N(6)-Methyladenosine RNA Modification Regulates Shoot Stem Cell Fate in Arabidopsis

Abstract

N(6)-Methyladenosine (m(6)A) represents the most prevalent internal modification on mRNA and requires a multicomponent m(6)A methyltransferase complex in mammals. How their plant counterparts determine the global m(6)A modification landscape and its molecular link to plant development remain unknown. Here we show that FKBP12 INTERACTING PROTEIN 37 KD (FIP37) is a core component of the m(6)A methyltransferase complex, which underlies control of shoot stem cell fate in Arabidopsis. The mutants lacking FIP37 exhibit massive overproliferation of shoot meristems and a transcriptome-wide loss of m(6)A RNA modifications. We further demonstrate that FIP37 mediates m(6)A RNA modification on key shoot meristem genes inversely correlated with their mRNA stability, thus confining their transcript levels to prevent shoot meristem overproliferation. Our results suggest an indispensable role of FIP37 in mediating m(6)A mRNA modification, which is required for maintaining the shoot meristem as a renewable source for continuously producing all aerial organs in plants.

Figure 1.. Loss of Function of FIP37 Causes Massive Overproliferation of the SAM

(A) Schematic diagram shows the T-DNA insertion sites in fip37–1S and fip37–4. Exons and other sequences are represented by black boxes and lines, respectively. (B) Dissected siliques from plants with various genetic backgrounds. Asterisks indicate fip37–4 homozygous seeds that appear in abnormal white color and fail to develop further. Scale bars, 0.5 mm. (C) qPCR analysis of FIP37expression in 15-day-old seedlings with various genetic backgrounds. Error bars, mean ± SD; asterisks indicate statistically significant differences of FIP37 expression between mutants and wild-type plants (two-tailed paired Student’s t test, p < 0.001); n = 3 biological replicates. (D) Loss of function of FIP37 at the post-embryonic stage results in enlarged SAMs. Upper panels show top views of 15-day-old seedlings, while lower panels show the corresponding SAMs examined by scanning electron microscopy (SEM). Red asterisks indicate SAMs. Double-headed arrows indicate the parameters for comparing meristem size in (E). Scale bars, 1 mm (upper panels) and 50 μm (lower panels). (E) Statistically significant differences in the SAM size between 15-day-old wild-type (WT) and fip37–4 LEC1:FIP37 seedlings. Error bars, mean ± SD; **p < 0.001, two-tailed paired Student’s t test; n = 15. (F) Median longitudinal sections of the SAMs of wild-type (upper panels) and fip37–4 LEC1:FIP37 (lower panels) seedlings at 5 (D5), 10 (D10), and 20 days (D20) after germination. Asterisks indicate SAMs. Scale bars, 50 μm. See also Figures S1 and S2.

Figure 2.. Gene Expression Patterns and Protein Localization of FIP37

(A) GUS staining of 5-day-old and 10-day-old FIP37:GUS seedlings. Cotyledons (c) are labeled. Scale bars, 1 mm. (B) In situ localization of FIP37 in the SAM of a 10-day-old wild-type seedling. The inset shows a section of a 10-day-old fip37–4 LEC1:FIP37 SAM hybridized with the FIP37 antisense probe as a control. Scale bars, 50 μm. (C) Confocal analysis of FIP37-GFP in the SAM (upper panel) and root tip (lower panel) of a 10-day-old fip37–4 gFIP37-GFP seedling. Inset shows an enlarged view of the nucleus of a root tip cell. GFP, GFP fluorescence; BF, bright-field image; Merge, merge of GFP and bright-field images. Scale bars, 20 μm. (D) Immunolocalization of FIP37–4HA in the nucleus of a fip37–4 gFIP37–4HA root tip cell. BF, bright-field image; Merge, merge of anti-HA and bright-field images. Scale bars, 10 μm. See also Figure S3.

