schreiner EP
5. Yoneda Y
6. Yanagida M
7. Horinouchi S
8. Yoshida M
(1998) Leptomycin B inhibition of signal-mediated nuclear export by direct binding to CRM1
Experimental Cell Research 242:540–547.
https://doi.org/10.1006/excr.1998.4136
- PubMed
- Google Scholar
1. Kudo N
2. Matsumori N
3. Taoka H
4. Fujiwara D
5. Schreiner EP
6. Wolff B
7. Yoshida M
8. Horinouchi S
(1999) Leptomycin B inactivates crm1/exportin 1 by covalent modification at a cysteine residue in the central conserved region
PNAS 96:9112–9117.
https://doi.org/10.1073/pnas.96.16.9112
- PubMed
- Google Scholar
1. la Cour T
2. Kiemer L
3. Mølgaard A
4. Gupta R
5. Skriver K
6. Brunak S
(2004) Analysis and prediction of leucine-rich nuclear export signals
Protein Engineering, Design & Selection 17:527–536.
https://doi.org/10.1093/protein/gzh062
- PubMed
- Google Scholar
1. Lu X
2. Zhu X
3. Li Y
4. Liu M
5. Yu B
6. Wang Y
7. Rao M
8. Yang H
9. Zhou K
10. Wang Y
11. Chen Y
12. Chen M
13. Zhuang S
14. Chen L-F
15. Liu R
16. Chen R
(2016) Multiple P-tefbs cooperatively regulate the release of promoter-proximally paused RNA polymerase II
Nucleic Acids Research 44:6853–6867.
https://doi.org/10.1093/nar/gkw571
- PubMed
- Google Scholar
(2018) The RNA polymerase II factor RPAP1 is critical for mediator-driven transcription and cell identity
Cell Reports 22:396–410.
https://doi.org/10.1016/j.celrep.2017.12.062
- PubMed
- Google Scholar
(2016) Transcription elongation control by the 7SK snrnp complex: releasing the pause
Cell Cycle 15:2115–2123.
https://doi.org/10.1080/15384101.2016.1181241
- PubMed
- Google Scholar
(2021) Rna polymerase II speed: a key player in controlling and adapting transcriptome composition
The EMBO Journal 40:e105740.
https://doi.org/10.15252/embj.2020105740
- PubMed
- Google Scholar
(2021) Post-Translational modification and subcellular compartmentalization: emerging concepts on the regulation and physiopathological relevance of RhoGTPases
Cells 10:1990.
https://doi.org/10.3390/cells10081990
- PubMed
- Google Scholar
1. Nelson GM
2. Ahlborn GJ
3. Allen JW
4. Ren H
5. Corton JC
6. Waalkes MP
7. Kitchin KT
8. Diwan BA
9. Knapp G
10. Delker DA
(2009) Transcriptional changes associated with reduced spontaneous liver tumor incidence in mice chronically exposed to high dose arsenic
Toxicology 266:6–15.
https://doi.org/10.1016/j.tox.2009.10.004
- PubMed
- Google Scholar
1. Neugebauer KM
(2019) Nascent rna and the coordination of splicing with transcription
Cold Spring Harbor Perspectives in Biology 11:ea032227.
https://doi.org/10.1101/cshperspect.a032227
- PubMed
- Google Scholar
(2021) Causes and consequences of RNA polymerase II stalling during transcript elongation
Nature Reviews. Molecular Cell Biology 22:3–21.
https://doi.org/10.1038/s41580-020-00308-8
- PubMed
- Google Scholar
1. Osman S
2. Cramer P
(2020) Structural biology of RNA polymerase II transcription: 20 years on
Annual Review of Cell and Developmental Biology 36:1–34.
https://doi.org/10.1146/annurev-cellbio-042020-021954
- PubMed
- Google Scholar
1. Proudfoot NJ
(2016) Transcriptional termination in mammals: stopping the RNA polymerase II juggernaut
Science 352:aad9926.
https://doi.org/10.1126/science.aad9926
- PubMed
- Google Scholar
1. Raices M
2. D’Angelo MA
(2022) Structure, maintenance, and regulation of nuclear pore complexes: the gatekeepers of the eukaryotic genome
Cold Spring Harbor Perspectives in Biology 14:a040691.
https://doi.org/10.1101/cshperspect.a040691
- PubMed
- Google Scholar
1. Ran FA
2. Hsu PD
3. Wright J
4. Agarwala V
5. Scott DA
6. Zhang F
(2013) Genome engineering using the CRISPR-Cas9 system
Nature Protocols 8:2281–2308.
https://doi.org/10.1038/nprot.2013.143
- PubMed
- Google Scholar
1. Rea MA
2. Gregg JP
3. Qin Q
4. Phillips MA
5. Rice RH
(2003) Global alteration of gene expression in human keratinocytes by inorganic arsenic
Carcinogenesis 24:747–756.
https://doi.org/10.1093/carcin/bgg010
- PubMed
- Google Scholar
1. Sansó M
2. Fisher RP
(2013) Pause, play, repeat: CDKs push RNAP II’s buttons
Transcription 4:146–152.
https://doi.org/10.4161/trns.25146
- PubMed
- Google Scholar
1. Savas JN
2. Toyama BH
3. Xu T
4. Yates JR
5. Hetzer MW
(2012) Extremely long-lived nuclear pore proteins in the rat brain
Science 335:e1217421.
https://doi.org/10.1126/science.1217421
- PubMed
- Google Scholar
1. Schier AC
2. Taatjes DJ
(2020) Structure and mechanism of the RNA polymerase II transcription machinery
Genes & Development 34:465–488.
https://doi.org/10.1101/gad.335679.119
- PubMed
- Google Scholar
1. Sun Y
2. Hamilton K
3. Tong L
(2020) Recent molecular insights into canonical pre-mRNA 3’-end processing
Transcription 11:83–96.
https://doi.org/10.1080/21541264.2020.1777047
- PubMed
- Google Scholar
1. Tam LM
2. Price NE
3. Wang Y
(2020) Molecular mechanisms of arsenic-induced disruption of DNA repair
Chemical Research in Toxicology 33:709–726.
https://doi.org/10.1021/acs.chemrestox.9b00464
- PubMed
- Google Scholar
1. Tan K
2. Zhang X
3. Cong X
4. Huang B
5. Chen H
6. Chen D
(2017) Tumor suppressor RYBP harbors three nuclear localization signals and its cytoplasm-located mutant exerts more potent anti-cancer activities than corresponding wild type
Cellular Signalling 29:127–137.
https://doi.org/10.1016/j.cellsig.2016.10.011
- PubMed
- Google Scholar
1. Toyama BH
2. Savas JN
3. Park SK
4. Harris MS
5. Ingolia NT
6. Yates JR
7. Hetzer MW
(2013) Identification of long-lived proteins reveals exceptional stability of essential cellular structures
Cell 154:971–982.
https://doi.org/10.1016/j.cell.2013.07.037
- PubMed
- Google Scholar
1. Xu D
2. Farmer A
3. Chook YM
(2010) Recognition of nuclear targeting signals by karyopherin-β proteins
Current Opinion in Structural Biology 20:782–790.
https://doi.org/10.1016/j.sbi.2010.09.008
- PubMed
- Google Scholar
1. Xu D
2. Farmer A
3. Collett G
4. Grishin NV
5. Chook YM
(2012) Sequence and structural analyses of nuclear export signals in the nesdb database
Molecular Biology of the Cell 23:3677–3693.
https://doi.org/10.1091/mbc.E12-01-0046
- PubMed
- Google Scholar
1. Yurko NM
2. Manley JL
(2018) The RNA polymerase II CTD “ orphan ” residues: emerging insights into the functions of tyr-1, thr-4, and Ser-7
Transcription 9:30–40.
https://doi.org/10.1080/21541264.2017.1338176
- PubMed
- Google Scholar
1. Zhou Q
2. Li T
3. Price DH
(2012) Rna polymerase II elongation control
Annual Review of Biochemistry 81:119–143.
https://doi.org/10.1146/annurev-biochem-052610-095910
- PubMed
- Google Scholar

Decision letter

Irwin Davidson

Reviewing Editor; Institut de Génétique et de Biologie Moléculaire et Cellulaire, France
Kevin Struhl

Senior Editor; Harvard Medical School, United States

Our editorial process produces two outputs: i) public reviews designed to be posted alongside the preprint for the benefit of readers; ii) feedback on the manuscript for the authors, including requests for revisions, shown below. We also include an acceptance summary that explains what the editors found interesting or important about the work.

