Autophagy induction promoted by m6A reader YTHDF3 through translation upregulation of FOXO3 mRNA

Autophagy is crucial for maintaining cellular energy homeostasis and for cells to adapt to nutrient deficiency, and nutrient sensors regulating autophagy have been reported previously. However, the role of eiptranscriptomic modifications such as m6A in the regulation of starvation-induced autophagy is unclear. Here, we show that the m6A reader YTHDF3 is essential for autophagy induction. m6A modification is up-regulated to promote autophagosome formation and lysosomal degradation upon nutrient deficiency. METTL3 depletion leads to a loss of functional m6A modification and inhibits YTHDF3-mediated autophagy flux. YTHDF3 promotes autophagy by recognizing m6A modification sites around the stop codon of FOXO3 mRNA. YTHDF3 also recruits eIF3a and eIF4B to facilitate FOXO3 translation, subsequently initiating autophagy. Overall, our study demonstrates that the epitranscriptome regulator YTHDF3 functions as a nutrient responder, providing a glimpse into the post-transcriptional RNA modifications that regulate metabolic homeostasis.

. Immunoblot analyses of YTHDF3 expression in MEF stable cell lines, as well as wild-type and YTHDF3 -/-MEFs. a Immunoblot analyses of YTHDF3 in MEFs infected with either nonspecific shRNA (shNS) or two independent shRNAs targeting YTHDF3 (shYTHDF3 and shYTHDF3-2). b Immunoblot analyses of YTHDF3 in shYTHDF3 MEFs ectopically expressing either YTHDF3 (shYTHDF3+YTHDF3) or its control vector (shYTHDF3+Con). c Immunoblot analyses of YTHDF3 overexpression and control MEFs. d Immunoblot analyses of YTHDF3 in wild-type and YTHDF3 -/-MEFs following nutrient starvation for the indicated time periods. GAPDH are shown as loading controls. Source data are provided as a Source Data file. Supplementary Fig.2. p62 mRNA expression in shNS and shYTHDF3 MEFs. a qRT-PCR analyses of p62 in shNS and shYTHDF3 MEFs following nutrient starvation for the indicated time periods. b shNS and shYTHDF3 MEFs were treated with Act.D (5μg/mL) for the indicated time periods, with or without nutrient starvation, respectively. The expression of p62 was examined with qRT-PCR. c Relative levels of p62 mRNAs in each ribosome fraction were quantified and plotted as a percentage relative to the total input. Data from three independent experiments are presented as mean values ± SEM. Source data are provided as a Source Data file. Supplementary Fig.4. Detection of the m 6 A writer and eraser proteins in MEFs during nutrient deficiency. a Cytoplasm fractions from MEFs following nutrient starvation for the indicated time periods were subjected to immunoblotting. b MEFs following nutrient starvation for the indicated time periods were subjected to fluorescence microscopy. Endogenous METTL3, METTL14, ALKBH5, and FTO were stained. Nuclei were stained with DAPI. Scale bar, 20μm. c, d The indicated cells were starved in either the glucose-deprived medium or HBSS for the indicated time periods, followed by immunoblot analyses of METTL3. Source data are provided as a Source Data file. Supplementary Fig.5. Representative confocal images of m 6 A fluorescence localization in shNS and shMETTL3 MEFs. a Immunoblot analyses of METTL3 in MEFs infected with shNS or two independent METTL3 shRNAs (shMETTL3 and shMETTL3-2). b Representative confocal images of m 6 A fluorescence localization were obtained in shNS and shMETTL3 MEFs, with or without nutrient deprivation. Nuclei were stained with DAPI. Scale bar, 20μm. Source data are provided as a Source Data file. Supplementary Fig.6. Transient activation of METTL3 favors cell survival and metabolism. a Immunoblot analysis of LC3-II in MEFs following nutrient starvation for the indicated time periods with Baf.A1 treatment(20nM). GAPDH was used as a loading control. b Percentage of surviving cells after nutrient starvation for the indicated time periods. c Representative phasecontrast images of MEFs following nutrient starvation for the indicated time periods. Scale bar, 100μm. d Measurement of cellular ATP levels in MEFs following nutrient starvation for the indicated time periods. For b, d, data from