Author response:
Public Reviews:
Reviewer #1 (Public review):
Summary:
In this manuscript, Dixit and colleagues investigate the role of FRG1 in modulating nonsense-mediated mRNA decay using human cell lines and zebrafish embryos. They present data from experiments that test the effect of normal, reduced or elevated levels of FRG1 on NMD of a luciferase-based NMD reporter and on endogenous mRNA substrates of NMD. They also carry out experiments to investigate FRG1's influence on UPF1 mRNA and protein levels, with a particular focus on the possibility that FRG1 regulates UPF1 protein levels through ubiquitin-mediated proteolysis of UPF1. The experiments described also test whether DUX4's effect on UPF1 protein levels and NMD could be mediated through FRG1. Finally, the authors also present experiments that test for physical interaction between UPF1, the spliceosome and components of the exon junction complex.
Strengths:
A key strength of the work is its focus on an intriguing model of NMD regulation by FRG1, which is of particular interest as FRG1 is positively regulated by DUX4, which has been previously implicated in subjecting UPF1 to proteosome-mediated degradation and thereby causing NMD inhibition. The data that shows that DUX4-mediated effect on UPF1 levels is diminished upon FRG1 depletion suggests that DUX4's regulation of NMD could be mediated by FRG1.
Weaknesses:
A major weakness and concern is that many of the key conclusions drawn by the authors are not supported by the data, and there are also some significant concerns with experimental design. More specific comments below describe these issues:
(1) Multiple issues lower the confidence in the experiments testing the effect of FRG1 on NMD.
(a) All reporter assays presented in the manuscript are based on quantification of luciferase activity, and in most cases, the effect on luciferase activity is quite small. This assay is the key experimental approach throughout the manuscript. However, no evidence is provided that the effect captured by this assay is due to enhanced degradation of the mRNA encoding the luciferase reporter, which is what is implied in the interpretation of these experiments. Crucially, there is also no control for the reporter that can account for the effects of experimental manipulations on transcriptional versus post-transcriptional effects. A control reporter lacking a 3'UTR intron is described in Barid et al, where the authors got their NMD reporter from. Due to small effects observed on luciferase activity upon FRG1 depletion, it is necessary to not only measure NMD reporter mRNA steady state levels, but it will be equally important to ascertain that the effect of FRG1 on NMD is at the level of mRNA decay and not altered transcription of NMD substrates. This can be accomplished by testing decay rates of the beta-globin reporter mRNA.
We thank the reviewer for raising these points and for the careful evaluation of our experimental approach. Here we provide our response to comment (a) in three parts
Reliance on luciferase-based reporter assays
While luciferase-based NMD reporter assays represent an important experimental component of this study, our conclusions do not rely exclusively on this approach. The reporter-based findings are independently supported by RNA sequencing analyses of FRG1-perturbed cells, which demonstrate altered abundance of established PTC-containing NMD target transcripts. This genome-wide analysis provides an unbiased and physiologically relevant validation of FRG1 involvement in NMD regulation.
All reporter assays presented in the manuscript are based on quantification of luciferase activity, and in most cases, the effect on luciferase activity is quite small.
We respectfully disagree with the comment that the magnitude of the luciferase effects is low. Increased expression of FRG1, which leads to reduced UPF1 levels, results in a ~3.5-fold increase in relative luciferase activity (Fig. 1C), indicating a robust effect. Furthermore, in the in vivo zebrafish model, FRG1 knockout causes a pronounced decrease in relative luciferase activity (Fig. 1H), consistent with elevated UPF1 levels and enhanced NMD activity.
It is also important to note that FRG1 functions as a negative regulator of UPF1; therefore, its depletion is expected to increase UPF1 levels. However, excessive elevation of UPF1 is likely constrained by additional regulatory mechanisms, which may limit the observable effects of FRG1 knockdown or knockout. In line with this, our previous study (1) demonstrated that FRG1 positively regulates multiple NMD factors while exerting an inverse regulatory effect on UPF1. This dual role suggests that FRG1 may act as a compensatory modulator of the NMD machinery, which likely explains the relatively subtle net effects observed in FRG1 knockdown/knockout conditions in vitro (Fig. 1A and 1B). This interpretation is explicitly discussed in the manuscript (Discussion, paragraph para 4).
