PMC

Coimmunoprecipitation via anti-METTL16 or IgG selection in HeLa nuclear-enriched lysates followed by Western blot analysis of METTL16, DHX9, ILF3, ILF2, and NPM1. RNase A treatment was performed before immunoprecipitation () to test for an RNA-mediated interaction with METTL16. DHX9, ILF3, ILF2, and NPM1 mostly interact with METTL16 in an RNA-independent manner.

Quantitation of MALAT1 and METTL16 molecules in cells. (A) Northern blot of full-length (FL) MALAT1 in HeLa and HEK293T cells next to a known amount of an in vitro-transcribed MALAT1 fragment (nucleotides 8121–8355) that includes the triple helix. We calculate ∼3,500 and ∼600 molecules of MALAT1 in HeLa and HEK293T cells, respectively. (B) Western blot of METTL16 in HeLa and HEK293T cells next to a known amount of recombinant METTL16-myc-DDK protein. There are ∼500,000 and ∼750,000 molecules of METTL16 in HeLa and HEK293T cells, respectively.

Sequences and structures of RNAs used for selection of binding proteins (). (A–C) RNAs 9–11 were annealed to a 3′-biotinylated oligonucleotide (blue nucleotides) and coupled to streptavidin-coated DynaBeads. RNA10 has WT MALAT1 ENE+A sequence, whereas RNA11 has a GC-to-AA mutation, which interferes with triplex formation. (D and E) WT and Mut2 RNAs were used to elute proteins bound to RNAs 9–11. The biotin modification is in brown, nucleotide mutations are in red, and nucleotide interactions are represented as in .

PLA evidence for the association of METTL16 with the 3′ end of MALAT1. PLA was performed in HeLa cells by using DNA probes that were sense or antisense to nucleotides 8216–8260 of MALAT1, which are immediately upstream of the triple-helical ENE+A structure (nucleotides 8263–8355), and anti-METTL16 or anti-NPM1 antibodies. Nuclei were stained with DAPI, and white arrows point to the green PLA signal. (Scale bars: 10 µm.) To increase visibility of the PLA signal, Bottom shows a zoom-in view of the regions outlined in pink dashed boxes.

METTL16 binds to MALAT1 ENE+A RNA in vitro. (A) Native gel-shift assays were performed by incubating HEK293T lysate or recombinant C-terminal myc-DDK–tagged human METTL16 with 5′-[³²P]–labeled MALAT1 ENE+A RNA (1 nM; ) in the absence or presence of 50 nM WT or Mut2 () competitor RNA. (B) Cell lysate was prepared from HEK293T cells treated with either negative control siRNA (siControl) or siRNA against METTL16 (siMETTL16). Western blots (Lower) confirmed knockdown; GAPDH was used as a loading control. The siControl or siMETTL16 lysate was then incubated with 5′-[³²P]–labeled MALAT1 ENE+A RNA (1 nM) before separation on native PAGE (Upper). (C) A supershift assay was performed by adding IgG control or anti-METTL16 antibodies to a binding mixture containing HEK293T lysate and 5′-[³²P]–labeled MALAT1 ENE+A RNA.

Identification of a MALAT1 ENE+A-interacting protein. (A) An ∼70-kDa protein in HEK293T cell lysate UV cross-links to radiolabeled MALAT1 ENE+A RNA. WT or Mut2 competitor RNAs were added to identify protein binding specific for the triple helix. (B) Depiction of RNAs 9–11 used in the pulldown scheme in C. A 3′-biotinylated oligonucleotide was annealed to its complement (RNA9) or to its complement fused to the 5′ end of a WT (RNA10) or a triplex-disrupting mutant (RNA11) MALAT1 ENE+A RNA (see for details). The dsRNA added for pulldown is in blue, the biotin-streptavidin interaction is a brown circle, the ENE+A structure is black, and the mutated region is in red. (C) Pulldown strategy used to isolate the MALAT1 ENE+A RNP. Folded RNAs were coupled to streptavidin-coated DynaBeads before incubation in HEK293T cell lysate. Bound proteins were eluted with either the WT or Mut2 MALAT1 ENE+A RNA shown in and subjected to mass spectrometric analysis. (D) Eluted proteins were resolved by using SDS/PAGE and silver stained. Protein bands (*) identified by mass spectrometry are listed to the right.

METTL16 associates with MALAT1 in vivo. (A) Subcellular localization of METTL16 in HeLa cells using immunofluorescence. Nuclei were stained with DAPI. (Scale bars: 20 µm.) (B) HeLa nuclear-enriched lysates were subjected to anti-METTL16 or control IgG selection and subsequently verified by Western blot. (C) RT-qPCR quantified the amounts of endogenous RNAs relative to input for both IgG (light gray bars) and METTL16 (dark gray bars) IPs. Error bars represent SD from three biological replicates. (D) Schematic diagram of the intronless β-globin (βΔ1,2) plasmid constructs with the ENE (green), A-rich tract (purple), and tRNA-like sequence (orange, representing mascRNA or menRNA) from MALAT1 or MENβ. β-globin expression is driven by the CMV promoter, the RNase P cleavage site is indicated by an arrowhead, and BGH pA is the bovine growth hormone polyadenylation signal. (E) HEK293T cells were transfected with plasmids expressing βΔ1,2-MALAT1 ENE+A or βΔ1,2-MENβ ENE+A reporter mRNA, and METTL16 was immunoprecipitated from prepared nuclear-enriched lysates. (F) RT-qPCR quantified the expression level of the βΔ1,2 reporter mRNA (normalized to the NeoR transfection control) in the inputs for βΔ1,2-MALAT1 ENE+A relative to the βΔ1,2-MENβ ENE+A reporter mRNA (Left). (Right) RT-qPCR quantified the fold enrichment in the METTL16 IPs for the βΔ1,2-MALAT1 ENE+A relative to the βΔ1,2-MENβ ENE+A reporter mRNA.

EMSAs reveal assembly of a MALAT1 ENE+A RNP and protein binding dependent on triple helix formation. (A) Schematic diagrams for the ENE+A structures from MALAT1, MENβ, and KSHV PAN RNAs. Nucleotides in the U-rich internal loop are green, and nucleotides in the A-rich tract are purple. Nucleotide interactions are represented as follows: single dash for Watson–Crick base pair, black circle for noncanonical base pair, double dash for A-minor interaction, and Leontis–Westhof’s circle-square notation for Hoogsteen interaction. The blue asterisk marks A8290. (B and C) EMSAs were performed by adding increasing amounts of HEK293T (B) or HeLa (C) cell lysate to a 5′-[³²P]–labeled MALAT1, MENβ, or KSHV PAN ENE+A RNA. Observed RNPs are arbitrarily labeled RNP1–3 to the left. (D) Schematic diagram of the mutant MALAT1 ENE+A RNAs used as unlabeled competitors in E. Note, the 5′-UCUAGCU₂A₁₃GCA₄-3′ sequence is deleted (Δ) in Mut1. (E) Competitive gel-shift assays were performed by incubating HEK293T lysate and 5′-[³²P]–labeled MALAT1 ENE+A RNA in the presence of 10- or 50-fold molar excess of the appropriate unlabeled competitor RNA. (F) The relative amount of RNP3 formation was quantified for each competitor, with RNP3 in the absence of a competitor set arbitrarily to 1. Fraction bound is the average of at least three independent experiments; error bars represent SD.