Decision letter after peer review:

Thank you for submitting your article "Lifting the ban on nuclear import activates Gdown1-mediated modulation of global transcription and facilitates adaptation to cellular stresses" for consideration by eLife. Your article has been reviewed by 2 peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Kevin Struhl as the Senior Editor. The reviewers have opted to remain anonymous.

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

The referees agreed that this study should be of general interest to readers of eLife following an adequate revision that takes into account the issues raised by the referees.

The authors should pay specific attention to the following points.

1). The BiFC results indicate that GDOWN1 appears to interact with all the tested cytoplasmic proteins and moreover most on the deletion mutants also still interact with their nuclear targets. The authors must provide negative controls to demonstrate the specificity of the interactions detected in this way. The referees also question the utility of these experiments that in any case to not add to the conclusions concerning the sequences that regulate GDOWN1 intra-cellular localization.

2). On page 9 the authors claim that M1 localizes mainly in the nucleus. It is more accurate to say partitions between the cytoplasm and nucleus. If fact there are cells where it seems also to remain cytoplasmic. Similarly, in Figure 3C there are cells on the left of the upper panel that also seem to show diffuse cytoplasmic staining. The authors should perform quantitative image analyses. This has been done in some figures but should be added for all.

3). The manuscript requires a major revision by a native English speaker to correct a large number of grammatical errors.

4). The style of the title could also be changed. Lifting the ban sounds too familiar. For example 'Overcoming cytoplasmic retention of GDOWN1 by the combinatorial action of nuclear export and cytoplasmic anchoring signals modulates transcription and facilitates adaption to cellular stress.

Reviewing editor comments.

1). Figure 3D is rather difficult to visualize, could the authors show a higher magnification?

2). In Figure 4B the Drosophila Cas mutant appears to accumulate more in the nucleus that the mutated zebrafish protein. This is rather surprising as given that Drosophila is more distantly related than zebrafish, but may reflect the fact that Drosophila CAS motif seems better conserved. Can the authors comment on this? Quantification analyses of the images would strengthen the results.

3). In Figure 5B, rather than simply using DAPI in the flow cytometry, can the authors use activated caspase 3 or Annexin V to look for apoptosis? Is apoptosis the mechanism of cell death?

4). In Figure 6B there seems to be a perinuclear staining on the iNLS-GDOWN in the bottom right Venus panel. Can the authors comment on this?

5). Many of the experiments in this study are performed with stable or inducible ectopic expression. It is difficult to assess the expression level of the exogenous proteins versus the endogenous by the immunofluorescence. In the supplemental data, the authors performed immunoblots that often showed the exogenous proteins to be expressed at much higher levels that endogenous. To some extent, this reinforces some of the conclusions when the overexpressed mutants remain cytoplasmic as it indicates the robustness of the retention mechanism. However, the authors should mention the potential caveats that overexpression may introduce.

In addition to the above, the authors should also address the various issues raised by the referees and listed below.

Reviewer #1 (Recommendations for the authors):

On p. 3, lines 42-43: The general Pol II transcription factors are listed; TFIIS should not be included in that group.

Figure 1D, for example. The authors use BiFC signals to track Gdown1; these rely on the presence in the cytoplasm of proteins that should be nuclear, like Spt4/5. I found this confusing, and the approach did not seem to add anything to straight detection of tagged Gdown1.

On p. 9, lines 198-199: The authors refer to the structurally flexible regions of Gdown1 without any indication of how that determination was made. They should refer to the structural predictions in Figure S4A.

On p. 14, lines 342-345: It is asserted that the function of the CAS motif depends partially on structural support from NES2. "Structural support" seems like an overstatement- I don't think any particular mechanistic connection can be inferred.

P. 17, first paragraph: Do the authors have any sense for the relative levels of the 10M and NLS^-10M versions of Gdown1, post dox exposure, as compared to the endogenous Gdown1?

Figure 6B and S6: The overall effect of Gdown1 access to the nucleus is clear, but more could have been learned with a quantitative comparison of the relative loss of total Pol II, ser5P and ser2P. For example, if ser2 is reduced more than ser5 this could imply an effect on elongation. Probing a panel C blot with the appropriate antibodies could have addressed this point.

I could not understand the result in panel 6C. If nuclear export is blocked by LMB, how can most of HeLa Pol II be cytoplasmic after LMB treatment?

Points in the Discussion:

How is Gdown1 acting in the nucleus? The earlier biochemical work provided a number of possibilities but those aren't investigated here or extensively discussed. Is Gdown1 mostly acting to tie up Pol II so that it cannot make PICs, or is it also/mostly acting downstream of initiation to block elongation? As noted above, determining whether ser5 and ser2 Pol II are affected differently might have shed light on this. Additionally, it would have been useful to know if nuclear Gdown1 is chromatin associated (probably talking to elongation) or free in the nucleoplasm (probably bound to Pol II). The point about effects on elongation seems particularly important when trying to understand how nuclear Gdown1 reduced overall Pol II levels. In particular, if elongation were severely affected polymerases could be stalled for extended periods at difficult-to-traverse locations, provoking the ubiquitination and clearance of those complexes (and thus reducing overall Pol II levels).

Finally, lines 604-607: The authors assert that EU labeling in nucleoli is drastically reduced upon Gdown1 entry into nuclei, showing that Gdown1 interferes with both Pol I and Pol III activity. There is no quantitation provided for the assertion that nucleolar labeling is reduced, but regardless such reduction would not show a direct effect on Pol I (or Pol III) activity- strong reductions in mRNA synthesis should feed back into reductions in rRNA synthesis.

Reviewer #2 (Recommendations for the authors):

We have a few suggestions to increase the clarity of this manuscript:

1. The current title is a bit complicated. Can the authors make the title simpler?

2. The authors have some factually incorrect statements in their introduction. They should try to improve the content and presentation of the introduction. Examples:

a. In the introduction (line 57-64), they discuss in vitro work done on GDOWN1, but they do not mention the 2018 structure from Roeder (doi: 10.1038/s41594-018-0118-5). Can the authors please add this paper, as it also clarifies what is known about GDOWN1-Pol II interactions?

b. In the introduction (line 89), they mention ChIP-seq data of Gdown1 and elongating Pol II. Can they please clarify this sentence? It is confusing to understand.

c. Line 38: RNA polymerase II is in fact not the only polymerase that synthesis protein coding genes in eukaryotes (e.g., mitochondrial RNA polymerase also synthesizes protein coding mRNA).

3. The figures are very complicated with many panels. Could the authors consider making more subpanels? For example, Figure 1D is showing two different things; what factors GDOWN1 interacts with, and then various factors interacting with other factors. We recommend separating these figures to increase the clarity of the paper.

4. We have a few suggestions for improving Figure 4A. It is of particular interest to see how NES1, NES2, and then the double NES mutation influences nuclear levels. Could they more directly compare these three mutation-types by putting quantitative analysis of all constructs next to each other?

5. While the evolutionary analysis in Figure 4B+C is exciting, we feel it takes away from the figures purpose. We suggest moving figure 4B+C to the supplementals and then expanding on the analysis of the different NES in the main figure.

6. Please make a cartoon describing that 10M is a triple mutation of all three identified domains in figure 5. We generally to produce an additional supplementary figure panel that summarizes all used constructs in a graphical representation.

