ABSTRACT
HemK2 is a highly conserved methyltransferase, but the identification of its genuine substrates has been controversial, and its biological importance in higher organisms remains unclear. We elucidate the role of HemK2 in the methylation of eukaryotic Release Factor 1 (eRF1), a process that is essential for female germline development in Drosophila melanogaster. Knockdown of hemK2 in the germline cells (hemK2-GLKD) induces apoptosis, accompanied by a pronounced decrease in both eRF1 methylation and protein synthesis. Overexpression of a methylation-deficient eRF1 variant recapitulates the defects observed in hemK2-GLKD, suggesting that eRF1 is a primary methylation target of HemK2. Furthermore, hemK2-GLKD leads to a significant reduction in mRNA levels in germline cell. These defects in oogenesis and protein synthesis can be partially restored by inhibiting the No-Go Decay pathway. In addition, hemK2 knockdown is associated with increased disome formation, suggesting that disruptions in eRF1 methylation may provoke ribosomal stalling, which subsequently activates translation-coupled mRNA surveillance mechanisms that degrade actively translated mRNAs. We propose that HemK2-mediated methylation of eRF1 is crucial for ensuring efficient protein production and mRNA stability, which are vital for the generation of high-quality eggs.
INTRODUCTION
Post-translational modifications (PTMs), such as methylation, phosphorylation, glycosylation, acetylation and ubiquitylation, are chemical alterations that occur after protein synthesis. PTMs profoundly influence protein function, interaction, stability and enzymatic activity (Ramazi and Zahiri, 2021; Ryšlavá et al., 2013). Among the over 400 types of PTMs, methylation is one of the most pervasive. Methylation reactions are catalyzed by S-adenosylmethionine-dependent methyltransferases (MTases), which transfer methyl groups to a variety of biomolecules, including proteins, DNA and RNA, targeting cytosines, and adenines (Chen et al., 2013; Dai et al., 2021; Ping et al., 2014; Wang et al., 2016). In the Drosophila genome, 127 MTase genes have been annotated, and these MTases have been further classified based on their substrate specificities. Lysine MTases and arginine MTases constitute substantial groups among protein MTases. Previous studies on histone methylation have highlighted the importance of lysine/arginine modifications in transcriptional regulation (Gates et al., 2017; Zhao and Garcia, 2015). In addition to lysine and arginine residues, other amino acid residues, such as histidine (Clarke, 2013), aspartic acid (Sprung et al., 2008), asparagine (Klotz et al., 1990) and glutamine (Heurgué-Hamard et al., 2002), are also subject to methylation.
HemK2 is a member of the unclassified protein MTases, distinct due to its broad substrate class. HemK2 has been previously shown to catalyze the methylation of histone H4 lysine 12 (H4K12), DNA N6-adenine (6mA) and eukaryotic translation release factor 1 (eRF1). HemK2-mediated H4K12 methylation is implicated in regulating the proliferation of prostate tumor cells (Metzger et al., 2019), while diminished genomic DNA 6mA levels are associated with tumorigenesis in individuals with cancer (Xiao et al., 2018). However, the enzymatic activity of HemK2 towards genomic DNA 6mA remains controversial (Kweon et al., 2019; Ratel et al., 2006; Woodcock et al., 2019). Conversely, the role of HemK2 as an MTase for eRF1 has been extensively investigated in yeast and mammalian cells. HemK2 (N6AMT1 in humans; YDR140w or Mtq2p in yeast) methylates the glutamine residue within the GGQ motif of eRF1 (Figaro et al., 2008; Heurgué-Hamard et al., 2006; Liu et al., 2010). Depletion of HemK2 in yeast (Mtq2p) results in sensitivity to translation-fidelity antibiotics (Polevoda et al., 2006). In mice, HemK2 (also known as N6AMT1) deficiency leads to early embryonic lethality and impaired post-implantation development (Liu et al., 2010). However, the methylation of these potential substrates by HemK2 under physiological conditions in multicellular organisms is not well defined.
Drosophila oogenesis serves as an exemplary model for studying the genetic regulation of egg production. The female fly harbors a pair of ovaries, each comprising 18-20 ovarioles, which contain the germarium at the anterior tip, succeeded by sequentially maturing egg chambers (Avilés-Pagán and Orr-Weaver, 2018; Gleason et al., 2018; Rastegari et al., 2020). Each egg chamber encompasses one oocyte and 15 nurse cells that support oocyte growth, surrounded by a monolayer of somatic follicle cells. In nurse cells, the production of mRNAs and proteins for maternal deposition into eggs is crucial for proper oogenesis. Consequently, Drosophila oogenesis is an energetically demanding process involving cellular proliferation and growth that necessitates extensive protein synthesis (Lasko, 2012).
The role of HemK2 as a methyltransferase (MTase) has been extensively investigated in yeast and mammalian cell cultures. Nonetheless, its function within multicellular organisms, particularly in the reproductive system – characterized by active gene expression and protein synthesis – has not been fully elucidated. In this study, we delineate the role of HemK2 in eRF1 methylation during Drosophila oogenesis. Our findings indicate that HemK2 is crucial for egg production; knockdown of hemK2 expression arrested oogenesis at an intermediate stage and induced apoptosis in germline cells. The disruption of hemK2 expression resulted in a substantial decrease in eRF1 methylation, whereas methylation of DNA 6mA and histone H4 lysine 12 (H4K12) was not significantly affected. Moreover, the lack of eRF1 methylation impaired translation efficiency, which in turn reduced mRNA levels via the No-Go Decay pathway. Collectively, these results highlight the indispensable function of HemK2-mediated eRF1 methylation in promoting effective protein synthesis and mRNA stability, thus modulating gene expression throughout Drosophila oogenesis.
RESULTS
Germline knockdown of hemK2 resulted in female sterility marked by the developmental arrest in the mid-oogenesis
To investigate the role of HemK2, a highly conserved methyltransferase encoded by CG9960 in Drosophila melanogaster, we employed RNA interference to repress its expression. This knockdown was conducted using shRNA targeted to the 3′UTR (BL#40837) using the germline-specific driver NGT40; NosGal4-VP16 (Grieder et al., 2000). In normal development, the oocyte accrues proteins and RNAs from 15 nurse cells through cytoplasmic channels (Fig. 1A,B). Females with germline-specific knockdown of hemK2 (hemK2-GLKD) showed a ∼65% decrease in mRNA levels (Fig. 1C, Fig. S1A) and exhibited sterility. These hemK2-GLKD females contained small ovaries and were incapable of laying eggs (Fig. 1B,D), indicative of developmental arrest at stages 7/8 before the vitellogenic stages. Conversely, hemK2-GLKD males, despite a similar reduction in its mRNA level to ∼25%, did not exhibit notable fertility impairments (Fig. S1F,G).
HemK2 is essential for Drosophila germline development. (A) Diagram of Drosophila oogenesis. Each ovariole is composed of a germarium at the anterior end, followed by progressively developing egg chambers. A 16-cell germline cyst enclosed with somatic follicle cells buds out of the germarium, forming an egg chamber. Of those 16 germline cells, one differentiates into the oocyte, while others become nurse cells. (B) Representative images of ovarian morphology from control and hemK2 germline knockdown (GLKD) samples. hemK2-GLKD leads to female sterility, characterized by arrested development of egg chambers at stages 7-8 and impaired vitellogenesis. Scale bar: 250 µm. (C) Quantitative RT-PCR confirmation of knockdown efficacy in hemK2-GLKD ovaries. Expression levels are normalized to rp49 with tubulin serving as an internal control. Error bars indicate s.d. (n=3). (D) Analysis of egg laying and hatching rate. Infertility observed in hemK2-GLKD is rescued by expression of the wild-type FLAG-tagged HemK2 transgene using the germline-specific driver NGT40; nosGal4-VP16. Daily egg production by three females was recorded (n=3). Error bars indicate s.d. (E) Immunofluorescence of FLAG-tagged HemK2 (green) and DNA (DAPI, blue) in ovaries of the indicated genotypes corresponding to D. Arrows indicate degenerated egg chambers in hemK2-GLKD and later-stage egg chambers after FLAG-HemK2 expression. Scale bar: 50 µm. (F) Western blot analysis detecting FLAG epitope to demonstrate germline expression of FLAG-HemK2 wild type in ovarian extracts. Tubulin is used as a loading control.
