Resource Identification of the m6 Am Methyltransferase PCIF1 Reveals the Location and Functions of m6 Am in the Transcriptome Graphical Abstract Highlights d PCIF1 is the N6-adenosine methylase that produces m6 Am in an m7 G cap-dependent manner d PCIF1 depletion allows transcriptome-wide mapping of m6 A and m6 Am d m6 Am mapping identifies alternative ‘‘internal’’ transcription start sites d m6 Am increases stability of a subset of mRNAs and has no effect on translation Authors Konstantinos Boulias, Diana Toczyd1owska-Socha, Ben R. Hawley, ..., L. Aravind, Samie R. Jaffrey, Eric Lieberman Greer Correspondence srj2003@med.cornell.edu (S.R.J.), eric.greer@childrens.harvard.edu (E.L.G.) In Brief m6 Am is a prevalent mRNA modification occurring adjacent to the m7 G cap. Boulias, Toczydlowska-Socha, Hawley et al. identify PCIF1 as the m6 Am methyltransferase and perform transcriptome-wide mapping to distinguish m6 Am from m6 A and identify ‘‘internal’’ TSSs. m6 Am increases stability of a subset of mRNAs but has minimal effects on translation. Boulias et al., 2019, Molecular Cell 75, 631–643 August 8, 2019 ª 2019 Elsevier Inc. https://doi.org/10.1016/j.molcel.2019.06.006 Molecular Cell Resource Identification of the m6 Am Methyltransferase PCIF1 Reveals the Location and Functions of m6 Am in the Transcriptome Konstantinos Boulias,1,2,7 Diana Toczyd1owska-Socha,3,4,7 Ben R. Hawley,3,7 Noa Liberman,1,2 Ken Takashima,1,2 Sara Zaccara,3 The´ o Guez,5 Jean-Jacques Vasseur,5 Franc¸ oise Debart,5 L. Aravind,6 Samie R. Jaffrey,3,* and Eric Lieberman Greer1,2,8,* 1Division of Newborn Medicine, Boston Children’s Hospital, 300 Longwood Avenue, Boston, MA 02115, USA 2Department of Pediatrics, Harvard Medical School, Boston, MA 02115, USA 3Department of Pharmacology, Weill Cornell Medicine, Cornell University, New York, NY 10065, USA 4Laboratory of Bioinformatics and Protein Engineering, International Institute of Molecular and Cell Biology in Warsaw, 02-109 Warsaw, Poland 5IBMM, Universite´ de Montpellier, CNRS, ENSCM, Montpellier, France 6NCBI, National Library of Medicine, NIH, Bethesda, MD 20894, USA 7These authors contributed equally 8Lead Contact *Correspondence: srj2003@med.cornell.edu (S.R.J.), eric.greer@childrens.harvard.edu (E.L.G.) https://doi.org/10.1016/j.molcel.2019.06.006 SUMMARY mRNAs are regulated by nucleotide modifications that influence their cellular fate. Two of the most abundant modified nucleotides are N6 -methyladenosine (m6 A), found within mRNAs, and N6 ,20 -O-dimethyladenosine (m6 Am), which is found at the first transcribed nucleotide. Distinguishing these modifications in mapping studies has been difficult. Here, we identify and biochemically characterize PCIF1, the methyltransferase that generates m6 Am. We find that PCIF1 binds and is dependent on the m7 G cap. By depleting PCIF1, we generated transcriptome-wide maps that distinguish m6 Am and m6 A. We find that m6 A and m6 Am misannotations arise from mRNA isoforms with alternative transcription start sites (TSSs). These isoforms contain m6 Am that maps to ‘‘internal’’ sites, increasing the likelihood of misannotation. We find that depleting PCIF1 does not substantially affect mRNA translation but is associated with reduced stability of a subset of m6 Am-annotated mRNAs. The discovery of PCIF1 and our accurate mapping technique will facilitate future studies to characterize m6 Am’s function. INTRODUCTION The most prevalent regulated methyl modifications in mRNA occur on two similar nucleotides: adenosine (A) and 20 -O-methyladenosine (Am) (Perry et al., 1975; Wei et al., 1975). METTL3 catalyzes the methylation on the N6 position of the adenine ring to form N6 -methyladenosine (m6 A) at internal sites in mRNA (Bokar et al., 1997). At least 25% of mRNAs contain at least one m6 A (Dominissini et al., 2012; Meyer et al., 2012). N6 methylation also occurs on Am to form a dimethylated adenosine: N6 ,20 -O-dimethyladenosine (m6 Am) (Keith et al., 1978; Wei et al., 1975). Am is primarily located at the first transcribed nucleotide position in mRNAs, adjacent to the m7 G cap. Nucleotides located at the first transcribed nucleotide position in an mRNA are typically methylated on the ribose at the 20 -hydroxyl position. However, if this nucleotide is Am, it can undergo further N6 methylation to m6 Am. Because m6 Am is present at the first transcribed nucleotide in $30% of all cellular mRNAs, m6 Am can affect the fate of a large subset of the transcriptome (Wei et al., 1975). Recent studies have started to reveal the functions of m6 Am. m6 Am is enriched in mRNAs with high stability and translation efficiency (Mauer et al., 2017). Mechanistically, m6 Am was shown to impair mRNA decapping mediated by DCP2, leading to increased stability of at least some m6 Am-modified mRNAs (Mauer et al., 2017). However, DCP2 does not regulate the stability of most mRNAs in the cell (Li et al., 2011, 2012). Instead, DCP2 is important for specific mRNA degradation pathways, such as nonsense-mediated decay, microRNA-mediated mRNA degradation, and mRNA degradation in response to interferon (Li et al., 2011, 2012). Therefore, it is not clear whether m6 Am has a general role in promoting the high stability of m6 Am-containing transcripts or whether m6 Am confers mRNA stability to a subset of mRNAs that are degraded though selective decapping pathways. Predicting the function of m6 Am is complicated by the difficulty in determining whether an mRNA contains m6 A or m6 Am. Transcriptome-wide mapping of m6 A and m6 Am uses antibodies that bind the 6-methyladenine (6mA) nucleobase portion found in both of these methylated adenosine nucleotides. The two mapping methods, i.e., MeRIP-seq (methyl RNA immunoprecipitation followed by sequencing) (Dominissini et al., 2012; Meyer et al., 2012) and miCLIP (m6 A individual-nucleotide-resolution Molecular Cell 75, 631–643, August 8, 2019 ª 2019 Elsevier Inc. 631 crosslinking and immunoprecipitation) (Linder et al., 2015), both map sites of 6mA rather than m6 A or m6 Am. The 6mA ‘‘peaks’’ are then interpreted to be either m6 A or m6 Am using a variety of criteria. For example, if the 6mA peak is in the 50 UTR, this suggests that the 6mA peak is caused by m6 Am because this nucleotide is exclusively found as the transcription start nucleotides. Nevertheless, it can be difficult to distinguish m6 Am from m6 A located within the 50 UTR of mRNAs. As a result, previous maps of m6 Am may have inaccuracies, which may make it difficult for predicting its function. To definitively distinguish m6 A and m6 Am in transcriptomewide maps, depletion of either m6 A or m6 Am would be required. m6 A depletion cannot be readily achieved, as Mettl3 is essential for survival in nearly all 341 cell lines that were screened (Tsherniak et al., 2017). The methyltransferase that generates m6 Am is not known, but its depletion could enable the identification of the sites that are m6 Am, because the remaining sites would be m6 A. Here, we describe the identification of PCIF1 as the methyltransferase that is responsible for generating m6 Am in mRNA. We show that PCIF1 methylates Am in the context of the m7 G cap and has negligible ability to methylate adenosine in mRNA outside this context. By mapping m6 A in the transcriptome of PCIF1-deleted cells, we distinguish between m6 Am and 50 UTR m6 A. We find numerous examples where previously annotated m6 Am sites reflect m6 A and vice versa. We show that transcript isoforms with alternative transcription start sites (TSSs) account for many of these discrepancies facilitating the identification of these ‘‘internal’’ TSSs. Using this new high-confidence map of m6 Am sites, we characterize the fate of m6 Am-modified mRNAs in PCIF1 knockout cells and show that m6 Am has negligible effects on translation under basal conditions but is associated with increased stability of a subset of m6 Am-initiated transcripts. Overall, our studies identify PCIF1 as the methyltransferase that generates m6 Am in the transcriptome and provides revised transcriptome-wide maps that distinguish between m6 A and m6 Am. RESULTS Identification of PCIF1 as a Candidate m6 Am-Forming Methyltransferase Studies in the 1970s provided initial characterization of an enzymatic activity in HeLa cells that synthesizes m6 Am (Keith et al., 1978). This enzyme selectively methylates Am adjacent to an m7 G cap in synthetic RNA substrates (Keith et al., 1978). In order to identify the m6 Am-forming enzyme, we performed a comparative bioinformatic analysis of orphan adenosine methyltransferases. These enzymes contain the [DNSH]PP[YFW] motif, which is present in all adenine N6-methyltransferases (Iyer et al., 2016). Among these putative adenine methyltransferases, PCIF1 is notable because it evolved at the same time that the 50 cap emerged in mRNA (Iyer et al., 2016). It has been hypothesized that the 50 cap emerged with eukaryotic evolution to replace the Shine-Dalgarno sequence and direct ribosomes to mRNAs and to protect from 50 exoribonucleases to distinguish selfversus-foreign mRNAs (Furuichi et al., 1977; Shimotohno et al., 1977; Shuman, 2002). The PCIF1 methyltransferase family is derived from the prokaryotic M.EcoKI/M.TaqI methyltransferases of the bacterial restriction-modification systems (Iyer et al., 2016). All of these methyltransferases contain helices before and after the conserved core strand-3, which display partial or complete degeneration into coil elements. Another feature of these methyltransferases is the addition of a conserved residue from a helix N-terminal to the core methyltransferase catalytic domain. A PCIF1 crystal structure revealed that its putative methyltransferase domain indeed adopts the classical Rossmann fold of many RNA methyltransferases (Akichika et al., 2019). PCIF1 also contains a WW domain that interacts with the C-terminal domain of RNA polymerase II (Fan et al., 2003; Figure 1A), suggesting that its function is linked to transcription. Based on this, we asked whether PCIF1 is an adenine N6-methyltransferase in mRNA. PCIF1 N6-Methylates 20 -O-Methyladenosine in an m7 G Cap-Dependent Manner In Vitro To identify potential PCIF1-dependent nucleotide methyltransferase activity, we bacterially expressed and purified glutathione S-transferase (GST)-tagged PCIF1. To test whether PCIF1 can methylate the cap-adjacent adenosine of mRNAs, we performed in vitro methyltransferase assays with an RNA oligonucleotide containing a 50 m7 G cap followed by 20 -O-methyladenosine (m7 G-ppp-Am-N16) (Figure 1B). We found that PCIF1 methylates Am in this RNA to produce m6 Am, as assessed by ultra-high-performance liquid chromatography coupled with triple-quadrupole tandem mass spectrometry (UHPLC-MS/MS; Figure 1C). Interestingly, we did not detect m6 A formation in these methylation reactions despite the presence of 5 internal adenosines in the RNA sequence (Figure 1C). Although PCIF1 may methylate an internal adenosine in a currently unknown sequence context, these findings suggest that PCIF1 preferentially N6-methylates 20 -O-methyladenosine rather than internal adenosines. As a control, we generated predicted catalytically inactive PCIF1 by mutating both asparagine 553 and phenylalanine 556 to alanines (NPPF/APPA) or to a serine and a glycine (NPPF/SPPG). The corresponding mutations inactivate the EcoKI and Dam N6 methyltransferases (Guyot et al., 1993; Willcock et al., 1994). Neither the APPA nor SPPG mutant was able to methylate RNA (Figure 1C), suggesting that the PCIF1 catalytic domain is required for Am methylation in vitro. We next asked whether PCIF1-mediated N6 methylation of the m7 G-adjacent A requires 20 -O-methyl modification on the A. To test this, we used an RNA substrate with a 50 m7 G cap followed by adenosine (m7 G-ppp-A-N16) rather than Am. Using in vitro methylation assays, we found that wild-type PCIF1, but not the SPPG or APPA PCIF1 mutant, was able to N6-methylate adenosine to m6 A (Figure 1D). We next examined the rate and substrate preference of PCIF1 using a serial dilution of the m7 G-ppp-Amand the m7 G-ppp-A-capped oligonucleotides and tritiated S-adenosyl methionine [3 H]-SAM as the methyl donor (Figure 1E). Michaelis-Menten analysis yielded a KM = 82 ± 18.2 nM for the capped 20 -O-methylated RNA and a KM = 630 ± 84.2 nM for the capped unmethylated A RNA (Figure 1F), suggesting that PCIF1 has an $7.6-fold higher preference for binding the 20 -Omethylated adenosine substrate. Notably, the m7 G moiety was required for methylation as PCIF1 efficiently methylated Am to m6 Am in an m7 G capped RNA (m7 G-ppp-Am-N16) but was unable to methylate Am in an 632 Molecular Cell 75, 631–643, August 8, 2019 A C B D E 0 2 4 6 8 10 12 14 m6 Amorm6 A(nM) m6 Am m6 A SPPG WT PCIF1: APPA m7 G-ppp-AmGUGGACUAACCACCAU *** *** ns ns 0 5 10 15 20 25 30 35 m6 A(nM) SPPG WT PCIF1: APPA m7 G-ppp-AGUGGACUAACCACCAU *** *** G H Nterm Cterm 704aa amino acids WW domain 42-79 NLS 109-113, 669-684 Predicted catalytic region 381-662 NPPF catalytic residues 553-556 sgRNA1 sgRNA2 PCIF1 F m7 G-ppp-AmGUGGACUAACCACCAU m7 G-ppp-AGUGGACUAACCACCAU ppp-AmGUGGACUAACCACCAU 1C, 1E, 1F, 1G 1D, 1F 1G Oligo used Figure Used in: 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 0 1 2 3 4 5 6 3HCPM Minutes 2 μM 1 μM 500 nM 250 nM 125 nM 62.5 nM 31.25 nM 2 μM m7G-ppp-Am oligo Km = 82nM Km = 630nM m7G-ppp-Am m7G-ppp-A Km = 82 nM Km = 630 nM mm7G-ppp-Am m77G-ppp-A Kmm = 82 nM KKm = 630 nM m7G-ppp-A… or Am... (µM) 0 1000 2000 3000 4000 5000 6000 m6 Am(pM) m7 GpppAm… …AmGUGGACUAACCACCAU pppAm… *** WT PCIF1 G-ppp-A Totalcelllysate αFlagαeIF4G HeLa Flag::PCIF1 IP beads eluate αeIF4E m7G-ppp-A m7G-ppp-A Vehicle G-ppp-A Vehicle Figure 1. PCIF1 N6 Methylates 20 -O-Methyladenosine In Vitro in an m7 G Cap-Dependent Manner (A) Schematic of PCIF1 indicating the position of predicted functional domains. The location of the sites of mutations used in the study is shown. The catalytic domain includes a four-amino-acid motif, NPPF, which is predicted to be essential for mediating methylation (Iyer et al., 2016). The location of the site guide RNAs (gRNAs) (50 -CGGUUGAAAGACUCCCGUGG-30 and 50 -ACUUAACAUAUCCUGCGGGG-30 ) used in Figure 2 is indicated. (B) Oligonucleotide sequences used in methyltransferase assays. (C) PCIF1 methylates m7 G-ppp-Am-N16 RNA. GST-PCIF1 (50 nM), but not the catalytically inactive mutants APPA or SPPG, efficiently converts m7 G-ppp-Am (4 mM) to m7 G-ppp-m6 Am as assessed by UHPLC-MS/MS. Under the same conditions (SAM, 160 mM, 10 min), PCIF1 does not convert any of the 5 internal adenosines to m6 A. Each bar represents the mean ± SEM of 3 independent experiments. n.s, not significant; ***p < 0.001; as assessed by unpaired Student’s t tests. (D) PCIF1 methylates cap-adjacent adenosine regardless of 20 -O-ribose methylation. GST-PCIF1, but not the APPA or SPPG PCIF1 mutants, efficiently converts m7 G-ppp-A-N16 (4 mM) to m7 G-ppp-m6 A-N16. Assays were performed as in (C). Each bar represents the mean ± SEM of 3 independent experiments. ***p < 0.001, as assessed by unpaired t tests. (E) PCIF1 enzyme kinetics. m7 G-ppp-Am-N16 (at indicated concentration) was incubated with GST-PCIF1 (20 nM) for the indicated times in the presence of 1.33 mM 3 H-SAM and 10 mM SAM. Methylation was determined by the presence of 3 H in the RNA, as assessed by scintillation counting. Each point represents the mean ± SEM of 3 independent experiments. (F) Michaelis-Menten kinetics of PCIF1 methylytransferase activity toward m7 G-ppp-Am and m7 G-ppp-A. Each point represents the mean ± SEM of 3 independent experiments. (G) PCIF1 activity depends on the presence of the m7 G cap. m7 G-ppp-Am-N16 or ppp-Am-N16 (4 mM) was incubated with GST-PCIF1 as in (C). PCIF1 converted Am to m6 Am specifically in the m7 G capped RNA. Each bar represents the mean ± SEM of 3 independent experiments. ***p < 0.001, as assessed by unpaired t tests. (H) PCIF1 directly binds the m7 G cap. Anti-FLAG immunoblotting was used to detect binding of 33FLAG-PCIF1 from HeLa cell extracts to m7 GTP-conjugated beads. The beads were eluted with m7 G-ppp-A or G-ppp-A. eIF4E and eIF4G were used to control for binding to m7 G. Molecular Cell 75, 631–643, August 8, 2019 633 A B C D E F G H I Figure 2. PCIF1 N6 Methylates 20 -O-Methyladenosine in Cells (A) CRISPR-mediated PCIF1 knockout (KO) in HEK293T cells was assessed by anti-PCIF1 immunoblotting. The upper band represents endogenous PCIF1, whereas the lower band is a non-specific band. b-actin, loading control. (legend continued on next page) 634 Molecular Cell 75, 631–643, August 8, 2019 RNA that lacked the m7 G cap (ppp-Am-N16; Figure 1G). Overall, these biochemical assays suggest that PCIF1 methyltransferase activity toward Am depends on the presence of the m7 G cap but does not require 20 -O-methylation on the adenosine. We next asked whether the PCIF1 preference for m7 G-capped RNA was due to an ability to bind the m7 G cap. We performed cap-binding assays with PCIF1 using 7-methylguanosine-5triphosphate (m7 G-ppp)-coupled Sepharose beads. In these experiments, we used lysates from HeLa cells expressing FLAGtagged wild-type PCIF1. As expected, cap-binding proteins eIF4E and eIF4G were bound to m7 G-ppp beads and were efficiently eluted using m7 G-ppp-A, but not G-ppp-A (Figure 1H). Similarly, PCIF1 bound to the m7 G-ppp beads and was eluted with m7 G-ppp-A, but not G-ppp-A (Figure 1H). Together, these data suggest that PCIF1 binds directly to the m7 G cap, which may account for its specificity toward adenosine adjacent to the m7 G. These results are consistent with the crystal structure of PCIF1, which shows specific interactions with m7 G (Akichika et al., 2019). PCIF1 Knockout Abolishes m6 Am Levels without Affecting m6 A in RNA To determine the ability of PCIF1 to generate m6 Am in cells, we used CRISPR to delete PCIF1 in various cell lines and examined levels of m6 Am and m6 A in RNA (Figures 2A and S1A). To measure m6 Am, we used a two-dimensional thin-layer chromatography (2D-TLC)-based method that can measure both m6 Am and Am, allowing the ratio of these modified forms of adenosine to be calculated in mRNA (Kruse et al., 2011). In this assay, mRNA is decapped, and the 50 nucleotide is selectively radiolabeled with [32 P]-ATP by polynucleotide kinase (PNK). Thus, the first transcribed nucleotide in RNA samples can be selectively detected and quantified. As expected, all the known nucleotides located at the first transcribed nucleotide in mRNA were detected, i.e., m6 Am, Am, Gm, Cm, and Um. However, in PCIF1 knockout cells, a selective and complete loss of m6 Am was detected (Figures 2B and S1B). A similar effect was seen using UHPLC-MS/MS to quantify m6 Am (Figure 2C). Thus, PCIF1 is required for the presence of m6 Am at the first transcribed nucleotide in mRNA. We next asked whether PCIF1 deletion affects m6 A levels in mRNA. To test this, we used a 2D-TLC-based method that selectively detects m6 A in the G-A-C context (Zhong et al., 2008) and complemented this with UHPLC-MS/MS to quantify all m6 A in mRNA. m6 A was readily detected in mRNA in control cells, and no reduction was seen in PCIF1 knockout cells (Figures 2D and 2E). To confirm that the loss of m6 Am in the PCIF1 knockout cells was due to a loss of PCIF1 itself, we performed rescue experiments. In these experiments, we used wild-type or the SPPG catalytically inactive PCIF1 mutant (Figures S1C and S1D). We found that re-expression of the wild-type, but not the catalytically inactive, PCIF1 restored m6 Am levels in mRNA of HEK293T PCIF1 knockout cells as assessed by 2D-TLC (Figure 2F) and by UHPLC-MS/MS (Figure 2G). We next asked whether PCIF1 was sufficient to increase m6 Am levels in cells. We found that PCIF1 overexpression in HEK293T cells (Figure 2H) led to a $3-fold increase in the m6 Am-to-Am ratio (Figure 2I). This increase in m6 Am levels was dependent on the catalytic activity of PCIF1, as overexpression of a catalytically inactive PCIF1 mutant had no effect on m6 Am levels (Figure 2I). Together, these data suggest that PCIF1 is both necessary and sufficient to generate m6 Am in mRNA in cells. miCLIP Analysis of PCIF1 Knockout Cells Distinguishes m6 Am from 50 UTR m6 A Residues Next, we used the PCIF1 knockout cells to distinguish m6 Am and m6 A in transcriptome-wide 6mA maps. We performed miCLIP, a method that produces narrow peaks, and nucleotide transitions at and adjacent to the m6 A (Linder et al., 2015). m6 A is nearly universally followed by cytosine in mRNA (Wei et al., 1976). This C is frequently observed to undergo a C-to-T transition as a result of antibody crosslinking in miCLIP, which can then be used to identify m6 A (Linder et al., 2015). Because m6 Am can also be followed by cytosine, C-to-T transitions alone are not sufficient to distinguish m6 A from m6 Am. Peaks caused by m6 Am display a unique shape that exhibits a marked drop off of reads at an annotated A-starting TSS, and this feature can be used to identify m6 Am (Linder et al., 2015). However, because m6 A occurring near the TSS would also produce a similar shaped peak, this approach may result in false-positive m6 Am identifications. (B) PCIF1 is required for formation of m6 Am in mRNA in cells. m6 Am and Am levels in poly(A) RNA were detected by radiolabeling the 50 nucleotide after decapping. RNA hydrolysates were resolved by 2D-TLC. Representative images are shown from 3 biological replicates. The bar graph on the right represents the mean ± SEM of 3 independent experiments. ***p < 0.001; Student’s t test. (C) m6 Am is depleted in PCIF1 KO HEK293T cells as assessed by UHPLC-MS/MS. Each bar represents the mean ± SEM of 3 independent experiments. ****p < 0.0001 as assessed by paired t tests. (D) PCIF1 does not affect the level of internal m6 A. 2D-TLC analysis of poly(A) RNA from HEK293T (wild-type [WT]) and PCIF1 KO HEK293T show no effect on the level of m6 A. (E) Internal m6 A is not affected in PCIF1 KO HEK293T cells as assessed by UHPLC-MS/MS. Each bar represents the mean ± SEM of 3 independent experiments. ns as assessed by paired t tests. (F) Wild-type, but not a catalytically inactive PCIF1 mutant, restores m6 Am levels in PCIF1 knockout cells as assessed by 2D-TLC. Each bar represents the mean ± SEM of two independent experiments. (G) Wild-type, but not catalytically inactive PCIF1 mutant, restores m6 Am levels in PCIF1 KO cells as assessed by UHPLC-MS/MS. Each bar represents the mean ± SEM of two independent experiments. *p < 0.05 as assessed by paired t tests. (H) Western blot analysis demonstrates equivalent expression of wild-type and catalytically inactive FLAG-tagged PCIF1. b-actin, loading control. (I) Overexpression of wild-type, but not catalytically inactive, PCIF1 increases m6 Am levels in HEK293T cells as assessed by 2D-TLC. Upper and left panels show representative images of 3 independent experiments. The bar graph represents the mean ± SEM of 3 independent experiments. ****p % 0.0001, unpaired t tests. See also Figure S1. Molecular Cell 75, 631–643, August 8, 2019 635 Furthermore, these approaches are highly dependent on transcript annotations that may not have accurate TSS information for the cell type investigated. For example, annotated TSSs produced by RefSeq and ENSEMBL differ frequently for the same gene (Zhao and Zhang, 2015). Therefore, true m6 Am peaks may have been discarded or thought to be m6 A based on their location away from a TSS. We therefore performed miCLIP in control and PCIF1 knockout cells to distinguish m6 Am and m6 A. In control cells, reads were enriched in the vicinity of the stop codon as well as the TSS, which is generally assumed to reflect m6 A and m6 Am, respectively (Figure 3A). PCIF1 knockout cells exhibited fewer reads mapping near the annotated TSS (Figure 3A; $55% decrease in 50 UTR), suggesting these reads derive from an m6 Am residue. A motif analysis of significant peaks showed the DRACH m6 A consensus (D = A, G, U; R = A, G; H = A, C, U) as the most common motif in each dataset (Figure 3B). This suggests that m6 A is the most common modification mapped in both datasets, as expected. To identify m6 Am marked transcripts, we next examined the 6mA peaks that showed differences in the control and PCIF1 knockout miCLIP datasets. As expected, we detected a loss of peaks near the TSS of certain genes in the PCIF1 knockout. For example, RPL35 and KDELR2 show peaks near the annotated TSS as well as at internal sites (Figure 3C). The