However, no evidence is provided that the effect captured by this assay is due to enhanced degradation of the mRNA encoding the luciferase reporter, which is what is implied in the interpretation of these experiments. Crucially, there is also no control for the reporter that can account for the effects of experimental manipulations on transcriptional versus post-transcriptional effects. A control reporter lacking a 3'UTR intron is described in Barid et al, where the authors got their NMD reporter from. Due to small effects observed on luciferase activity upon FRG1 depletion, it is necessary to not only measure NMD reporter mRNA steady state levels, but it will be equally important to ascertain that the effect of FRG1 on NMD is at the level of mRNA decay and not altered transcription of NMD substrates. This can be accomplished by testing decay rates of the beta-globin reporter mRNA.
Thank you for your suggestion. We will test decay rates of the beta-globin reporter mRNA.
(b) It is unusual to use luciferase enzymatic activity as a measurement of RNA decay status. Such an approach can at least be justified if the authors can test how many-fold the luciferase activity changes when NMD is inhibited using a chemical inhibitor (e.g., SMG1 inhibitor) or knockdown of a core NMD factor.
We respectfully disagree that the use of luciferase enzymatic activity as a readout for NMD is unusual. Multiple prior studies have successfully employed identical or closely related luciferase-based/fluorescence-based reporters to quantify NMD activity (2–5). Importantly, the goal of our study was not to measure RNA decay kinetics per se, but rather to assess how altered FRG1 levels influence the functional efficiency of the NMD pathway. Given that FRG1 is a structural component of the spliceosome C complex (6) and is previously indirectly linked to NMD regulation (1,7) this approach was well-suited to address our central question.
As suggested by the reviewer, we will also assess luciferase activity following pharmacological inhibition of NMD to further validate the reporter system's responsiveness.
(c) The concern about the direct effect of FRG1 on NMD is further amplified by the small effects of FRG1 knockout on steady-state levels of endogenous NMD targets (Figure 1A and B: ~20% reduction in reporter mRNA in MCF7 cells; Figure 1M, only 18 endogenous NMD targets shared between FRG1_KO and FRG1_KD).
The modest changes observed upon FRG1 loss do not preclude a direct role in NMD. As detailed in our response to comment (a) and discussed in paragraph 4 of the Discussion, limited effects on steady-state levels of endogenous NMD targets are expected given the buffering capacity of the NMD pathway and the contribution of compensatory regulatory mechanisms.
(d) The question about transcriptional versus post-transcriptional effects is also important in light of the authors' previous work that FRG1 can act as a transcriptional regulator.
We agree that distinguishing between transcriptional and post-transcriptional effects is important, particularly in light of our previous work demonstrating that FRG1 can function as a transcriptional regulator of multiple NMD genes (1). Consistent with this, the current manuscript shows that FRG1 influences the transcript levels of UPF1. In addition, we demonstrate that FRG1 regulates UPF1 at the protein level. We therefore conclude that FRG1 regulates UPF1 dually, at both transcriptional and post-transcriptional levels, supporting a dual role for FRG1 in the regulation of NMD.
This conclusion is further supported by prior studies indicating post-transcriptional functions of FRG1. FRG1 is a nucleocytoplasmic shuttling protein(8), interacts with the NMD factor ROD1 (7), and has been identified as a component of the spliceosomal C complex (6). FRG1 has also been reported to associate with the hnRNPK family of proteins (8), which participate in extensive protein–protein interaction networks. Collectively, these observations are consistent with a role for FRG1 in regulating NMD components at multiple levels.
(2) In the experiments probing the relationship between DUX4 and FRG1 in NMD regulation, there are some inconsistencies that need to be resolved.
(a) Figure 3 shows that the inhibition of NMD reporter activity caused by DUX4 induction is reversed by FRG1 knockdown. Although levels of FRG1 and UPF1 in DUX4 uninduced and DUX4 induced + FRG1 knockdown conditions are similar (Figure 5A), why is the reporter activity in DUX4 induced + FRG1 knockdown cells much lower than DUX4 uninduced cells in Figure 3?