HemK2 is essential for Drosophila germline development. (A) Diagram of Drosophila oogenesis. Each ovariole is composed of a germarium at the anterior end, followed by progressively developing egg chambers. A 16-cell germline cyst enclosed with somatic follicle cells buds out of the germarium, forming an egg chamber. Of those 16 germline cells, one differentiates into the oocyte, while others become nurse cells. (B) Representative images of ovarian morphology from control and hemK2 germline knockdown (GLKD) samples. hemK2-GLKD leads to female sterility, characterized by arrested development of egg chambers at stages 7-8 and impaired vitellogenesis. Scale bar: 250 µm. (C) Quantitative RT-PCR confirmation of knockdown efficacy in hemK2-GLKD ovaries. Expression levels are normalized to rp49 with tubulin serving as an internal control. Error bars indicate s.d. (n=3). (D) Analysis of egg laying and hatching rate. Infertility observed in hemK2-GLKD is rescued by expression of the wild-type FLAG-tagged HemK2 transgene using the germline-specific driver NGT40; nosGal4-VP16. Daily egg production by three females was recorded (n=3). Error bars indicate s.d. (E) Immunofluorescence of FLAG-tagged HemK2 (green) and DNA (DAPI, blue) in ovaries of the indicated genotypes corresponding to D. Arrows indicate degenerated egg chambers in hemK2-GLKD and later-stage egg chambers after FLAG-HemK2 expression. Scale bar: 50 µm. (F) Western blot analysis detecting FLAG epitope to demonstrate germline expression of FLAG-HemK2 wild type in ovarian extracts. Tubulin is used as a loading control.
HemK2 is expressed from a dicistronic transcript that is co-transcribed with the upstream open reading frame, snapin, in a single mRNA (Fig. S1A) (Wall et al., 2005). To confirm that the observed defects in hemK2-GLKD were due to diminished HemK2 protein expression, individual transgenes encoding N-terminal FLAG-tagged HemK2 or Snapin, lacking the shRNA target sequence, were expressed in the germline after knockdown of the dicistronic mRNA (Fig. 1E, Fig. S1E). Females expressing the hemK2 transgene alone in their germline exhibited egg laying and hatching rates comparable with controls (Fig. 1D), thus rescuing the sterility defect. In contrast, expression of snapin did not restore these defects (Fig. S1D). These results confirm that the sterility observed was specifically due to compromised hemK2 function resulting from knockdown of the dicistronic transcript, establishing the vital role of hemK2 in germline cells during oogenesis.
To investigate the function of hemK2 in somatic cells, we knocked down its expression in ovarian somatic cells, including follicle cells, using the somatic driver traffic jam-Gal4 (tj-Gal4). hemK2-STKD females became sterile with rudimentary ovaries (Fig. S1B,C). These ovaries often presented with fused egg chambers and disorganized follicle cells (Fig. S1C). Remarkably, the fertility and developmental disruptions were nearly fully rescued by expressing the hemK2 transgene with tj-Gal4 (Fig. S1B,C), suggesting that hemK2 is crucial in both germline and somatic cell functions during oogenesis.
Pleiotropic phenotypes in hemK2-GLKD ovaries indicate essential roles in oogenesis
Ovaries with hemK2-GLKD exhibited multiple phenotypic aberrations: hyperplasia of the posterior follicle cells, chromosomal dispersion failure and extensive apoptosis in germline cells (Fig. 2). All egg chambers in the control females were encapsulated by a monolayer of follicle cells, whereas the majority of egg chambers in hemK2-GLKD females demonstrated multi-layered follicle cells in the posterior region (Fig. 2A; 94%). In addition, an increased presence of mitotic cells, marked by phospho-histone H3 (pH3), was observed in the hyperplastic multi-layered follicle cells of hemK2-GLKD during stages 7/8 (Fig. 2A). This hyperplasia phenotype might be due to the altered Notch-Delta signaling pathway (see Discussion).
Knockdown of hemK2 leads pleiotropic defects in germline development. (A) Left: ovaries displaying mitotic marker phospho-histone H3 (pH3) (red) and DAPI (blue). Dotted rectangles highlight hyperplastic posterior follicle cells in hemK2-GLKD (upper right panel). Insets show enlarged views of the areas outlined. Scale bars: 20 µm. Right: quantification of ovarioles with single versus multi-layered follicle cells is provided (n=100). (B) Immunostaining for nucleolar marker Fibrillarin and DNA (DAPI) in stage 7/8 egg chambers. Insets show enlarged views of nurse cell nuclei with either dispersed or undispersed chromosomes. Scale bars: 10 µm. (C) Quantitative RT-PCR analysis of gene expression involved in chromosomal dispersion, with error bars indicating s.d. (n=3). (D) Left: immunostaining for apoptosis marker p53 and DNA (DAPI) in ovaries. Arrows indicate p53 signal accumulation in nurse cell nuclei. Scale bars: 20 µm. Right: percentage of ovarioles displaying p53 signals is quantified (right panel, n=100).
Knockdown of hemK2 leads pleiotropic defects in germline development. (A) Left: ovaries displaying mitotic marker phospho-histone H3 (pH3) (red) and DAPI (blue). Dotted rectangles highlight hyperplastic posterior follicle cells in hemK2-GLKD (upper right panel). Insets show enlarged views of the areas outlined. Scale bars: 20 µm. Right: quantification of ovarioles with single versus multi-layered follicle cells is provided (n=100). (B) Immunostaining for nucleolar marker Fibrillarin and DNA (DAPI) in stage 7/8 egg chambers. Insets show enlarged views of nurse cell nuclei with either dispersed or undispersed chromosomes. Scale bars: 10 µm. (C) Quantitative RT-PCR analysis of gene expression involved in chromosomal dispersion, with error bars indicating s.d. (n=3). (D) Left: immunostaining for apoptosis marker p53 and DNA (DAPI) in ovaries. Arrows indicate p53 signal accumulation in nurse cell nuclei. Scale bars: 20 µm. Right: percentage of ovarioles displaying p53 signals is quantified (right panel, n=100).
Moreover, nurse cells in hemK2-GLKD displayed defective chromosomal dispersion (Fig. 2B). Typically, wild-type nurse cells underwent a distinct polytene chromosome condensation into a ‘5-blob’ structure, which was fully dispersed by stage 6 (Dej and Spradling, 1999). In stark contrast, almost all hemK2-GLKD ovarioles contained nurse cell nuclei with chromosomal configurations retained in a blob-like state beyond this stage (Fig. 2B, Fig. S2A; 99%). To investigate the possible causes of this chromosomal defect, we examined the expression levels of genes known to be involved in chromosomal dispersion (Van Buskirk and Schüpbach, 2002; Goodrich et al., 2004; Klusza et al., 2013). qRT-PCR analysis showed that the mRNA level of otu was significantly reduced in hemK2-GLKD ovaries (Fig. 2C), suggesting a correlation between otu expression and the chromosomal dispersion defect. The previous studies have reported that one of the isoforms of Otu, Otu104, can rescue the chromosome dispersion defects observed in sqd and hfp mutant nurse cells (Van Buskirk and Schüpbach, 2002; Goodrich et al., 2004). In the same way, the expression of the N-terminal FLAG-tagged Otu104 isoform in hemK2-GLKD ovaries substantially recovered the dispersion of chromosomes (Fig. S2A), suggesting that hemK2 supports chromosomal dispersion during oogenesis, partly by ensuring otu expression.
Additionally, hemK2-GLKD led to pronounced degeneration of egg chambers during mid-oogenesis, a process that is typically regulated by germline cell viability in response to physiological and environmental cues (Drummond-Barbosa and Spradling, 2001; McCall, 2004). Immunostaining for p53, an apoptotic inducer, revealed a marked increase in p53 signal in the nurse cell nuclei of most hemK2-GLKD ovarioles (Fig. 2D; 93%), suggesting cell death in mid-oogenesis. Consistently, the simultaneous knockdown of p53 and hemK2 significantly restored the progression of oogenesis in hemK2-GLKD ovaries (Fig. S2B). Furthermore, expression of p35, an inhibitor of caspases DrICE and Dcp-1 (Fuchs and Steller, 2015; Werz et al., 2005), also suppressed the cell death in hemK2-GLKD ovaries (Fig. S2B). Although females with either p53 knockdown or ectopic p35 expression in hemK2-GLKD did not fully recover egg-laying to control levels, the hatching rates of the resulting eggs were comparable with those of controls (Fig. S2C). Collectively, these data suggest that p53-dependent apoptosis occurs during mid-oogenesis in hemK2-GLKD, although the precise initiating mechanisms remain to be determined. Notably, chromosomal dispersion remained impaired in hemK2-GLKD ovaries even after cell death suppression (Fig. S2B), indicating that apoptosis and chromosomal dispersion defects are mediated by distinct pathways affected by hemK2 downregulation.
eRF1 methylation by the catalytic activity of HemK2 is required for oogenesis
Previous studies have identified HemK2 as a methyltransferase acting on three different substrates: DNA N6-adenine (6mA) in human cultured cells (Xiao et al., 2018), histone H4 lysine 12 (H4K12) in A549 lung adenocarcinoma cells (Metzger et al., 2019), and a glutamine residue within the GGQ motif of eRF1 in yeast (Heurgué-Hamard et al., 2005) and human (Figaro et al., 2008). Nonetheless, the genuine substrate and functional role of HemK2 in multicellular organisms have remained elusive. First, we investigated the 6mA modification in genomic germline DNA of hemK2-GLKD through immunostaining with an anti-6mA antibody. No alteration in the pattern or signal intensity of 6mA staining was observed in hemK2-GLKD germline cells (Fig. S3A). Owing to the limitations of fluorescence detection in providing accurate quantification, we cannot definitively exclude a potential role for hemK2 in DNA methylation. However, these results suggest that hemK2 does not influence the pattern of genomic N6-adenine methylation in the Drosophila germline. Moreover, levels of methylated H4K12 were similar between hemK2-GLKD and control ovaries (Fig. S3B), thereby excluding H4K12 as a primary substrate in the Drosophila germline.