TSS-proximal peaks were absent in the PCIF1 knockout miCLIP dataset. These data are consistent with the idea that the TSS peaks predominantly reflect m6 Am. However, in some cases, the peaks near the TSS were not affected in the PCIF1-knockout dataset. For example, peaks near the TSSs of RACK1 and RPS5, which were previously annotated as m6 Am in HEK293T cells based on their location, peak shape, and lack of C-to-T transitions (Mauer et al., 2017), persist in the PCIF1 knockout dataset (Figures 3D and S2A). These peaks contain a canonical DRACH m6 A consensus motif, and C-to-T transitions are detected for RACK1 (Figure 3D), suggesting that these sites are actually m6 A. The variability in C-to-T transitions reflects the low transition rate induced by the antibody adduct on this transcript. Altogether, these data indicate that PCIF1 depletion can be used to determine the identity of an m6 A peak. Overall, only 60.2% of genes that had previously been annotated as m6 Am (Mauer et al., 2017) were validated as m6 Am based on their loss in PCIF1 knockout cells. In some cases, A B C D E F Figure 3. Depletion of PCIF1 Distinguishes m6 A and m6 Am in Transcriptome-wide 6mA Maps (A) Metagene of miCLIP reads in wild-type and PCIF1 KO HEK293T cells. Shown is a metagene analysis of reads from the wild-type or PCIF1 KO miCLIP dataset. The first nucleotide of each read (with respect to the RNA strand) was extracted and plotted. Reads in the 50 UTR were lost in the PCIF1 knockout, suggesting a complete loss of m6 Am in the PCIF1 knockout cells. (B) DREME motif search within called peaks show the DRACH motif as the most enriched in all datasets, consistent with m6 A as the most abundant 6mA-containing nucleotide mapped by miCLIP. (C) miCLIP peaks can be identified as m6 Am or m6 A based on their decrease in PCIF1 KO cells. Genome tracks were plotted for RPL35 and KDELR2 with called m6 A sites (false discovery rate [FDR] < 0.1) and m6 Am sites indicated by red circles and blue triangle, respectively. Zoomed insets show m6 Am peaks can be distinguished from nearby m6 A sites. (D) The previously annotated m6 Am site in RACK1 is actually a 50 UTR m6 A. The TSS-proximal peak in RACK1 is not affected in the PCIF1 knockout and overlaps with a DRACH motif. (E) Metagene analysis of PCIF1 KO-validated m6 Am sites shows m6 Am sites throughout the 50 UTR and in thetranscriptbody.Shownisametageneoftheexact sites of m6 Am within the PCIF1-dependent peaks as determined by A-to-T transitions and the read dropoff method. The metagene reveals an overall enrichment at the TSS, with some sites that appear to be within the coding sequence (CDS) and 30 UTR. (F) DREME motif search of the nucleotides surrounding each m6 Am was performed, confirms the previously reported BCA motif, and shows that the promoter sequence upstream of the m6 Am is GC enriched. See also Figure S2. 636 Molecular Cell 75, 631–643, August 8, 2019 this could be explained by peaks being below the threshold for detection in one or both replicates. Nevertheless, this difference highlights the importance of depleting cells of PCIF1 to reduce false-positive m6 Am identification. A High-Confidence Transcriptome-wide Map of m6 A and m6 Am Based on PCIF1 Depletion To create a high-confidence map of all m6 Am sites in the transcriptome, we searched for all peaks that exhibit a marked reduction in miCLIP signal in the PCIF1 knockout dataset. The majority of peaks showed no substantial difference between control and PCIF1 knockout miCLIP datasets, suggesting that they are m6 A (Figure S2B). However, 2,360 peaks overlapping 2,291 genes exhibited a significant reduction in both PCIF1 knockout datasets (Figure S2B). In contrast, only 11 sites appeared to increase, suggesting a very low incidence of false positives. We next identified the exact m6 Am residue within each of these peaks. In our previous approach, we used a ‘‘pile up’’ of reads that drop off at the 50 end of these read clusters in A-starting genes to predict the m6 Am site (Linder et al., 2015). In some cases, the drop off is not easily detected or several of these were found in close proximity. This appears to occur when (1) the total reads are too few or (2) the reads terminate before the TSS, possibly due to impaired reverse transcription through the 20 -O-methyl modifications (Maden et al., 1995) in the cap-proximal nucleotides or due to non-templated nucleotide addition that occurs at the ends of cDNAs generated by reverse transcriptases (Chen and Patton, 2001). Therefore, we wanted to develop an alternative approach to identify m6 Am within the PCIF1-dependent peaks. Previously, we observed antibody-induced A-to-T transitions at the m6 A site in miCLIP (Linder et al., 2015). We confirmed that A-to-T transitions are readily detected at known m6 Am and m6 A throughout the transcriptome (Figures S2C and S2D). Therefore, we used a 10% A-to-T transition rate to identify the m6 Am within PCIF1-dependent peaks. The drop-off approach was used when the A-to-T transition rate did not meet these criteria (Figure S2E). There was high similarity in the m6 Am sites that were called when using these methods separately (Figure S2F). Overall, the 2,350 m6 Am sites mapped based on their dependence on PCIF1 (Table S1) were primarily located throughout the 50 UTR ($94%), with a prominent enrichment at the annotated TSS (Figure 3E). Motif analysis of the genomic context of the exact m6 Am nucleotide revealed the BCA motif, with A representing the m6 Am and BC representing upstream genomic nucleotides (B = C, G, or T), as reported previously for m6 Am (Linder et al., 2015). Additionally, motif analysis shows the upstream promoter sequence is GC enriched (Figure 3F). Motifs downstream and including the transcription-start adenosine were also enriched, suggesting that m6 Am occurs in specific sequence contexts within mRNA transcripts (Figure S2G). Next, we mapped m6 A in the 50 UTR. C-to-T transitions in a DRACH consensus were used to call on average 44,025 m6 A sites in 10,383 genes based on the miCLIP protocol (Linder et al., 2015). This identified 399 50 UTR m6 A sites that were robustly called across all datasets (Table S2). We next asked whether mRNAs with 50 UTR m6 A and mRNAs that contain m6 Am are linked to different cellular processes, based on our updated m6 A and m6 Am sites. Functional annotation using DAVID (Database for Annotation, Visualization and Integrated Discovery) shows that transcripts containing these distinct modified nucleotides are linked to different cellular processes, with 50 UTR m6 A associated with processes such as transcription and cell division, and m6 Am is primarily associated with splicing (Figures S2H and S2I; Table S3). ATF4 Contains a m6 Am Rather Than m6 A in Its 50 UTR We next wanted to understand whether our revised map of m6 Am and m6 A can identify transcripts with misannotated modified nucleotides. A 50 UTR m6 A site has been described as mediating the unusual stress-regulated translation of ATF4 (Zhou et al., 2018). ATF4 has two upstream open reading frames (uORFs) in its 50 UTR. In unstressed cells, the uORFs are translated, which prevents translation of the main open reading frame, which encodes the ATF4 protein (Vattem and Wek, 2004). However, during stress, the second uORF is skipped, and the ribosome scans to the main open read frame after translating the first uORF. This allows the ATF4 protein to be translated during stress. m6 A was mapped to the second open reading frame and was described as disappearing in a stress-dependent manner, thus causing a stress-regulated switch in ATF4 translation (Zhou et al., 2018). However, using miCLIP, it is apparent that the 6mA peak in the 50 UTR of ATF4 is not located within the second open reading frame (Figure S3A). Instead, the peak is located at the transcription-start nucleotide and does not overlap with the position of the putative m6 A. Based on the location of the peak, we asked whether it instead reflects m6 Am rather than m6 A. To test this, we examined ATF4 in the PCIF1 knockout miCLIP dataset. Here, we observed a complete loss of this peak, further confirming that this site is m6 Am (Figure S3A). The role ofm6 A incontrolling stress-induced Atf4 translationwas described in mouse embryonic fibroblast cells (Zhou et al., 2018) rather than the HEK293T cells used here. Human cells appear to have lost the DRACH consensus sequence surrounding the putative m6 A site (Figure S3B). Conceivably, human cells exhibit stress-induced regulation of ATF4 translation through an m6 A-independent pathway and mouse cells utilize an m6 A-dependent pathway. Therefore, we mapped 6mA in mouse embryonic fibroblasts using miCLIP (Figure S3C). Again, the 6mA peak was at the TSS, not at a position corresponding to the second uORF (Figure S3D). These data further show that this peak derives from a m6 Am residue. In comparison, there were low levels of 6mA reads throughout the transcript body, suggesting either background reads or low stoichiometry m6 A sites (Figure S3D). Overall, these data suggest that ATF4 contains a m6 Am at the TSS but no prominent m6 A site within the 50 UTR as previously reported (Zhou et al., 2018). Thus, a role for a 50 UTR m6 A in regulating ATF4 translation seems unlikely. Overall, these data demonstrate the ease with which m6 A and m6 Am sites can be confused for each other. Identification of Internal 6mA Sites that Reflect Transcription-Start m6 Am We noticed two unusual features in our mapping results. First, not all m6 Am sites mapped to regions within annotated mRNA Molecular Cell 75, 631–643, August 8, 2019 637 transcripts. Second, the m6 Am metagene showed that, although 94% of m6 Am sites were located in the 50 UTR, many were not directly at the annotated TSS and, in some cases, further downstream within the transcript body (Figure 3E). We considered that these findings could be due to m6 Am that occurs in mRNA isoforms with alternate TSSs upstream or downstream of the TSS in the RefSeq-annotated transcript. To test this, we created an m6 Am metaplot relative to RefSeq-annotated TSSs (Figure 4A). We observed 16.7% of m6 Am sites mapping within 250 nt upstream of annotated TSSs, suggesting that some m6 Am occurs in isoforms with upstream TSSs. We similarly observed m6 Am upstream of TSSs using GENCODE (Frankish et al., 2019) transcript annotations (Figure 4B). The FANTOM5 promoter-level expression atlas (Abugessaisa et al., 2017) uses a set of TSSs specifically mapped across multiple tissues using the cap analysis gene expression (CAGE) approach. Using FANTOM5, we observed a marked overlap with our m6 Am sites, supporting the idea that these m6 Am sites are indeed TSSs (Figure 4C). TSS heterogeneity likely explains why some m6 Am sites map within the 50 UTR rather than being solely located at the annotated TSS. In the case of YBX1, a 6mA peak is mapped to the 50 UTR and is lost in the PCIF1 knockout miCLIP dataset, suggesting that this peak is due to m6 Am (Figure 4D). This m6 Am likely reflects an isoform with a TSS located at this m6 Am site, based on its overlap with a CAGE peak (Figure 4D). Thus, the presence of m6 Am within the 50 UTR likely reflects TSS heterogeneity rather than ‘‘internal’’ m6 Am nucleotides. We next wanted to understand why $6% of m6 Am sites (121 sites) appear to map to coding sequences or 30 UTR regions (Figure 4E). For example, YOD1 shows an internal m6 A peak in the first exon that is lost in the PCIF1 knockout miCLIP dataset (Figure 4F). As with YBX1, we observed a TSS that overlapped with the m6 Am site. Thus, this internal site, which would normally have been assumed to be m6 A using MeRIP-seq and possibly miCLIP, derives from an isoform starting with m6 Am. To test this idea further, we performed a metagene analysis on m6 Am sites mapping to coding sequences or the 30 UTR. Here, we plotted the distance to the nearest CAGE sites (Figure 4G). This analysis shows that many m6 Am sites in the coding sequence and 30 UTR are located at or near CAGE sites. Additionally, these m6 Am sites show the BCA motif, which resembles the transcription initiation site (initiator element [Inr]) motif of RNA polymerase II (Figure S2G; Yang et al., 2007). This motif was also previously found for m6 Am mapped to canonical TSSs (Linder et al., 2015). 