We appreciate the reviewer’s comment. Figures 3 and 5A represent independent experiments in which FRG1 knockdown was achieved by transient transfection. As such, variability in transfection efficiency is expected and likely accounts for the quantitative difference. We want to highlight that compared to DUX4_induced lane (Fig. 5A, lane 2), when we knock down FRG1 on the DUX4_induced background, it shows a clear increase in the UPF1 level (Fig. 5A, lane 3). We will add one more replicate to 5 A with better FRG1_KD transfection to the experiment.
(b) In Figure 3, it is important to know the effect of FRG1 knockdown in DUX4 uninduced conditions.
We thank the reviewer for this thoughtful suggestion. The effect of FRG1 knockdown under DUX4-uninduced conditions is presented in Figure 1A, where FRG1 levels are reduced without altering DUX4 expression. In contrast, Figure 3 is specifically designed to assess the rescue effect—namely, how reduction of FRG1 expression under DUX4-induced conditions influences NMD efficiency. Therefore, inclusion of an FRG1 knockdown–only group in Figure 3 was not relevant to the objective of this experiment.
(c) On line 401, the authors claim that MG132 treatment leads to "time-dependent increase in UPF1 protein levels" in Figure 5C. However, upon proteasome inhibition, UPF1 levels significantly increase only at 8h time point, while the change at 12 and 24 hours is not significantly different from the control.
We thank the reviewer for this observation and agree that the statement of a “time-dependent increase in UPF1 protein levels” was inaccurate. A significant increase is observed only at the 8 h time point following MG132 treatment, with no significant changes at 12 h or 24 h. The text will be revised accordingly to reflect Figure 5C.
(3) There are multiple issues with experiments investigating ubiquitination of UPF1:
(a) Ubiquitin blots in Figure 6 are very difficult to interpret. There is no information provided either in the text or figure legends as to which bands in the blots are being compared, or about what the sizes of these bands are, as compared to UPF1. Also, the signal for Ub in most IP samples looks very similar to or even lower than the input.
We agree that the ubiquitin blots in Figure 6 require clearer presentation. In the revised figure, we will annotate the ubiquitin immunoblots to indicate the region corresponding to UPF1 (~140 kDa), which is the relevant molecular weight for interpretation. Because UPF1 is polyubiquitinated, ubiquitinated species are expected to appear as multiple bands rather than a single discrete signal; therefore, ubiquitination was assessed across the full blot. Importantly, interpretation is based on comparisons between UPF1 immunoprecipitated samples within each panel (Fig. 6C–F), rather than between input and IP lanes. For example, in Figure 6 C UPF1 IP FRG1_KD compared to UPF1 IP FRG1_Ex, in Figure 6 D UPF1 IP FRG1_WT compared to UPF1 IP FRG1_KO, in Figure 6 E UPF1 IP FRG1_KO compared to UPF1 IP FRG1_KO+FRG1_Ex, and in Figure 6 F UPF1 IP FRG1_Ex compared to UPF1 IP FRG1_Ex+MG132 TRT.
(b) Western blot images in Figure 6D appear to be adjusted for brightness/contrast to reduce background, but are done in such a way that pixel intensities are not linearly altered. This image appears to be the most affected, although some others have also similar patterns (e.g., Figure 5C).
We thank the reviewer for raising this point. The appearance noted in Figure 6D was not due to non-linear alteration of pixel intensities, but rather resulted from the poor quality of the ubiquitin antibody, which required prolonged exposure times. To address this, we replaced the antibody and repeated the ubiquitin immunoblots shown in Figures 6D, 6E, and 6F.
For Figure 5C, only uniform contrast adjustment was applied for clarity. Importantly, all adjustments were performed linearly and applied to the entire image. Raw, unprocessed images for all blots are provided in the Supplementary Information. Updated versions of Figures 5 and 6 will be included in the revised manuscript.
