Increased abundance of the nuclear long noncoding RNA (lncRNA) Malat1 drives metastatic progression and is a strong predictor of poor patient prognosis. Although the mechanism that stabilizes Malat1 through processing of its 3′ terminus is well-characterized, the pathways governing its turnover remain poorly understood. Here, we show that upon exit from mitosis, Malat1 localizes to the cytoplasm, where it is degraded during early G1, resetting its abundance at the start of each cell cycle. Mechanistically, we demonstrate that Malat1 turnover is mediated by a translation- and Smg1-dependent decay pathway and triggered by redundant elements. Importantly, failure to reset Malat1 levels in early G1, due to decay inhibition or in the absence of progression through mitosis, results in Malat1 accumulation. These findings uncover a cell cycle–dependent mechanism that harnesses the translation machinery to regulate Malat1 abundance and identify cancer cell dormancy as a potential mechanism underlying the widespread overexpression of Malat1 in cancer.
Introduction
Widespread transcription of long noncoding RNAs (lncRNAs) is a defining feature of mammalian genomes (Carninci et al., 2005). Over the past two decades, hundreds of lncRNAs spanning a broad range of expression levels and subcellular localization patterns have been identified (Mattick et al., 2023). Although the physiological functions and mechanisms of most lncRNAs remain incompletely understood, a subset distinguished by evolutionary sequence conservation, high expression levels, and defined subcellular localization patterns has emerged as critical regulators of diverse biological processes (Winkler and Dimitrova, 2021). Among these, metastasis-associated lung adenocarcinoma transcript 1 (Malat1) has attracted significant attention due to its strong evolutionary conservation and enrichment in nuclear speckles (Hutchinson et al., 2007).
Malat1 was originally identified as a transcript overexpressed in metastatic lung adenocarcinoma, where elevated expression was strongly associated with poor patient prognosis (Ji et al., 2003; Muller-Tidow et al., 2004). Subsequent studies in murine cancer models demonstrated that increased Malat1 abundance is not merely correlated with aggressive disease but is sufficient to drive metastasis through epigenetic reprogramming of tumor cells and remodeling of the tumor microenvironment (Arun et al., 2016; Gutschner et al., 2013; Martinez-Terroba et al., 2024). These findings established aberrant upregulation of Malat1 as a potent oncogenic driver of tumor progression.
Interestingly, Malat1 is ubiquitously and highly expressed in both normal and transformed cells (Arun et al., 2020). Its exceptionally high abundance reflects both robust transcription and a long transcript half-life. Transcript stability is conferred by an unconventional RNase P–dependent 3′-end processing mechanism that cleaves nascent Malat1 to generate a highly conserved expression and nuclear retention element (ENE) triple-helix structure, which protects the 3′ terminus from exonucleolytic degradation (Brown et al., 2014; Wilusz et al., 2008; Wilusz et al.,2012).
Although the mechanism that confers stability to Malat1 has been extensively characterized, the pathways responsible for the turnover of ENE-stabilized Malat1 remain poorly understood. Several studies have suggested that Malat1 is subject to miRNA-mediated regulation (Leucci et al., 2013; Macias et al., 2012) or interacts with components of the nonsense-mediated decay (NMD) and Staufen-mediated decay (SMD) pathways (Xiao et al., 2024; Zund et al., 2013). However, these observations have been difficult to reconcile with the predominantly nuclear localization of Malat1, as these pathways function primarily in the cytoplasm. More recently, studies have identified RNA degradation pathways that are antagonized by the Malat1 ENE (Che et al., 2025), yet the molecular pathways responsible for the turnover of ENE-protected Malat1 have remained elusive.
To elucidate the mechanisms that regulate Malat1 levels and may contribute to its dysregulation in cancer, we investigated Malat1 abundance and subcellular localization during the cell cycle. Our findings reveal an unexpected cell cycle–coordinated mechanism for the turnover of mitotically inherited Malat1 and explain the frequent dysregulation of its abundance in cancer.
Results and discussion
Malat1 abundance correlates negatively with proliferation rate
We hypothesized that the frequent overexpression of Malat1 in cancer may be linked to a common cancer cell feature, such as high proliferation rate. To test this, we compared Malat1 levels in two independent early-passage primary WT E13.5 mouse embryonic fibroblast (MEF) lines (WT1 and WT2), which are highly proliferative, to late-passage MEFs, which are largely senescent (Fig. S1, A and B). Unprocessed and total Malat1 were measured using primer sets specific to the 3′ (E) and 5′ (A–D) regions of the transcript, respectively (Fig. 1 A). Contrary to our hypothesis, we observed that the levels of Malat1 progressively rose from passage 2 to 10 (Fig. 1 B and Fig. S1 C). This increase was predominantly due to accumulation of processed Malat1, as the levels of unprocessed Malat1 accounted for <0.5% of total Malat1 (Fig. S1 D). We also observed that processed Malat1 levels increased approximately two to threefold in response to serum deprivation and upon acute exposure to the DNA damage-inducing agent, doxorubicin, in primary WT MEFs (Fig. 1 C). We concluded that Malat1 abundance is inversely correlated to proliferation status.