50 RACE confirmed that our called m6 Am sites are indeed transcription start nucleotides (Figure S3E). Overall, these data further suggest that m6 Am is not internally located within transcripts but is instead found at the TSSs. Approximately 8% of m6 Am sites that mapped to the coding sequence or 30 UTR also contained an adjacent C-to-T transition. As a result, these peaks would likely have been called as an m6 A. These data highlight the value of using PCIF1 depletion to validate the transcriptome-wide m6 Am and m6 A maps. m6 Am Correlates with Enhanced Translation, Expression, and Stability of mRNAs In our previous studies, we found that m6 Am is correlated with transcripts that are highly expressed and have long half-lives in cells (Mauer et al., 2017). We therefore wanted to re-examine this correlation based on the high-confidence m6 Am annotation based on peaks that were depleted in the PCIF1 knockout miCLIP dataset. In some cases, mRNAs that had been previously annotated as beginning with Am, Cm, Gm, or Um were re-annotated as m6 Am for this analysis, and mRNAs previously annotated as m6 Am were re-annotated as Am based on our revised mapping data. Analysis of mRNA expression and half-lives showed that transcripts that begin with m6 Am are indeed more highly expressed and stable than mRNAs with other start nucleotides. Notably, m6 Am appears to be the predominant start nucleotide of the mRNAs that are the most abundant and have annotated half-lives greater than 24 h (Figures 5A–5D). Overall, these data suggest that the presence of m6 Am correlates with an overall increase in mRNA stability and that m6 Am is the predominant starting nucleotide on ‘‘outlier’’ mRNAs with unusually high stability and expression. To determine whether the N6-methyl in m6 Am was required for the unique properties of these outlier mRNAs, we examined mRNA stability in PCIF1 knockout HEK293T cells. mRNA stability was measured using SLAM-seq (thiol(SH)-linked alkylation for the metabolic sequencing of RNA) (Herzog et al., 2017; Figure S4A). To examine the outlier mRNAs, which are highly expressed, we separately examined mRNAs in the lower and upper half of gene expression. We only used transcripts that exhibited a minimum threshold of transitions required for mRNA half-life quantification. For mRNAs in the lower half of gene expression, we observed a marked decrease in mRNA half-life upon PCIF1 depletion (Figure 5E). We confirmed this effect by examining the stability of individual mRNAs after treatment of control and PCIF1 knockout HEK293T cells with actinomycin D. Both NBR1 and AKAP12 transcripts exhibited decreased expression after 8 h of actinomycin D treatment (Figure S4B). This effect was more prominent in PCIF1 knockout cells, consistent with a stabilizing effect of m6 Am (Figure S4B). However, when we examined the more abundant mRNAs, which are enriched in the outlier transcripts, these transcripts did not show a substantial change in mRNA half-life (Figures 5F and S4C). We observed a slight reduction in stability relative to Am-annotated transcripts, but compared to all mRNAs (Am, Cm, Gm, and Um), these mRNAs appeared to show small but nonsignificant increase in mRNA stability in PCIF1 knockout cells. Thus, although m6 Am is highly enriched in these outlier transcripts, the N6 methyl does not appear to account for their unusual stability. In contrast, mRNAs in the lower half of gene expression appear to utilize m6 Am for transcript stability. Previously, we found that m6 Am-containing transcripts exhibit a subtle increase in translation relative to mRNAs with other start nucleotides (Mauer et al., 2017). To more directly test the role of m6 Am on translation, we compared the translation efficiency of transcripts in control and PCIF1 knockout cells by ribosome profiling (Figures S4D and S4E). Here, we found that transcripts that contained m6 Am as the transcription-start nucleotide did not show a substantial change in translation efficiency upon PCIF1 depletion (Figure S4F). Rather than showing a decrease in translation, we observed a slight increase in translation upon loss of m6 Am compared to transcripts annotated to begin with other nucleotides (Figure S4F). In agreement with the modest 638 Molecular Cell 75, 631–643, August 8, 2019 A B C D F G E Figure 4. Internally Mapped m6 Am Sites Reflect m6 Am in mRNA Isoforms with Alternative TSSs (A) A metaplot centered on the closest RefSeq TSS for each called m6 Am site shows most m6 Am sites are found downstream of the annotated start site, but not at the annotated start site. The proportion of m6 Am directly at the annotated TSS or up- or downstream is shown. (B) A metaplot analysis of m6 Am locations using GENCODE TSS annotations shows higher overlap with TSSs. GENCODE annotations include more transcript isoforms and TSSs than RefSeq. (C) A metaplot of the distance from each m6 Am site to the closest CAGE peak in the FANTOM5 database shows that m6 Am sites are indeed TSSs. Here, the overlap of m6 Am was highest, suggesting that m6 Am sites are selectively localized to TSSs and not internal nucleotides within mRNA. (D) The m6 Am mapping to the annotated 50 UTR of the YBX1 transcript reflects a transcript isoform. The PCIF1-dependent 6mA peak in YBX1 maps within the annotated 50 UTR of YBX1. However, this peak overlaps with a CAGE site (orange triangles), indicating the existence of a transcript isoform that initiates at this 6mA site. m6 Am peaks that appear within the 50 UTR reflect m6 Am in transcript isoforms with alternative TSSs. The exact m6 Am site (blue triangle) was determined using the A-to-T transition within the PCIF1-dependent peak. (E) Most m6 Ams are found in the annotated 50 UTR of transcripts. (F) The internally mapping m6 Am in YOD1 derives from a TSS of a YOD1 transcript isoform. The m6 Am peak in YOD1 begins beyond the start codon of both annotated isoforms. CAGE peaks (orange triangles) suggest this is indeed a TSS. (G) A metaplot analysis of CDS and 30 UTR mapping m6 Am sites show overlap with CAGE data, indicating that m6 Am occurs at TSSs. The closest CAGE peak to each of the 6% of sites that appeared to not map to the 50 UTR (E) was calculated and plotted. See also Figure S3. Molecular Cell 75, 631–643, August 8, 2019 639 effects of PCIF1 depletion on translation rates, we found that levels of proteins encoded by several m6 Am-modified mRNAs remained largely unchanged in PCIF1 KO cells (Figure S4G). Together, these experiments suggest that, under the conditions used in these experiments, N6 methylation does not mediate the increase in translation efficiency of m6 Am-initiated mRNAs in HEK293T cells. DISCUSSION A major challenge when mapping m6 A and m6 Am is that both nucleotides are recognized by 6mA-specific antibodies and both can produce peaks in the 50 UTR of mRNA transcripts. Here, by identifying PCIF1 as the m6 Am-forming methyltransferase and by depleting PCIF1 to definitively identify m6 Am sites, we present a revised annotation of m6 Am and m6 A in the transcriptome. We find that previous annotations contain errors that reflect the existence of mRNA isoforms that differ by TSSs. In some cases, the isoforms contain TSSs that map to internal sites within the annotated transcripts, resulting in the appearance of peaks that would otherwise be attributed to m6 A. The identification and characterization of PCIF1 coupled with precise m6 Am annotations generated by PCIF1 depletion will facilitate the identification of functions for m6 Am. Our studies provide insights into the function of m6 Am. Using our new high-confidence m6 Am map, we find that m6 Am is found on unusually stable and highly abundant transcripts in cells. However, depletion of m6 Am by PCIF1 knockout does not markedly impair the stability of these unusual transcripts under basal conditions. This suggests that m6 Am does not account for the stability of these unusual mRNAs. m6 Am therefore is likely to co-occur with other transcript features that confer these unusual properties to these mRNAs. However, m6 Am does promote the stability of other mRNAs. When we examined mRNAs in the lower half of gene expression, we found a marked drop in mRNA stability in PCIF1-depleted cells. Why might some mRNAs be stabilized by m6 Am and others may not be affected? First, m6 Am is enriched in different sequence contexts. This contrasts with m6 A, which is nearly always found in a single sequence context. m6 Am is therefore Figure 5. mRNA Expression Level and mRNA Half-Life Depending on TSS (A) mRNAs with an annotated m6 Am start nucleotide show higher mRNA expression than other mRNAs. mRNA expression level in wild-type HEK293T cells was based on the first annotated nucleotide and an earlier m6 Am map (Mauer et al., 2017). Transcripts that start with m6 Am are significantly upregulated. ****p < 2.2 3 10À16 ; Student’s t test. Cumulative distribution plot and boxplot represent the expression for mRNAs starting with m6 Am, Am, Cm, Gm, and Um. Data shown are the average gene expression measured from two replicates for HEK293T cells. (B) m6 Am mRNAs annotated using the PCIF1 KO miCLIP dataset show increased expression compared to mRNAs with other start nucleotides. Cumulative distribution plots were prepared as in (A) using the high-confidence m6 Am dataset. The transcripts starts with m6 Am are significantly upregulated as in (A). ****p < 2.2 3 10À16 ; Student’s t test. (C) mRNAs with an annotated m6 Am start nucleotide show higher mRNA half-life than other mRNAs. Annotated mRNA half-lives were based on the first annotated nucleotide and an earlier m6 Am map (Mauer et al., 2017). mRNAs with an annotated m6 Am exhibit a significantly elevated mRNA half-life than mRNAs with other annotated start nucleotides. ****p % 2.2 3 10À16 ; Student’s t test. (D) m6 Am mRNAs annotated using the PCIF1 KO miCLIP dataset show increased half-life compared to mRNAs with other start nucleotides. Transcripts with m6 Am have significantly longer half-life with similar p value. ****p % 2.2 3 10À16 ; Student’s t test. (E) Influence of PCIF1 depletion on mRNA half-life for transcripts in the lower half of gene expression. Transcripts with m6 Am have significantly shorter half-life in comparison to mRNAs with other annotated start nucleotides. *p = 0.0258 by Student’s t test. (F) Influence of PCIF1 depletion on mRNA half-life for highly expressed transcripts. Transcripts with m6 Am show no significant decrease in mRNA half-life in comparison to mRNAs with other annotated start nucleotides. n.s., Student’s t test. See also Figure S4. 