Hakai is required for stabilization of core components of the m6A mRNA methylation machinery

Abstract

N⁶-methyladenosine (m⁶A) is the most abundant internal modification on mRNA which influences most steps of mRNA metabolism and is involved in several biological functions. The E3 ubiquitin ligase Hakai was previously found in complex with components of the m⁶A methylation machinery in plants and mammalian cells but its precise function remained to be investigated. Here we show that Hakai is a conserved component of the methyltransferase complex in Drosophila and human cells. In Drosophila, its depletion results in reduced m⁶A levels and altered m⁶A-dependent functions including sex determination. We show that its ubiquitination domain is required for dimerization and interaction with other members of the m⁶A machinery, while its catalytic activity is dispensable. Finally, we demonstrate that the loss of Hakai destabilizes several subunits of the methyltransferase complex, resulting in impaired m⁶A deposition. Our work adds functional and molecular insights into the mechanism of the m⁶A mRNA writer complex.

Fig. 1. Hakai interacts with MACOM components.

a Table indicating the names of MAC and MACOM components in Vertebrates and Drosophila. b Schematic of the Hakai short and long protein isoforms depicting the RING domain (pink), the Zn-finger (grey) and the HYB domain (black). The green domain indicates the alternative region of the second exon that is included in isoforms 311 aa and 473 aa. c Immunostaining of Myc-tagged Hakai short and long protein isoforms overexpressed in S2R+ cells. GFP-tagged Barentsz protein served as a cytoplasmic marker. DAPI staining is shown in blue. The short Hakai isoform localizes strictly to the cytoplasm, whereas the long isoform localizes to both cellular compartments with the enrichment in the nucleus. Scale bars, 10 μm. d Identification of proteins interacting with GFP-Hakai-long in S2R+ cells based on label-free analysis of two replicate experiments analyzed by quantitative MS-based proteomics. HAKAI immunoprecipitates alongside of all MACOM components. MACOM component proteins are highlighted in red. A complete list of quantified proteins can be found in Supplementary Data 2. e, f Co-immunoprecipitation experiments were carried out with lysates prepared from S2R+ cells transfected with Myc-tagged Hakai (long isoform) and HA-tagged Nito or Fl(2)d. In control lanes, S2R+ cells were transfected with Myc alone and an identical HA-containing protein. Extracts were immunoprecipitated with anti-Myc antibody and immunoblotted using anti-Myc and anti-HA antibodies. Two percent of input was loaded. The same experiment was repeated in the presence of RNase T1. Nito and Fl(2)d interact with Hakai in an RNase-independent manner. Blots shown are representative of one biological replicate. g Yeast-two-hybrid assay to investigate Hakai interaction with Nito and Fl(2)d. Proteins were cloned in yeast expression vectors and fused with either Gal4-DNA binding domain (BD) or Gal4-DNA activation domain (AD). Indicated combinations of vectors were co-expressed in yeast and empty vectors encoding only activation or binding domain were used as control (Ctr). Recovered colonies were spotted on plates lacking Leucine and Tryptophan (-Leu, -Trp) as well as selection plates lacking amino acids Leucine, Tryptophan and Histidine (-Leu, -Trp, -His). AD-Hakai long isoform interacts with BD-Fl(2)d and BD-Nito. Source data for Western blots and the yeast-two-hybrid assay are provided as a Source Data file.

Fig. 2. Virilizer acts as a scaffold between Hakai and Fl(2)d.