Subsequently, we evaluated the methylation status of endogenous eRF1 in hemK2-GLKD ovaries. Owing to the ineffectiveness of the anti-methyl-eRF1 antibody in western blot analyses, we performed eRF1 immunoprecipitation followed by immunoblotting with an antibody against N5-methyl-glutamine (methyl-Q). The eRF1 methylation in hemK2-GLKD ovaries was significantly reduced to 35% of control, likely reflecting the contribution from unaffected somatic cells (Fig. 3A). Immunostaining with the anti-methyl-eRF1 antibody showed a substantial reduction of eRF1 methylation specifically in germline cells of hemK2-GLKD ovaries, while the expression level of eRF1 remained consistent across cell types (Fig. 3B). This finding was corroborated by the immunoprecipitation of FLAG-tagged eRF1 expressed in the germline, which exhibited minimal methylation in hemK2-GLKD ovaries compared with control (Fig. 3C), confirming the role of Drosophila HemK2 in eRF1 methylation. It is well documented that a glutamine residue within the GGQ motif of eRF1, which is crucial for peptide release, is consistently methylated by HemK2 across various species (Dinçbas-Renqvist et al., 2000; Figaro et al., 2008; Gao et al., 2020; Kusevic et al., 2016; Svidritskiy and Korostelev, 2018). To determine whether the GGQ motif of eRF1 is the methylation target in Drosophila, we mutated the corresponding glutamine (Q185) to alanine (eRF1 Q185A) and examined their methylation in S2 cells (Fig. 3D). Consistent with observations in other systems, methylation of the eRF1 Q185A mutant was nearly undetectable, irrespective of hemK2 knockdown, indicating that this mutation disrupts eRF1 methylation (Fig. 3D). Collectively, these results suggest that Q185 is the predominant methylation site of eRF1 by Drosophila HemK2.
HemK2 functions for eRF1 methylation in Drosophila. (A) Left: immunoprecipitation of endogenous eRF1 and subsequent western blot analysis for N5-methyl-glutamine (Methyl-Q) from ovarian lysates, indicating a significant reduction in eRF1 methylation in hemK2-GLKD compared with control ovaries (bottom panel). Right: quantification of methyl-Q band intensity normalized to eRF1 signal is shown, with error bars indicating s.d. (n=3) and P-value indicated (unpaired t-test). (B) Immunostaining of methylated eRF1 (green) and total eRF1 (red) in control and hemK2-GLKD ovaries. Arrows indicate the cytoplasm of germline cells. Scale bar: 20 µm. (C) Left: immunoprecipitation of germline-expressed FLAG-tagged eRF1 from ovaries. Western blot for methyl-Q illustrates a notable decrease in methylation of FLAG-eRF1 in hemK2-GLKD. Nonspecific bands are marked with asterisks. Right: quantification of methyl-Q band intensity normalized to FLAG-eRF1, as in A, with error bars indicating s.d. (n=3) and P-value provided (unpaired t-test). (D) Immunoprecipitation of Myc-tagged eRF1 (WT or Q185A mutant) in S2 cells, with western blot detection for methyl-Q. Quantification of methyl-Q band intensities normalized to Myc-eRF1 are shown with error bars indicating s.d. (n=3).
HemK2 functions for eRF1 methylation in Drosophila. (A) Left: immunoprecipitation of endogenous eRF1 and subsequent western blot analysis for N5-methyl-glutamine (Methyl-Q) from ovarian lysates, indicating a significant reduction in eRF1 methylation in hemK2-GLKD compared with control ovaries (bottom panel). Right: quantification of methyl-Q band intensity normalized to eRF1 signal is shown, with error bars indicating s.d. (n=3) and P-value indicated (unpaired t-test). (B) Immunostaining of methylated eRF1 (green) and total eRF1 (red) in control and hemK2-GLKD ovaries. Arrows indicate the cytoplasm of germline cells. Scale bar: 20 µm. (C) Left: immunoprecipitation of germline-expressed FLAG-tagged eRF1 from ovaries. Western blot for methyl-Q illustrates a notable decrease in methylation of FLAG-eRF1 in hemK2-GLKD. Nonspecific bands are marked with asterisks. Right: quantification of methyl-Q band intensity normalized to FLAG-eRF1, as in A, with error bars indicating s.d. (n=3) and P-value provided (unpaired t-test). (D) Immunoprecipitation of Myc-tagged eRF1 (WT or Q185A mutant) in S2 cells, with western blot detection for methyl-Q. Quantification of methyl-Q band intensities normalized to Myc-eRF1 are shown with error bars indicating s.d. (n=3).
The above-mentioned rescue experiments validated the indispensable role of HemK2 in vivo for egg development (Fig. 1 and Fig. S1). We proceeded to evaluate the significance of amino acid residues that are crucial for the methyltransferase activity of HemK2 during oogenesis. The NPPY motif of HemK2, which is highly conserved and recognized as a substrate-binding pocket, is vital for its methylation function in human cell lines (Fig. 4A) (Gao et al., 2020; Metzger et al., 2019). It has been reported that mutations in residues proximal to the substrate, specifically Asn122 and Tyr125 within the NPPY motif of human HemK2, abolish its catalytic activity for eRF1 methylation in vitro (Fig. 4B) (Metzger et al., 2019). Correspondingly, we introduced mutations at analogous positions in Drosophila HemK2, Asn116 and Tyr119, substituting them with alanine (N116A and Y119A, respectively), and expressed these mutants in the germline of hemK2-GLKD ovaries (Fig. 4C, Fig. S4A). The Y119A mutant did not recover the observed defects, such as egg chamber degeneration, egg laying and hatching rate; however, the N116A mutant restored these parameters to levels comparable with the control (Fig. 4C,D). These findings suggest that Tyr119, as opposed to Asn116, is crucial for the function of HemK2 within Drosophila germline cells.
The conserved NPPY motif of HemK2 is crucial for eRF1 methylation. (A) Illustration of the methyltransferase (MTase) domain within Drosophila HemK2, highlighting the highly conserved NPPY motif across various species. Asn116 and Tyr119 of Drosophila HemK2 are mutated to Ala, generating N116A and Y119A, respectively. (B) Crystal structure depiction of human HemK2-TRMT112 complexed with methyl-glutamine (Methyl-Q) and S-adenosylhomocysteine (SAH) (Gao et al., 2020), focusing on the active site proximal to methyl-Q. HemK2 is depicted in gray, methyl-Q in green with hydrogen bonds around the methyl-Q. N122 and Y125 of human HemK2 correspond to N116 and Y119 in Drosophila HemK2, respectively. (C) Immunostaining for FLAG-tagged HemK2 variants against a DNA counterstain (DAPI) in ovaries of the respective genotypes. Perturbation of proper oogenesis in hemK2-GLKD ovaries is restored by the expression of the N116A HemK2 variant but not the Y119A variant. Arrows indicate either degenerating or maturating egg chambers in each genotype. Scale bar: 50 µm. (D) Analysis of egg laying and hatching rate for each genotype described in C. The number of eggs were counted daily for groups of three females (n=3), with error bars indicating s.d. (E) Immunoprecipitation followed by western blot analysis for methyl-Q in S2 cells. Co-transfection of FLAG-tagged HemK2 and Myc-tagged eRF1, with or without hemK2 knockdown (KD), demonstrates that eRF1 methylation loss in hemK2-KD is restored by wild-type HemK2 and the N116A mutant but not by the Y119A mutant. The lower panel provides quantification of methyl-Q intensity relative to Myc-eRF1 with error bars indicating s.d. (n=3).