Cell cycle–dependent regulation of Malat1 abundance
Visualization of Malat1 by single-molecule RNA FISH (smRNA-FISH) (Raj et al., 2008) in WT1 and WT2 MEFs at passage 10 revealed uniform intercellular nuclear staining that overlapped with the nuclear speckle marker, Srrm2, detected by immunofluorescence (IF) (Fig. 1, D and E; and Fig. S1 E). In contrast, primary WT MEFs at passage 4 displayed markedly greater cell-to-cell variation in nuclear Malat1 pattern and intensity, independent of speckles (Fig. 1, D and E; and Fig. S1 E). In particular, early G1 cells—identified by the close proximity of the daughter nuclei and the presence of cytoplasmic Malat1, previously observed in cells exiting mitosis (Blower et al., 2023; Hutchinson et al., 2007)—consistently exhibited lower nuclear Malat1 signal than cells at other stages of interphase (Fig. 1, D and F). This observation suggested that increased heterogeneity of nuclear Malat1 staining in proliferating cells may reflect differences in cell cycle stage distribution.
To independently assess Malat1 abundance across the cell cycle, we introduced the GFP-tagged Geminin component of the fluorescent ubiquitination-based cell cycle indicator (FUCCI) system, which is stabilized in the S/G2/M stages of the cell cycle (Fig. 1 G) (Sakaue-Sawano et al., 2008). Consistent with cell cycle–dependent regulation, qRT-PCR and smRNA-FISH both showed significantly lower Malat1 abundance in GFP-negative G1 cells than in GFP-positive S/G2/M cells (Fig. 1, G and H). Conversely, arrest in late G2 or mitosis using the CDK1 inhibitor, RO-3306, or the microtubule polymerization inhibitor, Nocodazole (Noc), respectively, increased Malat1 levels approximately two to threefold relative to asynchronous cells (Fig. 1 I). Importantly, similar cell cycle–dependent pattern was observed for human MALAT1 in retinal pigment epithelium (RPE) and patient-derived EGFR-mutant lung adenocarcinoma (PC9) cells (Fig. S1, F–H and Fig. 1 J). Together, these findings demonstrate that the abundance of murine Malat1 and human MALAT1 fluctuates during the cell cycle, reaching its lowest levels in G1 and highest levels in G2/M.
Malat1 is degraded in the postmitotic early G1 cytoplasm
To determine why Malat1 abundance is lower in G1 than G2/M cell cycle stages, we examined its localization and levels during the M-to-G1 transition by smRNA-FISH. Consistent with previous reports, we observed that Malat1/MALAT1 dissociates from chromatin during mitosis and disperses throughout the nucleoplasm in MEFs, RPE, and PC9 cells (Fig. 2 A; and Fig. S1, I and J) (Blower et al., 2023; Sharp et al., 2020). Also consistent with prior reports, upon mitotic exit, inherited Malat1/MALAT1 was found to remain in the cytoplasm of daughter cells as hundreds of discrete foci corresponding to individual RNA molecules (Fig. 2 A; and Fig. S1, I and J) (Blower et al., 2023; Hutchinson et al., 2007). In contrast, we observed that the nuclei of early G1 cells contain, on average, only 4.1 Malat1 foci (Fig. 2 A, arrows, and Fig. S1, I and J), of which 67% co-colocalize with Neat1, a lncRNA transcribed in genomic proximity to Malat1, indicating that they primarily represent sites of nascent Malat1 transcription (Fig. 2 B). Thus, in early G1 cells, which comprise 8.5 ± 4.3% and 12.5 ± 2.6% of asynchronous MEFs and RPE populations, respectively, mitotically inherited Malat1 remains in the cytoplasm, while newly synthesized Malat1 accumulates in the nucleus. As cells progress through interphase, nuclear Malat1 signal was observed to increase and adopt its characteristic speckle pattern, while cytoplasmic Malat1 foci were no longer detectable (Fig. 1 D, Fig. 2 A, and Fig. S1, F–J).
We confirmed these observations using cell cycle synchronization. p53-deficient MEFs were arrested in mitosis with Noc, collected by mitotic shake-off, and synchronously released into the next cell cycle (Fig. 2, C and D). qRT-PCR analysis of subcellular fractions at 1 h postmitotic release (pmr) revealed significant enrichment of cytoplasmic Malat1 relative to asynchronous cells, followed by a significant depletion at 2 and 4 h pmr (Fig. 2 E). These results confirmed that mitotically inherited Malat1 transiently localizes to the cytoplasm in early G1.