640 Molecular Cell 75, 631–643, August 8, 2019 likely to have different sensitivities to decapping or bind to different m6 Am readers in a context-dependent manner. Thus, unlike traditional approaches, where mRNAs are binned and bioinformatically analyzed based on the presence or absence of a modification, m6 Am functions are more likely to be revealed by analysis of mRNAs binned based on m6 Am sequence contexts. Another important factor that might affect whether m6 Am has a destabilizing effect is whether the mRNA utilizes DCP2 or potentially other m6 Am-sensitive decapping mechanisms. Previous studies showed that m6 Am confers stability of mRNAs to DCP2-mediated decapping (Mauer et al., 2017). DCP2 is not the major decapping enzyme in cells, but DCP2 targets mRNAs with specific 30 UTR features. Thus, m6 Am may stabilize mRNAs when DCP2-dependent pathways are activated. Overall, our studies show that m6 Am acts in a transcript-selective manner rather than a general mRNA-stabilizing modification. Although our study focused on mRNA stability, another study examined m6 Am mRNA abundance and found no effects upon PCIF1 depletion (Akichika et al., 2019). This might reflect compensatory upregulation of m6 Am mRNAs. Notably, Akichika et al. examined all mRNAs rather than specific subsets of m6 Amannotated mRNAs. Two recent studies also reported PCIF1 as the m6 Am capdependent methyltransferase (Akichika et al., 2019; Sun et al., 2019). Akichika et al. found that m6 Am slightly enhances translation relative to the Am form of the transcript in PCIF1 knockout cells, based on ribosome profiling (Akichika et al., 2019). Our ribosome profiling analysis of PCIF1 knockout cells showed a slight repressive effect of m6 Am on translation. Therefore, mRNAs modified with m6 Am are efficiently translated, but Ammodified mRNAs are even slightly more efficiently translated. Regardless, both our study and the Akichika et al. study are consistent in finding a very minor effect on translation of m6 Am-annotated mRNAs upon PCIF1 depletion. It should be noted that effects of m6 Am on both mRNA translation and mRNA stability are likely to depend on the sequence motif following m6 Am and may be different during signaling or stress conditions that were not examined in our study. STAR+METHODS Detailed methods are provided in the online version of this paper and include the following: d KEY RESOURCES TABLE d LEAD CONTACT AND MATERIALS AVAILABILITY d METHOD DETAILS B Synthesis and characterization of synthetic oligonucle- otides B Cell culture B Antibodies B Generation of PCIF1 CRISPR knockout cells and overexpression cell lines B Protein expression and purification B In Vitro methyltransferase assays B UHPLC-MS/MS analysis B Cap-binding assay B Immunofluorescence B Determination of relative m6 Am, Am, and m6 A levels by thin layer chromatography B miCLIP B miCLIP bioinformatic analyses B Transcript 50 end cloning B SLAM-seq B Real-time PCR assay to determine transcript stability B Ribosome profiling B SLAM-seq bioinformatic analysis B Statistics and software d DATA AND CODE AVAILABILITY SUPPLEMENTAL INFORMATION Supplemental Information can be found online at https://doi.org/10.1016/j. molcel.2019.06.006. ACKNOWLEDGMENTS We thank J. Lipton for reagents and A.O. Olarerin-George for assistance with data analysis. This work was supported by NCN 2017/24/T/NZ1/00170 (D.T.-S.) and NIH grants R00AG043550 and DP2AG055947 (E.L.G.) and R01DA037755 (S.R.J.). AUTHOR CONTRIBUTIONS K.B. performed biochemical analysis of PCIF1 and generated PCIF1 KO/OE cell lines; D.T.-S. and K.B. performed assays of PCIF1 activity in cells; K.B., D.T.-S., and S.Z. performed and analyzed ribosome profiling data; D.T.-S. performed and analyzed SLAM-seq experiments; B.R.H. performed and analyzed miCLIP experiments; N.L. performed cap-binding experiments; K.T. performed experiments assessing the translational effect of PCIF1 KO; T.G., J.-J.V., and F.D. synthesized capped and uncapped RNA; L.A. identified PCIF1 as putative m6 Am methyltransferase; and E.L.G. and S.R.J. wrote the manuscript with input from all authors. DECLARATION OF INTERESTS S.R.J. is scientific founder of, advisor to, and owns equity in Gotham Therapeutics. Received: November 29, 2018 Revised: April 8, 2019 Accepted: June 2, 2019 Published: July 3, 2019 REFERENCES Abugessaisa, I., Noguchi, S., Hasegawa, A., Harshbarger, J., Kondo, A., Lizio, M., Severin, J., Carninci, P., Kawaji, H., and Kasukawa, T. (2017). FANTOM5 CAGE profiles of human and mouse reprocessed for GRCh38 and GRCm38 genome assemblies. Sci. Data 4, 170107. Akichika, S., Hirano, S., Shichino, Y., Suzuki, T., Nishimasu, H., Ishitani, R., Sugita, A., Hirose, Y., Iwasaki, S., Nureki, O., et al. (2019). Cap-specific terminal N6-methylation of RNA by an RNA polymerase II-associated methyltransferase. Science 363, eaav0080. Bailey, T.L., Boden, M., Buske, F.A., Frith, M., Grant, C.E., Clementi, L., Ren, J., Li, W.W., and Noble, W.S. (2009). MEME SUITE: tools for motif discovery and searching. Nucleic Acids Res. 37, W202–W208. Barral, K., Sallamand, C., Petzold, C., Coutard, B., Collet, A., Thillier, Y., Zimmermann, J., Vasseur, J.J., Canard, B., Rohayem, J., et al. (2013). Development of specific dengue virus 20 -O- and N7-methyltransferase assays for antiviral drug screening. Antiviral Res. 99, 292–300. Molecular Cell 75, 631–643, August 8, 2019 641 Bokar, J.A., Shambaugh, M.E., Polayes, D., Matera, A.G., and Rottman, F.M. (1997). Purification and cDNA cloning of the AdoMet-binding subunit of the human mRNA (N6-adenosine)-methyltransferase. RNA 3, 1233–1247. Chen, D., and Patton, J.T. (2001). Reverse transcriptase adds nontemplated nucleotides to cDNAs during 50 -RACE and primer extension. Biotechniques 30, 574–580, 582. Dodt, M., Roehr, J.T., Ahmed, R., and Dieterich, C. (2012). FLEXBAR-flexible barcode and adapter processing for next-generation sequencing platforms. Biology (Basel) 1, 895–905. Dominissini, D., Moshitch-Moshkovitz, S., Schwartz, S., Salmon-Divon, M., Ungar, L., Osenberg, S., Cesarkas, K., Jacob-Hirsch, J., Amariglio, N., Kupiec, M., et al. (2012). Topology of the human and mouse m6A RNA methylomes revealed by m6A-seq. Nature 485, 201–206. Fan, H., Sakuraba, K., Komuro, A., Kato, S., Harada, F., and Hirose, Y. (2003). PCIF1, a novel human WW domain-containing protein, interacts with the phosphorylated RNA polymerase II. Biochem. Biophys. Res. Commun. 301, 378–385. Frankish, A., Diekhans, M., Ferreira, A.M., Johnson, R., Jungreis, I., Loveland, J., Mudge, J.M., Sisu, C., Wright, J., Armstrong, J., et al. (2019). GENCODE reference annotation for the human and mouse genomes. Nucleic Acids Res. 47, D766–D773. Furuichi, Y., LaFiandra, A., and Shatkin, A.J. (1977). 50 -terminal structure and mRNA stability. Nature 266, 235–239. Guyot, J.B., Grassi, J., Hahn, U., and Guschlbauer, W. (1993). The role of the preserved sequences of Dam methylase. Nucleic Acids Res. 21, 3183–3190. Herzog, V.A., Reichholf, B., Neumann, T., Rescheneder, P., Bhat, P., Burkard, T.R., Wlotzka, W., von Haeseler, A., Zuber, J., and Ameres, S.L. (2017). Thiollinked alkylation of RNA to assess expression dynamics. Nat. Methods 14, 1198–1204. Ingolia, N.T., Brar, G.A., Rouskin, S., McGeachy, A.M., and Weissman, J.S. (2012). The ribosome profiling strategy for monitoring translation in vivo by deep sequencing of ribosome-protected mRNA fragments. Nat. Protoc. 7, 1534–1550. Iyer, L.M., Zhang, D., and Aravind, L. (2016). Adenine methylation in eukaryotes: Apprehending the complex evolutionary history and functional potential of an epigenetic modification. BioEssays 38, 27–40. Jia, G., Fu, Y., Zhao, X., Dai, Q., Zheng, G., Yang, Y., Yi, C., Lindahl, T., Pan, T., Yang, Y.G., and He, C. (2011). N6-methyladenosine in nuclear RNA is a major substrate of the obesity-associated FTO. Nat. Chem. Biol. 7, 885–887. Keith, J.M., Ensinger, M.J., and Moss, B. (1978). HeLa cell RNA (20 -O-methyladenosine-N6-)-methyltransferase specific for the capped 50 -end of messenger RNA. J. Biol. Chem. 253, 5033–5039. Kruse, S., Zhong, S., Bodi, Z., Button, J., Alcocer, M.J., Hayes, C.J., and Fray, R. (2011). A novel synthesis and detection method for cap-associated adenosine modifications in mouse mRNA. Sci. Rep. 1, 126. Lavergne, T., Bertrand, J.R., Vasseur, J.J., and Debart, F. (2008). A base-labile group for 20 -OH protection of ribonucleosides: a major challenge for RNA synthesis. Chemistry 14, 9135–9138. Lavergne, T., Janin, M., Dupouy, C., Vasseur, J.J., and Debart, F. (2010). Chemical synthesis of RNA with base-labile 20 -o-(pivaloyloxymethyl)-protected ribonucleoside phosphoramidites. Curr. Protoc. Nucleic Acid Chem. Chapter 3, Unit3.19. Li, H., Handsaker, B., Wysoker, A., Fennell, T., Ruan, J., Homer, N., Marth, G., Abecasis, G., and Durbin, R. (2009). The sequence alignment/map format and SAMtools. Bioinformatics 25, 2078–2079. Li, Y., Song, M., and Kiledjian, M. (2011). Differential utilization of decapping enzymes in mammalian mRNA decay pathways. RNA 17, 419–428. Li, Y., Dai, J., Song, M., Fitzgerald-Bocarsly, P., and Kiledjian, M. (2012). Dcp2 decapping protein modulates mRNA stability of the critical interferon regulatory factor (IRF) IRF-7. Mol. Cell. Biol. 32, 1164–1172. Li, W., Wang, W., Uren, P.J., Penalva, L.O.F., and Smith, A.D. (2017). Riborex: fast and flexible identification of differential translation from Ribo-seq data. Bioinformatics 33, 1735–1737. Linder, B., Grozhik, A.V., Olarerin-George, A.O., Meydan, C., Mason, C.E., and Jaffrey, S.R. (2015). Single-nucleotide-resolution mapping of m6A and m6Am throughout the transcriptome. Nat. Methods 12, 767–772. Maden, B.E., Corbett, M.E., Heeney, P.A., Pugh, K., and Ajuh, P.M. (1995). Classical and novel approaches to the detection and localization of the numerous modified nucleotides in eukaryotic ribosomal RNA. Biochimie 77, 22–29. Mauer, J., Luo, X., Blanjoie, A., Jiao, X., Grozhik, A.V., Patil, D.P., Linder, B., Pickering, B.F., Vasseur, J.J., Chen, Q., et al. (2017). Reversible methylation of m6 Am in the 50 cap controls mRNA stability. Nature 541, 371–375. McGlincy, N.J., and Ingolia, N.T. (2017). Transcriptome-wide measurement of translation by ribosome profiling. Methods 126, 112–129. Meyer, K.D., Saletore, Y., Zumbo, P., Elemento, O., Mason, C.E., and Jaffrey, S.R. (2012). Comprehensive analysis of mRNA methylation reveals enrichment in 30 UTRs and near stop codons. Cell 149, 1635–1646. Olarerin-George, A.O., and Jaffrey, S.R. (2017). MetaPlotR: a Perl/R pipeline for plotting metagenes of nucleotide modifications and other transcriptomic sites. Bioinformatics 33, 1563–1564. Perry, R.P., Kelley, D.E., Friderici, K., and Rottman, F. (1975). The methylated constituents of L cell messenger RNA: evidence for an unusual cluster at the 50 terminus. Cell 4, 387–394. Quinlan, A.R., and Hall, I.M. (2010). BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics. 26, 841–842. Ramirez, F., Ryan, D.P., Gruning, B., Bhardwaj, V., Kilpert, F., Richter, A.S., Heyne, S., Dundar, F., and Manke, T. (2016). deepTools2: a next generation web server for deep-sequencing data analysis. Nucleic Acids Res. 44, W160–W165. Shah, A., Qian, Y., Weyn-Vanhentenryck, S.M., and Zhang, C. (2017). CLIP Tool Kit (CTK): a flexible and robust pipeline to analyze CLIP sequencing data. Bioinformatics. 33, 566–567. Shimotohno, K., Kodama, Y., Hashimoto, J., and Miura, K.I. (1977). Importance of 50 -terminal blocking structure to stabilize mRNA in eukaryotic protein synthesis. Proc. Natl. Acad. Sci. USA 74, 2734–2738. Shuman, S. (2002). What messenger RNA capping tells us about eukaryotic evolution. Nat. Rev. Mol. Cell Biol. 3, 619–625. Sun, H., Zhang, M., Li, K., Bai, D., and Yi, C. (2019). Cap-specific, terminal N6 methylation by a mammalian m6 Am methyltransferase. Cell Res. 29, 80–82. Thillier, Y., Decroly, E., Morvan, F., Canard, B., Vasseur, J.J., and Debart, F. (2012). Synthesis of 50 cap-0 and cap-1 RNAs using solid-phase chemistry coupled with enzymatic methylation by human (guanine-N7 )-methyl transferase. RNA 18, 856–868. Tsherniak, A., Vazquez, F., Montgomery, P.G., Weir, B.A., Kryukov, G., Cowley, G.S., Gill, S., Harrington, W.F., Pantel, S., Krill-Burger, J.M., et al. (2017). Defining a cancer dependency map. Cell 170, 564–576.e16. Vattem, K.M., and Wek, R.C. (2004). Reinitiation involving upstream ORFs regulates ATF4 mRNA translation in mammalian cells. Proc. Natl. Acad. Sci. USA 101, 11269–11274. Webb, S., Hector, R.D., Kudla, G., and Granneman, S. (2014). PAR-CLIP data indicate that Nrd1-Nab3-dependent transcription termination regulates expression of hundreds of protein coding genes in yeast. Genome Biol. 15, R8. Wei, C., Gershowitz, A., and Moss, B. (1975). N6, O20 -dimethyladenosine a novel methylated ribonucleoside next to the 50 terminal of animal cell and virus mRNAs. Nature 257, 251–253. Wei, C.M., Gershowitz, A., and Moss, B. (1976). 50 -terminal and internal methylated nucleotide sequences in HeLa cell mRNA. Biochemistry 15, 397–401. Willcock, D.F., Dryden, D.T., and Murray, N.E. (1994). A mutational analysis of the two motifs common to adenine methyltransferases. EMBO J. 13, 3902–3908. Yang, C., Bolotin, E., Jiang, T., Sladek, F.M., and Martinez, E. (2007). Prevalence of the initiator over the TATA box in human and yeast genes and identification of DNA motifs enriched in human TATA-less core promoters. Gene 389, 52–65. 