7. In Figure 6, the authors provide evidence for a decrease in EU RNA incorporation in the context of the triple mutant. It would strengthen the manuscript if the authors could demonstrate a transcriptional defect by providing data from a complementary technique.

8. Could the authors provide additional quantification of Figure S6B?

9. The authors should also provide error bars for Figure S6C.

10. As the authors are well aware, Pol II antibodies show sub-optimal specificity. Could the authors provide data from additional Pol II antibodies (e.g., S2P, S5P, RPB3) for Figure S6C?

11. In lines 455-456, the authors mention in vitro transcription assays, but do not cite anything. Please cite the study being referred to.

12. We ask the authors to discuss why GDOWN1 apparently interacts with so many transcriptional regulators in the cytoplasm. Is GDOWN1 regulating these factors by interacting with them in the cytoplasm?

13. The Alphafold structure in supplement figure 4 – 1A is difficult to see. Can the authors increase clarity by changing to a white background and focusing more on the regions they are showing? Please also add the same cartoon that is used throughout the paper to further help clarifications.

14. The authors should also carefully check their manuscript for missing prepositions etc.

https://doi.org/10.7554/eLife.79116.sa1

Author response

The authors should pay specific attention to the following points.

1). The BiFC results indicate that GDOWN1 appears to interact with all the tested cytoplasmic proteins and moreover most on the deletion mutants also still interact with their nuclear targets. The authors must provide negative controls to demonstrate the specificity of the interactions detected in this way. The referees also question the utility of these experiments that in any case to not add to the conclusions concerning the sequences that regulate GDOWN1 intra-cellular localization.

Thank you for the Reviewers’ comments! The Reviewers concerned about the interacting specificity between GDOWN1 mutants and NELF-E. We agree this part of data is not very easy to be fully understood. GDOWN1 was reported to be directly interacting to multiple components of Pol II (Hu et al., 2006; Jishage et al., 2018). We also found GDOWN1 played an important role in the cytoplasmic assembly of Pol II (our unpublished data), which led us hypothesize that the GDOWN1 (both the full length and the truncated mutants) & NELF-E interactions seen in our BiFC assays were possibly mediated by other factors such as the Pol II components. The binding specificity of our BiFC results have been well proven. Quite a lot of factors were screened in our BiFC assays and some of them did show negative results. To further strengthen the specificity, we added more negative controls in Figure 1—figure supplement 1D. This new collection of negative controls includes three cytoplasmic proteins, VDAC1 (a mitochondrial protein), GALNT2 (a Golgi protein) and PDIA3 (an endoplasmic reticulum protein). These results, together with the two nuclear factors previously shown in Figure 1D (MED1 and MED26), emphasize that GDOWN1 only interacts to its real binding partners.

We prefer to keep the BiFC data in the manuscript since it is beneficial of using BiFC assays to identify key localization-controlling motifs. The detailed reasons are listed as followed:

1) BiFC assays have the advantages of directly exhibiting the interactions between GDOWN1 and its potential binding partners in live cells. Originally, we employed these assays as a quick way to verify the interaction of GDOWN1 to Pol II components, and meanwhile to test our hypothesis that GDOWN1 might physically interact to transcription elongation factors due to their functional coordination (Cheng et al., 2012). Indeed, the BiFC results indicated very strong signals between GDOWN1 and DSIF (SPT4 and SPT5) or NELF (NELF-E and NELF-A etc.). However, these interactions were unexpectedly present in the cytoplasm. As shown in Figure 1E, the BiFC signals of SPT4•SPT5 interaction, and NELF-E•NELF-A interaction were both present in the nucleus as expected. These results actually fit quite well with the fact that GDOWN1 is primarily present in the cytoplasm. Under the BiFC conditions, a portion of the overexpressed nuclear factors are present in the cytoplasm, so that their interactions to the cytoplasmic GDOWN1 become well captured, while these interactions are typically occurring only in the nucleus upon the appearance of GDOWN1.

2) BiFC system has the advantage to stably capture GDOWN1’s binding partner(s). The irreversibility of BiFC has been considered as its drawback in many circumstances, while in our case, it was actually turned it into an advantage for efficiently identifying any possible and even transient interactions between GDOWN1 and other factors. Our BiFC results well supported the data from our immunofluorescence and cell fractionation assays about GDOWN1’s subcellular localization on its own. In addition, they were able to capture protein/protein interactions between GDOWN1 and other factors, even when they only occurred transiently.

3) BiFC system provides better efficiency and contrast in mutant screening for the key localization motifs of GDOWN1. We observed that when being transiently transfected into the cells, BiFC signals could be well detectable very rapidly (as early as 6 hours post transfection), comparing to expressing GDOWN1 tagged with a fluorescent protein (more than 12 hours). Thus, it allowed us to acquire the data at an early time point, which was especially useful when testing the mutants with a potential of triggering cell death. In addition, since GDOWN1 by itself does not contain any strong NLS, its truncation mutants do NOT enter the nucleus actively even when its cytoplasmic retention signal is chopped off. To overcome this potential problem, we employed NELF-E as a potent nuclear partner of GDOWN1 in our BiFC assays to facilitate GDOWN1’s nucleus entry. This difference is clearly demonstrated in Figure 2B, when data from both IF and BiFC are shown side by side. For example, the m2 mutant was uniformly localized in both cytoplasm and nucleus in the IF result, while it was strictly present in the nucleus in the BiFC result. Similar examples also include Figure 4D and Figure 4—figure supplement 1D. Overall, this better contrast of BiFC data facilitated our mutant screening and simplified our statistical analyses.

The functional efficiency of those regulatory motifs identified by BiFC assays was further confirmed by a battery of tests including IF assays and direct detections of the mutants via monitoring the fused fluorescent tags. Taken together, we believe that the main conclusions from these BiFC results in the manuscript are solid.

2). On page 9 the authors claim that M1 localizes mainly in the nucleus. It is more accurate to say partitions between the cytoplasm and nucleus. If fact there are cells where it seems also to remain cytoplasmic. Similarly, in Figure 3C there are cells on the left of the upper panel that also seem to show diffuse cytoplasmic staining. The authors should perform quantitative image analyses. This has been done in some figures but should be added for all.

Thank you for your pointing out our inaccurate claim here. We corrected it by rewording it into ‘M1 localizes partially in the nucleus’.

We added the quantitative image analyses for Figure 3—figure supplement 1A,1C and the rest of figures in the revised manuscript, including Figure 2B, Figure

2—figure supplement 1B, 1C, Figure 3—figure supplement 1A, 1B, 1D, Figure 4B, Figure 4D (Figure 4—figure supplement 1C), Figure 4—figure supplement 1D, Figure 6—figure supplement 1C etc. Overall, we think these additions improved the quality of our manuscript, thank you!

3). The manuscript requires a major revision by a native English speaker to correct a large number of grammatical errors.

Thank you for pointing out our weakness of writing in English. We now carefully modified our manuscript and corrected those grammatical mistakes.

4). The style of the title could also be changed. Lifting the ban sounds too familiar. For example 'Overcoming cytoplasmic retention of GDOWN1 by the combinatorial action of nuclear export and cytoplasmic anchoring signals modulates transcription and facilitates adaption to cellular stress.

Thank you for this really good suggestion! This title accurately describes the main findings and the key points of our manuscript while it seems a little long. We shortened this title and decided to use the new title shown below. ‘Overcoming the cytoplasmic retention of GDOWN1 modulates global transcription and facilitates stress adaptation’.

Reviewing editor comments.

1). Figure 3D is rather difficult to visualize, could the authors show a higher magnification?

Thank you for pointing out this issue! We think the visualization difficulty for Figure 3D is mainly due to the brightness, not the magnification. The actual intensity of BiFC signals were variable across samples. The main point here is to demonstrate the interactions between GDOWN1 and some representative nuclear pore components, and compare their distinct subcellular localizations qualitatively. Thus, for better demonstration, we adjusted the brightness of the images to make the differences of their subcellular localization easier to be visualized.