To determine the fate of cytoplasmic Malat1, we analyzed Malat1 levels by qRT-PCR at different time points pmr. Strikingly, we observed that Malat1 levels dropped by over 80% at 2 h pmr compared with M phase–arrested cells, indicating that the majority of inherited Malat1 is degraded in early G1 (Fig. 2 F). The sharp reduction of Malat1 levels was followed by gradual accumulation of Malat1 at consecutive time points pmr, consistent with new transcription as cells progressed through the cell cycle (Fig. 2 F). Together, these findings demonstrated that mitotically inherited Malat1 is degraded in the postmitotic cytoplasm, explaining its lower abundance in G1 compared with G2/M.
Moreover, Malat1 turnover appeared to be limited to early G1, as Malat1 levels did not significantly change over a 16-h treatment with the transcription inhibitor actinomycin D (ActD), which suppresses mitotic progression (Fig. 2 G). In contrast, when we used a 5-ethynyluridine (EU) pulse to label newly synthesized RNA in the absence of cell cycle perturbations, we determined that the half-life of the Malat1 transcript is 6.5 h, consistent with the cell cycle length in p53-deficient MEFs (Fig. 2 G).
Translation- and Smg1-dependent decay of cytoplasmic Malat1
As MALAT1 has been reported to associate with translating ribosomes (Wilusz et al., 2012) and interact with components of the NMD and SMD pathways (Xiao et al., 2024; Zund et al., 2013), we asked whether Malat1 is degraded by a translation-dependent cytoplasmic decay pathway in early G1. To test this, we treated asynchronous cells or cells released from mitotic arrest with the translation inhibitor cycloheximide (Chx) for 2 h (Fig. 3 A). FACS analysis confirmed that cells progressed from mitosis into G1 despite inhibition of new protein synthesis, consistent with prior reports (Fig. S2 A) (Verbin and Farber, 1967). Importantly, Chx treatment increased Malat1/MALAT1 levels ∼1.8-fold in asynchronous MEFs and RPE cells (Fig. 3, B and C; and Fig. S2 B) and rescued postmitotic Malat1/MALAT1 degradation following mitotic release (Fig. 3, B and C; and Fig. S2 B). These findings revealed a translation-dependent mechanism for Malat1/MALAT1 degradation in early G1. In contrast, Neat1_2/NEAT1_2, the other mammalian lncRNA terminating in an ENE protective structure, did not undergo significant degradation in early G1, suggesting that this mechanism is specific to Malat1/MALAT1 (Fig. S2, C and D).
Treatment with Smg1 11e, a selective inhibitor of the shared NMD/SMD kinase Smg1 (Gopalsamy et al., 2012), similarly increased Malat1 abundance approximately twofold at 2 h pmr, supporting a role for NMD or SMD in Malat1 degradation (Fig. 3 D). Moreover, at 2 h pmr, cytoplasmic Malat1 was detected in 100 ± 0% and 94.4 ± 1.2% of MEFs, treated with Chx or Smg1 11e, respectively, compared with 14.0 ± 1.4% of untreated cells (Fig. 3 E). Together, these findings demonstrate that cytoplasmic Malat1 is degraded in early G1 through a translation- and Smg1-dependent pathway.
Malat1 degradation is triggered by redundant elements
To identify the region of Malat1 that triggers decay, we complemented Malat1 knockout (KO) MEFs, immortalized by p53 inactivation, with constructs expressing doxycycline (Doxy)-inducible empty vector (KO+EV), full-length Malat1 (KO+FL), or Malat1 deletion mutants lacking 1.6-kb regions A–D (KO+ΔA–D) (Fig. 1 A and Fig. S3 A). All Malat1 constructs contained region E to enable posttranscriptional processing of Malat1 into an ENE-terminating transcript (Fig. 1 A and Fig. S3 A). Analyses of KO+FL and KO+ΔA–D cells revealed that deletion of region A did not affect speckle localization or cell cycle progression (Fig. S3, B and C) but led to significantly increased abundance of ΔA relative to endogenous and FL Malat1, indicating a role for region A in Malat1 destabilization (Fig. S3 D). Notably, region A has been reported to strongly associate with translating ribosomes and encode a small ORF (sORF), making it a candidate for an NMD trigger (Fig. S3 A) (Wilusz et al., 2012; Xiao et al., 2024).