642 Molecular Cell 75, 631–643, August 8, 2019 Zhao, S., and Zhang, B. (2015). A comprehensive evaluation of ensembl, RefSeq, and UCSC annotations in the context of RNA-seq read mapping and gene quantification. BMC Genomics 16, 97. Zhong, S., Li, H., Bodi, Z., Button, J., Vespa, L., Herzog, M., and Fray, R.G. (2008). MTA is an Arabidopsis messenger RNA adenosine methylase and interacts with a homolog of a sex-specific splicing factor. Plant Cell 20, 1278–1288. Zhou, J., Wan, J., Shu, X.E., Mao, Y., Liu, X.M., Yuan, X., Zhang, X., Hess, M.E., Br€uning, J.C., and Qian, S.B. (2018). N6 -methyladenosine guides mRNA alternative translation during integrated stress response. Mol. Cell 69, 636–647.e7. Zlatev, I., Lavergne, T., Debart, F., Vasseur, J.J., Manoharan, M., and Morvan, F. (2010). Efficient solid-phase chemical synthesis of 50 -triphosphates of DNA, RNA, and their analogues. Org. Lett. 12, 2190–2193. Molecular Cell 75, 631–643, August 8, 2019 643 STAR+METHODS KEY RESOURCES TABLE REAGENT or RESOURCE SOURCE IDENTIFIER Antibodies mouse anti-FLAG M2 Sigma F1804, RRID: AB_262044 rabbit anti-PCIF1 Abcam ab205016, RRID: AB_2753142 mouse anti-b actin Sigma A5441, RRID: AB_476744 anti-eIF4E Cell Signaling 2067, RRID: AB_2097675 anti-eIF4G Cell Signaling 2498, RRID: AB_2096025 rabbit anti-GAPDH Abcam ab181602, RRID: AB_2630358 mouse anti-KAP1 Abcam ab22553, RRID: AB_447151 rabbit anti-EEF2 Abcam ab75748, RRID: AB_1310165 mouse anti-RACK1 Santa Cruz B-3, RRID: AB_2247471 mouse anti-HSP70/72 Enzo Life Sciences ADI-SPA-810-D, RRID: AB_2039260 rabbit anti-PARP1 Cell Signaling 9542, RRID: AB_10616513 rabbit anti-HSPA8 Cell Signaling 8444, RRID: AB_10831837 rabbit anti-m6 A Abcam ab151230, RRID: AB_2753144 rabbit anti-ATF5 Abcam ab60126, RRID: AB_940375 Bacterial and Virus Strains T7 Express lysY New England Biolabs C3010 Chemicals, Peptides, and Recombinant Proteins Cycloheximide Sigma Aldrich C4859 iodoacetamide Sigma Aldrich I1149 4-thiouridine (s4 U) Sigma Aldrich T4509 Actinomycin D Sigma Aldrich A1410 Nuclease P1 Sigma Aldrich N8630 Nuclease P1 Wako USA 145-08221 RppH New England Biolabs M0356 Fast AP Thermo EF0654 T4 PNK New England Biolabs M0201 rSAP New England Biolabs M0371 Apyrase New England Biolabs M0398 RNase I Epicenter N6901K RNase T1 Thermo AM2283 T4 ligase 2, truncated K227Q New England BioLabs M0351 CircLigase II ssDNA Ligase Lucigen CL9021K AccuPrime SuperMix I Thermo 12342010 SuperScript III Reverse Transcriptase Thermo 18080044 S1 Nuclease Thermo EN 0321 Terminator exonuclease Epicenter TER51020 CIP New England Biolabs M0290 T4 RNA ligase 1 New England Biolabs M0437 RNase H New England Biolabs M0297 Phusion Master Mix with HF buffer New England Biolabs M0531 SuperScript IV Reverse Transcriptase Thermo 18090010 iQ SYBR Green supermix Bio-Rad 1708880 (Continued on next page) e1 Molecular Cell 75, 631–643.e1–e8, August 8, 2019 Continued REAGENT or RESOURCE SOURCE IDENTIFIER Critical Commercial Assays oligo d(T)25 Magnetic mRNA isolation kit New England Biolabs S1550 Dynabeads oligo d(T)25 Thermo 61005 Quant-iT RiboGreen RNA Assay Kit Thermo R11490 Ribo-Zero Gold rRNA Removal Kit (Human, Mouse, Rat) Illumina MRZG12324 NEBNext Ultra Directional RNA Library Prep Kit for Illumina New England Biolabs E7420 Hydrophilic streptavidin magnetic beads New England Biolabs S1421 Deposited Data Raw and analyzed data This Paper GEO: GSE122948 Unprocessed and uncompressed imaging data This Paper https://doi.org/10.17632/rnpfzjd7mj.1 Experimental Models: Cell Lines HEK293T ATCC CRL-3216 HeLa ATCC CCL-2 Oligonucleotides m7GpppAmGUGGACUAACCACCAU This Paper and Trilink N/A m7GpppAGUGGACUAACCACCAU This Paper and Trilink N/A pppAmGUGGACUAACCACCAU This Paper N/A sgRNA1; 50 - CGGUUGAAAGACUC CCGUGG-30 This Paper N/A sgRNA2; 50 - ACUUAACAUAUCCU GCGGGG-30 This Paper N/A Biotin 50 adaptor; biotin-GTTCAGAGT TCTACAGTCCGACGATC This Paper N/A RT_ACOT7: cccgttctggctgttgcaatgc This Paper N/A RT_BNIP2: agcctggattcaatgtcaactac This Paper N/A RT_YOD1: cggctggacagcccctgcaaaac This Paper N/A PCR_50 adaptor: taatacgactcactataggg tctcGGATCCcgacGTTCAGAGTTCTA CAGTCCGACGATC This Paper N/A PCR_ACOT7: GTAAAACGACGGCC AGTaagaGAATTCggaacccgttctggct gttgcaatgc This Paper N/A PCR_BNIP2: GTAAAACGACGGC CAGTaagaGAATTCggaaagcctgg attcaatgtcaactac This Paper N/A PCR_YOD1: GTAAAACGACGGCCAG TaagaGAATTCggaacggctggacagccc ctgcaaaac This Paper N/A qPCR_ACTB_Fw: GAAGATCAAGA TCATTGCTCCTC This Paper N/A qPCR_ACTB_Rv: ATCCACATCTG CTGGAAGG This Paper N/A qPCR_RPS28_Fw: ATCAAGCTGGC TAGGGTAACC This Paper N/A qPCR_RPS28_Rv: GGCCTTTGACAT TTCGGATGA This Paper N/A qPCR_AKAP12_Fw: CATTGTCACAGAG GTTGGA This Paper N/A (Continued on next page) Molecular Cell 75, 631–643.e1–e8, August 8, 2019 e2 LEAD CONTACT AND MATERIALS AVAILABILITY Please contact E.L.G. (eric.greer@childrens.harvard.edu) or S.R.J. (srj2003@med.cornell.edu) for reagents and resources generated in this study. METHOD DETAILS Synthesis and characterization of synthetic oligonucleotides The sequences of all the oligonucleotides used in this study are shown in Figure 1B. The synthetic RNA oligonucleotides, used in Figure 1E, were chemically assembled on an ABI 394 DNA synthesizer (Applied Biosystems) from commercially available long chain alkylamine controlled-pore glass (LCAA-CPG) solid support with a pore size of 1000 A˚ derivatized through the succinyl linker with 50 -O-dimethoxytrityl-20 -O-Ac-uridine (Link Technologies). All RNA sequences were prepared using phosphoramidite chemistry at 1-mmol scale in Twist oligonucleotide synthesis columns (Glen Research) from commercially available 20 -O-pivaloyloxymethyl amidites (50 -O-DMTr-20 -O-PivOM-[U, CAc , APac or GPac ]-30 -O-(O-cyanoethylN,N-diisopropylphosphoramidite)(Lavergne et al., 2010) (Chemgenes). The 50 -terminal adenosine was methylated in 20 -OH (Am). The 50 -O-DMTr-20 -O-Me-APac -30 -O-(O-cyanoethyl-N,N-diisopropylphosphoramidite) (Chemgenes) was used to introduce Am at the 50 end of RNA. All oligoribonucleotides were synthesized using standard protocols for solid-phase RNA synthesis with the PivOM methodology(Lavergne et al., 2008). Continued REAGENT or RESOURCE SOURCE IDENTIFIER qPCR_AKAP12_Rv: GCTGACTTAGTAG CCATCTC This Paper N/A qPCR_NBR1_Fw: GTCTAATACCCTGAT GCTCCC This Paper N/A qPCR_NBR1_Rv: GCAAATTCTCATCCA CAAATGC This Paper N/A Recombinant DNA pSpCas9n(BB)-2A-Puro (PX459) V2.0 vector Addgene 62988 pBABE-puro Addgene 1764 pBABE-puro-PCIF1 WT This Paper N/A pBABE-puro-PCIF1 SPPG This Paper N/A pMD2.G Addgene 12259 pUMVC Addgene 8449 p3xFLAG-CMV-10 Sigma E7658 p3xFLAG-CMV-10-PCIF1 WT This Paper N/A p3xFLAG-CMV-10-PCIF1 SPPG This Paper N/A Software and Algorithms Prism 5 Graphpad N/A MassHunter Suite Agilent N/A NIS-Elements AR Nikon N/A ImageQuantTL software GE Healthcare N/A ImageJ (V2.0.0-rc-24/1.49 m) NIH N/A Flexbar v2.5 Dodt et al., 2012 https://github.com/seqan/flexbar pyCRAC Webb et al., 2014 https://bitbucket.org/sgrann/pycrac bwa v0.7.17 Li et al., 2009 http://bio-bwa.sourceforge.net/ CTK v1.1.3 Shah et al., 2017 https://github.com/chaolinzhanglab/ctk deeptools v3.1.3 Ramirez et al., 2016 https://deeptools.readthedocs.io/en/ develop/ bedtools v2.27.1 Quinlan and Hall, 2010 https://bedtools.readthedocs.io/en/latest/ MetaPlotR Olarerin-George and Jaffrey, 2017 https://github.com/olarerin/metaPlotR DREME/MEME v5.0.2 Bailey et al., 2009 http://meme-suite.org/ e3 Molecular Cell 75, 631–643.e1–e8, August 8, 2019 After RNA assembly, the 50 -hydroxyl group of the 50 -terminal adenosine Am of RNA sequences, still anchored to solid support, was phosphorylated and the resulting H-phosphonate derivative was oxidized and activated into a phosphoroimidazolidate derivative to react with either pyrophosphate (for pppAm-RNA synthesis) (Zlatev et al., 2010) or guanosine diphosphate (for G-ppp-Am-RNA synthesis) (Thillier et al., 2012). After deprotection and release from the solid support upon basic conditions (DBU then aqueous ammonia treatment for 4h at 37 C), all RNA sequences were purified by IEX-HPLC(Barral et al., 2013), they were obtained with high purity (> 95%) and they were unambiguously characterized by MALDI-TOF spectrometry. N7 methylation of the purified G-ppp-Am-RNAs to give m7 G-ppp-Am-RNAs was carried out quantitatively using human mRNA guanine-N7 methyltransferase and S-adenosylmethionine as previously described(Thillier et al., 2012). The oligonucleotides used in Figures 1C, 1D, 1G, and 1H were synthesized by Trilink. Cell culture HEK293T and HeLa cells were maintained in DMEM (11995-065, ThermoFisher Scientific) with 10% FBS and antibiotics (100 units/ml penicillin and 100 mg/ml of streptomycin) under standard tissue culture conditions. Cells were split using TrypLE Express (Life Technologies) according to the manufacturer’s instructions. Mycoplasma contamination in cells were routinely tested by Hoechst staining. Antibodies Antibodies used for western blot analysis or immunostaining were as follows: mouse anti-FLAG M2 (F1804, Sigma), rabbit anti-PCIF1 (ab205016, Abcam), mouse anti-b actin (A5441, Sigma), anti-eIF4E (2067, Cell Signaling), anti-eIF4G (2498, Cell Signaling), rabbit anti-GAPDH (ab181602, Abcam), mouse anti-TRIM28 (ab22553, Abcam), rabbit anti-ATF5 (ab60126, Abcam), rabbit anti-EEF2 (ab33523), mouse anti-RACK1 (B-3, Santa Cruz), rabbit anti-PARP1 (9542, Cell Signaling), rabbit anti-HSPA8 (8444, Cell Signaling), mouse anti-HSP70/72 (C92F3A-5, Enzo Life Sciences). For m6 A individual-nucleotide-resolution cross-linking and immunoprecipitation (miCLIP), rabbit anti-m6 A (ab151230, Abcam) was used. Generation of PCIF1 CRISPR knockout cells and overexpression cell lines HEK293T and HeLa PCIF1-knockout cell lines were generated by CRISPR/Cas9 technology using two guide RNAs (gRNAs; 50 - CGGUUGAAAGACUCCCGUGG-30 and 50 - ACUUAACAUAUCCUGCGGGG-30 ) designed to target the PCIF1 genomic region between exon 8 and exon 17, that corresponds to the C-terminal catalytic domain. Double-stranded DNA oligonucleotides corresponding to the gRNAs were inserted into the pSpCas9n(BB)-2A-Puro (PX459) V2.0 vector (62988, Addgene). Equal amounts of the two gRNA plasmids were mixed and transfected into HEK293T and HeLa cells using FuGENE 6 (Promega). The transfected cells were then subjected to puromycin selection for three days and viable cells were used for serial dilution to generate single-cell clones. The genomic deletion was screened by PCR and was confirmed by Sanger sequencing. HEK293T and HeLa PCIF1-knockout lines used in this study contained a 4655 or 4656 nt homozygous deletion that removed the region between exon 8 and exon 17, including the stop codon, resulting in the disruption of PCIF1 protein after P229 (aa 230-704). Loss of PCIF1 protein expression was confirmed by western blot with anti-PCIF1 antibody (Abcam). Stable cell lines overexpressing PCIF1 WT or catalytically inactive mutant proteins were generated through retroviral infection. The coding sequence of human PCIF1 fused to a N-terminal 3X FLAG tag sequence that was cloned into the pBABE-puro retroviral vector (Addgene, 1764). Retroviral particles were generated in HEK293T cells through co-transfection of the packaging vectors pMD2.G (12259, Addgene) and pUMVC (8449, Addgene) with the appropriate pBABE-puro vectors. HEK293T and HeLa cells were infected with retroviral particles of pBABE-puro-3X-FLAG-PCIF1 WT or pBABE-puro-3X-FLAG-PCIF1 SPPG or control pBABE-puro empty vector, followed by puromycin selection (1 mg/ml). Cells were maintained at 70%–80% confluency before harvesting for mRNA purification. Two rounds of poly(A) mRNA isolation from mammalian cells was performed using oligo d(T)25 Magnetic mRNA isolation kit (NEB), according to the manufacturer’s instructions. Protein expression and purification The coding sequence of human PCIF1 was cloned as an in-frame fusion to the GST tagged vector pGEX-4T1. The catalytic site NPPF was mutated to APPA or SPPG thru site-directed mutagenesis using the Q5 mutagenesis kit (NEB), according to the manufacturer’s instructions. Recombinant GST-PCIF1 wild-type and catalytically inactive mutant proteins were expressed in E. coli T7 Express lysY. Overnight induction of protein expression was carried out with 0.5 mM IPTG at 18 C. Bacteria were harvested at 4000 rpm, 4 C and the cell pellet was resuspended in protein purification lysis buffer (50 mM Tris-HCl pH 7.5, 0.25 M NaCl, 0.1% Triton-X, 1 mM PMSF, 1 mM DTT, and protease inhibitors). The lysate was sonicated 6 times in 30 s on/off cycles and then centrifuged at 12,000 rpm for 20 minutes. Lysates were incubated with glutathione Sepharose 4B beads (Sigma). Proteins and beads were washed 3 times with protein purification lysis buffer before incubating the beads with elution buffer (12 mg/ml Glutathione in protein purification lysis buffer, pH 8.0) for 30 minutes. Eluates were dialyzed overnight at 4 C with enzyme storage buffer (40 mM Tris-HCl pH 8.0, 110 mM NaCl, 2.2 mM KCl, 1 mM DTT, 20% glycerol) and were subsequently stored at À80 C. Bradford assays and SDS-page gel electrophoresis followed by Coomassie staining was performed to determine integrity and quantity of purified proteins. Molecular Cell 75, 631–643.e1–e8, August 8, 2019 e4 In Vitro methyltransferase assays In vitro methylation reactions (50 ml) assaying PCIF1 activity against the m7 G capped RNA oligonucleotides were performed in methylation reaction buffer (50 mM Tris pH 8.0, 1 mM EDTA, 1 mM DTT, 5% glycerol) supplemented with 160 mM SAM (NEB) using 50 nM GST-PCIF1 protein and 4 mM m7 G capped oligonucleotide. Reactions were incubated for 10 minutes at 37 C, followed by heat inactivation for 20 minutes at 65 C and subsequent clean up and buffer exchange using Biospin P6 columns (Biorad). RNA oligonucleotides were decapped using 25 Units of RppH (NEB) in ThermoPol buffer for 3 hours at 37 C, followed by clean up and buffer exchange with Biospin P6 columns. Decapped RNA oligonucleotides were digested to nucleosides with 2 units of Nuclease P1 (Wako USA) at 37 C for 3 hours in a buffer containing 10 mM ammonium acetate pH 5.3, 2mM ZnCl2 followed by treatment with 2 units of Fast Alkaline Phosphatase (FastAP, Thermo Scientific) in FastAP reaction buffer for 1 hour at 37 C. After digestion the sample volume was brought to 100 mL with ddH2O followed by filtration using 0.22 mm Millex Syringe Filters (EMD Millipore). 