2). In Figure 4B the Drosophila Cas mutant appears to accumulate more in the nucleus that the mutated zebrafish protein. This is rather surprising as given that Drosophila is more distantly related than zebrafish, but may reflect the fact that Drosophila CAS motif seems better conserved. Can the authors comment on this? Quantification analyses of the images would strengthen the results.

This is an interesting point! In this study, we identified that the three arginines as a whole in the CAS motif were very important for the cytoplasmic retention of human GDOWN1. However, we did not further narrow it down to dissect the contribution of each individual R. From our data, it appeared that fly CAS played a very similar localization-regulatory role to its human counterpart. We followed the Reviewing editor’s suggestion to quantify these images and provided the data in Figure 4─figure supplement 1C. The quantitative data were consistent with the images that the Drosophila CAS mutant was indeed accumulated more in the nucleus than the zebrafish mutant upon LMB treatment.

Our experiments of ectopically expressing either fly or fish Gdown1 in human cells mainly tested the conservation of CAS motifs qualitatively but not quantitatively. Overall, we think the data emphasized that the CAS motifs (and some of the associated arginines) were very conserved from the low animals to mammals. Further experiments are necessary to figure out the accurate evolutionary orders across species.

3). In Figure 5B, rather than simply using DAPI in the flow cytometry, can the authors use activated caspase 3 or Annexin V to look for apoptosis? Is apoptosis the mechanism of cell death?

Thank you for this very valuable suggestion! We have been very interested in the type and the mechanism of cell death upon the nuclear accumulation of GDOWN1 as well. Previously, we transiently transfected a GDOWN1 truncation mutant and observed severe cell death within a few hours upon its nuclear accumulation (Author response image 1A). These cells were rapidly dying with the appearance of apoptotic-like bodies. Recently, we carried out Flow cytometry experiments using Annexin-V to check any possible signs of early-stage apoptosis. Two cell lines shown in Figure 5B, iGDOWN1(WT)-Venus expressing in the cytoplasm and iNLS-GDOWN1(10M) expressing in the nucleus, were compared in parallel. The two cell lines were incubated with Dox for 3 days to induce the expression of the GDOWN1 proteins and then subjected to Flow cytometry analyses. As seen in Author response image 1B, the cells expressing GDOWN1-Venus did not show much Annexin-V⁺ signals (1.49% in Quadrant 4), while the cell expressing NLS-GDOWN1(10M) indeed showed increased Annexin-V⁺ signals (8.83%). Due to some loss of dead cells during the cell harvest process, the ratio of Annexin-V⁺ cells might be underestimated. These data suggest that the nuclear GDOWN1 may cause apoptosis to some extent.

Author response image 1

Download asset Open asset

The nuclear accumulation of GDOWN1 may trigger apoptosis.

(A) Transient transfection of a human GDOWN1 truncation mutant resulted in cell death within hours post LMB-induced nuclear translocation of GDOWN1. In this case, we observed the appearance of apoptotic body-like extracellular vesicles. (B) Flow cytometry analyses for detection of the apoptotic signals comparing the cells expressing the either the cytoplasmic GDOWN1-Venus [iGDOWN1(WT)-Venus] or the nuclear GDOWN1 [iNLS-GDOWN1(*10M*)-Venus]. The X-axis showed the phosphatidylserine (PS) signals of the outer membrane labeled with APC fluorescence, and the Y-axis marked the π signals.

Although the transient expression of the nuclear GDOWN1 resulted in a severe cell death within hours as mentioned above, in the cell line stably expressing NLS-GDOWN1(10M), we observed that a significant portion of these cells were able to survive the constant expression of the nuclear GDOWN1 for several days at least. This cell population actually experienced a constant and slow process of cell death, indicating that some other unknown cell-death mechanisms were also existing along with apoptosis. Since further elucidation of these complicated cell death mechanisms is out of the scope of this manuscript, we decided not to include this piece of data or draw any immature conclusion in this manuscript. Further analyses are currently ongoing in our laboratory.

4). In Figure 6B there seems to be a perinuclear staining on the iNLS-GDOWN in the bottom right Venus panel. Can the authors comment on this?

Thank you for pointing this out! We think this observation was consistent to the rest of the findings in this manuscript quite well and happened exactly as we expected! The massive accumulation of the perinuclear signal of NLS-GDOWN1 was contributed by the combinatory effects of two layers of actions. One was the intrinsic nature of cytoplasmic retention determined by GDOWN1’s CAS motif, the other was the artificial driving force for its accumulation in the nucleus enforced by NLS. Apparently, NLS lost this game, as indicated by the enhanced retention of GDOWN1 at the entrance of the nucleus. Overall, this observation emphasized the unusually potent capability of GDOWN1’s retention in the cytoplasm.

5). Many of the experiments in this study are performed with stable or inducible ectopic expression. It is difficult to assess the expression level of the exogenous proteins versus the endogenous by the immunofluorescence. In the supplemental data, the authors performed immunoblots that often showed the exogenous proteins to be expressed at much higher levels that endogenous. To some extent, this reinforces some of the conclusions when the overexpressed mutants remain cytoplasmic as it indicates the robustness of the retention mechanism. However, the authors should mention the potential caveats that overexpression may introduce.

Thank you for this valuable suggestion! In this revised manuscript, we provided immunoblots to directly compare the expression levels of the ectopically expressed GDOWN1 to their endogenous counterpart (shown in Figure 5─figure supplement 1A) and the result is briefly mentioned in the RESULTS session of the manuscript. As seen in this figure, this experimental system often resulted in overexpression of the exogenous protein(s) of interest, as the Reviewing Editor concerned. Theoretically, this might generate non-objective observations and conclusions. Luckily, in our case, the three stable cell lines with GDOWN1 overexpressed (~5 fold or more comparing to the endogenous GDOWN1) did not cause much cell death or other cell growth defects, while the one that triggered severe cell death (iNLS-GDOWN1-10M) only expressed the exogenous protein at a level comparable to the endogenous. Thus, the general concerns and the potential caveats about studying overexpressed exogenous proteins did not apply here so that the basic observation and the conclusions would not be interfered much. Overall, we believe that our conclusions stated in the manuscript are objective and solid. We also discuss this point in the DISCUSSION session.

In addition to the above, the authors should also address the various issues raised by the referees and listed below.

Reviewer #1 (Recommendations for the authors):

On p. 3, lines 42-43: The general Pol II transcription factors are listed; TFIIS should not be included in that group.

Thank you for your suggestion! It was our mistake and it is reasonable to distinguish TFIIS from other general transcription factors. This factor is now removed from the original description ‘including general transcription factors (TFIID, TFIIA, TFIIB, TFIIF, TFIIE, TFIIH and TFIIS)’ in the manuscript (line 43).

Figure 1D, for example. The authors use BiFC signals to track Gdown1; these rely on the presence in the cytoplasm of proteins that should be nuclear, like Spt4/5. I found this confusing, and the approach did not seem to add anything to straight detection of tagged Gdown1.

Thank you for your comments! Since both reviewers brought up this question, we provided our answers to this question above in the “Common concerns by the Reviewers” session (Q#1).

On p. 9, lines 198-199: The authors refer to the structurally flexible regions of Gdown1 without any indication of how that determination was made. They should refer to the structural predictions in Figure S4A.

Thank you pointing out the unclear statement in our manuscript! We add a Supplementary figure in Figure 2─figure supplement 1A, providing the secondary structural prediction of the entire human GDOWN1 using PSIPRED

(http://bioinf.cs.ucl.ac.uk/psipred/). The three mutants (m1-m3) used in Figure 2 are labeled and as can be seen in this figure, the two positions selected to split GDOWN1 (amino acids 110 and 250) are both located within the coiled region, which we believe to have minimal impact on the overall structure of GDOWN1. We now add a brief description in the main text for clarification (line 205).