a Schematic representation of proteins and protein fragments used for Drosophila co-immunoprecipitation assays. b–e Co-immunoprecipitation experiments were carried out with lysates prepared from S2R+ cells transfected with Myc-GFP-tagged Virilizer (full length or fragments) and HA-tagged Hakai-long (b), Myc-GFP-tagged Hakai-long (full length or fragments) and HA-tagged Virilizer (c), Myc-GFP-tagged Virilizer (full length or fragments) and HA-tagged Fl(2)d (d), Myc-GFP-tagged Fl(2)d (full length or fragments) and HA-tagged Virilizer (e). In control lanes, S2R+ cells were transfected with Myc-GFP alone and an identical HA-containing protein. Extracts were incubated with magnetic agarose GFP binder beads and immunoblotted using anti-Myc and anti-HA antibodies (b–e), as indicated. Two percent of input was loaded. The experiment was performed in the presence of RNase A. Images shown are representative of two biological replicates. f Co-immunoprecipitation experiments were carried out with lysates prepared from S2R+ cells transfected with Myc-GFP-tagged Hakai-long and HA-tagged Fl(2)d upon control (LacZ) or vir KDs. In control lanes, S2R+ cells were transfected with Myc-GFP alone and an identical HA-containing protein. Extracts were incubated with magnetic agarose GFP binder beads and immunoblotted using anti-GFP and anti-HA antibodies. Two percent of input was loaded. g, h Schematic representing the interaction domains between Hakai, Vir and Fl(2)d derived from co-IP experiments in Drosophila (g) or HEK393T cells (h). Orange dotted lines indicate the second interaction domain between Human WTAP and VIRMA. Source data for western blots are provided as a Source Data file.

Fig. 3. Hakai is required for m⁶A methylation and sex determination by regulating Sex-lethal alternative splicing.

a Schematic of the Hakai genomic locus depicting the transposon (black triangle) used to generate the deletion in the Hakai allele and the premature stop codon present in the Hakai allele, and the epitope-tagged UAS constructs of the short and long isoforms. Schematic diagram of a 2D thin-layer chromatography (TLC) depicting standard and methylated nucleotides (b), and TLCs depicting m⁶A in control (c) and Hakai/Df(2L)Exel8041 pupae (d). e Viability of female flies from a cross of the indicated genotypes mated with Sxl^7BO males. The loss of one copy of Mettl3 or Hakai significantly reduces female survival in a genetic background where one copy of Sxl and da are absent. Viability was calculated from the numbers of females compared with males, and statistical significance was determined by a χ² test (Graphpad Prism). ^∗∗∗P ≤ 0.0001). Unpaired two-tailed Student’s t-test for unequal variances. f Viability of Hakai/Df(2L)Exel8041 flies. g The viability of female flies with homozygous vir^2F mutation can be rescued by the loss of a single copy of Hakai. Viability was calculated from the numbers of homozygous vir^2F females compared with heterozygous balancer-carrying siblings, and statistical significance was determined by a χ² test (Graphpad Prism). ^∗∗∗P ≤ 0.0001). h–n External sexual differentiation (h, j–n) and Sxl alternative splicing in abdomen and head/thorax fraction (i) of control, vir^2F/vir^ts and Hakai¹ vir^2F/vir^ts female flies. The gel shown in (i) is a representative of two biological replicates. The marker is a 100 bp DNA ladder with 500 bp indicated on top. Note that Sxl alternative splicing in Hakai vir^2F/vir^ts female abdomens is switched to the male mode and that these females display male pigmentation (h, m), but no male sex combs (h). Source data for TLC, fly numbers and RT-PCR gels are provided as a Source Data file.

Fig. 4. Hakai regulates m⁶A levels and m⁶A-dependent splicing events.

a LC–MS/MS quantification of m⁶A levels in either control samples or mRNA extracts depleted for the indicated proteins in S2R+ cells. The bar chart shows the mean with standard error (SE) of three biological replicates and three technical measurements. KD of Hakai results in substantial reduction of m⁶A levels. ^∗∗∗P = 7.49E−04 (Hakai KD), ^∗∗∗∗P = 7.51E−06 (Mettl3, Mettl14 KD). Unpaired two-tailed Student’s t-test for unequal variances. b Relative isoform quantification of m⁶A-regulated genes (fl(2)d and Hairless) upon depletion of the indicated components. The bar chart shows the mean with standard error (SE) of three biological replicates and three technical measurements. Hakai is required for m⁶A-dependent splicing regulation. c Number of differentially spliced events upon knockdown of the indicated proteins (left) and common differentially spliced targets (right) (FDR < 0.1). d Radar charts display relative distribution of differentially spliced events upon knockdown of Fl(2)d and Hakai. Alternative 5’splice site (A5SS) selection and intron retention (RI) are overrepresented events upon loss of m⁶A writer components. Source data for m⁶A measurement, qPCR and RNA-seq are provided as a Source Data file.