three independent experiments are presented as mean values ± SEM. Two-tailed unpaired t-tests were used to estimate significance. P-values are indicated in the figure. Source data are provided as a Source Data file. Supplementary Fig.7. Immunoblot analyses and m 6 A dot blot analyses of YTHDF3-OE MEFs infected with shNS or two independent METTL3 shRNAs (shMETTL3 and shMETTL3-2). a Immunoblot analyses of YTHDF3-OE MEFs infected with shNS or two independent METTL3 shRNAs (shMETTL3 and shMETTL3-2). b Poly(A)+ RNA from samples in a were subjected to dot blot analysis with an antibody recognizing m 6 A. Loading control was methylene blue. c Immunoblot analyses of METTL3-silenced YTHDF3-OE MEFs transfected with lentiviral vectors (Con), wild-type METTL3 (METTL3-WT), or a catalytic mutant of METTL3 (METTL3-Mut). d Poly(A)+ RNA from samples in c were subjected to dot blot analysis with an antibody recognizing m 6 A. Loading control was methylene blue. Source data are provided as a Source Data file. Supplementary Fig.8. Detection of the METTL3's impact on autophagy flux in MEFs without YTHDF3 overexpression. a Immunoblot analyses of LC3-II and p62 in shNS and shMETTL3 MEFs following nutrient starvation for the indicated time periods, with or without Baf.A1 treatment (20nM). GAPDH was used as a loading control. b Immunoblot analyses of LC3-II and p62 in MEFs ectopically expressing either METTL3 or its control vector (Con). GAPDH is shown as a loading control. c Immunoblot analyses of METTL3-silenced MEFs transfected with lentiviral vectors (Con), wild type METTL3 (METTL3-WT), or a catalytic mutant of METTL3 (METTL3-Mut) following nutrient starvation for the indicated time periods, with or without Baf.A1 treatment (20nM). GAPDH is used as a loading control. d, e Measurement of autophagy flux and quantification of autophagosomes (yellow) and autolysosomes (red) by a tandem mCherry-GFP-LC3 reporter assay in shNS and shMETTL3 MEFs. Scale bar, 20μm. f, g Measurement of autophagy flux and quantification of autophagosomes (yellow) and autolysosomes (red) by a tandem mCherry-GFP-LC3 reporter assay in METTL3-silenced MEFs transfected with METTL3-WT, METTL3-Mut, or Con, with or without nutrient deficiency. h, i Measurement of autophagy flux and quantification of autophagosomes (yellow) and autolysosomes (red) by a tandem mCherry-GFP-LC3 reporter assay in MEFs ectopically expressing either METTL3 or its control vector (Con) with or without nutrient deficiency. Scale bar, 20μm. For e, g, i, mean numbers of puncta per cell from each randomly selected field over three independent experiments were plotted (dots). Bars represent mean values ± SEM. Two-tailed unpaired multiple t-tests with two-stage step-up correction (Benjamini, Krieger, and Yekutieli) were used to estimate significance. P-values are indicated in the figure. Source data are provided as a Source Data file.
Supplementary Fig.9. Immunoblot analysis of the impact of METTL14 depletion on autophagy flux. a Immunoblot analyses of METTL14 in shNS and shMETTL14 MEFs. b Immunoblot analyses of LC3-II and p62 in shNS and shMETTL14 MEFs following nutrient starvation for the indicated time periods, with and without Baf.A1 treatment (20nM). GAPDH is used as a loading control. Source data are provided as a Source Data file.
Supplementary Fig.10. MeRIP-seq data analyses. a Metagene plots depict the distribution of m 6 A peaks in MEFs, with or without nutrient starvation. The consensus motifs within m 6 A peaks were generated by DREME database. b Violin box plots of log2 fold enrichment of all the m 6 A peaks in each group (n=2 biological replicates). Violin plot range denotes the minima and maxima, boxes represent 25th and 75th percentile, centre depicts the median, and whiskers indicate 1.5×interquartile range. P=3.2963E-19 by two-tailed Wilcoxon signed-rank tests. c Venn diagrams show overlap between hyper-m 6 A genes from our MeRIP-seq data and METTL3-bound genes from starBase human CLIP-seq data; and overlap between the resultant METTL3-bound genes which were not hyper-m 6 A-methylated upon nutrient starvation and FTO-or ALKBH5-bound genes from starBase human CLIP-seq data.