5 mL of the filtered solution was analyzed by UHPLC-MS/MS. Enzyme kinetics assaying PCIF1 activity against the m7 G-Am and m7 G-A RNA oligonucleotides were performed in methylation reaction buffer supplemented with 1.33 mM [3 H]-SAM (Perkin Elmer) and 10 mM SAM (NEB), using 20 nM GST-PCIF1 protein and a range of concentrations of m7 G-Am oligonucleotide for 2-4 min at 37 C in 50 mL reactions. The reactions were stopped with 0.1% TFA followed by removal of unincorporated [3 H]-SAM with Biospin P30 columns (Biorad). The purified RNA oligonucleotide samples were then subjected to scintillation counting using a Perkin Elmer scintillation counter. The Michaelis-Menten curve and KM value were determined using Graphpad Prism software. UHPLC-MS/MS analysis For the detection and quantification of internal m6 A in mRNA, 500 ng of poly(A) mRNA was denatured at 70 C for 5 minutes followed by digestion to nucleotides using 20 units of S1 Nuclease (Thermo Scientific) in S1 Nuclease buffer for 2 hours at 37 C in 25 mL reactions. Nucleotides were then dephosphorylated to nucleosides by the addition of 2 units of Fast Alkaline Phosphatase (NEB) in FastAP reaction buffer for 1 hour at 37 C. After digestion the sample volume was brought to 100 mL with ddH2O followed by filtration using 0.22 mm Millex Syringe Filters (EMD Millipore). 5 mL of the filtered solution was analyzed by LC-MS/MS. For the detection and quantification of cap-adjacent m6 Am in mRNA, 500 ng of poly(A) mRNA was decapped using 25 Units of RppH (NEB) in ThermoPol buffer for 3 hours at 37 C, followed by clean up and buffer exchange with Biospin P30 columns. Subsequently decapped RNA was denatured at 70 C for 5 minutes followed by digestion to nucleotides using 2 units of Nuclease P1 (Wako USA) in a buffer containing 10 mM ammonium acetate pH 5.3, 2mM ZnCl2 for 3 hours at 37 C. Nucleotides were then dephosphorylated to nucleosides by the addition of 2 units of Fast Alkaline Phosphatase (NEB) in FastAP reaction buffer for 1 hour at 37 C. After digestion the sample volume was brought to 100 mL with ddH2O followed by filtration using 0.22 mm Millex Syringe Filters. 5 mL of the filtered solution was analyzed by LC-MS/MS. The separation of nucleosides was performed using an Agilent 1290 UHPLC system with a C18 reversed-phase column (2.1 3 50 mm, 1.8 m). The mobile phase A was water with 0.1% (v/v) formic acid and mobile phase B was methanol with 0.1% (v/v) formic acid. Online mass spectrometry detection was performed using an Agilent 6470 triple quadrupole mass spectrometer in positive electrospray ionization mode. Quantification of each nucleoside was accomplished in dynamic multiple reaction monitoring (dMRM) mode by monitoring the transitions of 268/136 (A), 282/136 (Am), 282/150 (m6 A), 296/150 (m6 Am), 244/112 (C). The amounts of A, C, Am, m6 A and m6 Am in the samples were quantified using corresponding calibration curves generated with pure standards. m6 Am and m6 A levels in the RNA oligonucleotides after in vitro methylation reactions were normalized by cytidine concentration. Cap-binding assay Cells were lysed in buffer B (20 mM HEPES-KOH pH 7.6, 100 mM KCl, 0.5 mM EDTA, 0.4% NP-40, 20% glycerol) supplemented with protease and phosphatase inhibitors (Roche), 1 mM dithiothreitol (DTT) and 80 units/ml RNasin (Promega). For pull down, 1-2.5 mg of total protein extract was first pre-cleared on Agarose beads (Jena Bioscience) followed by incubation with 25 mL m7 GTP conjugated Agarose beads (Jena Bioscience) for 1 hour at 4 C. Following pull-down the beads were washed three times and the supernatant was removed and replaced by lysis buffer. Beads were incubated with 0.25 mM cap analog, m7 G-ppp-A, or G-ppp-A, or water (mock) for 1 hour at 4 C. Supernatant (Eluate) was removed and diluted with Laemmli sample buffer. Beads were washed three times and resuspended in Laemmli sample buffer. Samples were resolved on a 4%–15% Tris-HCl gradient gel (BioRad) and analyzed by western blotting using specific antibodies. Immunofluorescence Cells were grown on poly-L-lysine pre-coated coverslips that were sterilized under UV light for 30 minutes - 1 hour. Cells were rinsed in 1X phosphate-buffered saline (PBS) solution followed by fixation in ice-cold methanol at À20 C for 10 minutes. Coverslips were then washed 3 times with 1X PBS before being blocked for 30 minutes in 1% BSA in 1X PBS. Primary antibody was diluted 1/200 in 1% BSA 1X PBS and incubated for 1 hour at room temperature in a humidified chamber. Slides were subsequently rinsed 3 times and washed 2 times for 15 minutes with 1% BSA in 1X PBS at room temperature before incubation with secondary antibody, diluted 1/200 in 1% BSA in 1X PBS, in a dark humidified chamber for 30 minutes at room temperature. Coverslips were then rinsed 3 times and washed 3 times for 15 minutes with 1% BSA in 1X PBS in the dark before being rinsed 3 times with ddH2O. Coverslips were e5 Molecular Cell 75, 631–643.e1–e8, August 8, 2019 mounted using mounting medium containing DAPI. Image acquisition was carried out on a Nikon Eclipse Ti microscope (Nikon), using NIS-Elements AR software. Determination of relative m6 Am, Am, and m6 A levels by thin layer chromatography Levels of internal m6 A in mRNA were determined by 2D-TLC essentially as previously described (Zhong et al., 2008). In brief, poly(A) RNA (100 ng) was digested with 2 units ribonuclease T1 (ThermoFisher Scientific) for 2h at 37 C in the presence of RNasin RNase Inhibitor (Promega). T1 cuts after every guanosine and exposes the 50 -hydroxyl of the following nucleotide, which can be A, C, U, or m6 A. Thus, this method quantifies m6 A in a GA sequence context. 50 ends were subsequently labeled with 10 units T4 PNK (NEB) and 0.4 mBq [g-32 P] ATP at 37 C for 30 min followed by removal of the g-phosphate of ATP by incubation with 10 units Apyrase (NEB) at 30 C for 30 min. After phenol-chloroform extraction and ethanol precipitation, RNA samples were resuspended in 10 ml of DEPC-H2O and digested to single nucleotides with 2 units of P1 nuclease (Sigma) for 1h at 60 C. 1 ml of the released 50 monophosphates from this digest were then analyzed by 2D-TLC on glass-backed PEI-cellulose plates (MerckMillipore) as described previously (Kruse et al., 2011). The protocol to detect the m6 Am:Am ratio was based on the protocol developed by Fray and colleagues (Kruse et al., 2011), with some modifications. 300ng of poly(A) RNA was decapped with 15 units of RppH (NEB) for 3 h at 37 C. 50 monophosphates in the resulting RNA were removed by addition of 5 units of rSAP phosphatase (NEB) for 1 h at 37 C. Up to this point, all enzymatic reactions were performed in the presence of SUPERase In RNase Inhibitor (ThermoFisher Scientific). After phenol-chloroform extraction and ethanol precipitation, RNA samples were resuspended in 10 ml of DEPC-H2O and 50 ends were labeled using 30 units T4 PNK and 0.8 mBq [g-32 P] ATP at 37 C for 30 min. PNK was heat inactivated at 65 C for 20 min and the reaction was passed through a P-30 spin column (Bio-Rad) to remove unincorporated isotope. 8 ml of labeled RNA were then digested with 2 units of P1 nuclease (Sigma) for 1 h at 60 C. 2 ml of the released 50 monophosphates from this digest were then analyzed by 2D-TLC on glass-backed PEI-cellulose plates (MerckMillipore) as described previously (Kruse et al., 2011). Signal acquisition was carried out using a storage phosphor screen (GE Healthcare Life Sciences) at 200 mm resolution and ImageQuantTL software (GE Healthcare Life Sciences). Quantification was carried out with ImageJ (V2.0.0-rc-24/1.49 m). For m6 Am experiments, the m6 Am:Am ratio was calculated. The use of this ratio has been described previously (Kruse et al., 2011). We confirmed that this assay is linear by spotting twice the sample material and confirming that the signal intensity doubles for the unmodified nucleotides (A, C, and U). Furthermore, exposure time of the TLC plates to the phosphor screen was chosen so the signal was not saturated. For m6 A quantification, m6 A was calculated as a percent of the total of the A, C, and U spots, as described previously (Jia et al., 2011). The use of relative ratios for each individual sample is important since it reduces the error derived from possible differences in loading. To minimize the effects of culturing conditions on the measured m6 Am:Am ratios of each experimental group (e.g., control versus knockout), all replicates were processed in parallel to minimize any source of variability between samples being compared. miCLIP Total RNA from wild-type and PCIF1 knockout HEK293T cells, and wild-type mouse embryonic fibroblasts, was extracted using TRIzol following the manufacturer’s protocol. Any contaminating genomic DNA was degraded using DNase I and poly(A) RNA was isolated using two rounds of Dynabeads Oligo(dT)25 capture. 10 mg poly(A) RNA was then used as input for single nucleotide-resolution m6 A mapping using the miCLIP protocol, as previously reported (Linder et al., 2015). Final libraries were amplified and subjected to 50-cycle paired-end sequencing on an Illumina HiSeq2500 at the Weill Cornell Medicine Epigenetic Core facility. miCLIP bioinformatic analyses The initial processing of raw FASTQ files was done as in the miCLIP protocol. Adapters and low quality nucleotides were first trimmed from paired reads using flexbar v2.5. The trimmed FASTQ file was then de-multiplexed using the pyBarcodeFilter.py script from the pyCRAC suite. The remainder of the random barcode was moved to the headers of the FASTQ reads using an awk script and PCR duplicates were removed using the pyCRAC pyDuplicateRemover.py script. Reads were aligned to hg38/mm10 using bwa v0.7.17 with the option ‘‘-n 0.06’’ as recommended in the CTK package. To identify m6 A within the DRACH consensus, C to T transitions were extracted and the CIMS pipeline from the CTK package was used. Due to a high transition frequency in this dataset, putative m6 A residues with an FDR < 0.1 and in a DRACH consensus were used as the final list of m6 A in this study. To identify putative m6 Am sites, coverage of wild-type and PCIF1 knockout samples were compared genome-wide using the bamCompare tool from deeptools v3.1.3. In short, the genome was binned into 50 nt non-sliding windows and the coverage of reads in each was counted for each strand, discarding zero-coverage bins. This was normalized to the total number of reads in bins, per million (BPM) and the log2 ratio of BPM+1 for wild-type to PCIF1 knockout was calculated. A log2 ratio threshold of 2 was chosen as the cutoff for each replicate. Adjacent bins passing threshold were merged using bedtools v2.27.1. The intersection of putative m6 Am regions across replicates was taken using bedtools intersect, resulting in 2360 high-confidence m6 Am peaks. To determine the precise m6 Am nucleotide within these peaks, a combination of A to T transitions and a read pileup/drop-off method was used. In PCIF1-dependent peaks with an A to T transition occurring at a frequency of 10% or greater, this A was selected as the m6 Am. For the remainder, a pileup/drop-off approach similar to the previous miCLIP criteria (Linder et al., 2015) was utilized. Here, the start nucleotide of each read (with respect to strand, i.e., the leftmost coordinate for + strand features and rightmost for – strand features) was Molecular Cell 75, 631–643.e1–e8, August 8, 2019 e6 extracted and piled up using the tag2cluster.pl script of the CTK package with the options ‘‘-s -v -maxgap À1.’’ Clusters of less than 5 reads were discarded, as were those that did not map to an A. When there was a single A-cluster in a PCIF1-dependent peak, this was selected as m6 Am. When more than one occurred, the most piled-up cluster of the two closest to the beginning of the peak (with respect to strand) was selected. To generate metagenes, MetaPlotR (Olarerin-George and Jaffrey, 2017) was used. In all cases, the longest GENCODE transcript isoform for each gene was selected. For metaplots centered on reference annotations, the closest m6 Am to each feature was measured using bedtools closest and these distances were plotted as a histogram. Aligned reads in bigwig format and BED files with coordinates for m6 A, m6 Am, and CAGE peaks were used to generate genome tracks using pyGenomeTracks v1.0. Motif searches were performed using either DREME v5.0.2 or MEME v5.0.2. For functional annotation analyses of m6 Am and 50 UTR m6 A genes, DAVID v6.8 was used specifying a background of all genes covered with at least 20 reads. Transcript 50 end cloning To validate called m6 Am as transcription start nucleotides, an adapted 50 RACE method was used. 10 mg poly(A) RNA was treated with Terminator exonuclease (Epicenter) and then CIP and rSAP (NEB) to remove 50 -monophosphate RNA following the manufacturer’s recommendations. Half of this was then decapped using RppH (NEB) at a final concentration of 5 U per 100 ng RNA in 1X Thermopol buffer (NEB). The remaining half was incubated without RppH as a capped control, to control for any residual 50 -monophosphates due to RNA degradation. 75 pmol of a biotinylated adaptor (biotin-GTTCAGAGTTCTACAGTCCGACGATC) was then ligated onto the 50 end of the processed RNA using T4 RNA ligase 1 (NEB) in a 30 ml reaction containing 1X T4 RNA ligase buffer, 1 mM ATP, and 20% PEG-8000 for 3 hours at 23 C. This was then diluted to 200 ml in streptavidin bead wash buffer (SA wash; 20 mM Tris-HCl pH 7.5, 0.5 M NaCl, 1 mM EDTA) and incubated with 0.5 mg hydrophilic streptavidin magnetic beads (NEB) for 15 minutes and room temperature. Beads were washed twice in SA wash buffer, then twice in annealing buffer (10 mM Tris-HCl pH 7.5, 50 mM NaCl, 1 mM EDTA). 12.5 pmol of each gene-specific reverse transcription primers (STAR Methods) were annealed to RNA in 150 ml annealing buffer by heating to 90 C for 5 minutes and cooled to room temperate over 20 minutes. Beads were then resuspended in a 50 ml SuperScript III reverse transcription reaction (ThermoFisher Scientific) according to the manufacturer’s protocol. Beads were washed and resuspended in a 50 ml RNase H (NEB) reaction and incubated for 30 minutes at 37 C. All reactions on beads were performed in a thermoshaker (15 s on 1100 RPM, 30 s off) to ensure beads remained in suspension. 100 ml wash buffer was added, heated at 75 C for 2 minutes, placed on beads and supernatant containing eluted cDNA immediately transferred to a fresh tube. The beads were resuspended in 50 ml and the elution repeated. Following an ethanol precipitation, 5% of the cDNA was used in 20 ml PCR reactions containing 1X Phusion master mix (NEB), 60% DMSO, 250 nM adaptor primer, and 250 nM genespecific primer (see STAR Methods for primer sequences). 5 ml of this was then loaded on a 6% TBE-PAGE gel and visualized. Bands of the correct size that were absent in the capped control were then identified as m6 Am starting transcripts. SLAM-seq SLAM-seq was performed as described previously (Herzog et al., 2017) with minor modifications. HEK293T (WT and PCIF1 KO) cells (at 60% confluency) were incubated with cell culture growth medium supplemented with 25 mM 4-thiouridine (s4 U) for 24 h (pulse phase). s4 U incorporation was confirmed by HPLC analysis, as previously described (Herzog et al., 2017). The uridine chase was initiated by changing media containing 2.5 mM uridine (Sigma) and cells were collected for RNA extraction after 6 and 12 h. The 0 h sample were the cells that have completed the pulse with s4 U, but without uridine-chase. Total RNA was extracted using RNAzol reagent (MRC) according to the manufacturer’s instructions, maintaining reducing conditions to prevent oxidation of s4 U (0.1 mM DTT final concentration). For thiol alkylation, a master mix (10 mM iodoacetamide, 50 mM NaPO4 pH 8 and 50% DMSO) was prepared, centrifuged, and added to 20 mg of total RNA at 50 C for 15 min and then purified by ethanol precipitation. After that, two rounds of poly(A) mRNA enrichment was carried out with oligo d(T)25 Magnetic Beads (NEB). Standard RNA-seq libraries were prepared using NEBNext Ultra Directional RNA Library Prep Kit for Illumina (NEB) following the instructions of the manufacturer. Sequencing was performed on a HiSeq2500 (Illumina) with 50 nucleotide reads. Real-time PCR assay to determine transcript stability Wild-type or PCIF1 knockout HEK293T cells were transfected with either empty vector or wild-type or SPPG mutant PCIF1 vectors for 48 hours and then treated with 5 mg/ml actinomycin D or vehicle (DMSO) for 8 hours. Total RNA was extracted using Trizol and 2 mg of this reverse transcribed using random hexamers and SuperScript IV (ThermoFisher Scientific) according to the manufacturer’s protocol. RT-PCR was performed in 20 ml reactions containing 250 nM forward and reverse primers and iQ SYBR Green supermix (Bio-Rad) on an Eppendorf RealPlex2 RT-PCR machine. A delta cycle threshold (Ct) was calculated using the average Ct values across technical triplicates, by subtracting the geometric mean of two control genes (RPS28 and ACTB). A delta-delta Ct was then calculated by subtracting the vehicle control delta Ct value for each sample and untransformed to obtain relative abundances. Fold changes were tested for p < 0.1 by Student’s t test. Ribosome profiling Ribosome profiling was performed as described previously (McGlincy and Ingolia, 2017). In brief, wild-type and PCIF1 knockout HEK293T cells were grown to $70% confluence, washed twice with ice cold PBS supplemented with 50 mg/ml of cycloheximide e7 Molecular Cell 75, 631–643.e1–e8, August 8, 2019 (CHX) and collected by scraping. After pelleting, cells were resuspended in 400 mL lysis buffer (20 mM Tris-HCl pH 7.4, 150 mM NaCl, 5 mM MgCl2, 1 mM DTT and 100 mg/ml CHX) After incubation on ice for 10 min, lysate was triturated 5 times through a 25-gauge needle and then lysate was centrifuged at 20,000 x g for 10 min. 5 mL of lysate was flash frozen and saved as input. To generate ribosomeprotected fragments the lysates (30 mg) were first mixed with 200 mL DEPC-H2O then incubated with 15 U RNase I for 45 min at room temperature. The reaction was stopped with 10 mL SUPERase*In RNase inhibitor. 0.9 mL of sucrose-supplemented lysis buffer was added to the digestion mixture and ultracentrifuged at 100,000 rpm, 4 C for 1 h. Pellets were resuspended in 300 mL of water and after phenol-chloroform extraction, precipitated with ethanol. The RNA was then run on a 15% 8 M urea TBE gel, stained with SYBR Gold, and a gel fragment between 17-34 nucleotides corresponding to ribosome-protected RNA was excised. RNA was eluted for 2 h at 37 C in 300 mL RNA extraction buffer (300 mM NaOAc pH 5.5, 1 mM EDTA, 0.25%v/v SDS) after crushing the gel fragment. RNA was ethanol precipitated and resuspended in 26 mL water and treated with RiboZero Gold kit. Libraries from RNA-protected fragments were generated as previously described in the protocol (Linder et al., 2015). In brief, the RNA fragments were dephosphorylated with T4 PNK for 1 h at 37 C in dephosphorylation buffer (70 mM Tris, pH 6.5, 10 mM MgCl2, 1 mM DTT). The 30 adaptor was ligated using T4 RNA Ligase 2, truncated K227Q ligase (New England BioLabs) for 3h at 22 C. Ligated sRNAs were purified by ethanol precipitation, and reverse transcribed using the primers complementary to the 30 adaptor containing specific barcodes. After circularization with CircLigase II, cDNAs were relinearized by BamHI digestion and in the next step, PCR-amplified and subjected to Illumina HiSeq 2500 platform. Due to the similarity in size between ligated and unligated adapters, the libraries were gel purified. RNA-Seq analysis was conducted using the ribosome profiling input material. Ribosomal RNAs were removed from the input RNA using the NEBNext rRNA Depletion Kit (NEB). Input RNA libraries were generated using the NEBNext Ultra Directional RNA library prep kit for Illumina (NEB). Libraries were sequenced using an Illumina HiSeq 2500 platform with 50 nt reads. Ribosome footprint reads and corresponding RNA-Seq reads were processed essentially as described (Ingolia et al., 2012). Adaptors and short reads (< 17nt) were trimmed using FLEXBAR v2.5, demultiplexed using pyBarcodeFilter.py (pyCRAC software). PCR duplicates were collapsed by pyFastqDuplicateRemover.py script. Ribosomal RNA reads were removed by STAR aligner38. Remaining reads were then aligned to the hg38 genome with STAR v2.5.2a in a splicing-aware manner and using UCSC refSeq as a transcript model database (version from June 02/2014 downloaded from Illumina iGenomes). Two mismatches were allowed and only unique alignments were reported. Aligned reads were then counted on transcript regions using custom R scripts considering only transcripts with annotated 50 and 30 UTRs. Gene count tables generated from STAR were normalized using DESeq2 (R-Bioconductor). Translation efficiency was calculated using Riborex (Li et al., 2017), with pre-filtering for transcripts that had at least ten counted reads. SLAM-seq bioinformatic analysis Raw sequencing data were trimmed of adaptor sequences and filtered of reads with uncalled bases and reads < 17 nucleotides in length using Flexbar. Duplicate reads were further removed using pyFastqDuplicateRemover.py script and remaining reads were aligned to the human genome (GrCh38) using the STAR aligner. To identify T/C conversions, aligned reads were analyzed using Rsamtools Pileup (version 1.27.16). This program was used to determine the frequency of each of the four nucleotides present in mapped reads at every genomic position with read coverage. After summation of all nucleotide mapped to each transcript, we selected only those with at least 100 T/C conversions at time point 0 h. Additionally, to select for those transcripts with a longer half-life, transcripts were filtered for those with at least 50 T/C conversions at time point 6h. The mRNA half-life for each transcript was calculated based on the equation: tð1=2Þ = À ln2 ln  NðtÞ No  t Statistics and software P-values were calculated with a two-tailed unpaired Student’s t test or, for the comparison of more than two groups, with a one- or two-way ANOVA followed by Bonferroni’s or Tukey’s post-test. Reproducibility of half-life and translation efficiency measurements was assessed by calculating the Spearman correlation coefficient between replicates. Significance of list overlaps was calculated using hypergeometric probability. DATA AND CODE AVAILABILITY The accession number for the RNA-sequencing and ribosome profiling data reported in this paper is NCBI GEO: GSE122948. Unprocessed and uncompressed imaging data is available at https://doi.org/10.17632/rnpfzjd7mj.1. Molecular Cell 75, 631–643.e1–e8, August 8, 2019 e8