We observed that ΔA transcripts were 5.5-fold more abundant than FL at 2 h pmr, indicating that ΔA is not degraded to the same extent as FL Malat1 in early G1 (Fig. 4 A). However, we also observed that over 50% of ΔA was degraded at 2 h pmr relative to non-released ΔA-expressing cells and that this degradation was reversed by Chx, indicating that ΔA is susceptible to translation-dependent degradation (Fig. 4 A). We concluded that region A contributes to but is not required for Malat1 degradation.
We also examined the relative contribution of sORFs in region A by generating a Δ9xATG Malat1 mutant, lacking initiation codons of 9 sORFs, including previously described M1 sORF (Fig. S3 E) (Xiao et al., 2024). Analysis of asynchronous cells revealed that Δ9xATG Malat1 was ∼1.5-fold more abundant compared with FL but significantly less abundant compared with ΔA (Fig. S3 E). This observation further supports the conclusion that degradation of Malat1 is dependent on redundant elements, with elements in region A and the 9 sORFs being necessary, but not sufficient, for degradation.
Cytoplasmic persistence and nuclear import of undegraded Malat1
Quantification of nuclear Malat1 abundance in early G1 cells revealed significantly higher mean signal intensity in ΔA than FL-expressing KO cells (Fig. 4. B and C). We reasoned that this could reflect either prolonged cytoplasmic persistence of ΔA Malat1 into later stages of the cell cycle or nuclear reimport of undegraded cytoplasmic transcripts. To distinguish between these possibilities, we transiently induced FL and ΔA Malat1 expression with Doxy, then synchronized cells in mitosis, and released them into the next cell cycle in the absence of Doxy (Fig. 4 D). Consistent with degradation of mitotically inherited Malat1 at the onset of each cell cycle, FL and ΔA Malat1 levels were reduced by 80% at 2 h pmr relative to non-released cells, and this reduction persisted at 4 h pmr in the absence of new transcription (Fig. 4. E and F). Notably, ΔA Malat1 remained ∼10-fold more abundant than FL at both time points, indicating incomplete degradation in the absence of region A, as reported above (Fig. 4, A and E). Quantification of the fraction of cells with cytoplasmic Malat1 by smRNA-FISH revealed that this increase was partly due to prolonged cytoplasmic persistence of ΔA Malat1 compared with FL (87.2 ± 9.0% vs. 36.8 ± 12.7% at 2 h pmr; 55.1 ± 14.0% vs. 3.9 ± 0.4 at 4 h pmr, Fig. 4 F).
Interestingly, we observed that nuclear Malat1 signal increased between 2 and 4 h pmr in both FL- and ΔA-expressing cells (Fig. 4 G). Because no new Malat1 transcription occurred during this period, this finding indicated that a fraction of cytoplasmic Malat1 escapes degradation and is reimported into the nucleus. Together, these analyses show that incomplete degradation of Malat1 in early G1 can result in both prolonged cytoplasmic persistence, particularly in ΔA mutants, and nuclear re-import of undegraded transcripts.
Malat1 accumulates in dormant cancer cell populations
Dormant cancer cell subpopulations that are transcriptionally active but non-proliferative are an established reservoir for metastasis, tumor relapse, and drug resistance (Klein, 2020; Russo et al., 2024). Because we determined that passage through mitosis is required for the cytoplasmic localization and early G1 degradation of Malat1, we wondered whether proliferative dormancy contributes to the overexpression of Malat1 in cancer. We analyzed MALAT1 abundance in EGFR-mutant PC9 cells treated with the tyrosine kinase inhibitor osimertinib (Osi), which enter a slow-cycling, drug-tolerant persister (DTP) state, where priming for drug resistance evolution occurs (Starble et al., 2025). Consistent with our model, MALAT1 levels progressively rose in DTP cells but declined once cells acquired Osi resistance and resumed proliferation in the continued presence of Osi (Fig. 5 A). These findings indicate that reduced proliferation is a direct mechanism driving MALAT1 overexpression.
Taken together, our work reveals a model for the dynamic regulation of Malat1 abundance that also explains its dysregulation in cancer. Whereas previous studies focused on Malat1 accumulation and stability in the interphase nucleus, we show that mitotically inherited Malat1 localizes to the cytoplasm, where it is degraded in the first 2 h of G1, resetting its abundance at the onset of each cell cycle (Fig. 5 B). We further demonstrate that non-proliferative cell states, including quiescence, senescence, cell cycle arrest, and dormancy, accumulate high levels of Malat1 because they do not progress through mitosis and therefore fail to execute this early G1 degradation program (Fig. 5 B). These findings raise the possibility that changes in proliferation state alone, without genetic or epigenetic rewiring, may contribute to the widespread overexpression of Malat1 observed across cancers. Given that both Malat1 overexpression (Arun et al., 2016; Gutschner et al., 2013; Martinez-Terroba et al., 2024) and tumor cell dormancy (Hinterleitner et al., 2026) promote metastatic progression, it will be important to determine whether they functionally converge.