The conserved NPPY motif of HemK2 is crucial for eRF1 methylation. (A) Illustration of the methyltransferase (MTase) domain within Drosophila HemK2, highlighting the highly conserved NPPY motif across various species. Asn116 and Tyr119 of Drosophila HemK2 are mutated to Ala, generating N116A and Y119A, respectively. (B) Crystal structure depiction of human HemK2-TRMT112 complexed with methyl-glutamine (Methyl-Q) and S-adenosylhomocysteine (SAH) (Gao et al., 2020), focusing on the active site proximal to methyl-Q. HemK2 is depicted in gray, methyl-Q in green with hydrogen bonds around the methyl-Q. N122 and Y125 of human HemK2 correspond to N116 and Y119 in Drosophila HemK2, respectively. (C) Immunostaining for FLAG-tagged HemK2 variants against a DNA counterstain (DAPI) in ovaries of the respective genotypes. Perturbation of proper oogenesis in hemK2-GLKD ovaries is restored by the expression of the N116A HemK2 variant but not the Y119A variant. Arrows indicate either degenerating or maturating egg chambers in each genotype. Scale bar: 50 µm. (D) Analysis of egg laying and hatching rate for each genotype described in C. The number of eggs were counted daily for groups of three females (n=3), with error bars indicating s.d. (E) Immunoprecipitation followed by western blot analysis for methyl-Q in S2 cells. Co-transfection of FLAG-tagged HemK2 and Myc-tagged eRF1, with or without hemK2 knockdown (KD), demonstrates that eRF1 methylation loss in hemK2-KD is restored by wild-type HemK2 and the N116A mutant but not by the Y119A mutant. The lower panel provides quantification of methyl-Q intensity relative to Myc-eRF1 with error bars indicating s.d. (n=3).
Furthermore, we investigated the activity of HemK2 in catalyzing the methylation of the glutamine residue within the GGQ motif of eRF1 in S2 cells. After co-transfecting each FLAG-tagged HemK2 variant with Myc-tagged eRF1 into S2 cells and knocking down endogenous hemK2, we assessed eRF1 methylation (Fig. 4E). Immunoprecipitation of Myc-eRF1 followed by immunoblotting with an anti-methyl-Q antibody showed significant reduction in eRF1 methylation upon endogenous hemK2 knockdown. This reduction was rescued by expressing wild-type HemK2 and the N116A mutant, but not by expressing the Y119A mutant (Fig. 4E). This demonstrates that Tyr119, but not Asn116, is indispensable for the methylation of eRF1 at Q185 by Drosophila HemK2.
Additionally, given that HemK1, a paralog of HemK2 with a conserved NPPY motif and 42% sequence homology, is known to methylate the glutamine in the GGQ motif of mitochondrial release factors in human cells (Fang et al., 2022; Ishizawa et al., 2008), we sought to determine the specificity of cytosolic eRF1 methylation by HemK2. In S2 cells, where hemK2 knockdown reduced eRF1 methylation, expression of HemK2, but not HemK1, rescued the methylation levels, signifying that cytosolic eRF1 methylation is an exclusive function of HemK2 (Fig. S4B).
HemK2-mediated eRF1 methylation is required for efficient translation and mRNA stability
We investigated the role of eRF1 methylation in oogenesis by knocking down eRF1 expression in germline cells. eRF1 germline knockdown (eRF1-GLKD) led to rudimentary ovaries that failed to produce eggs, a phenotype more severe than that of hemK2-GLKD (Figs 1B and 5A,B). Immunostaining with Vasa (Vas), a DEAD box RNA helicase and germline marker, indicated agametic germaria with a stark reduction in germline cells in eRF1-GLKD ovaries (Fig. 5C). The defects in eRF1-GLKD were successfully rescued by expressing RNAi resistant FLAG-tagged wild-type eRF1 (Fig. S5A,B), but not by that of RNAi resistant methylation-deficient Q185A mutant in germline cells (Fig. 5B,C). This highlights the crucial role of the Q185 residue, and likely its methylation, in germline cell viability and oogenesis progression. The less severe phenotype in hemK2-GLKD could result from incomplete disruption of eRF1 methylation, potentially due to residual hemK2 expression or compensation by other methyltransferase(s).
eRF1 methylation ensures proper progression of oogenesis and efficient translation. (A) Comparative morphological images of control and eRF1 germline knockdown (eRF1-GLKD) ovaries. Scale bar: 250 µm. (B) Analysis of egg production and hatching rate across different genotypes. The numbers of the eggs were counted daily for groups of three females (n=3), with error bars indicating s.d. (C) Immunostaining of wild-type and Q185A mutant FLAG-tagged eRF1 (green), with Vas (red) marking germline cells and DAPI (blue) for DNA, in ovaries corresponding to each genotype listed in B. Loss of germline cells by eRF1-GLKD is restored with wild-type eRF1 but not with the Q185A mutant. Scale bars: 20 µm. (D) Immunostaining for FLAG-tagged eRF1 Q185A (red) and DNA (DAPI, blue) in ovaries. Overexpression of FLAG-eRF1 Q185A in wild-type background recapitulates the phenotypic consequences of hemK2-GLKD. Arrows indicate degenerating egg chambers. Insets show enlarged views of nurse cell nuclei. Scale bar: 50 µm. (E) Left: protein synthesis assay using L-homopropargylglycine (HPG) in ovaries for each genotype indicated (green), with concurrent immunostaining for FLAG-HemK2 (red). Arrows indicate egg chambers with diminished HPG signal intensity. Scale bars: 20 µm. Right: quantification of HPG signal intensity, with measurements from representative images (n=5) and error bars indicating s.d.
eRF1 methylation ensures proper progression of oogenesis and efficient translation. (A) Comparative morphological images of control and eRF1 germline knockdown (eRF1-GLKD) ovaries. Scale bar: 250 µm. (B) Analysis of egg production and hatching rate across different genotypes. The numbers of the eggs were counted daily for groups of three females (n=3), with error bars indicating s.d. (C) Immunostaining of wild-type and Q185A mutant FLAG-tagged eRF1 (green), with Vas (red) marking germline cells and DAPI (blue) for DNA, in ovaries corresponding to each genotype listed in B. Loss of germline cells by eRF1-GLKD is restored with wild-type eRF1 but not with the Q185A mutant. Scale bars: 20 µm. (D) Immunostaining for FLAG-tagged eRF1 Q185A (red) and DNA (DAPI, blue) in ovaries. Overexpression of FLAG-eRF1 Q185A in wild-type background recapitulates the phenotypic consequences of hemK2-GLKD. Arrows indicate degenerating egg chambers. Insets show enlarged views of nurse cell nuclei. Scale bar: 50 µm. (E) Left: protein synthesis assay using L-homopropargylglycine (HPG) in ovaries for each genotype indicated (green), with concurrent immunostaining for FLAG-HemK2 (red). Arrows indicate egg chambers with diminished HPG signal intensity. Scale bars: 20 µm. Right: quantification of HPG signal intensity, with measurements from representative images (n=5) and error bars indicating s.d.
Moreover, expressing the FLAG-eRF1 Q185A mutant in germline cells of wild-type ovaries disrupted egg production (Fig. 5D). This expression led to developmental arrest at mid-oogenesis, with defects in chromosome dispersion and extensive apoptosis, mirroring the hemK2-GLKD phenotype (Fig. 5D, Fig. S5C,D). Over 80% of ovarioles expressing eRF1 Q185A exhibited p53-positive egg chambers (Fig. S5E). Therefore, the expression of eRF1 Q185A effectively recapitulates the hemK2-GLKD phenotype, supporting that the absence of eRF1 methylation primarily causes the observed defects.
Given that eRF1 GGQ motif methylation is essential for nascent peptide release and translational termination efficiency, and that its perturbation could lead to translational deficits by hindering ribosome recycling, we assessed translation activity in hemK2-GLKD ovaries using a HPG incorporation assay ex vivo. HPG, a methionine analog with an alkyne group, can be incorporated into newly synthesized proteins and is visualized through a ‘Click reaction’ with a fluorescent dye (Shen et al., 2021). A notable decrease in HPG signal in cytoplasm of the germline cell was observed in hemK2-GLKD ovaries (Fig. 5E), indicating a requirement for eRF1 methylation in maintaining translation efficiency. This translational reduction in hemK2-GLKD was completely restored by expressing either wild-type HemK2 or the N116A mutant, but not by the Y119A mutant (Fig. 5E). Similarly, germline expression of the methylation-deficient eRF1 Q185A mutant abrogated translation activity (Fig. S5F). Collectively, these results underscore the necessity of HemK2-mediated eRF1 methylation for efficient translation in the Drosophila ovary.