On p. 14, lines 342-345: It is asserted that the function of the CAS motif depends partially on structural support from NES2. "Structural support" seems like an overstatement- I don't think any particular mechanistic connection can be inferred.

Thanks for this valuable comment! We agree that more evidence is required in order to draw this conclusion. Our data indicated that the functions of NES2 and CAS are interconnected. Since these two regulatory motifs are very close at the primary structural level, we are even lacking of enough evidence that they are absolutely independent to each other. Therefore, we reword the statement here. ‘The above data from the intrinsic motif analyses demonstrate that each one of the two NES motifs of GDOWN1 acts as an independent CRM1-regulated element and the function of CAS motif depends partially on the existence of NES2.’

P. 17, first paragraph: Do the authors have any sense for the relative levels of the 10M and NLS^-10M versions of Gdown1, post dox exposure, as compared to the endogenous Gdown1?

To address reviewer’s concern about the expression levels across the stable cell lines used on Figure 5, we add the immunoblots of these cell lines in Figure 5─figure supplement 1A, showing the expression levels of both endogenous and exogenous GDOWN1 in the presence or absence of Dox induction. As seen in this figure, the expression levels of the exogenous GDOWN1 are all higher than the endogenous counterpart, ranging from a couple to several folds. When comparing the expressions in each cell line pairs, it can be seen that the level of the GDOWN1(10M) is similar to that of wild type GDOWN1, while the level of the NLS-GDOWN1(10M) is much less than that of NLS-GDOWN1(WT). However, our results from various experiments shown in Figure 5 clearly and consistently indicated that the equal or less expressions of the 10M mutants achieved more severe defects on global transcription and cellular growth. Therefore, we believe that the drastic defects observed in the cells with the 10M mutants being expressed are mainly attributed to the nuclear accumulation and functions of GDOWN1. It is worth to note that the expression level of NLSGDOWN1(10M) is at a level very close to that of the endogenous, which is the lowest among the four cell lines, while it exhibits the most severe defects. Thus, our major findings claimed in the manuscript were not generated due to the excessive expression of exogenous GDOWN1.

In the revised manuscript, besides adding the above WB data in the supplementary figure, we also pointed out this results in the RESULTS session (line 485-486).

Figure 6B and S6: The overall effect of Gdown1 access to the nucleus is clear, but more could have been learned with a quantitative comparison of the relative loss of total Pol II, ser5P and ser2P. For example, if ser2 is reduced more than ser5 this could imply an effect on elongation. Probing a panel C blot with the appropriate antibodies could have addressed this point.

Thank you for this good suggestion and we agree that checking the changes of Ser5P and Ser2P would provide more valuable information. We did this experiment and the quality of the WB data was ok. Since the stable cell lines used in these assays were cell pools with the expression levels of the total and the nuclear GDOWN1 varied from cell to cell, checking the total protein levels by WB might underestimate the real expression changes. Please refer to our response to your related question #8 and the Response Figure 2A-B. In addition, we carried out the IF experiments using various antibodies and carefully quantified the results (Figure 6C and Figure 6—figure supplement 1B) and the data indicated that upon the expression of the GDOWN1(10M) and NLS- GDOWN1(10M) (on day 4 post Dox addition), both Ser5P and Ser2P were reduced upon the expression of the nuclear GDOWN1, suggesting that both transcription initiation and elongation were inhibited. However, the statistical analyses shown in Figure 6—figure supplement 1C (better presented in Response Figure 2C-D) indicated that the S5P signal dropped more severe than the S2P signal, suggesting that the transcriptional initiation was more sensitive to the regulation of GDOWN1. The point was mentioned in both the RESULTS session.

I could not understand the result in panel 6C. If nuclear export is blocked by LMB, how can most of HeLa Pol II be cytoplasmic after LMB treatment?

This is an interesting question! Although Pol II is not a target of CRM1 by itself, it has been known that its assembly factors, such as RPAP2 and RPAP4/GPN1 are cargoes of CRM1(Forget et al., 2010; Forget et al., 2013). Upon LMB treatment, these factors will be accumulated in the nucleus and result in the dramatic reduction of Pol II assembly and its cytoplasm retention. This piece of data is now moved to the Figure 6—figure supplement 1A.

Points in the Discussion:

How is Gdown1 acting in the nucleus? The earlier biochemical work provided a number of possibilities but those aren't investigated here or extensively discussed. Is Gdown1 mostly acting to tie up Pol II so that it cannot make PICs, or is it also/mostly acting downstream of initiation to block elongation? As noted above, determining whether ser5 and ser2 Pol II are affected differently might have shed light on this. Additionally, it would have been useful to know if nuclear Gdown1 is chromatin associated (probably talking to elongation) or free in the nucleoplasm (probably bound to Pol II). The point about effects on elongation seems particularly important when trying to understand how nuclear Gdown1 reduced overall Pol II levels. In particular, if elongation were severely affected polymerases could be stalled for extended periods at difficult-to-traverse locations, provoking the ubiquitination and clearance of those complexes (and thus reducing overall Pol II levels).

Very good questions and suggestions, thank you! Our original goal was to dissect GDOWN1’s functional mechanisms in regulating transcription in the nucleus. However, the fact that it did not enter the nucleus redirected us to a new route of digging the molecular basis of its cytoplasmic retention. It was exciting to uncover that the native GDOWN1 was translocated into the nucleus upon the oxidative stresses, possibly opening a new avenue of identifying its physiological roles in the cellular adaptation to stresses.

When being translocated into the nucleus, GDOWN1 triggered the global repression of transcription. The IF signals of the un-phosphorylated Pol II (preferably detected by the Pol II-CTD antibody 8WG16, Abcam, ab817; please also refer to our response to Reviewer 2’s Q8 for some evidence about this antibody’s specificity), and the phosphorylated forms of Pol II on the Ser5 residues (detected by the Pol II-S5P-CTD antibody CTD4H8, Biolegend, 904001) and on Ser2 residues (detected by the Pol II-S2P-CTD antibody, Abcam, ab5095) were all reduced in the cells with the expression of the nuclear GDOWN1 at 4 days after the Dox induction (Figure 6B and Figure 6—figure supplement 1A).

To confirm these data, we carefully tested the changes of these various forms of Pol II via WB experiments using a series of Pol II antibodies (Author response image 2A). A newly ordered Pol II-NTD antibody (clone D8L4Y, CST) was employed to more accurately detect the total level of Pol II, including the un-phosphorylated and the phosphorylated forms (Figure 6B). As seen in Author response image 2A and the quantified data in Author response image 2B, the protein levels of the total, the un-phosphorylated, and the S5P form of Pol II all got reduced significantly upon the nuclear accumulation of GDOWN1 (the 10M mutant VS the wild type), while the change of the S2P-Pol II was relatively small. We also ordered another Ser5P antibody, while unfortunately, after a 3 months-waiting for its oversea shipping, it turned out it did not work well.

Author response image 2

Download asset Open asset

The nuclear accumulation of GDOWN1 causes reduction of Pol II.

(A) Using four different Pol II antibodies to detect the changes of Pol II in the indicated cell lines upon the expression of GDOWN1 on day 4 post Dox addition; [T: total, the whole cell lysate; C: cytosol; N: nuclei] (B) The quantification of the data shown in (A); (C) The nuclear accumulation of GDOWN1 in the iNLS-GDOWN1(*10M*) cells causes differential decreases of the distinct forms of Pol II in IF assays; (D) The quantification of the data shown in (C).

To confirm the change of S2P-Pol II shown in Figure 6—figure supplement 1A [Anti-RPB1 CTD (S2P), Abcam, ab5095], we repeated the IF experiments with another S2P specific antibody [Anti-RPB1 CTD (S2P), Abcam, clone3E10]. As seen in Author response image 2C (also contain a subset of images in Figure 6B for comparison), both S2P antibodies generated very similar results that the levels of the S2P-Pol II went down in the cells expressing the nuclear GDOWN1. However, the extent of the reduction for S2P was much less than that of the S5P. We recently noticed that at early time points, accumulation of the nuclear GDOWN1 seemed to primarily affect Ser5P levels (data not shown).