(4) The experiments probing physical interactions of FRG1 with UPF1, spliceosome and EJC proteins need to consider the following points:
(a) There is no information provided in the results or methods section on whether immunoprecipitations were carried out in the absence or presence of RNases. Each RNA can be bound by a plethora of proteins that may not be functionally engaged with each other. Without RNase treatment, even such interactions will lead to co-immunoprecipitation. Thus, experiments in Figure 6 and Figure 7A-D should be repeated with and without RNase treatment.
We thank the reviewer for this important point. The co-immunoprecipitation experiments shown in Figures 6 and 7A–D were performed in the absence of RNase treatment; this information was inadvertently omitted and will be added to the Methods section and the relevant figure legends. To directly assess whether the observed interactions are RNA-dependent, we will repeat the key co-immunoprecipitation experiments in the presence of RNase treatment and include these results in the revised manuscript.
(b) Also, the authors claim that FRG1 is a "structural component" of EJC and NMD complexes seems to be an overinterpretation. As noted in the previous comment, these interactions could be mediated by a connecting RNA molecule.
We thank the reviewer for this insightful comment. As noted, previous studies have suggested that FRG1 interacts with components of the EJC and NMD machinery. Specifically, Bertram et al. (6) identified FRG1 as a component of the spliceosomal C complex via Cryo-EM structural analysis, and pull-down studies have shown direct interaction between FRG1 and ROD1, a known EJC component (7). These findings support a protein-protein interaction rather than one mediated solely by RNA. To further address the reviewer’s concern, we will perform key co-immunoprecipitation experiments in the presence of RNase treatment to distinguish RNA-dependent from RNA-independent interactions.
(c) A negative control (non-precipitating protein) is missing in Figure 7 co-IP experiments.
We agree that including a non-precipitating protein as a negative control is important, and we will perform the co-IP experiment incorporating this control.
(d) Polysome analysis is missing important controls. FRG1 and EIF4A3 co-sedimentation with polysomes could simply be due to their association with another large complex (e.g., spliceosome), which will also co-sediment in these gradients. This possibility can at least be tested by Western blotting for some spliceosome components across the gradient fractions. More importantly, a puromycin treatment control needs to be performed to confirm that FRG1 and EIF4A3 are indeed bound to polysomes, which are separated into ribosome subunits upon puromycin treatment. This leads to a shift of the signal for ribosomal proteins and any polysome-associated proteins to the left.
As recommended, we will examine the distribution of a spliceosome component across the gradient fractions to assess potential co-sedimentation. Additionally, we will perform a puromycin treatment control to confirm that FRG1 and EIF4A3 are genuinely associated with polysomes.
Reviewer #2 (Public review):
Summary:
In this manuscript, Palo et al present a novel role for FRG1 as a multifaceted regulator of nonsense-mediated mRNA decay (NMD). Through a combination of reporter assays, transcriptome-wide analyses, genetic models, protein-protein interaction studies, ubiquitination assays, and ribosome-associated complex analyses, the authors propose that FRG1 acts as a negative regulator of NMD by destabilizing UPF1 and associating with spliceosomal, EJC, and translation-related complexes. Overall, the data, while consistent with the authors' central conclusions, are undermined by several claims-particularly regarding structural roles and mechanistic exclusivity. To really make the claims presented, further experimental evidence would be required.
Strengths:
(1) The integration of multiple experimental systems (zebrafish and cell culture).
(2) Attempts to go into a mechanistic understanding of the relationship between FGR1 and UPF1.
Weaknesses:
(1) Overstatement of FRG1 as a structural NMD component.
Although FRG1 interacts with UPF1, eIF4A3, PRP8, and CWC22, core spliceosomal and EJC interactions (PRP8-CWC22 and eIF4A3-UPF3B) remain intact in FRG1-deficient cells. This suggests that, while FRG1 associates with these complexes, this interaction is not required for their assembly or structural stability. Without further functional or reconstitution experiments, the presented data are more consistent with an interpretation of FRG1 acting as a regulatory or accessory factor rather than a core structural component.
We thank the reviewer for this clarification. We would like to emphasize that we do not claim FRG1 to be a core structural component of either the spliceosome or the EJC. Consistent with the reviewer’s interpretation, our data indicate that FRG1 deficiency does not disrupt the structural integrity of these complexes. Our intended conclusion is that FRG1 functions as a regulatory or accessory factor in NMD rather than being required for complex assembly or stability. We will carefully revise the manuscript to remove any language that could be interpreted as an overstatement. In addition, we are currently performing further experiments to better define the association of FRG1 with the EJC.