To elucidate the molecular deficiencies resulting from hemK2-GLKD and eRF1 dysfunction, we conducted ovarian mRNA-seq analysis using next-generation sequencing technology. Late egg chambers in control ovaries were selectively removed to minimize developmental discrepancies between hemK2-GLKD and control samples. Aligning with our RT-PCR data (Figs 1C and 2C), mRNA-seq confirmed the marked downregulation of hemK2 (log2FC=−1.27) and otu (log2FC=−1.45), although hfp and smn genes were not significantly reduced (Table S4). Additionally, we identified 478 upregulated and 325 downregulated genes (log2FC<−1 or >1; q-val<0.01). Gene Ontology (GO) term analysis (Thomas et al., 2022) of differentially expressed genes indicated a significant downregulation in genes associated with female reproductive processes, such as female gamete generation, oogenesis and germ cell development (Fig. 6A), whereas genes unrelated to oogenesis were comparatively upregulated (Fig. 6B).
Knockdown of hemK2 leads mRNA reduction and ribosome stalling. (A) Gene Ontology (GO) analysis of genes downregulated in hemK2-GLKD ovaries. Enrichment of GO biological process terms is analyzed by PANTHER (version 18.0). The fold enrichments are plotted and colored by their False Discovery Rate (FDR) values. (B) GO analysis for genes upregulated in hemK2-GLKD ovaries, conducted similarly to A. (C) Volcano plot representing the differential expression of germline-enriched genes in hemK2-GLKD ovaries, with significance threshold set at P<0.05 and log2FC<−1 or >1 (orange). Gene classification is based on single-cell mRNA-seq analysis of Drosophila ovaries (Jevitt et al., 2020). (D) Log2-fold changes of cell-type enriched genes in hemK2-GLKD ovaries relative to control, with mean values and s.d. indicated. (E) Ribosome profiling in ovarian lysates from control (blue), hemK2-GLKD (red) and double-germline knockdown of hemK2 and ski8 (green). Monosome and disome were separated using sucrose gradient sedimentation. Absorbance at 254 nm was recorded to generate elution profiles, normalized to the monosome peak for each sample. Inset highlights the disome peak and the baseline derived from the monosome. (F) Disome to monosome peak ratios were calculated across three experimental replicates. Bar graphs represent average ratios±s.d. (n=3) and P-values (unpaired t-test).
Knockdown of hemK2 leads mRNA reduction and ribosome stalling. (A) Gene Ontology (GO) analysis of genes downregulated in hemK2-GLKD ovaries. Enrichment of GO biological process terms is analyzed by PANTHER (version 18.0). The fold enrichments are plotted and colored by their False Discovery Rate (FDR) values. (B) GO analysis for genes upregulated in hemK2-GLKD ovaries, conducted similarly to A. (C) Volcano plot representing the differential expression of germline-enriched genes in hemK2-GLKD ovaries, with significance threshold set at P<0.05 and log2FC<−1 or >1 (orange). Gene classification is based on single-cell mRNA-seq analysis of Drosophila ovaries (Jevitt et al., 2020). (D) Log2-fold changes of cell-type enriched genes in hemK2-GLKD ovaries relative to control, with mean values and s.d. indicated. (E) Ribosome profiling in ovarian lysates from control (blue), hemK2-GLKD (red) and double-germline knockdown of hemK2 and ski8 (green). Monosome and disome were separated using sucrose gradient sedimentation. Absorbance at 254 nm was recorded to generate elution profiles, normalized to the monosome peak for each sample. Inset highlights the disome peak and the baseline derived from the monosome. (F) Disome to monosome peak ratios were calculated across three experimental replicates. Bar graphs represent average ratios±s.d. (n=3) and P-values (unpaired t-test).
To further dissect the gene expression changes after germline knockdown of hemK2, we referenced a previous study detailing gene expression across different ovarian cell types from single-cell mRNA analyses; many different cell types were distinguished based on gene expression patterns. Focusing on genes expressed in one germline and five somatic cell types representing early to mid-oogenesis, we observed a significant downregulation (average log2FC=−0.55) in hemK2-GLKD for genes enriched in germline 2 cell type (n=154), which are expressed from germarium region 2 onward (Fig. 6C,D) (Jevitt et al., 2020). In contrast, genes associated with the five somatic cell types did not show notable changes (Fig. 6D, Fig. S6A-E), suggesting that hemK2-GLKD specifically downregulates mRNAs in germline cells.
Previous research indicates that eRF1 methylation at the GGQ motif substantially enhances the release of nascent peptides, accelerating translation termination by 10- to 100-fold (Pierson et al., 2016). We hypothesize that impaired methylation on eRF1 could decrease the peptide release rate, leading to ribosome stalling at stop codons of actively translated mRNAs. Such stalling is known to activate the mRNA surveillance process, targeting mRNAs with engaged ribosomes for degradation via the No-Go Decay (NGD) pathway (Brandman and Hegde, 2016; Doma and Parker, 2006; Ikeuchi et al., 2019; Simms et al., 2017; Tomomatsu et al., 2023). To investigate this hypothesis, we analyzed ribosome stalling in hemK2-GLKD by performing a sucrose gradient sedimentation assay with ovarian lysates. The lysates, which were treated with emetine to halt translation and with RNase A to digest unprotected mRNA regions, revealed a predominance of monosomes and a negligible disome presence in the control, indicative of minimal ribosome stalling and collision (Fig. 6E). However, in hemK2-GLKD, the disome level increased to approximately 2.1-fold of that in the control (Fig. 6F, hemK2-GLKD), suggesting that ineffective nascent polypeptide release due to unmethylated eRF1 leads to ribosome collision.
Blockage of the NGD pathway restores mRNA destabilization and mid-oogenesis arrest in hemK2-GLKD ovaries
To ascertain whether the NGD pathway was responsible for the pronounced defects observed in hemK2-GLKD, we conducted a simultaneous knockdown of the mRNA surveillance pathway components along with hemK2 in germline cells (Fig. 7, Fig. S7A). In the double-germline knockdown (DGKD) of hemK2 and ski8, which is a component of the Ski complex that collaborates with the exosome and facilitates 3′ to 5′ mRNA degradation (Zinoviev et al., 2020), developmental impairments associated with hemK2-GLKD were partially restored (Fig. 7A). Females from the ski8-hemK2 DGKD had more advanced stages of egg chambers and laid a comparable number of eggs to controls, though their hatching rate was reduced to 40% of the control (Fig. 7A,B). Considering the role of the Ski complex in general RNA decay, we further investigated additional NGD factors, such as pelota (pelo), which is crucial for dissociating stalled ribosomes and promoting RNA decay within the NGD and NSD pathways (Becker et al., 2011; Jamar et al., 2017; Kobayashi et al., 2010; Tsuboi et al., 2012). Remarkably, pelo-hemK2 DGKD females exhibited a similar phenotype to that of the ski8-hemK2 DGKD, with advanced egg chamber development and a 60% hatching rate, comparable with the number of eggs laid (Fig. 7A,B). Furthermore, the mid-oogenesis arrest induced by expressing the dominant-negative eRF1 Q185A mutant was also partly rescued by the germline knockdown of either ski8 or pelo (Fig. S7B), reinforcing the notion that the eRF1 methylation defect triggers mRNA surveillance mechanisms.
Blockage of No-Go-decay pathway restores oogenesis defects in hemK2-GLKD ovaries. (A) Fluorescent staining with DAPI for DNA in ovaries from various genotypes. Rescue of the oogenesis defects observed in hemK2-GLKD is achieved through combined knockdown of ski8 or pelo, but not with CG7139 or upf3. Scale bar: 50 µm. Arrows indicate either degenerating or maturating egg chambers in each genotype. (B) Analysis of egg production and hatching rate for the indicated genotypes. The numbers of eggs counted for groups of three females each day (n=3), with error bars indicating s.d. (C) Assay of protein synthesis in ovaries using L-homopropargylglycine (HPG) across the specified genotypes. Scale bar: 20 µm. Arrows indicate egg chambers with diminished HPG signal intensity. (D) Quantitative analysis of HPG fluorescence intensity for each genotype, with measurements taken from representative images (n=5). Error bars indicate s.d. (E) Box and whisker plot depicting the log2-fold changes of genes enriched in germline 2 cell type in ovaries from hemK2-GLKD, ski8-hemK2 DGKD and pelo-hemK2 DGKD compared with control. The central line of the box represents the median; the top and bottom edges of the box represent the first quartile and third quartile, respectively. The whiskers indicate the minimum and maximum values within 1.5 times the interquartile range (IQR). Paired t-test P-values are presented, comparing hemK2 single knockdown with each respective double knockdown.