Mechanistically, our findings reveal an unexpected twist on the central dogma, where the translation machinery is co-opted to mediate the turnover of one of the most abundant and stable mammalian lncRNAs. Our work clarifies that the previously reported ribosomal occupancy at the 5′ end of Malat1 (Wilusz et al., 2012) likely reflects engagement with ribosomes in the postmitotic cytoplasm. Recent studies have also identified translated Malat1 sORFs, including the neuron-specific 33-aa M1 peptide (Rai et al., 2026; Xiao et al., 2024). This could result from specialized trafficking of Malat1 to the cytoplasm during neuronal interphase or from reduced postmitotic decay efficiency in neurons (Palou-Marquez and Supek, 2025), which prolongs cytoplasmic Malat1 residence and enables M1 translation. These two models can be distinguished by examining whether quiescent neurons harbor cytoplasmic Malat1 and express the M1 peptide.
Our work implicates an Smg1-dependent decay mechanism, such as NMD or SMD, in Malat1 turnover. Indeed, we observed cytoplasmic retention of Malat1 in over 90% of early G1 cells treated with an Smg1 inhibitor compared with 14% in control samples. How Malat1 engages these decay mechanisms remains unclear. NMD has been proposed to act primarily through the “exon-junction complex (EJC) model,” where the translating ribosomes rely on the EJC to distinguish premature from normal termination codons (Kim et al., 2001; Le Hir et al., 2001; Lykke-Andersen et al., 2001). While Malat1 is generally not considered to be spliced, a recent report has indicated that Malat1 can be spliced in the absence of association with the splicing regulator U2AF2 (Grammatikakis et al., 2026, Preprint). It will be interesting to determine whether cytoplasmic Malat1 is bound by U2AF2 in early G1. As an alternative, the “Faux 3′UTR model” has been proposed for NMD of unspliced mRNAs with unusually long unstructured 3′UTRs that limit productive interactions between the terminating ribosome at the PTC and the poly(A)-binding protein, PABPC1 (Amrani et al., 2004). In support of this model, processed Malat1 is a 7-kb noncoding transcript that is unlikely to effectively recruit Pabpc1 as it lacks a canonical poly(A) tail. Additionally, Malat1 was recently reported to co-localize with Staufen in neurons (Xiao et al., 2024), suggesting a potential role for the SMD pathway (Kim et al., 2005). Although region A and nine sORFs were found to contribute to Malat1 destabilization, they were not required, suggesting redundant elements, such as additional sORFs or secondary structures. Indeed, Malat1 contains 41 sORFs (>20 aa), 33 located outside region A, as well as multiple reported structured RNA elements (McCown et al., 2019; Monroy-Eklund et al., 2023; Xiao et al., 2026, Preprint). Lastly, our work does not exclude the complementary or redundant role of RNA interference (RNAi) in the regulation of cytoplasmic Malat1 in early G1. Further work will define the relative contribution of various functional elements and mechanisms to the regulation of Malat1 turnover.
Materials and methods
Cell lines and constructs
All cells were maintained at 37°C in a humidified incubator with 5% CO2. Primary Malat1 WT and KO MEFs were isolated from E13.5 littermate embryos from timed matings of Malat1 heterozygous mice (Malat1+/−), generously provided by Dr. David Spector (CSHL, Cold Spring Harbor, NY, USA) (Zhang et al., 2012). Primary MEF lines were established and passaged in DMEM (Gibco) supplemented with 15% fetal bovine serum, 50 U ml−1 penicillin-streptomycin, 2 mM L-glutamine, 0.1 mM nonessential aa, and 0.055 mM 2-mercaptoethanol. Experiments in primary MEFs were performed between passages 2 and 10. To immortalize MEFs, p53 was inactivated by CRISPR-Cas9 mutagenesis using p53-targeting sgRNA, sgp53, listed in Table S1, expressed from BRD005 (U6-sgRNA-spCas9-RFP) (a gift from the Broad Institute, MIT). Briefly, MEFs were transiently transfected with sgp53-expressing BRD005 using the Fast-forward Attractene (301005; Qiagen) protocol and passaged to select for immortalized clones. Immortalized MEFs were maintained in 10% fetal bovine serum, 50 U ml−1 penicillin-streptomycin, 2 mM L-glutamine, and 0.1 mM nonessential aa. hTERT RPE-1 cells, a gift from Dr. David Breslow (Yale University, New Haven, CT, USA), were maintained in DMEM:F12 (Gibco), supplemented with 10% fetal bovine serum, 50 U ml−1 penicillin-streptomycin, and 2 mM L-glutamine. PC9 cells, a gift from Dr. Katerina Politi (Yale University, New Haven, CT, USA), were maintained in RPMI (Gibco) supplemented with 10% fetal bovine serum and 50 U ml−1 penicillin-streptomycin.