(2) Causality between UPF1 depletion and NMD inhibition is not fully established.
While reduced UPF1 levels provide a plausible explanation for decreased NMD efficiency, the manuscript does not conclusively demonstrate that UPF1 depletion drives all observed effects. Given FRG1's known roles in transcription, splicing, and RNA metabolism, alterations in transcript isoform composition and apparent NMD sensitivity may arise from mechanisms independent of UPF1 abundance. To directly link UPF1 depletion to altered NMD efficiency, rescue experiments testing whether UPF1 re-expression restores NMD activity in FRG1-overexpressing cells would be important.
As suggested, to directly test causality, we will perform rescue experiments to determine whether UPF1 re-expression restores NMD activity in FRG1-overexpressing MCF7 cells.
(3) Mechanism of FRG1-mediated UPF1 ubiquitination requires clarification.
The ubiquitination assays support a role for FRG1 in promoting UPF1 degradation; however, the mechanism underlying this remains unexplored. The relationship between FRG1-UPF1 what role FRG1 plays in this is unclear (does it function as an adaptor, recruits an E3 ubiquitin ligase, or influences UPF1 ubiquitination indirectly through transcriptional or signaling pathways?).
We agree with the reviewer that the precise mechanism by which FRG1 promotes UPF1 ubiquitination remains to be defined. Our ubiquitination assays support a role for FRG1 in facilitating UPF1 degradation; however, whether FRG1 functions directly as an adaptor or E3 ligase, or instead influences UPF1 stability indirectly, is currently unclear. Notably, a prior study by Geng et al. reported that DUX4 expression alters the expression of numerous genes involved in protein ubiquitination, including multiple E3 ubiquitin ligases (9), and FRG1 itself has been reported to be upregulated upon DUX4 expression in muscle cells. We will expand the Discussion to address these potential mechanisms and place our findings in the context of indirect transcriptional or signaling pathways that may regulate UPF1 proteolysis. A detailed mechanistic dissection of FRG1-mediated ubiquitination is beyond the scope of the present study.
(4) Limited transcriptome-wide interpretation of RNA-seq data.
Although the RNA-seq data analysis relies heavily on a small subset of "top 10" genes. Additionally, the criteria used to define NMD-sensitive isoforms are unclear. A more comprehensive transcriptome-wide summary-indicating how many NMD-sensitive isoforms are detected and how many are significantly altered-would substantially strengthen the analysis.
We thank the reviewer for this comment and agree that the current presentation may place a disproportionate emphasis on a limited subset of genes. These genes were selected as illustrative examples from an isoform-level analysis performed using IsoformSwitchAnalyzeR (ISAR) (10); however, we acknowledge that this approach does not fully convey the transcriptome-wide scope of the analysis.
Using quantified RNA-seq data, ISAR was employed to identify significant isoform switches and transcripts predicted to be NMD-sensitive. Isoforms were annotated using GENCODE v47, and NMD sensitivity was assigned based on the established 50-nucleotide rule, as described in the Materials and Methods. To address the reviewer’s concern, we will revise the Results section to include a transcriptome-wide summary derived from the ISAR analysis.
(5) Clarification of NMD sensor assay interpretation.
The logic underlying the NMD sensor assay should be explained more clearly early in the manuscript, as the inverse relationship between luciferase signal and NMD efficiency may be counterintuitive to readers unfamiliar with this reporter system. Inclusion of a schematic or brief explanatory diagram would improve accessibility.
We agree with the reviewer and would provide a schematic as well as the experimental setup diagram to improve accessibility to the readers.
(6) Potential confounding effects of high MG132 concentration.
The MG132 concentration used (50 µM) is relatively high and may induce broad cellular stress responses, including inhibition of global translation (its known that proteosome inhibition shuts down translation). Controls addressing these secondary effects would strengthen the conclusion that UPF1 stabilization specifically reflects proteasome-dependent degradation would be essential.