Supplementary Fig.11. Identification of FOXO3 as the candidate of YTHDF3. a YTHDF3-RIP followed by qRT-PCR detected the interaction between YTHDF3 and the indicated mRNAs in shNS and shMETTL3 MEFs upon nutrient deficiency (n=3 biological replicates). Data are presented as mean values ± SEM. Two-tailed unpaired t-tests were used to estimate significance. P-values are indicated in the figure. b, c Immunoblot analyses of indicated proteins in shNS, shYTHDF3, control, and YTHDF3-OE MEFs, with or without nutrient starvation, respectively. Source data are provided as a Source Data file.
Supplementary Fig.12. YTHDF3-promoted FOXO3 targeted autophagic genes are dependent on METTL3. a qRT-PCR analysis of mRNA levels of FOXO3 target autophagyrelated genes in siNC, siFOXO3, control, and FOXO3-OE MEFs (n=3 biological replicates). Bars represent mean ± SEM. Two-tailed unpaired multiple t-tests with two-stage step-up correction (Benjamini, Krieger, and Yekutieli) were used to estimate significance. P-values are indicated in the figure. b Immunoblot analyses of protein levels of the indicated FOXO3 targeted autophagic genes in siNC, siFOXO3, control, and FOXO3-OE MEFs. c Immunoblot analyses of protein levels of FOXO3 targeted autophagic genes in METTL3-silenced YTHDF3-OE MEFs. Source data are provided as a Source Data file.
Supplementary Fig.14. The YTHDF3 -/mice show less sensitivity to starvation in vivo. a Representative image of 24 h-fasted wild-type and YTHDF3 -/mice. b, c Body weights were measured in wild-type and YTHDF3 -/mice before and after a 24 h fast (b), and the body weight changes were compared between the groups (c). d Representative image of livers from the fasted wild-type and YTHDF3 -/mice. e Liver weights in fasted wild-type and YTHDF3 -/mice were measured. f Representative images of H&E, PAS, and Oil Red O staining of liver sections from fasted wild-type and YTHDF3 -/mice. Scale bar, 50μm. g, h The percentages of the PAS staining positive cells (g) and Oil Red O staining positive areas (h) were compared between groups. i Left, immunoblot analysis of FOXO3 and LC3-II in liver tissues derived from wildtype and YTHDF3 -/mice. GAPDH is used as a loading control. Right, FOXO3 and LC3-II expressions were quantitatively defined, respectively. Data are presented as mean values ± SEM. Two-tailed unpaired t-tests (n=6 mice per group, male, 8 weeks old) were used to estimate significance. P-values are indicated in the figure. Source data are provided as a Source Data file. Supplementary Fig.15. Predicted m 6 A sites in FOXO3 mRNA by the SRAMP algorithm. Supplementary Fig.16. Detection of the impact of the m 6 A motifs in downstream 3' UTR on FOXO3 expression. a FOXO3-3′ UTR-WT-2 or FOXO3-3′ UTR-Mut-2 reporters were transfected into control and YTHDF3-OE MEFs for 72 hours. Firefly luciferase activity was measured and normalized to Renilla luciferase activity. b After transfecting FOXO3-3′UTR-WT-2 or FOXO3-3′ UTR-Mut-2, MEFs were