Blockage of No-Go-decay pathway restores oogenesis defects in hemK2-GLKD ovaries. (A) Fluorescent staining with DAPI for DNA in ovaries from various genotypes. Rescue of the oogenesis defects observed in hemK2-GLKD is achieved through combined knockdown of ski8 or pelo, but not with CG7139 or upf3. Scale bar: 50 µm. Arrows indicate either degenerating or maturating egg chambers in each genotype. (B) Analysis of egg production and hatching rate for the indicated genotypes. The numbers of eggs counted for groups of three females each day (n=3), with error bars indicating s.d. (C) Assay of protein synthesis in ovaries using L-homopropargylglycine (HPG) across the specified genotypes. Scale bar: 20 µm. Arrows indicate egg chambers with diminished HPG signal intensity. (D) Quantitative analysis of HPG fluorescence intensity for each genotype, with measurements taken from representative images (n=5). Error bars indicate s.d. (E) Box and whisker plot depicting the log2-fold changes of genes enriched in germline 2 cell type in ovaries from hemK2-GLKD, ski8-hemK2 DGKD and pelo-hemK2 DGKD compared with control. The central line of the box represents the median; the top and bottom edges of the box represent the first quartile and third quartile, respectively. The whiskers indicate the minimum and maximum values within 1.5 times the interquartile range (IQR). Paired t-test P-values are presented, comparing hemK2 single knockdown with each respective double knockdown.
Investigating the role of the NGD pathway in the defects seen in hemK2-GLKD ovaries, we assessed another putative NGD component, nonu-1/cue2, an endonuclease identified in C. elegans and S. cerevisiae that cleaves mRNA near a stalled ribosome during translation (D'Orazio et al., 2019; Glover et al., 2020) (Fig. S7C). Contrasting with the partial rescue observed with ski8 and pelo knockdowns, the simultaneous germline knockdown of CG7139, which encodes the Drosophila ortholog of nonu-1/cue2, and hemK2 did not ameliorate the observed phenotypic defects (Fig. 7A,B, Fig. S7A). This indicates that CG7139 may not be an essential component of the NGD pathway in Drosophila, or it may act redundantly with other endonucleases in the translation-coupled mRNA surveillance mechanism compromised by hemK2 knockdown. Further analysis revealed that a dual germline knockdown of hemK2 and upf3, a factor in Nonsense-Mediated Decay (NMD) (Tan et al., 2022), did not reverse the phenotypic abnormalities in hemK2-GLKD ovaries (Fig. 7A,B), emphasizing a specific role of NGD in mediating mRNA degradation. The efficiency of the knockdown for each component was verified by qRT-PCR (Fig. S7A; ranging from 10 to 40% relative to control), accounting for the contribution of non-knockdown somatic cells to the total ovarian RNA, confirming a significant reduction in expression levels. Consistent with their roles in rescuing developmental morphology, the HPG incorporation assay demonstrated that knockdown of ski8 or pelo, but not CG7139 or upf3, in hemK2-GLKD ovaries partially restored translational activity (Fig. 7C,D). These results support the hypothesis that disrupted eRF1 methylation in hemK2-GLKD leads to compromised translation termination and ribosome stalling, specifically triggering mRNA degradation via the NGD pathway.
In alignment with the recovery of oogenesis defects in hemK2-GLKD, the expression levels of downregulated protein-coding RNAs were restored by the additional germline knockdown of either ski8 or pelo (Fig. 7E, Fig. S6F,G), further validating the specific activation of NGD upon hemK2-GLKD. The concurrent germline knockdown of ski8 or pelo with hemK2 also reinstated general protein synthesis, despite the observation of ribosome stalling to a similar degree as in the single hemK2 knockdown (Fig. 6E,F). These findings suggest that when mRNAs are not eliminated by NGD, translation re-initiation may occur, enabling protein synthesis and partially recovering egg maturation (Fig. 7A,B). Overall, our results indicate that perturbations in eRF1 methylation due to hemK2-GLKD result in the inefficient release of synthesized peptides and the activation of the NGD pathway, likely due to ribosome stalling.
DISCUSSION
The reduction of hemK2 expression led to female sterility, marked by a significant interruption of oogenesis in both germline and somatic cells of the ovary, culminating in developmental arrest during the mid-stages (Fig. 1, Fig. S1). Notably, hemK2-GLKD ovaries exhibited morphological aberrations, such as follicle cell hyperplasia and chromosomal dispersion failure (Fig. 2A,B). It has been previously reported that the expression of the Delta ligand in germline cells activates Notch-Delta signaling pathway, which prompts the transition from the mitotic cell cycle to the endocycle in follicle cells (Deng et al., 2001; Schaeffer et al., 2004; Lopez-Schier and St Johnston, 2001). The hyperplasia observed in the posterior follicle cells of hemK2-GLKD ovaries might be due to diminished Delta ligand expression in germline cells, leading to an extended period of mitotic division in follicle cells beyond stage 6.
In egg chambers of stage 6 and beyond, hemK2-GLKD ovarioles predominantly contained nurse cell nuclei with unspread chromosomes (Fig. 2B, Fig. S2A). Inhibition of cell death in hemK2-GLKD ovaries significantly improved egg chamber formation, yet the chromosomal dispersion defect persisted (Fig. S2B). These findings imply that germline cell death in hemK2-GLKD is not caused by the failure of chromosomal dispersion but likely due to ribosome stalling, which is a consequence of eRF1 methylation disturbance (Wu et al., 2020). It has been reported that ribosome collisions can initiate two distinct stress response pathways: a transient collision may trigger a GCN2-mediated pathway that promotes cell survival, whereas more severe collisions could lead to the activation of MAPKK, inducing apoptosis in human cell lines (Wu et al., 2020). Our data suggest that perturbation of methylation due to hemK2 knockdown results in ribosome stalling and subsequent collisions, triggering apoptosis in the translationally active environment of Drosophila ovaries.
Although HemK2 in yeast and mammalian cells has been shown to mediate methylation of DNA 6mA, histone H4K12 and eRF1 (Figaro et al., 2008; Metzger et al., 2019; Xiao et al., 2018), our findings indicate that, in Drosophila, HemK2 does not affect the methylation of DNA 6mA or histone H4K12 (Fig. S3A,B). Instead, hemK2 knockdown led to a substantial decrease in eRF1 methylation (Fig. 3A-C), underscoring eRF1 as the specific substrate for Drosophila HemK2. Notably, Tyr119, but not Asn116, is essential for the methyltransferase activity of HemK2, which is pivotal for oogenesis and eRF1 methylation in Drosophila (Fig. 4C-E). This contrasts with the necessity of both corresponding residues in human HemK2 for eRF1 methylation in vitro (Metzger et al., 2019). However, recent structural analyses of human HemK2 revealed that the Asn122 side chain does not interact with the methylated glutamine within the substrate binding pocket (Gao et al., 2020), aligning with the dispensability of Asn116 in the methyltransferase activity of Drosophila HemK2.
Murine HemK2 substrate specificity studies using a peptide array have delineated a putative consensus motif: G-Q-[SRYKLGAMTC]-[ARFGLWYCSQKH]-[LARCQFYT]-R, where square brackets indicate any of the residues listed therein (Kusevic et al., 2016). Employing this motif, we have identified several Drosophila proteins, including CG17841, Spargel, Chd1, Ythdc1 and eRF1, as potential HemK2 substrates. These candidates may undergo methylation by HemK2 under certain conditions. However, the phenotypes associated with the overexpression of the methylation-deficient eRF1 Q185A mutant mirrored those observed in hemK2-GLKD (Fig. 5D), proposing that, in the context of Drosophila germline cells, eRF1 is the primary substrate for HemK2.
Consistent with previous reports, methylation of the conserved GGQ motif within release factors has been shown to significantly enhance nascent peptide release, accelerating the termination rate of protein synthesis by 10- to 100-fold (Pierson et al., 2016). Corroborating this, our study found that perturbation of eRF1 methylation by hemK2-GLKD led to a marked decrease in translation efficiency (Fig. 5E). Notably, the reduced protein synthesis, alongside the defects in oogenesis and the downregulation of mRNAs, were substantially rescued by inhibiting general RNA degradation and the NGD pathway, whereas NMD blockade did not yield similar results (Fig. 7A-D). These findings imply that the ribosome, even when associated with unmethylated eRF1, can synthesize minimal yet essential polypeptides, provided that the mRNAs are available for translation.