We acknowledge the reviewer’s concern regarding the relatively high concentration of MG132 used in this study. While proteasome inhibition can indeed induce global translation inhibition, our interpretation is based on the specific stabilization of UPF1 observed under these conditions. Since inhibition of global translation would generally reduce protein levels rather than cause selective accumulation, the observed increase in UPF1 is unlikely to result from translational effects. To address this point, we plan to repeat selected experiments using a lower MG132 concentration to further confirm that UPF1 stabilization reflects proteasome-dependent degradation.
(7) Interpretation of polysome co-sedimentation data.
While the co-sedimentation of FRG1 with polysomes is intriguing, this approach does not distinguish between direct ribosomal association and co-migration with ribosome-associated complexes. This limitation should be explicitly acknowledged in the interpretation.
We acknowledge that polysome co-sedimentation alone cannot definitively distinguish between direct ribosomal binding and co-migration with ribosome-associated complexes. Importantly, our interpretation does not rely solely on this assay; when combined with co-immunoprecipitation and proximity ligation assay results, the data consistently support an association of FRG1 with the exon junction complex. We are also conducting additional experiments with appropriate controls to further validate the specificity of FRG1’s association with ribosomes and to address the possibility of nonspecific co-migration.
(8) Limitations of PLA-based interaction evidence.
The PLA data convincingly demonstrate close spatial proximity between FRG1 and eIF4A3; however, PLA does not provide definitive evidence of direct interaction and is known to be susceptible to artefacts. Moreover, a distance threshold of ~40 nm still allows for proteins to be in proximity without being part of the same complex. These limitations should be clearly acknowledged, and conclusions should be framed accordingly.
We thank the reviewer for highlighting this important point. We agree that PLA indicates close spatial proximity but does not constitute definitive evidence of direct interaction and can be susceptible to artefacts. We will explicitly acknowledge this limitation in the revised manuscript. Importantly, our conclusions are not solely based on PLA data; they are supported by complementary co-immunoprecipitation and polysome co-sedimentation assays, which provide biochemical evidence consistent with an association between FRG1 and eIF4A3.
Reviewer #3 (Public review):
The manuscript by Palo and colleagues demonstrates identification of FRG1 as a novel regulator of nonsense-mediated mRNA decay (NMD), showing that FRG1 inversely modulates NMD efficiency by controlling UPF1 abundance. Using cell-based models and a frg1 knockout zebrafish, the authors show that FRG1 promotes UPF1 ubiquitination and proteasomal degradation, independently of DUX4. The work further positions FRG1 as a structural component of the spliceosome and exon junction complex without compromising its integrity. Overall, the manuscript provides mechanistic insight into FRG1-mediated post-transcriptional regulation and expands understanding of NMD homeostasis. The authors should address the following issues to improve the quality of their manuscript.
(1) Figure 7A-D, appropriate positive controls for the nuclear fraction (e.g., Histone H3) and the cytoplasmic fraction (e.g., GAPDH or α-tubulin) should be included to validate the efficiency and purity of the subcellular fractionation.
We thank the reviewer for the suggestion. We will include appropriate positive controls for the nuclear fraction (Histone H3) and the cytoplasmic fraction (GAPDH or α-tubulin) in Figure 7A–D to validate the efficiency and purity of the subcellular fractionation.
(2) To strengthen the conclusion that FRG1 broadly impacts the NMD pathway, qRT-PCR analysis of additional core NMD factors (beyond UPF1) in the frg1⁻/⁻ zebrafish at 48 hpf would be informative.
We appreciate the reviewer’s insightful comment. We will perform qRT-PCR analysis of additional core NMD factors in the frg1⁻/⁻ zebrafish at 48 hpf to further strengthen the conclusion that FRG1 broadly impacts the NMD pathway.
(3) Figure labels should be standardized throughout the manuscript (e.g., consistent use of "Ex" instead of mixed terms such as "Oex") to improve clarity and readability.
We thank the reviewer for noticing the inconsistency. We will ensure that all figure labels are standardized throughout the manuscript (e.g., using “Ex” consistently) to improve clarity and readability.