nutrient-starved. Firefly luciferase activity was measured and normalized to Renilla luciferase activity. Data from three independent experiments are expressed as means ± SEM. Two-tailed unpaired t-tests were used to estimate significance. P-values are indicated in the figure. Source data are provided as a Source Data file. Supplementary Fig.17. Detection of the impacts of YTHDF1 or YTHDF2 depletion on FOXO3 mRNAs. a Nuclear and cytoplasm fractions from shNS and shYTHDF1 MEFs following nutrient starvation for the indicated time periods were subjected to immunoblotting. b Nuclear and cytoplasm fractions from shNS and shYTHDF2 MEFs following nutrient starvation for the indicated time periods were subjected to immunoblotting. c Sucrose gradient-based polysome profiling of shNS and shYTHDF1 MEFs. d FOXO3 mRNAs in each ribosome fraction were quantified through qRT-PCR and plotted as percentages of the total input from shNS and shYTHDF1 MEFs. e Sucrose gradient-based polysome profiling of shNS and shYTHDF2 MEFs. f FOXO3 mRNAs in each ribosome fraction were quantified through qRT-PCR and plotted as percentages of the total input from shNS and shYTHDF2 MEFs. g shNS and shYTHDF1 MEFs were treated with Act. D (5μg/mL) for the indicated time periods. The expression of FOXO3 was examined with qRT-PCR. h shNS and shYTHDF2 MEFs were treated with Act. D (5μg/mL) for the indicated time periods. The expression of FOXO3 was examined with qRT-PCR. i YTHDF1-RIP followed by qRT-PCR detected the interaction between YTHDF1 and FOXO3 mRNA in MEFs before and after nutrient starvation. j YTHDF2-RIP followed by qRT-PCR detected the interaction between YTHDF2 and FOXO3 mRNA in MEFs before and after nutrient starvation. Data from three independent experiments are expressed as means ± SEM. Two-tailed unpaired ttests were used to estimate significance. P-values are indicated in the figure. Source data are provided as a Source Data file. Supplementary Fig.18. YTHDF3 protein interactome data analyses. a GO biological process enrichment analysis of the 1065 proteins that were differentially bound by YTHDF3 upon nutrient deprivation. b KEGG pathway enrichment analysis of the 1065 proteins that were differentially bound by YTHDF3 upon nutrient deprivation. c Schematic representation of the domain architecture of putative YTHDF3 RBP partner proteins. The domains are named and located according to the CCD protein domain database. Specific domains are shown with boxes of the indicated colors. d The consensus sequence motifs identified within the indicated YTHDF3 partner's target sequences, according to the starBase and CLIPdb databases. The qvalues in a and b were calculated using the ConsensusPathDB server based on hypergeometric test followed by false discovery rate (FDR) correction for multiple testing. Supplementary Fig.19. IGV tracks displaying the reads of YTHDF3 RIP-seq and MeRIPseq along the FOXO3 mRNAs. a IGV tracks displaying the reads of YTHDF3 RIP-seq (upper