Fig. 5. Hakai RING domain is required for interaction with MACOM components.

a Mass spectrometry analysis of Hakai-dependent ubiquitinated proteins in S2R+ cells. Scatter plot of normalized forward versus inverted reverse experiments plotted on a log2 scale. The threshold was set to a twofold enrichment or depletion (red dashed line). One protein in the top right quadrant is enriched in both replicates. Hakai depletion does not affect global ubiquitination levels in D. melanogaster S2R+ cells. A complete list of quantified ubiquitylation sites after HAKAI depletion can be found in Supplementary Data 3. b–d Co-immunoprecipitation experiments were carried out with lysates prepared from S2R+ cells transfected with Myc-GFP-tagged Hakai-long (WT or RING mutant) and HA-tagged Hakai-long (b), HA-tagged Fl(2)d (c) and HA-tagged Virilizer fragment (d). In control lanes, S2R+ cells were transfected with Myc-GFP alone and an identical HA-containing protein. Extracts were incubated with magnetic GFP binder beads and immunoblotted using anti-Myc and anti-HA antibodies, as indicated. Two percent of input was loaded. The experiment was performed in the presence of RNase A. Images shown are representative of two biological replicates. Source data for Western blots are provided as a Source Data file.

Fig. 6. Hakai regulates the stability of other MACOM components.

a Mass spectrometry analysis of Hakai-dependent proteome in S2R+ cells. Scatter plot of normalized forward versus inverted reverse experiments plotted on a log2 scale. The threshold was set to a 1.4-fold enrichment or depletion (red dashed line). Proteins in the bottom left quadrant are decreased in both replicates. Heat map of proteins whose levels were reduced by >1.4-fold in both replicates of the whole proteome analysis. Other components of the m⁶A writer complex and reader proteins are shown for comparison. A complete list of quantified proteins can be found in Supplementary Data 4. b Levels of endogenous Fl(2)d, Mettl3 and Mettl14 proteins from control cells and cells depleted for Hakai were analyzed by western blot. Tubulin was used as a loading control. Quantification of Fl(2)d, Mettl3 and Mettl14 levels from blots shown below (left). The bar chart shows the mean with standard error (SE) of four biological replicates. ^∗∗∗P = 2.48E−04 (Hakai KD). Unpaired two-tailed Student’s t-test for equal variances. Hakai depletion strongly destabilizes Fl(2)d, but not Mettl3 or Mettl14 proteins. Relative expression levels of Hakai are shown as a validation of its KD efficiency. The bar chart shows the mean with standard error (SE) of three technical measurements (Bottom right). c, d Analysis of Fl(2)d levels upon Hakai, Mettl3, nito (c) or Hakai, vir, Flacc (d) depletion. Protein lysates from control and depleted cells were analyzed by western blot for levels of endogenous Fl(2)d protein. Tubulin was used as a loading control. One representative experiment is shown and quantification of three biological replicates is shown below. The bar chart shows the mean with standard error (SE). (c) ^∗∗∗∗P = 4.74E−05 (Hakai KD), ^n.s.P = 0.80029 (Mettl3 KD) and ^n.s.P = 0.05398 (Nito KD). (d) ^∗∗P = 0.0057 (Hakai KD), P = 0.0012 (Vir KD) (Hakai KD), and ^n.s.P = 0.6061 (Flacc KD). Unpaired two-tailed Student’s t-test for equal variances. e Western blots were carried out with lysates prepared from HeLa and U2OS cells transfected with scrambled siRNA or siRNA against HAKAI or VIRMA. Extracts were immunoblotted using the indicated antibodies. Depletion of HAKAI reduced VIRMA levels while depletion of VIRMA had no impact on HAKAI levels. Images shown are representative of two biological replicates. Source data for western blots, measurement of protein levels and qPCR are provided as a Source Data file.

Fig. 7. Model showing the composition of the m⁶A methyltransferase complex in fly and human and the impact of Hakai on MACOM integrity.

Top left and right represent the human and fly m⁶A methyltransferase complexes, respectively. (Bottom) The depletion of Hakai leads to a reduction in protein levels for Vir, Flacc and Fl(2)d.