Malat1 PiggyBac expression vectors were prepared by cloning commercially synthesized Malat1 sequences (IDT) downstream of the Doxy-inducible promotor in the PB-TRE-EGFP-EF1a-rtTA vector (Addgene, 104454) after excision of the EGFP sequence. FL Malat1 contained sequences A–E, while ΔA, ΔB, ΔC, and ΔD deletion mutant constructs lacked regions A through D, respectively, and Δ9xATG contained 9 ATG->AAG mutations (NR_002847.3: T210A, T271A, T567A, T742A, T1512A, T1718A, T1901A, T2206A, and T3144A). The 9 sORFs were selected based on length exceeding 20 aa, presence of strong Kozak sequences, and location in the 5′ half of Malat1, which has a strong ribosomal footprint. All constructs were verified by restriction digest and Sanger sequencing. Malat1 FL, mutant, and deletion constructs were co-transfected with PiggyBac transposase vector into Malat1 KO MEFs using Attractene transfection reagent (301005; Qiagen) according to the Fast-Forward Protocol described by the manufacturer. Selection in 2 µg/ml puromycin (P9620; Millipore-Sigma) was initiated at 48 h after transfection and continued for 3–4 days. Exogenous Malat1 expression was initiated through treatment with 1 µg/ml doxycycline (D1822; Millipore-Sigma).
FUCCI mAG-hGem(1/110) (Sakaue-Sawano et al., 2008) was cloned into MSCV-puro (68469; Addgene) and transfected via calcium phosphate into the Phoenix-ECO retrovirus producer cell line (CRL-3214; ATCC). Cells were infected by three rounds of retroviral infection performed 8–16 h apart by passing viral-containing supernatant through a 0.45-µM filter (28145-505; VWR), supplemented with 4 µg/ml polybrene (107689; Millipore-Sigma). Cells were selected in 2 µg/ml puromycin (P9620; Millipore-Sigma).
Drug treatments
Cells were incubated for the indicated timepoints with the following drugs: 50 ng/ml Nocodazole (AC358240100; Thermo Fisher Scientific), 8 μM RO-3306 (15149; Cayman Chemical Company), 1 μg/ml Doxycycline (D1822; Millipore-Sigma), 0.5 μM Doxorubicin (D1515; Millipore-Sigma), 5 μg/ml Actinomycin D (C988H61; Millipore-Sigma), 50 μg/ml Cycloheximide (C974S42; Millipore-Sigma), 5 μM hSmg1 inhibitor 11e (HY-124760; MedChemExpress), and 100 nM Osimertinib (AZD9291, 16237; Cayman Chemical Company).
Cell cycle synchronization
Mitosis-enriched cell populations were generated by treatment with 50 ng/ml Nocodazole (AC358240100; Thermo Fisher Scientific) for 4 h, followed by manual shake-off of mitotic cells in PBS. To initiate recovery from mitotic arrest, cells were washed once in PBS and replated in normal growth medium for the indicated release time. At indicated time points, adherent and non-adherent cells were harvested for RNA and protein analyses or fixed for smRNA-FISH.
smRNA-FISH
smRNA-FISH was performed according to the manufacturer recommendations using commercially available probes, specific to murine Malat1 (Q570, SMF-3008-1, and Q670; Biosearch Technologies, custom made, same probes as SMF-3008-1), murine Neat1-5′ (Q570 SMF-3009-1; Biosearch Technologies), and human MALAT1 (Q570, SMF-2035-1; Biosearch Technologies). Briefly, cells were grown on coverslips and fixed for 10 min in 4% methanol-free formaldehyde (Thermo Fisher Scientific) diluted in RNase-free 1xPBS (Thermo Fisher Scientific) at RT, followed by 1xPBS washes. Cells were dehydrated overnight at 4°C in 70% EtOH diluted in DEPC-H2O and stored in 70% EtOH for up to a week at 4°C. Coverslips were transferred to a hybridization chamber and equilibrated for 5 min in Wash Buffer A (SMF-WA1-60; Biosearch Technologies) prepared with formamide (Millipore Sigma) according to the manufacturer’s instructions. Cells were incubated overnight at 30°C with the indicated probes diluted 1:50 in Hybridization solution (SMF-HB1-10; Biosearch Technologies). The next day, cells were washed two times for 30 min at 30°C in Wash Buffer A, incubated in Wash Buffer B (SMF-WB1-20; Biosearch Technologies) for 5 min at RT, mounted in antifade reagent (Vectashield Mounting medium with DAPI, Vector Laboratories), and sealed with nail polish.