Our findings support the role of HemK2 as an authentic protein methyltransferase in Drosophila, crucially involved in modulating eRF1 activity via glutamine methylation, a modification that is vital for appropriate germline development. Germline nurse cells efficiently produce proteins and mRNAs destined for transfer to the oocyte, necessitating stringent regulation of resource-intensive protein synthesis. Nonetheless, the specific regulatory mechanisms governing eRF1 methylation by HemK2 remain elusive and warrant additional exploration. It is plausible that this mechanism serves to optimize egg production under favorable physiological conditions, while downregulating it when conditions are less than ideal.
MATERIALS AND METHODS
Fly stocks and cultures
All Drosophila stocks and crosses were raised at room temperature or at 25°C on standard food [5% (w/v) dry yeast, 5% (w/v) corn flower, 2% (w/v) rice bran, 10% (w/v) glucose, 0.7% (w/v) agar, 0.2% (v/v) propionic acid and 0.05% (w/v) p-hydroxy butyl benzoic acid]. For RNAi-mediated knockdown experiments, the following transgenic RNAi Project (TRiP) lines from the Bloomington Drosophila Stock Center were used: hemK2 (HMS02003, 40837), eRF1 (HMS05812, 67900), upf3 (HMJ22158, 58181), cue2 (HMS02721, 44007), ski8 (HMC04664, 57377) and p53 (HMS02286, 41720). The attP40 (25709) and attP2 (25710) lines were employed for transgene injections. The following drivers were used to express transgenes or shRNAs: NGT40 and nosGal4-VP16 (Grieder et al., 2000), Matα-tub-Gal4 (7063), and traffic jam-Gal4 (Drosophila Genetic Resource Center, Japan, 104055) (Hayashi et al., 2002). The UASp-P35 transgenic line was generously provided by Dr. Bergmann (Werz et al., 2005). Controls for each experiment were either the y w strain or the corresponding heterozygous line.
Generation of transgenic fly lines
The transgenic fly lines generated were as follows: FLAG-HemK2 wild type at attP2, FLAG-HemK2 N116A at attP2, FLAG-HemK2 Y119A at attP2, FLAG-Snapin at attP2, FLAG-eRF1 wild type at attP2, FLAG-eRF1 Q185A at attP2 and FLAG-Otu104 at attP2. In each instance, the coding sequence or its mutant variant was amplified by PCR, using y w genomic DNA or cDNA synthesized from y w ovarian RNA as templates. To avoid the eRF1 transgenes being subjected to knockdown, the target sequence C AGC AAG ATG TTG GCC GAT GA within the eRF1 coding region was modified to C AGT AAA ATG TTA GCG GAC GA, without altering the amino acid sequence. A triple FLAG tag sequence was appended to the N-terminal region of the amplification products. The purified PCR fragments were subsequently cloned into the XbaI-linearized pUASp-K10-attB vector (Koch et al., 2009) via an In-Fusion reaction (639648, Takara Bio). Plasmids confirmed to be correct were injected into embryos of nos-phiC31;;P{CaryP} attP2 (#25710) following standard protocols.
The hemK2 TRiP construct at attP2 and pelo TRiP constructs at attP2 and attP40 sites were also created. These TRiP constructs were produced by ligating annealed oligonucleotides (TRiP-Fw/Rv) into the NheI- and EcoRI-linearized pWalium22 vector (Perkins et al., 2015). UASp and Walium22 plasmids were then injected into embryos of nos-phiC31; P{CaryP}attP40 (25709) or nos-phiC31;;P{CaryP} attP2 (25710). The sequences of the oligonucleotides used for annealing are provided in Table S1.
Fluorescent immunostaining
Immunostaining was performed as previously described (Lim et al., 2022). Ovaries were dissected and slightly teased apart in phosphate-buffered saline (PBS), then fixed with 5.3% (v/v) paraformaldehyde (PFA, Electron Microscopy Sciences) in PBS for 10 min at room temperature. The fixed ovaries were washed for 30 min with several changes of PBX [PBS containing 0.2% (v/v) Triton X-100]. This was followed by a blocking step involving incubation in PBX containing 4% (w/v) bovine serum albumin (BSA) for a minimum of 30 min. Ovaries were incubated with primary antibodies diluted in PBX containing 0.4% (w/v) BSA overnight at 4°C, then washed for 1 h with several changes of PBX. Secondary antibodies were applied for over 2 h at room temperature; after this, ovaries were washed for 1 h with PBX and stained with 1 μg/ml of 4′,6-diamidino-2-phenylindole (DAPI) (Sigma) for 10 min in PBX. After a final rinse with PBS, the ovaries were mounted onto slides using Fluoro-KEEPER Antifade (Nacalai Tesque).
The following primary antibodies were used at the indicated dilutions: mouse anti-FLAG [Sigma (M2), F1804, 1:200], guinea pig anti-Vas (1:2000) (Patil and Kai, 2010), rabbit anti-pH3 (Millipore, 06-570, 1:250), mouse anti-p53 [DSHB (25F4), 1:100], rabbit anti-Fibrillarin (Abcam, ab5821, 1:200), rabbit anti-DNA 6mA (Synaptic Systems, 202003, 1:1000), rabbit anti-eRF1 (Sigma, E8156, 1:200) and rabbit anti-methyl-eRF1 (1:100) (Lacoux et al., 2020). Secondary antibodies were Alexa Fluor 488- and 555-conjugated goat antibodies against rabbit, mouse and guinea pig IgG (Invitrogen, A11029, A11034, A11073, A21424, A21429 and A21435) at a 1:200 dilution in PBX containing 0.4% (w/v) BSA. Ovarian images were captured using a Zeiss LSM 900 confocal microscope and processed with Zen (Zeiss) and Fiji (Schindelin et al., 2012) software.
Protein synthesis assay
Protein synthesis was assessed using the Click-iT L-homopropargylglycine (HPG) Alexa Fluor 488 Protein Synthesis Assay Kit (Molecular Probes), which measures newly synthesized proteins through quantification of the fluorescent HPG signal. Freshly dissected ovaries were incubated in 1 mM HPG in PBS for 30 min and then fixed with 5.3% PFA in PBS for 10 min. Subsequently, the ovaries were permeabilized by washing three times in PBX for 30 minutes: the first wash used 3% BSA in PBX, followed by two additional washes in PBX alone. The ovaries were then incubated with a freshly prepared Click-iT reaction cocktail for 30 min. After washing with Click-iT reaction rinse buffer, ovaries were stained with HCS NuclearMask Blue Stain working solution (1:2000 in PBX) for 30 min as per the manufacturer's instructions. Subsequent immunostaining with antibodies and image acquisition were performed as described above.
Quantification of HPG signal intensity for translation activity
Fluorescence intensity of HPG signals in nurse cells of egg chambers (between stages 3 to 8) was quantified to assess translation activity. The region of interest (ROI) was selected within the cytosolic areas, avoiding the nucleus. The fluorescence intensity for each ROI was measured after background subtraction, with the background defined as a region lacking any cells or signals. For each egg chamber, two to five of these ROIs were analyzed. The average fluorescence intensity across all ROIs within an image was calculated to represent the translation activity. This process was repeated for a total of five images per genotype. Image processing and analysis were conducted using Fiji software.
Immunoprecipitation and western blotting
For immunoprecipitation, 100 ovaries were dissected in ice-cold PBS and homogenized using a pestle in PBS with 0.1% (v/v) Tween-20 (PBST), supplemented with a protease inhibitor cocktail (PIC, 11873580001, Roche). After centrifugation at 15,000 g for 10 min at 4°C, the supernatant was collected. Protein G magnetic beads (TAMAGAWA SEIKI, TAS8848N1173) were pre-incubated with anti-FLAG antibody [Sigma (M2), F1804] for more than 1 h at 4°C with gentle rotation before adding lysate. The antibody-conjugated beads were washed twice with ice-cold PBST and then incubated with the lysate for 3-4 h at 4°C. The proteins were washed three times with ice-cold PBST and eluted from the beads in 50 µl of sample buffer at 95°C for 5 min.