(4) The methods describing the generation of the frg1 knockout zebrafish could be expanded to include additional detail, and a schematic illustrating the CRISPR design, genotyping workflow, and validation strategy would enhance transparency and reproducibility.
We appreciate the reviewer’s suggestion and will expand the Methods section to provide additional detail on the generation of the frg1 knockout zebrafish. A schematic illustrating the CRISPR design, genotyping workflow, and validation strategy will also be included to enhance transparency and reproducibility.
(5) As FRG1 is a well-established tumor suppressor, additional cell-based functional assays under combined FRG1 and UPF1 perturbation (e.g., proliferation, migration, or survival assays) could help determine whether FRG1 influences cancer-associated phenotypes through modulation of the NMD pathway.
We thank the reviewer for this thoughtful and constructive suggestion. While FRG1 is indeed a well-established tumor suppressor, incorporating additional cell-based functional assays under combined FRG1 and UPF1 perturbation would significantly broaden the scope of the current study. The present work is focused on elucidating the molecular relationship between FRG1 and the NMD pathway. Investigation of downstream cancer-associated phenotypes represents an important and interesting direction for future studies, but is beyond the scope of the current manuscript.
(6) Given the claim that FRG1 inversely regulates NMD efficacy via UPF1, an epistasis experiment such as UPF1 overexpression in an FRG1-overexpressing background followed by an NMD reporter assay would provide stronger functional validation of pathway hierarchy.
We agree with the reviewer’s suggestion. To strengthen the functional validation of the proposed pathway hierarchy, we will perform an epistasis experiment by overexpressing UPF1 in an FRG1-overexpressing background and assess NMD activity using an established NMD reporter assay. The results of this experiment will be included in the revised manuscript.
References
(1) Palo A, Patel SA, Shubhanjali S, Dixit M. Dynamic interplay of Sp1, YY1, and DUX4 in regulating FRG1 transcription with intricate balance. Biochim Biophys Acta Mol Basis Dis. 2025 Mar;1871(3):167636.
(2) Sato H, Singer RH. Cellular variability of nonsense-mediated mRNA decay. Nat Commun. 2021 Dec 10;12(1):7203.
(3) Baird TD, Cheng KCC, Chen YC, Buehler E, Martin SE, Inglese J, et al. ICE1 promotes the link between splicing and nonsense-mediated mRNA decay. eLife. 2018 Mar 12;7:e33178.
(4) Chu V, Feng Q, Lim Y, Shao S. Selective destabilization of polypeptides synthesized from NMD-targeted transcripts. Mol Biol Cell. 2021 Dec 1;32(22):ar38.
(5) Udy DB, Bradley RK. Nonsense-mediated mRNA decay uses complementary mechanisms to suppress mRNA and protein accumulation. Life Sci Alliance. 2022 Mar;5(3):e202101217.
(6) Bertram K, El Ayoubi L, Dybkov O, Agafonov DE, Will CL, Hartmuth K, et al. Structural Insights into the Roles of Metazoan-Specific Splicing Factors in the Human Step 1 Spliceosome. Mol Cell. 2020 Oct 1;80(1):127-139.e6.
(7) Brazão TF, Demmers J, van IJcken W, Strouboulis J, Fornerod M, Romão L, et al. A new function of ROD1 in nonsense-mediated mRNA decay. FEBS Lett. 2012 Apr 24;586(8):1101–10.
(8) Sun CYJ, van Koningsbruggen S, Long SW, Straasheijm K, Klooster R, Jones TI, et al. Facioscapulohumeral muscular dystrophy region gene 1 is a dynamic RNA-associated and actin-bundling protein. J Mol Biol. 2011 Aug 12;411(2):397–416.
(9) Geng LN, Yao Z, Snider L, Fong AP, Cech JN, Young JM, et al. DUX4 activates germline genes, retroelements, and immune mediators: implications for facioscapulohumeral dystrophy. Dev Cell. 2012 Jan 17;22(1):38–51.
(10) Vitting-Seerup K, Sandelin A. The Landscape of Isoform Switches in Human Cancers. Mol Cancer Res MCR. 2017 Sep;15(9):1206–20.