Figure 3.. Effect of Loss of FIP37 on the Arabidopsis m⁶A RNA Methylome and Transcriptome

(A) Dot blot analysis of m⁶A levels in mRNA purified from 5-day-old wild-type and fip37–4 LEC1:FIP37 seedlings. (B) m⁶A percentage relative to adenosine (m⁶A/A ratio) determined by LC-MS/MS in mRNA purified from 5-day-old wild-type, LEC1:FIP37, and fip37–4 LEC1:FIP37 seedlings. LC-MS/MS was repeated with three biological replicates. Statistical analysis was performed using two-tailed paired Student’s t test. The results are considered statistically significant at p < 0.05. Error bars, mean ± SD. (C) Heatmap representing IP enrichment values for m⁶A peaks with statistically significant difference between fip37–4 LEC1:FIP37 and wild-type seedlings. (D) Cumulative distribution function of log2 peak intensity of m⁶A peaks in fip37–4 LEC1:FIP37 and wild-type seedlings. (E) Comparison of distribution of m⁶A peaks in different segments of wild-type (upper panel) and fip37–4 LEC1:FIP37 (lower panel) transcripts. Left panels show pie charts presenting the percentages of m⁶A peaks in different transcript segments, while right panels show relative enrichment of m⁶A peaks in different transcript segments. (F) Distribution of m⁶A peaks in transcript segments divided into 5′ UTR, CDS, and 3′ UTR in wild-type and fip37–4 LEC1:FIP37 seedlings. (G and H) Boxplot comparison of enrichment fold of m⁶A peaks in different transcript segments (G) or shared m⁶A peaks (H) in wild-type and fip37–4 LEC1:FIP37 seedlings. (I) Heatmap showing differentially expressed genes in wild-type and fip37–4 LEC1:FIP37 seedlings. (J) Cumulative distribution of changes in gene expression between fip37–4 LEC1:FIP37 and wild-type seedlings for non-m⁶A and m⁶A targeted genes. p < 10 × 10⁻¹⁶, two-sided Mann-Whitney test. (K) Boxplot comparison of expression levels of non-m⁶A genes and genes bearing m⁶A in different transcript segments between fip37–4 LEC1:FIP37 and wild-type seedlings. See also Figures S4 and S5, and Table S1, S2, and S3.

Figure 4.. Upregulation of STM and WUS Is Coupled with Their Loss of m⁶A in fip37–4 LEC1:FIP37

