Lamin A safeguards the m6A methylase METTL14 nuclear speckle reservoir to prevent cellular senescence

Abstract Mutations in LMNA gene are frequently identified in patients suffering from a genetic disorder known as Hutchison–Gilford progeria syndrome (HGPS), providing an ideal model for the understanding of the mechanisms of aging. Lamin A, encoded by LMNA, is an essential component of the subnuclear domain‒nuclear speckles; however, the functional significance in aging is unclear. Here, we show that Lamin A interacts with the m6A methyltransferases, METTL3 and METTL14 in nuclear speckles. Lamin A deficiency compromises the nuclear speckle METTL3/14 reservoir and renders these methylases susceptible to proteasome‐mediated degradation. Moreover, METTL3/14 levels progressively decline in cells undergoing replicative senescence. Overexpression of METTL14 attenuates both replicative senescence and premature senescence. The data reveal an essential role for Lamin A in safeguarding the nuclear speckle reservoir of the m6A methylase METTL14 to antagonize cellular senescence.


| INTRODUC TI ON
Lamin A is a major component of nuclear lamina and multiple subnuclear domains; it is first synthesized as a precursor (prelamin A) and then processed by ZMPSTE24 for maturation (Pendas et al., 2002). A de novo G608G mutation in LMNA gene causes a 50 amino acid truncation of prelamin A. The mutated form of prelamin A, also known as Progerin, is considered to be pathogenic in Hutchinson-Gilford progeria syndrome (HGPS) (Eriksson et al., 2003). Mice lacking Zmpste24 accumulate prelamin A and develop premature aging features that resemble HGPS (Pendas et al., 2002). Cultured fibroblasts derived from Zmpste24 −/− mice and HGPS patients undergo accelerated senescence and apoptosis, attributable to genomic instability, hyperactivation of the p53 pathway and epigenetic changes, etc. (Bridger & Kill, 2004;Krishnan et al., 2011;Liu et al., 2005Liu et al., , 2013Varela et al., 2005). Although it is extremely rare, HGPS offers an ideal model to understand the mechanisms of aging and age-related diseases.
Lamins serve as an anchor for proteins that shuttle between the nuclear lamina and the nucleoplasm, and as a platform for RNA metabolism (Ho, Jaalouk, Vartiainen, & Lammerding, 2013;Scaffidi & Misteli, 2008). Lamin A, Lamin B1, and Lamin C1/C2 and multiple proteins that are involved in RNA modification are detectable in the nuclear speckles, also called interchromatin granule clusters (IGCs) (Mintz, Patterson, Neuwald, Spahr, & Spector, 1999). Indeed, N6-methyladenosine (m 6 A) methylases METTL3 and METTL14 are constantly observed in the nuclear speckles, which form a methyltransferase complex that methylates adenosine residues at the N6 position. This methylation event is the most prevalent internal post-transcriptional modification to occur on mammalian mRNAs.
METTL3 and METTL14 interact with each other via an extensive hydrogen bond network, wherein METTL3 primarily functions as the catalytic core while METTL14 serves as the RNA-binding platform (Wang et al., 2017).
Here, we aimed to understand the functional relevance of Lamin A-containing nuclear speckles in cellular senescence. We confirmed Lamin A as an essential component of the nuclear speckles. Further, we found that Lamin A interacts with METTL3/14, thus to ensure their proper localization in the nuclear speckles and protein stability. Most importantly, METTL14 overexpression prevents cell senescence.
Indeed, we noticed that almost all SC-35 foci were co-localized with Lamin A and more than half of them were co-localized with METTL3/14 in wild-type (WT) mouse embryonic fibroblasts (MEFs) ( Figure S1a). In primary human skin fibroblasts (HSFs) co-transfected with FLAG-METTL3/14 and DsRed-Lamin A (LA-Red), most of the FLAG-METTL3/14-positive speckles also expressed LA-Red ( Figure S1b). We thus reasoned that Lamin A might interact with METTL3/14 in the nuclear speckles. To test the hypothesis, we did co-immunoprecipitation (Co-IP) in HEK293 cells overexpressing F I G U R E 1 Lamin A interacts with METTL3/14. (a-b) Co-immunoprecipitation (Co-IP) and Western blot analysis of the interactions between FLAG-Lamin A and HA-METTL14 (a); and FLAG-Lamin A HA-METTL3 (b) in HEK293 cells. (c-d) Co-IP and Western blot analysis of the endogenous interaction between Lamin A and METTL14 (c) and Lamin A and METTL3 (d) in MEFs FLAG-Lamin A with HA-METTL14 or HA-METTL3. As shown, HA-METTL14 and HA-METTL3 were present in anti-FLAG-Lamin A immunoprecipitates from HEK293 cells (Figure 1a,b). In addition, endogenous Lamin A was co-immunoprecipitated with METTL3 ( Figure 1c) and METTL14 (Figure 1d) in MEFs. Together, these results indicate that Lamin A interacts with METTL3/14 in the nuclear speckles.