panels) and MeRIP-seq (lower panels) along the FOXO3 mRNAs. The red squares mark significant hyper-m 6 A peaks from MeRIP-seq of two replicates, the green square marks significant YTHDF3 up-enriched peaks from RIP-seq of two replicates, the orange square marks the m 6 A peaks predicted by bioinformatics, identified hyper-m 6 A-methylated in MeRIP-seq of one replicate, and functionally verified by EMSA and dual-luciferase reporter assay. b IGV tracks displaying the indicated RBPs' binding sites on FOXO3 mRNAs (pink line) from CLIPdb and starBase databases mouse data. The light-blue range marks the corresponding regions to the nutrient starvation-induced hyper-m 6 A-methylated and YTHDF3 up-enriched peaks on mouse FOXO3 transcript. Supplementary Fig.20. IGV tracks displaying the indicated RBPs' binding sites on FOXO3 mRNAs (pink line) from the starBase human data. The light-blue ranges mark the corresponding regions to the mouse hyper-m 6 A-methylated and YTHDF3 up-enriched peaks in response to nutrient starvation on mouse FOXO3 transcript. mRNAs (pink line) from CLIPdb database human data. The light-blue ranges mark the corresponding regions to the mouse hyper-m 6 A-methylated and YTHDF3 up-enriched peaks in response to nutrient starvation on mouse FOXO3 transcript. Supplementary Fig.22. Flow cytometry gating strategy. Gating strategy for AO staining in shYTHDF3 MEFs and shNS MEFs, with or without 20nM Baf.A1 for 4 hours. The left panels show cells gated on forward scatter (FSC) and side scatter (SSC) to exclude the most debris and dead cells. The right panels show cells analyzed using FITC and PE-Texas Red channels. (2) FOXO3-CDS-Mut1 :   ATGGCAGAGGCACCAGCCTCCCCGGTCCCGCTCTCTCCGCTCGAAGTGGAGCTGGTCCCAGAGTTCGAGCCACA  GAGTCGGCCACGCTCCTGTACGTGGCCCCTGCAGAGGCCGGAGCTGCAGGCGAGCCCGGCCAAGCCCTCGGGGG  AGACGGCCGCAGTCTCCATGATCCCCGAGGAGGACGACGATGAAGACGACGAGGACGGCGGCGGCCGAGCCAGC  TCGGCCATGGTGATCGGTGGCGGCGTGAGCAGCACGCTGGGTTCCGGGCTGCTCCTCGAGGATTCGGCCATGCT  GCTGGCTCCAGGAGGGCAGGACCTCGGGTCGGGGCCAGCGTCCGCCGCAGGCGCTCTGAGTGGGGGCACGCCGA  CGCAGCTGCAGCCTCAGCAGCCACTGCCACAGCCGCAGCCGGGGGCGGCTGGGGGCTCTGGGCAACCAAGGAAA  TGCTCCTCGCGGCGGAATGCCTGGGGGAACCTGTCCTATGCCGACCTGATCACCCGCGCCATCGAGAGCTCCCC  GGTCAAACGGCTCACTTTGTCCCAGATCTACGAGTGGATGGTGCGCTGTGTGCCCTACTTCAAGGATAAGGGCG  ACAGCAACAGCTCTGCGGGCTGGAAGATCTCCATCCGGCACAACCTGTCCCTGCACAGCCGCTTCATGCGCGTT  CAGAATGAAGGCACGGGCAAGAGCTCTTGGTGGATCATCAACCCCGATGGGGGAAAGAGCGGGAAGGCCCCCCG  GCGGCGTGCGGTCTCCATGGTCAACAGCAACAAGTACACCAAGAGCCGAGGCCGGGCAGCCAAGAAGAAGGCGG  CCCTGCAGGCTGCCCCAGAGTCGGCAGACGACAGTCCTTCCCAGCTCTCCAAGTGGCCTGGCAGCCCCACGTCC  CGCAGCAGCGACGAGCTGGATGCGTGGTCCGACTTCCGCTCGCGCACCAATTCCAACGCCAGCACCGTGAGCGG  CCGCCTGTCGCCCATCCTGGCAAGCACGGAGCTGGATGACGTCCAGGATGATGATGGTCCCCTGTCCCCCATGC  TGTACAGCAGCTCTGCCAGCCTGTCGCCCTCCGTGAGCAAGCCGTGTACTGTGGAGCTTCCGCGGCTGACGGTC  ATGGCCGGCACCATGAATCTGAATGATGGGCTGGCCGAGAACCTCATGGACGACCTGCTGGATAACATCGCGCT  CCCGCCATCGCAGCCATCGCCTCCTGGCGGGCTTATGCAGCGGGGCTCCAGCTTCCCATATACCGCCAAGAGCT  