Combined IF-smRNA-FISH
Cells, grown on coverslips, were fixed in 4% methanol-free formaldehyde (Thermo Fisher Scientific) diluted in RNase-free 1xPBS (Thermo Fisher Scientific) for 10 min at RT, washed twice with 1xPBS, permeabilized with 0.5% Triton X-100 diluted in 1xPBS for 10 min at RT, and washed three times with 1xPBS. Next, cells were blocked for 10 min in 4% BSA (Jackson ImmunoResearch) in 1xPBS prior to incubation with anti-Srrm2 primary antibody (1:500, NBP2-55691; Novus) diluted in 1xPBS for 2 h at RT. Following three 1xPBS washes, cells were incubated with Alexa Fluor 488 AffiniPure Goat Anti-Rabbit IgG (H+L) secondary antibody (1:500, Jackson ImmunoResearch) diluted in 1xPBS for 1 h at RT, followed by two 1xPBS washes. Cells were fixed in 4% formaldehyde for 10 min at RT, followed by two 1xPBS washes, prior to performing smRNA-FISH, as described above.
Image acquisition and analysis
smRNA-FISH and IF-smRNA-FISH imaging was performed on an Axio Imager 2 microscope system (Zeiss) with a Plan-Apochromat 63×1.4 oil DIC objective lens (Zeiss) and a Plan-Apochromat 40× 1.3 oil DIC (UV) VIS-IR objective lens (Zeiss) using Immersol 518F Immersion Oil (Zeiss). Images were captured with an Axiocam 503 mono camera using Zen 2.6 Pro software (Zeiss) with the following settings: (Binning—2 × 2; filter [target]: exposure—Cy5 [Malat1 Q670]: 1,000 ms; rhodamine [Malat1, MALAT1, and Neat1-5′ Q570]: 1,000 ms; EGFP (Srrm2): 200 ms; DAPI [DNA]: 20 ms). Quantification of Malat1/MALAT1 nuclear mean intensity (MI) was performed in Zen software using the measurement tool within a circle with 14.6 μm diameter in a.u. MI of an equivalent area was subtracted as a background in each field. Cells were scored as positive for cytoplasmic Malat1 if they had 50 or more Malat1 cytoplasmic foci, representative of single molecules. This quantification was appropriate as the presence or absence of cytoplasmic Malat1 molecules was largely binary. 50 was selected as the cutoff to account for the number of specific foci commonly captured in focus in a single plane. Images were exported as TIFF files and edited in Adobe Photoshop.
RNA isolation and qRT-PCR
Total RNA was isolated with the RNeasy Mini Kit (Qiagen). 1 μg of total RNA was reverse transcribed using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems). For ActD experiments, on-column digestion with DNase was performed using the RNase-Free DNase Set (Qiagen, 79254). SYBR Green PCR master mix (Applied Biosystems) was used for quantitative PCR in triplicate reactions with primers listed in Table S1. Relative RNA expression levels were calculated using the ddCt method relative to Gapdh or Atp5f1e (for ActD experiments) and normalized to control samples, when indicated.
Immunoblotting
Cells were collected, counted, and lysed in 2 × Laemmli buffer (100 mM Tris-HCl pH 6.8, 200 mM DTT, 3% SDS, 20% (wt/vol) glycerol, and 0.01% bromophenol blue) at 1 × 104 cells/μl. Samples were heated at 95°C for 7 min and passed through an insulin syringe. Protein from 1 × 105 cells was separated on a 10% SDS-polyacrylamide gel and transferred to a nitrocellulose membrane (Bio-Rad). After blocking (5% milk, PBST), membranes were incubated overnight at 4°C in primary antibodies anti-cyclin B1 (1:1,000, 4138S; Cell Signaling Technology) and anti-Hsp90 (1:1,000, 610419; BD Biosciences), then 1 h at RT in secondary antibody (1:10,000, Peroxidase AffiniPure Donkey Anti-Rabbit IgG [H+L] and Peroxidase AffiniPure Goat Anti-Mouse IgG [H+L], Jackson ImmunoResearch). Protein bands were visualized using Pierce ECL Prime Western Blotting Substrate (Thermo Fisher Scientific).
Cell cycle analysis by FACS
Trypsinized cells were pelleted, washed in PBS once, fixed by dropwise resuspension in ice-cold 70% ethanol, and incubated overnight at 4°C. Fixed cells were stained in PBS with 0.5% BSA (A7638; Millipore Sigma), 25 µg/ml propidium iodide (P3566; Thermo Fisher Scientific), and 100 µg/ml RNase A (19101; Qiagen) at RT for 4 h. Propidium iodide signal was detected in the PE Texas Red channel with an LSRFortessa X-20 flow cytometer (BD Biosciences).