| Lamin A safeguards the proper nuclear localization of METTL3/14
To examine whether the nuclear speckle localization of METTL3/14 requires Lamin A, we first performed immunofluorescence microscopy to assess the localization of METTL3/14 in Lmna −/− MEFs. We DsRed-Lamin A (DsRed-LA) into Lmna −/− MEFs. As shown, normal METTL14 localization was restored by Lamin A overexpression ( Figure 2b). By contrast, this rescue effect was not achieved when these MEFs were reconstituted with a mutant Lamin A protein, in which the Lamin A nuclear localization sequence (NLS) was deleted.
We next examined the localization of METTL14 in Zmpste24 −/− MEFs, in which prelamin A accumulates. Interestingly, the number of METTL14 foci that co-localized with SC35 expression was sig-

| Lamin A abnormalities destabilize METTL3/14
We noticed that the levels of METTL3/14 were substantially de- Given that METTL14 expression seemed to decrease in Lamin A-deficient cells, we presumed that WT Lamin A might enhance the protein stability of METTL14. Indeed, after the treatment of cycloheximide (CHX) together with or without MG132, which inhibited protein synthesis and proteasome activity, respectively, we found that the protein degradation rate of METTL14 was accelerated MEFs compared with that in WT MEFs (Figure 3f,g and Figure S3).

Moreover, the ubiquitination level of METTL14 was increased in
Zmpste24 −/− MEFs ( Figure S4). Collectively, the data suggest that Lamin A protects METTL14 protein from proteasomal degradation and the absence of Lamin A or the presence of prelamin A induces protein degradation of METTL14.

| A decline in METTL14 expression accelerates cellular senescence
During our analyses, we noticed that the protein levels of METTL3/14 declined with increasing passage of HGPS and healthy control cells ( Figure 3d,e, P18 vs P15). As Progerin expression increases with passaging in human cells (Scaffidi & Misteli, 2006), we therefore asked whether METTL3/14 decline is a general feature of cellular senescence. Indeed, METTL3/14 expression levels were progressively reduced with passaging in both HSFs and MEFs (Figure 4a,b).
To examine whether METTL3/14 decline is a trigger or a consequence of senescence, we knocked down METTL3/14 and exam-

| METTL14 overexpression ameliorates cellular senescence
In our final analyses, we asked whether METTL14 overexpression could ameliorate senescence. To this end, we generated lentivirus particles overexpressing METTL14 (lenti-M14) and used them to overexpression significantly alleviated nuclear membrane abnormalities (Figure 5d,e) and restored H3K9me3 levels (Figure 5f,g). We then tested the effect of AAV2/9-mediated METTL14 overexpression in Zmpste24 −/− MEFs at P6. Here, we found that upon METTL14 infection, p21 and p16 levels decreased (Figure 5h) and the SA-β-Gal level reduced (Figure 4i,j). Of note, the overexpression of METTL14 merely affected the apoptotic level of HGPS cells compared with the empty vector control ( Figure S6). Thus, METTL14 overexpression seems to delay replicative senescence in normal fibroblasts and can rescue premature senescence in progeria.