For western blotting, ovary lysates or immunoprecipitates were separated by SDS-PAGE and transferred onto a 0.45 µm Clear Trans PVDF membrane (Wako). The membrane was blocked in PBST with 3% (w/v) skim milk (Nacalai Tesque) and incubated with primary antibodies diluted in HIKARI signal enhancer (Nacalai Tesque). The primary antibodies used included rabbit anti-eRF1 (Sigma, E8156, 1:1000), rabbit anti-methyl-Q (Millipore, ABS2185, 1:1000), mouse anti-FLAG [Sigma (M2), F1804, 1:1000], mouse anti-c-Myc [Wako (9E10), 017-21871, 1:1000], mouse anti-histone H4 (MAB Institute, 388-09171, 1:2000), rabbit anti-H4K12met (PTM BIO, PTM-685RM, 1:2000), mouse anti-Tubulin [Santa Cruz (DM1A), sc-32293, 1:3000], rabbit anti-HA [Cell Signaling Technology (C29F4), 3724, 1:1000]. Horseradish peroxidase (HRP)-conjugated secondary antibodies used were goat anti-rabbit IgG or anti-mouse IgG (Bio-Rad, 170-6515 or 170-6516, 1:3000). To prevent signal overlap from IgG heavy chain with eRF1, VeriBlot-HRP (Abcam, ab131366, 1:1000) was used in western blot analysis for immunoprecipitation of endogenous eRF1 or FLAG-eRF1. Chemiluminescent signals were detected using Chemi-Lumi One (Nacalai Tesque, 07880-70) and imaged by the ChemiDoc Touch MP system (Bio-Rad). Band intensities were quantified using Image Lab software (Bio-Rad) from independent experiments.
Quantitative RT-PCR
Total RNA was extracted from ovaries using TRIzol LS Reagent (Invitrogen) strictly according to the manufacturer's instructions. The purified RNA (1 µg) was then treated with DNase I (Invitrogen) to remove any genomic DNA contamination. This was followed by inactivation of the DNase I with EDTA. Reverse transcription to synthesize cDNA was conducted using the Superscript III First-Strand Synthesis System (Invitrogen), as per the manufacturer's protocol. Quantitative PCR (qPCR) was performed using the KAPA SYBR Fast qPCR Master Mix (KAPA Biosystems) on the StepOnePlus Real-Time PCR System (Applied Biosystems). Relative expression levels were normalized to those of the reference gene rp49 across three biological replicates. The primer sequences used to determine gene expression levels are listed in Table S2.
S2 cell experiments
Drosophila S2 cells were maintained at 28°C in Schneider's medium, supplemented with 10% (v/v) fetal bovine serum (FBS) and antibiotics (penicillin and streptomycin). Knockdown of endogenous hemK2 was achieved using double-stranded RNA (dsRNA) targeting a 186 bp region from the ATG start codon of the snapin gene (refer to Fig. S1). The dsRNA was synthesized using the AmpliScribe T7 High Yield Transcription Kit (Lucigen). Primer sequences for the PCR amplification are provided in Table S3.
For transfection, plasmids employed were Myc-HemK2 wild type, Myc-HemK2 N116A, Myc-HemK2 Y119A, FLAG-eRF1 wild type, FLAG-eRF1 Q185A and Myc-HemK1. These constructs were prepared using the Gateway Cloning System (Life Technologies) and transfected into S2 cells using HilyMax (Dojindo Molecular Technologies) according to the manufacturer's protocol.
Fertility test
To assess female fertility, y w males were paired with three eclosed females of each genotype. Mating was allowed to proceed for several days in vials containing standard food. Subsequently, the flies were transferred to apple juice plates supplemented with a small quantity of moist yeast paste. The total count of eggs laid by the mated females was recorded. After a 24 h period, the number of unhatched eggs was tallied to calculate the hatching rates. For the evaluation of male fertility in the hemK2-GLKD line, three y w virgin females were mated with hemK2-GLKD males over an extended duration. The fertility metrics for the males were documented in a manner analogous to that described above. Each fertility assessment was replicated across three independent biological samples.
mRNA-seq analyses
Total RNA for each genotype was extracted in duplicate from dissected ovaries using TRIzol LS (Invitrogen) following the manufacturer's instructions. To avoid morphological impacts, RNA extraction was performed using only ovaries up to mid-oogenesis from all genotypes, as hemK2-GLKD excludes late-stage egg chambers. Poly(A)-tailed RNAs were enriched and purified using oligo-dT beads from the NEBNext Poly(A) mRNA Magnetic Isolation Module (New England Biolabs). These samples then underwent fragmentation, reverse transcription, adapter ligation and PCR amplification to prepare cDNA libraries using the NEBNext Ultra II Directional RNA Library Prep Kit (New England Biolabs). Library preparation and next-generation sequencing were outsourced to Rhelixa (Japan), yielding approximately 13 million paired-end sequences per sample. The fastp program (Chen et al., 2018) was used to remove the sequences of low quality and adapter. Alignment to the Drosophila melanogaster reference genome (dm6) was conducted using the STAR program (Dobin et al., 2013). Read quantification at the gene level was performed with featureCounts (Liao et al., 2014). Differential expression analysis was executed via DESeq2 (Love et al., 2014), selecting upregulated (log2FoldChange>1, P-adj<0.01) and downregulated (log2FoldChange<−1, P-adj<0.01) genes for further GO term enrichment analysis using PANTHER (Thomas et al., 2022).
Monosome/disome profiling
Ovaries were dissected in polysome lysis buffer [20 mM Tris-Cl (pH 7.5), 5 mM MgCl2 and 100 mM NaCl] with the addition of 0.1% NP-40, 200 µg/ml emetine (Sigma, E2375) and PIC. Dissected ovaries were snap frozen in liquid nitrogen and stored at −80°C. Homogenization was performed with a plastic pestle in 500 µl of polysome lysis buffer supplemented with 0.1% CHAPS (Nacalai Tesque), emetine and PIC. The homogenate was subjected to sequential centrifugations at 15,000 g for 5 min, then for 15 min, both at 4°C. A small aliquot of the clear supernatant was taken for RNA quantification using the Qubit RNA HS Assay Kit (Invitrogen, Q32852). The RNA concentration was adjusted to 0.05 µg/µl with additional lysis buffer containing CHAPS, emetine and PIC. The supernatant was treated with RNase A (1 µg/ml) at 25°C for 10 min to cleave unprotected RNA regions. After RNase treatment, 200 µl of sample was applied to a 10-50% sucrose gradient prepared in polysome lysis buffer and ultracentrifuged in a SW41Ti rotor at 30,000 g for 3 h at 4°C. After centrifugation, gradients were fractionated from the top using a piston gradient fractionator (BioComp). The monosome peak was normalized to 1.0 for each profile. The baseline derived from the monosome was subtracted from the disome peak. The height of the remaining disome peak was quantified.
Acknowledgements
We thank Dr Valérie Heurgué-Hamard (IBPC, Paris) and Dr Andreas Bergmann (UMASS Chan Medical School, Massachusetts) for generous gifts of antibody against the methyl-eRF1 and UASp-P35 transgenic fly line, respectively. We acknowledge the Bloomington Drosophila Stock Centre and the Kyoto Stock Center for the fly stocks. We also thank the FBS Core Facility in Osaka University for providing access to the LSM 900 and ChemiDoc Touch. Assistance for sucrose gradient analysis was provided by the Center for Medical Research and Education, Graduate School of Medicine, Osaka University. We appreciate the insightful discussion and suggestions from all the members of T.K.’s laboratory.
Footnotes
Author contributions
Conceptualization: F.X., R.S., T.I., S.K., T.K.; Methodology: F.X., R.S., T.I., S.K.; Software: F.X., R.S., S.K.; Validation: F.X., S.K.; Formal analysis: F.X., R.S., S.K.; Investigation: F.X., S.K.; Resources: F.X., R.S., S.K., T.K.; Data curation: F.X., R.S., S.K.; Writing - original draft: F.X., R.S., S.K., T.K.; Writing - review & editing: F.X., S.K., T.K.; Visualization: F.X., R.S., S.K.; Supervision: T.I., S.K., T.K.; Project administration: S.K., T.K.; Funding acquisition: S.K., T.K.
Funding
This work was supported by a Takeda Science Foundation bioscience research grant (J191503009 to T.K.), by the Japan Society for the Promotion of Science Grant-in-Aid for Scientific Research on Innovative Areas (20H05390 to T.K.) by a Japan Society for the Promotion of Science Grant-in-Aid for Transformative Research Areas (A) (21H05275 to T.K.) and by the Institute for Datability Science, Osaka University ‘Transdisciplinary Research Project’ (Na22990007 to T.K. and S.K.).
Data availability
The next generation sequencing data shown in this study have been deposited in the DNA Data Bank of Japan (DDBJ) under BioProject ID PRJDB17326.
References
Competing interests
The authors declare no competing or financial interests.