Subcellular fractionation
Cytoplasmic and nuclear RNA were isolated by subcellular fractionation. Briefly, cell pellets were resuspended in 175 μl of ice-cold Buffer RLN (50 mM Tris-HCl, pH 8.0, 140 mM NaCl, 1.5 mM MgCl2, and 0.5% [vol/vol] Nonidet P-40) containing 1,000 U/ml RNase inhibitor (M0314S; NEB) and 1 mM DTT and incubated on ice for 5 min to lyse the plasma membrane. The lysate was centrifuged, and the supernatant containing the cytoplasmic fraction was transferred to a new tube for cytoplasmic RNA isolation. The remaining pellet containing intact nuclei was washed with ice-cold Buffer RLN to remove residual cytoplasmic contamination and processed for nuclear RNA isolation using the RNeasy Mini Kit (Qiagen). Relative RNA expression levels were calculated using the ddCt method for Malat1, Rn7sl (cytoplasmic marker), or Kcnq1ot1 (nuclear marker) relative to Gapdh.
RNA stability assay
RNA stability assay was performed using the Click-iT Nascent RNA Capture Kit (C10365; Invitrogen) according to the manufacturer’s protocol. 200,000 MEFs were plated in 35-mm dishes. The following day, cells were incubated for 1 h with 0.5 mM 5-EU, then media containing EU was removed, and cells were rinsed and then grown in fresh media until harvest at 0, 4, 8, and 24 h following EU removal. Total RNA was extracted using TRIzol and nascent RNA capture was performed according to the manufacturer’s recommendations using the following specifications: EU biotinylation was performed using 1 μg total RNA and 0.5 mM biotin azide. Biotinylated RNA was precipitated overnight at −80°C and 0.5 μg RNA was captured using 50 μl of Dynabeads MyOne Streptavidin T1 magnetic beads. RT-PCR was performed with the SuperScript VILO cDNA synthesis kit (11756050; Invitrogen) using bead-bound RNA, and qRT-PCR was done using the SsoAdvanced Universal SYBR Green Supermix (1725274; Bio-Rad) according to the manufacturer’s instructions.
Quantification and statistical analysis
Data are represented as mean ± SD of the indicated number of replicates. Paired or unpaired t test, as indicated, was performed to test for statistical significance between pairs of samples and indicated as follows: ns, not significant; *P < 0.05; **P < 0.01; ***P < 0.001.
Online supplemental material
Fig. S1 describes copy number quantification of Malat1 abundance and demonstrates similar cell cycle–dependent Malat1/MALAT1 localization patterns in murine and human cell lines, related to Figs. 1 and 2. Fig. S2 provides supporting information for translation inhibition studies, described in Fig. 3, as well as demonstrates that Neat1_2/NEAT1_2 is not subject to the cell cycle– and translation-mediated regulatory mechanism that controls Malat1/MALAT1 abundance. Fig. S3 shows characterization of the localization and expression levels of the Malat1 WT and mutant complementation constructs used in Fig. 4. Table S1 shows primers and sgRNA used in this study.
Declaration of generative AI and AI-assisted technologies
Authors declare that generative AI and AI-assisted technologies were not used in this manuscript.
Data availability
The data are available from the corresponding author upon request.
Acknowledgments
We are grateful to David Spector (CSHL, Cold Spring Harbor, NY, USA) for sharing the Malat1 knockout mice.
This work was funded by the National Institutes of Health (NIH) R01CA262286 (N. Dimitrova) and by a Developmental Research Program Grant from the Yale SPORE in Lung Cancer (NIH P50CA196530 [N. Dimitrova]). R.V. Gupta was supported by the Predoctoral Training Program in Genetics (NIH T32GM007499). E.A. Dangelmaier is supported by an NIH Predoctoral Individual National Research Service Award (NIH F31CA306101).
Author contributions: Leah M. Plasek-Hegde: conceptualization, formal analysis, investigation, methodology, supervision, and writing—review and editing. Yeolhoe Kim: conceptualization, formal analysis, investigation, methodology, visualization, and writing—review and editing. Rahul V. Gupta: resources. Emily A. Dangelmaier: formal analysis and investigation. Sigrid Nachtergaele: investigation, and writing—review and editing. Nadya Dimitrova: conceptualization, formal analysis, funding acquisition, investigation, methodology, project administration, resources, supervision, visualization, and writing—original draft, review, and editing.
References
Author notes
Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. S. Nachtergaele reported, “S.N. holds equity and is a member of the scientific advisory board of RNA Connect, Inc.” No other disclosures were reported.