CCGGCCTGGGCTCCCCAACCGGCTCCTTCAACAGTACCGTGTTTGGACCTTCGTCTCTGAACTCCTTGCGTCAG  TCACCCATGCAGACTATCCAGGAGAACAGACCAGCCACCTTCTCTTCCGTGTCACACTACGGCAACCAGACACT  CCAAGACCTGCTTGCTTCAGACTCACTCAGCCACAGCGACGTCATGATGACCCAGTCGGACCCCTTGATGTCTC  AGGCTAGCACCGCCGTGTCCGCCCAGAATGCCCGCCGGAACGTGATGCTTCGCAACGATCCAATGATGTCCTTT  GCTGCCCAGCCTACCCAGGGGAGTTTGGTCAATCAGAACTTGCTCCACCACCAGCACCAAACCCAGGGCGCTCT  TGGTGGCAGCCGTGCCTTGTCAAATTCTGTCAGCAACATGGGCTTGAGTGACTCCAGCAGCCTTGGCTCAGCCA  AACACCAGCAGCAGTCTCCCGCCAGCCAGTCTATGCAAACCCTCTCGGACTCTCTCTCAGGCTCCTCACTGTAT  TCAGCTAGTGCAAACCTTCCCGTCATGGGCCACGATAAGTTCCCCAGTGACTTGGACCTGGACATGTTCAATGG  GAGCTTGGAATGTGACATGGAGTCCATCATCCGTAGTGAACTCATGGATGCTGACGGGTTGGATTTTAACTTTG  ACTCCCTCATCTCCACACAGAACGTTGTTGGTTTGAATGTGGGGAACTTCACTGGTGCTAAGCAGGCCTCATCT  CAAAGCTGGGTACCAGGCGACTACAAAGACCATGACGGTGATTATAAAGATCATGACATCGATTACAAGGATGA  CGATGACAAGTGA (3) FOXO3-CDS-Mut2 :   ATGGCAGAGGCACCAGCCTCCCCGGTCCCGCTCTCTCCGCTCGAAGTGGAGCTGGACCCAGAGTTCGAGCCACA  GAGTCGGCCACGCTCCTGTACGTGGCCCCTGCAGAGGCCGGAGCTGCAGGCGAGCCCGGCCAAGCCCTCGGGGG  AGACGGCCGCAGACTCCATGATCCCCGAGGAGGACGACGATGAAGACGACGAGGACGGCGGCGGCCGAGCCAGC  TCGGCCATGGTGATCGGTGGCGGCGTGAGCAGCACGCTGGGTTCCGGGCTGCTCCTCGAGGATTCGGCCATGCT  GCTGGCTCCAGGAGGGCAGGACCTCGGGTCGGGGCCAGCGTCCGCCGCAGGCGCTCTGAGTGGGGGCACGCCGA  CGCAGCTGCAGCCTCAGCAGCCACTGCCACAGCCGCAGCCGGGGGCGGCTGGGGGCTCTGGGCAACCAAGGAAA  TGCTCCTCGCGGCGGAATGCCTGGGGGAACCTGTCCTATGCCGACCTGATCACCCGCGCCATCGAGAGCTCCCC  GGACAAACGGCTCACTTTGTCCCAGATCTACGAGTGGATGGTGCGCTGTGTGCCCTACTTCAAGGATAAGGGCG  ACAGCAACAGCTCTGCGGGCTGGAAGAACTCCATCCGGCACAACCTGTCCCTGCACAGCCGCTTCATGCGCGTT  CAGAATGAAGGCACGGGCAAGAGCTCTTGGTGGATCATCAACCCCGATGGGGGAAAGAGCGGGAAGGCCCCCCG  GCGGCGTGCGGTCTCCATGGACAACAGCAACAAGTACACCAAGAGCCGAGGCCGGGCAGCCAAGAAGAAGGCGG  CCCTGCAGGCTGCCCCAGAGTCGGCAGACGACAGTCCTTCCCAGCTCTCCAAGTGGCCTGGCAGCCCCACGTCC  CGCAGCAGCGACGAGCTGGATGCGTGGACCGACTTCCGCTCGCGCACCAATTCCAACGCCAGCACCGTGAGCGG  CCGCCTGTCGCCCATCCTGGCAAGCACGGAGCTGGATGACGTCCAGGATGATGATGGACCCCTGTCCCCCATGC  TGTACAGCAGCTCTGCCAGCCTGTCGCCCTCCGTGAGCAAGCCGTGTACTGTGGAGCTTCCGCGGCTGACGGAC  ATGGCCGGCACCATGAATCTGAATGATGGGCTGGCCGAGAACCTCATGGACGACCTGCTGGATAACATCGCGCT  CCCGCCATCGCAGCCATCGCCTCCTGGCGGGCTTATGCAGCGGGGCTCCAGCTTCCCATATACCGCCAAGAGCT  CCGGCCTGGGCTCCCCAACCGGCTCCTTCAACAGTACCGTGTTTGGTCCTTCGTCTCTGATCTCCTTGCGTCAG  TCACCCATGCAGTCTATCCAGGAGAACAGACCAGCCACCTTCTCTTCCGTGTCACACTACGGCAACCAGACACT  CCAAGACCTGCTTGCTTCAGTCTCACTCAGCCACAGCGACGTCATGATGACCCAGTCGGTCCCCTTGATGTCTC  AGGCTAGCACCGCCGTGTCCGCCCAGAATGCCCGCCGGAACGTGATGCTTCGCAACGATCCAATGATGTCCTTT  GCTGCCCAGCCTACCCAGGGGAGTTTGGTCAATCAGATCTTGCTCCACCACCAGCACCAAACCCAGGGCGCTCT  TGGTGGCAGCCGTGCCTTGTCAAATTCTGTCAGCAACATGGGCTTGAGTGTCTCCAGCAGCCTTGGCTCAGCCA  AACACCAGCAGCAGTCTCCCGCCAGCCAGTCTATGCAAACCCTCTCGGTCTCTCTCTCAGGCTCCTCACTGTAT  TCAGCTAGTGCAAACCTTCCCGTCATGGGCCACGATAAGTTCCCCAGTGTCTTGGTCCTGGTCATGTTCAATGG  GAGCTTGGAATGTGACATGGAGTCCATCATCCGTAGTGATCTCATGGATGCTGACGGGTTGGATTTTAACTTTG  ACTCCCTCATCTCCACACAGAACGTTGTTGGTTTGAATGTGGGGATCTTCACTGGTGCTAAGCAGGCCTCATCT  CAAAGCTGGGTACCAGGCGACTACAAAGACCATGACGGTGATTATAAAGATCATGACATCGATTACAAGGATGA  CGATGACAAG