| DISCUSS ION
Lamin A is extensively investigated as a component of nuclear lamina, and mounting evidences support its essential roles in regulating various nuclear activities such as nuclear shape, cargo transportation, chromatin structure, and gene expression (Prokocimer et al., 2009). Lamins are also constantly identified in multiple subnuclear domains; however, the related functions are rather overlooked.
We found that Lamin A interacts with the m 6 A methyltransferase METTL3/14 in nuclear speckles. Lamin A deficiency compromises the nuclear speckle METTL3/14 reservoir and renders them susceptible to proteasome-mediated degradation. Thus, Lamin A dictates a novel role of nuclear speckles and we propose a schematic model: Lamin A anchors METTL14 to the nuclear speckles via direct interaction, thus to safeguard METTL14 protein stability; without Lamin A, METTL14 loses its anchorage to the nuclear speckles and undergoes accelerated degradation ( Figure S7). Future work remains to investigate how the Lamin A-containing nuclear speckles protect METTL3/14 from proteasomal degradation. One possibility is that the compromised interaction between METTL3/14 and Lamins likely leaves it vulnerable by exposing ubiquitination sites.
The accumulation of abnormal Lamin A−prelamin A/Progerin disrupts nuclear lamina integrity, which is thought to be the cause of HGPS. Based on this rationale, treatment with farnesyltransferase inhibitors (FTIs) restore the nuclear structure abnormalities and alleviate premature aging in progeria murine models (Fong et al., 2006).
The first clinical trial of FTI-lonafarnib for HGPS treatment is now ongoing. However, only a subtle improvement of health status, reduction in mortality rate, and extension of life span (about 1-2 years) were expected (Gordon et al., 2018). Thus, parallel mechanisms and additional treatment strategies are needed. Here, we for the first time demonstrate that abnormal Lamin A not only jeopardizes the nuclear shape, but also compromises subnuclear domains exemplified by the nuclear speckles. Consequently, METTL14 is significantly downregulated in HGPS cells and the late passage of normal cells.
Most importantly, METTL14 overexpression attenuates senescence in both normal cells and HGPS cells. Indeed, it has been shown that the alternative splicing of LMNA exon 11 may also occur in normally senescent cells, which causes the accumulation of Progerin (Scaffidi & Misteli, 2006). It is thus speculated that the accumulation of Progerin might render METTL14 susceptible to proteasomal degradation during replicative senescence.
In summary, Lamin A safeguards the proper localization of METTL3/14 in nuclear speckles. Abnormal Lamin A mislocalizes METTL3/14 from the nuclear speckles and induces proteasome-mediated degradation. The data reveal a novel mechanism by which Lamin A maintains the METTL3/14 reservoir and highlights the importance of m 6 A RNA methylation in senescence.

| Cell transfection and RNA interference
Plasmid transfections were performed using Lipofectamine ® 3000 (Invitrogen) and siRNA transfections were performed using Lipofectamine ® RNAiMAX (Invitrogen), following the manufacturer's instructions. Specific custom siRNAs were synthesized by GenePharma. The siRNA sequences are listed in Table S1.

| Immunofluorescence staining
Cells were fixed with 4% paraformaldehyde on ice and permeabilized with PBS containing 0.1% Triton X-100 for 15 min. Then, the cells were blocked with 1% bovine serum in PBS for 30 min at room temperature. The coverslips were first incubated with primary antibody overnight at 4°C and then detected by Alexa Fluor-conjugated secondary antibodies (Alexa 488, Alexa 594; 1:500, Life Technologies) for 1 h at room temperature in the dark and mounted with DAPIcontaining mounting medium. Images were captured under an immunofluorescence confocal microscope (Zeiss). A representative image for each condition is shown.

| RNA isolation and quantitative RT-PCR
Total RNA was isolated with TRIzol ® reagent (Invitrogen) from WT or transiently transfected cells. DNase I-treated total RNA was used to synthesize cDNA using an iScript cDNA Synthesis Kit (Bio-Rad) according to the manufacturer's protocols. Real-time quantitative polymerase chain reaction (RT-qPCR) was performed using 2 × SYBR Green Mix (Takara) in a Bio-Rad detection system. Each sample was run in triplicate, and the gene expression levels were normalized to β-actin. The primer sequences are listed in Table S1.
Briefly, cells were seeded in 6-well plates, fixed with 4% paraformaldehyde at room temperature for 15 min, and then washed with 1× PBS, stained with 2 ml freshly prepared 1× β-gal detection solution at 37°C overnight in the dark. The cells were washed twice with 1× PBS, overlaid with 70% glycerol/PBS, and images were captured under a microscope. The number of blue-stained cells was counted from >250 randomly chosen cells. The data were analyzed by twotailed Student's t test.

| m 6 A dot blotting
mRNA was denatured at 75°C for 5 min, spotted, and cross-linked to a positively charged nylon membrane 2× in UV Stratalinker with 1800 μJ/cm 2 at 254 nm. The membrane was probed with m 6 A antibody (No. 202003, 1:1000; Synaptic Systems) overnight at 4°C and then incubated with goat anti-rabbit IgG-HRP (1:10,000 dilution) in 10 ml dilution buffer for 1 h at room temperature with gentle shaking. After washing four times, immunoreactive products were visualized using an Enhanced Chemiluminescence Kit (Pierce) and a Bio-Rad imaging system.

| Statistical analyses
Statistical analyses were performed in GraphPad Prism 7 (GraphPad Software Inc., USA) using a two-tailed t test. Statistical significance was considered as *p < 0.05, **p < 0.01, and ***p < 0.001. The data represent the means ± SEM of three independent experiments.

ACK N OWLED G M ENTS
This study was supported by grants from the National Natural Science Jessica Tamanini (Shenzhen University and ETediting) for editing the manuscript prior to submission.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

AUTH O R CO NTR I B UTI O N S
J.Z. and Y.A. conducted the experiments; Z.Z., Y.M., L.P., and Y.J.
provided technical support and analyzed data; and J.Z., Z.W., and B.L. designed the study and wrote the manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.