We also confirmed that the loss of Pol II signals was attributed to protease-dependent protein degradation (-/+ MG132 treatment, data not shown), while the mechanisms behind remain to be uncovered. We added some of the above information in the DISCUSSION session.

All of our current experimental evidence supports the hypothesis that once entering the nucleus, GDOWN1 preferentially binds to and affects the unphosphorylated Pol II and the S5P-Pol II engaged in the initiation stage, and causes their degradation. At later stage, its long-lasting accumulation in the nucleus also leads to the severe reduction of S2P-Pol II. More thorough analyses are definitely necessary to dissect the mechanistic details of these processes, and we are working on it to make it clear.

Finally, lines 604-607: The authors assert that EU labeling in nucleoli is drastically reduced upon Gdown1 entry into nuclei, showing that Gdown1 interferes with both Pol I and Pol III activity. There is no quantitation provided for the assertion that nucleolar labeling is reduced, but regardless such reduction would not show a direct effect on Pol I (or Pol III) activity- strong reductions in mRNA synthesis should feed back into reductions in rRNA synthesis.

Thank you for helping us find this inaccurate statement in our manuscript. Actually, it was shown in the literature that the transcription rate of Pol II was much higher than Pol I and Pol III (Jackson et al., 2000), and ~80% of the EU labeled transcripts via a short pulse (within one hour) were Pol II products (Nozawa et al., 2017). Therefore, we partially missed the detection about the dynamic changes of Pol I and Pol III transcripts in our 20 min-EU pulse labeling assays, and the major signals and the associated changes shown in our IF results were mainly representing Pol II transcripts. Yes, you are right, the reduction of the nucleolar labeling might be directly or indirectly affected by GDOWN1 and the associated Pol II activity, while we are currently lacking of evidence. Based on the above considerations, we removed this statement from the DISCUSSION session.

Reviewer #2 (Recommendations for the authors):

We have a few suggestions to increase the clarity of this manuscript:

1. The current title is a bit complicated. Can the authors make the title simpler?

Thank you for your suggestion! We simplified this title as followed, “Overcoming cytoplasmic retention of GDOWN1 modulates global transcription and facilitates adaptation to cellular stresses”.

2. The authors have some factually incorrect statements in their introduction. They should try to improve the content and presentation of the introduction. Examples:

a. In the introduction (line 57-64), they discuss in vitro work done on GDOWN1, but they do not mention the 2018 structure from Roeder (doi: 10.1038/s41594-018-0118-5). Can the authors please add this paper, as it also clarifies what is known about GDOWN1-Pol II interactions?

b. In the introduction (line 89), they mention ChIP-seq data of Gdown1 and elongating Pol II. Can they please clarify this sentence? It is confusing to understand.

c. Line 38: RNA polymerase II is in fact not the only polymerase that synthesis protein coding genes in eukaryotes (e.g., mitochondrial RNA polymerase also synthesizes protein coding mRNA).

Thank you for the Reviewer’s comments about the INTRODUCTION session. We carefully revised this part and corrected or clarified all the points mentioned by the reviewer and some others.

a. Professor Roeder’s paper in 2018 really is a very important paper demonstrating that the GDOWN1/Pol II interactions via the structural point of view, and further providing evidence about the associated effects on transcription. We now added this reference in the INTRODUCTION session and pointed out its main findings. b. We clarified this point in the revised manuscript.

c. This mistake was corrected.

3. The figures are very complicated with many panels. Could the authors consider making more subpanels? For example, Figure 1D is showing two different things; what factors GDOWN1 interacts with, and then various factors interacting with other factors. We recommend separating these figures to increase the clarity of the paper.

Thank you for the Reviewer’s suggestion! We now split Figure 1D into two panels, D and E. All the other main figures are also carefully checked. We agree that they are complicated, while we find logically the current organization seems to be reasonable. Therefore, we would like to keep them as they are. If the Reviewers do think some of the other figures need to be further split or they are organized inappropriately, please specify the figure number(s) and we would like to try our best to improve, thank you!

4. We have a few suggestions for improving Figure 4A. It is of particular interest to see how NES1, NES2, and then the double NES mutation influences nuclear levels. Could they more directly compare these three mutation-types by putting quantitative analysis of all constructs next to each other?

We thank the Reviewer for this good suggestion! We now reorganized the quantitative data analyses of all the constructs shown in Figure 4A and show them side by side for better comparison (Figure 4B) and the detailed description was added in the RESULTS Session. Upon LMB treatment, the N/C ratios of the three GDOWN1 mutants are ordered as mCAS (g) > mNES2 (e) >> mNES1 (c), which indicates that the CRM1-independent cytoplasmic localization of GDOWN1 is mainly determined by CAS, while NES2 also plays a role on this regulation. In the presence of LMB, there is no significant difference between mNES2 (e) and mNES1+2 (f), indicating that NES1 is not involved in regulating GDOWN1’s localization via the CRM1-independent manner. The significantly enhanced N/C ratio of NES1+CAS and NES2+CAS double mutants comparing to the NES single mutants (d to c, b to e) in the presence of LMB further confirms that CAS is the core motif in controlling GDOWN1’s cytoplasmic localization.

5. While the evolutionary analysis in Figure 4B+C is exciting, we feel it takes away from the figures purpose. We suggest moving figure 4B+C to the supplementals and then expanding on the analysis of the different NES in the main figure.

Thank you for this suggestion! We reorganized the data arrangement in Figure 4 and further expanded the data analyses of various GDOWN1 mutants (Figure 4B). We agree that the part of the evolutionary analysis is slightly far away from the main route of the topic, while these important observations provide a strong support for the conservation of these key amino acids in controlling the subcellular localization of Gdown1 during evolution. Logically, it is suitable to be in this main figure. Therefore, we would like to keep it as it is to emphasize the importance of these amino acids.

6. Please make a cartoon describing that 10M is a triple mutation of all three identified domains in figure 5. We generally to produce an additional supplementary figure panel that summarizes all used constructs in a graphical representation.

It’s a good suggestion. We have added a panel of cartoons on the top of Figure 5B. In addition, we summarized all the constructs in a table of the MATERIALS AND METHODS session.

7. In Figure 6, the authors provide evidence for a decrease in EU RNA incorporation in the context of the triple mutant. It would strengthen the manuscript if the authors could demonstrate a transcriptional defect by providing data from a complementary technique.

Thank you for your suggestion! We agree that providing more evidence in addition to the EU staining would be better. In this case, when global transcription is dramatically reduced, technically it is challenging to quantify the changes in RT-qPCR assays or RNA-Seq analyses. We now provide some normalized RT-qPCR data in Author response image 3. The total RNAs were extracted from equal numbers of the four cell lines shown in Figure 6A on day 3 after the mock treatment or Dox incubation, respectively. We found that the overall quantities of the total RNA were generally not changed, suggesting that the bulk of cellular rRNAs may not be significantly affected, at least on the third day after the nuclear GDOWN1 expression (data not shown). In order to accurately quantify the expression changes in qPCR assays, we introduced the total RNA from Arabidopsis thaliana as an internal reference. The reverse transcription reactions were performed using the mixed samples composed of 700 ng of the total RNA from each type of HeLa stable cell line (-/+Dox) and 300 ng of the total RNA from Arabidopsis. The resultant cDNA samples were further used as templates in qPCR reactions and an Arabidopsis unique amplicon (in Actin2 gene) was used as an “internal control” for data quantification (primers used are listed in Response Table 1).

Author response image 3

Download asset Open asset

The changes of the indicated transcripts after 3 days of induction in cells expressing the indicated GDOWN1 variants.

Total RNA samples from HeLa cells expressing GDOWN1 (or its mutant) for mock or 3 days of Dox induction were extracted. 700 ng RNA from the indicated cell lines and 300 ng RNA from Arabidopsis were mixed and used to do reverse transcription. The cDNA samples were further prepared and used in qPCR assays and the Actin2 gene of *Arabidopsis* was used as a reference gene for data normalization.

The results shown that the levels of the highly expressed mRNA species, such as GAPDH and H2B, were significantly reduced in the two cell lines expressing the nuclear-localized GDOWN1(10M) mutants, indicating that the nuclear GDOWN1 may play an important role in repressing Pol II-catalyzed global transcription of mRNAs (and some lncRNAs). The level of the 28S rRNA didn’t change much after 3 days of induction in the cells expressing GDOWN1, 10M or NLS-GDOWN1, but reduced in the cells expressing NLS^-10M to some extent, which might be correlated to the strong reduction of mRNA synthesis. Taken together, these results and the data shown in Figure 6 demonstrate that the massive nuclear accumulation of GDOWN1 results in a dramatic reduction of the global transcription of Pol II.

Author response table 1

The sequences of primers used in RT-qPCR assays.

qPCR Primers	Sequences
AtActin2-F	TGTGCCAATCTAGGAGGGTTT
AtActin2-R	TTTCCCGCTCTGCTGTTGT
hGAPDH-F	GTCAGCCGCATCTTCTTTTG
hGAPDH-R	GCGCCCAATACGACCAAATC
h28S rRNA-F	GACCCGAAAGATGGTGAACTATG
h28S rRNA-R	CGATTTGCACGTCAGGACCG
hH2B-F	CCTGAGCCAGCCAAGTCTGC
hH2B-R	TGCGGCTGCGCTTGCGCTTC

8. Could the authors provide additional quantification of Figure S6B?

We guess the Reviewer meant ‘Figure 6C’ since there was no ‘Figure S6C’ in our original manuscript submitted. However, the bar graph shown on the right in Figure 6C is the quantitative analyses of WB result. We think the error bars are not essential for the quantification analyses of the WB results.

In Figure 6C of our first-time submitted manuscript, we used the Pol II CTD antibody [clone 8WG16, (Abcam)] to indicate the changes of the total Pol II, while the specificity of this antibody was somewhat controversial. Previously, we actually confirmed in our in vitro kinase assays, this antibody preferentially recognized the unphosphorylated Pol II (only one major band shown in the WB data), although it could also recognize the S2P-Pol II when it became dominant in the sample{Cheng, 2007 #5106}. We borrowed that piece of data (the supplemental figure 2 in that JBCpaper) and provided it in Response Figure 4A.. Thus, we made some corrections in the revised manuscript and considered it as the antibody specifically recognizes the un-phosphorylated Pol II. The WB experiments using this antibody were repeated and very similar results were obtained (both sets of data are shown in parallel in Author response image 4). To simplify the data for publication, we only provided one set in Figure 6—figure supplement 1A in the revised manuscript.

Author response image 4

Download asset Open asset

The changes of the level of the un-phosphorylated RPB1 and its nuclear fraction upon the nuclear accumulation of GDOWN1 or the treatment of LMB. A. The target specificity test of the Pol II-CTD antibody, 8WG16. in vitro kinase assays were carried out using the purified P-TEFb as the kinase, and either the purified factors or the HeLa nuclear extract (HNE) as the substrates. The 8WG16 antibody was employed for Pol II detection in the WB experiment. This data was published as the Supplementary Figure 2 in Cheng et al., *JBC*, 2007, 282(30): 21901-21912. B-E. The indicated cell lines were cultured and subjected to cell fractionation and WB analyses. Two independent experiments were performed and the results are shown in B&C, and D&E, respectively. Antibodies used for detection were labeled on the left side. The relative changes of the RPB1 level as a whole and the ratio of its nuclear fraction were shown in the bar graphs on the right. As a very long-life protein, histone H3 was used as the internal control for data normalization. TUB: TUBULIN; the nuclear fraction (N), and the total cell lysate (T) .

To improve the accuracy and reliability of our result for the total Pol II detection, we employed a new Pol II NTD antibody [Anti-RPB1 NTD, clone D8L4Y, (CST)], and the result also demonstrated that the level of Pol II was significantly decreased when the nuclear GDOWN1 was expressed. The original Figure 6C was thus replaced with this new set of data, and became the new Figure 6B in the revised manuscript.

9. The authors should also provide error bars for Figure S6C.

Figure S6B in the previously submitted manuscript demonstrated the controls for

the WB experiment shown in the main Figure 6 to support the effectiveness and purity of our cell fractionation experiment. Both TUBULIN and FBL were used as the reference proteins to represent the cytoplasmic protein and the nuclear protein, respectively, and the purpose of showing these controls was mainly for the qualitative use, but not for the quantitative use. We understand why the Reviewer ask for data quantification here since the level of these controls (especially FBL) apparently reduced quite a lot upon the expression of the nuclear GDOWN1. Because GDOWN1 causes a global transcriptional repression, the effect on the level of each protein actually varies based on the half time of the protein. Therefore, for data quantification, we refer to use histone 3 because histones proteins are the well-known, most stable proteins in cells.

10. As the authors are well aware, Pol II antibodies show sub-optimal specificity. Could the authors provide data from additional Pol II antibodies (e.g., S2P, S5P, RPB3) for Figure S6C?

We thank the reviewer for this good suggestion! We ordered several new Pol II antibodies and provided some new IF and WB data. Please refer our responses to Reviewer 1’s Q8 and to Reviewer 2’s Q8. Overall the results were consistent with our previous data, and the data quantification suggest that the unphospho-Pol II and the S5P-Pol II were more severely and rapidly affected than the S2P-Pol II.

11. In lines 455-456, the authors mention in vitro transcription assays, but do not cite anything. Please cite the study being referred to.

Thanks for your reminding! We cited two corresponding references (Cheng et al., 2012; Jishage et al., 2012).

12. We ask the authors to discuss why GDOWN1 apparently interacts with so many transcriptional regulators in the cytoplasm. Is GDOWN1 regulating these factors by interacting with them in the cytoplasm?

Very good point! We checked the subcellular localization of some transcriptional regulators, such as NELF-E etc., and found they were only present in the nucleus. Thus, the BiFC signals about the ‘GDOWN1•transcriptional regulator’ interactions may only represent their interaction potential in the nucleus instead of the actual occurrence in the cytoplasm between their endogenous counterparts. We hypothesize that under the physiological conditions, GDOWN1 may be transiently and unstably interacting to the newly translated transcriptional regulators in the cytoplasm while it does NOT trap them in the cytoplasm or affect their functions. When both factors are overexpressed in cells in our BiFC assays, GDOWN1 is able to trap a fraction of its nuclear partners in the cytoplasm and the transient interaction can be stabilized by BiFC (the BiFC signal is irreversible because of the covalent bonds set up in between the two split parts of a fluorescent protein upon the BiFC interaction occurs). On the other hand, from the mass spectrometry data and some of our GST-pull down or IP data, we confirmed that GDOWN1 interacted to Pol II components and their assembly factors in the cytoplasm. In addition, we observed that in GDOWN1 KO cells, the nuclear/cytoplasmic ratio of RPB1 decreased (data not shown). Therefore, GDOWN1 may contribute to the assembly and also the nuclear import of Pol II. More thorough analyses are needed to dissect its roles in these processes.

13. The Alphafold structure in supplement figure 4 – 1A is difficult to see. Can the authors increase clarity by changing to a white background and focusing more on the regions they are showing? Please also add the same cartoon that is used throughout the paper to further help clarifications.

Thank you for your suggestion! We have changed the background to white and focused more on the two NESs and the CAS motif. We also added a cartoon to further facilitate clarification according to your advice. The corresponding changes have been made in Figure 4—figure supplement 1A.

14. The authors should also carefully check their manuscript for missing prepositions etc.

Thank you for your kind reminder! We carefully checked and corrected the grammatical mistakes throughout our manuscript.

References

Cheng, B., Li, T., Rahl, P.B., Adamson, T.E., Loudas, N.B., Guo, J., Varzavand, K., Cooper, J.J., Hu, X., Gnatt, A.,etal.(2012). Functional association of Gdown1 with RNA polymerase II poised on human genes. Mol Cell 45, 38-50. Doi: http://doi.org/10.1016/j.molcel.2011.10.022, Accession number: 22244331

Forget, D., Lacombe, A.A., Cloutier, P., Al-Khoury, R., Bouchard, A., Lavallee-Adam, M., Faubert, D., Jeronimo, C., Blanchette, M., and Coulombe, B. (2010). The protein interaction network of the human transcription machinery reveals a role for the conserved GTPase RPAP4/GPN1 and microtubule assembly in nuclear import and biogenesis of RNA polymerase II. Mol Cell Proteomics 9, 2827-2839. Doi: http://doi.org/10.1074/mcp.M110.003616, Accession number: 20855544

Forget, D., Lacombe, A.A., Cloutier, P., Lavallee-Adam, M., Blanchette, M., and Coulombe, B. (2013). Nuclear import of RNA polymerase II is coupled with nucleocytoplasmic shuttling of the RNA polymerase II-associated protein 2. Nucleic Acids Res 41, 6881-6891. Doi: http://doi.org/10.1093/nar/gkt455, Accession number: 23723243

Hu, X., Malik, S., Negroiu, C.C., Hubbard, K., Velalar, C.N., Hampton, B., Grosu, D., Catalano, J., Roeder, R.G., and Gnatt, A. (2006). A Mediator-responsive form of metazoan RNA polymerase II. Proc Natl Acad Sci U S A 103, 9506-9511. Doi: http://doi.org/10.1073/pnas.0603702103, Accession number: 16769904

Jackson, D.A., Pombo, A., and Iborra, F. (2000). The balance sheet for transcription: an analysis of nuclear RNA metabolism in mammalian cells. FASEB J 14, 242-254. Doi: Accession number: 10657981

Jishage, M., Malik, S., Wagner, U., Uberheide, B., Ishihama, Y., Hu, X., Chait, B.T., Gnatt, A., Ren, B., and Roeder, R.G. (2012). Transcriptional regulation by Pol II(G) involving mediator and competitive interactions of Gdown1 and TFIIF with Pol II. Mol Cell 45, 51-63. Doi: http://doi.org/10.1016/j.molcel.2011.12.014, Accession number: 22244332

Jishage, M., Yu, X., Shi, Y., Ganesan, S.J., Chen, W.Y., Sali, A., Chait, B.T., Asturias, F.J., and Roeder, R.G. (2018). Architecture of Pol II(G) and molecular mechanism of transcription regulation by Gdown1. Nat Struct Mol Biol 25, 859-867. Doi:

http://doi.org/10.1038/s41594-018-0118-5, Accession number: 30190596

Nozawa, R.S., Boteva, L., Soares, D.C., Naughton, C., Dun, A.R., Buckle, A., Ramsahoye, B., Bruton, P.C., Saleeb, R.S., Arnedo, M.,etal. (2017). SAF-A Regulates Interphase Chromosome Structure through Oligomerization with Chromatin-Associated RNAs. Cell 169, 1214-1227 e1218. Doi: http://doi.org/10.1016/j.cell.2017.05.029, Accession number: 28622508

https://doi.org/10.7554/eLife.79116.sa2

Article and author information

Author details

Zhanwu Zhu

School of Life Sciences, Lanzhou University, Lanzhou, China

Contribution
Conceptualization, Methodology, Writing – original draft, Project administration

Competing interests
No competing interests declared

"This ORCID iD identifies the author of this article:" 0000-0001-8494-4348
Jingjing Liu

School of Life Sciences, Lanzhou University, Lanzhou, China

Contribution
Formal analysis, Investigation, Visualization, Project administration

Competing interests
No competing interests declared
Huan Feng

School of Life Sciences, Lanzhou University, Lanzhou, China

Contribution
Conceptualization, Validation, Visualization, Methodology, Project administration

Competing interests
No competing interests declared
Yanning Zhang

School of Life Sciences, Lanzhou University, Lanzhou, China

Contribution
Project administration

Competing interests
No competing interests declared
Ruiqi Huang

Cuiying Honors College, Lanzhou University, Lanzhou, China

Contribution
Project administration

Competing interests
No competing interests declared
Qiaochu Pan

Cuiying Honors College, Lanzhou University, Lanzhou, China

Contribution
Project administration

Competing interests
No competing interests declared
Jing Nan

School of Life Sciences, Lanzhou University, Lanzhou, China

Contribution
Supervision, Writing - review and editing

Competing interests
No competing interests declared
Ruidong Miao

School of Life Sciences, Lanzhou University, Lanzhou, China

Contribution
Supervision, Validation, Writing - review and editing

Competing interests
No competing interests declared
Bo Cheng
1. School of Life Sciences, Lanzhou University, Lanzhou, China
2. Key Laboratory of Cell Activities and Stress Adaptations, Ministry of Education, Lanzhou University, Lanzhou, China
Contribution
Conceptualization, Supervision, Funding acquisition, Validation, Investigation, Methodology, Writing – original draft, Writing - review and editing

For correspondence
bocheng@lzu.edu.cn

Competing interests
No competing interests declared

"This ORCID iD identifies the author of this article:" 0000-0002-7060-1616

Funding

The National Natural Science Foundation of China (31771447)

Bo Cheng

The Gansu Provincial Outstanding Graduate Student "Innovation Star" project (2021CXZX-107)

Jingjing Liu

The Gansu Provincial Outstanding Graduate Student "Innovation Star" Project (2022CXZX-096)

Zhanwu Zhu

The Foundation of the Ministry of Education of China, the Fundamental Research Funds for the Central Universities (lzujbky-2020-it16)

Zhanwu Zhu

The Foundation of the Ministry of Education of China, Key Laboratory of Cell Activities and Stress Adaptations Grant (lzujbky-2021-kb05)

Bo Cheng

The National Natural Science Foundation of China (31471233)

Bo Cheng

The National Natural Science Foundation of China (31970624)

Bo Cheng

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

The authors acknowledge Dr. David Price (the University of Iowa) for providing Gdown1 antibody made in sheep (Cheng et al., 2012) and for the helpful comments and discussion from his group; Dr. Ruichuan Chen (Xiamen University), Dr. Qintong Li (Sichuan University), Dr. Yingmei Zhang and Bingtao Niu (Lanzhou University) for providing necessary plasmids, cells and animals; many members of Cheng laboratory for their helpful discussion; Yaojia Wang for her help of revising our model diagram. We also thank the core facilities in the School of Life Sciences, Lanzhou University for providing high-quality instruments and service for our confocal microscopy experiments. We apologize to our colleagues in transcription field whose work was not discussed adequately owing to space constraints. Funding This work was financially supported by the National Natural Science Foundation of China (31771447, 31471233, 31970624 to Bo Cheng), the Foundation of the Ministry of Education Key Laboratory of Cell Activities and Stress Adaptations grant (lzujbky-2021-kb05 to Bo Cheng), the Fundamental Research Funds for the Central Universities (lzujbky-2020-it16 to Zhanwu Zhu), the Gansu Provincial Outstanding Graduate Student "Innovation Star" Project (2022CXZX-096 to Zhanwu Zhu), and the Gansu Provincial Outstanding Graduate Student "Innovation Star" Project (2021CXZX-107 to Jingjing Liu).

Senior Editor

Kevin Struhl, Harvard Medical School, United States

Reviewing Editor

Irwin Davidson, Institut de Génétique et de Biologie Moléculaire et Cellulaire, France

Publication history

Preprint posted: March 30, 2022 (view preprint)
Received: March 31, 2022
Accepted: November 21, 2022
Version of Record published: December 7, 2022 (version 1)

Copyright

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.