(A) m⁶A modification in STM mRNA is abolished in fip37–4 LEC1:FIP37. Upper panel shows m⁶A peaks revealed by m⁶A-seq In STM mRNA from wild-type and fip37–4 LEC1:FIP37. The STM transcript structure is shown beneath, with thick boxes and lines representing exons and introns, respectively. Lower panel shows m⁶A-IP-qPCR results, in which cDNA fragments amplified are indicated below the STM transcript structure. Shoot apices of 5-day-old wild-type and fip37–4 LEC1:FIP37 seedlings were harvested for m⁶A-IP-qPCR in (A and B). Error bars, mean ± SD; *p < 0.05 in (A and B), two-tailed paired Student’s t test; n = 3 biological replicates. (B) m⁶A modification in WUS mRNA is abolished in fip37–4 LEC1:FIP37. Upper panel shows the WUS transcript structure labeled with cDNA fragments amplified in m⁶A-IP-qPCR assays shown in the lower panel. Error bars, mean ± SD. (C) Western blot analysis using anti-HA antibody shows the expression of FIP37–4HA in protein extracts (Input) or immunoprecipitated fractions (Eluate) from shoot apices of 5-day-old fip37–4 gFIP37–4HA seedlings. The arrowhead indicates the FIP37–4HA band. (D) RIP assay shows direct binding of FIP37–4HAto STM and WUS transcripts. Shoot apices of 5-day-old fip37–4 gFIP37–4HA seedlings were harvested for RIP. Enrichment of ACTIN2 (ACT) serves as a negative control. Error bars, mean ± SD. Asterisks indicate statistically significant differences in FIP37–4HA enrichment with STM or WUS transcripts compared with that with ACTIN2 (two-tailed paired Student’s t test, p < 0.05); n = 3 biological replicates. (E) qPCR analysis of mRNA expression of STM and WUS in fip37–4 LEC1:FIP37 versus wild-type seedlings. Gene expression levels in wild-type seedlings are set as 1, which is indicated by a red dotted line. Error bars, mean ± SD; asterisks indicate statistically significant differences in gene expression between fip37–4 LEC1:FIP37 and wild-type seedlings (two-tailed paired Student’s t test, p < 0.05). n = 3 biological replicates. (F) In situ localization of STM and WUS in wild-type (WT) and fip37–4 LEC1:FIP37 SAMs. Arrow indicates an incipient leaf primordium in a wild-type SAM. Arrowheads indicate STM expression in the regions where lateral organs emerge in a fip37–4 LEC1:FIP37 SAM. Scale bars, 50 μm. (G and H) The FIP37-GR fusion protein is biologically functional. One-day-old fip37–4gFIP37-GR seedlings were mock treated (0.03% ethanol and 0.012% Silwet L-77) (Mock) or treated with 10 μM dexamethasone (Dex) and 0.012% Silwet L-77 once a week. At 3 weeks after the first treatment, a mock-treated plant shows similar phenotypes to fip37–4 LEC1:FIP37 (G) (upper panel, a top view of the seedling; lower panel, an enlarged view of the SAM shown in the upper panel), whereas a Dex-treated plant exhibits a rescued SAM phenotype with continuous generation of leaves (H). Scale bars, 1 mm. (I) Single Dex treatment of fip37–4 gFIP37-GR is able to rescue the developmental defects of fip37–4. Upper panel, a 2-week-old fip37–4 gFIP37-GR plant before Dex treatment shows an overproliferated SAM and a dramatic arrest in leaf production. Lower panel, a 4-week-old fip37–4 gFIP37-GR plant shows continuous generation of leaves from the SAM after a single Dex treatment at 2 weeks old. Scale bars, 1 mm. (J) Induced FIP37 activity upregulates m⁶A levels in STM and WUS mRNA. Time-course m⁶A-IP-qPCR was performed using shoot apices of fip37–4 gFIP37-GR mock treated (Mock) or treated with 10 μM Dex for 0, 4, and 8 hr. Error bars, mean ± SD; asterisks indicate statistically significant differences in m⁶A enrichment levels between mock- and Dex-treated seedlings (two-tailed paired Student’s t test, p < 0.05); n = 3 biological replicates. (K) Induced FIP37 activity does not affect m⁶A levels in CLV3 and KNAT1 mRNA. m⁶A-IP-qPCR was performed using shoot apices of fip37–4 gFIP37-GR mock treated (Mock) or treated with 10 μM Dex for 4 hr. Error bars, mean ± SD; n = 3 biological replicates. (L) Induced FIP37 activity reduces the mRNA levels of WUS and STM. Time-course expression of WUS and STM in 5-day-old fip37–4 gFIP37-GR shoot apices mock treated (Mock) or treated with 10 μM Dex for 0, 4, and 8 hr. The relative expression of each gene at different time points was calculated by normalizing the gene expression in Dex-treated samples against that in mock-treated samples. Error bars, mean ± SD; n = 3 biological replicates. (M) Induced FIP37 activity does not affect CLV3 and KNAT1 expression. Expression of CLV3 and KNAT1 was examined in 5-day-old fip37–4 gFIP37-GR shoot apices mock treated (Mock) or treated with 10 μM Dex for 4 hr. Error bars, mean ± SD; n = 3 biological replicates. See also Figures S6.

Figure 5.. Downregulation of MTA Causes Various Developmental Defects and a Reduction in m⁶A Levels

(A) Phenotypic comparison of developing wild-type and AmiR-mta seedlings. Scale bars, 0.1 cm. (B) A strong AmiR-mta line at 1 month old exhibits overproliferation of the SAM and abnormal leaf development. Red arrowheads indicate the formation of meristematic-like structure. Scale bars, 0.1 cm. (C) Dot blot analysis shows a reduction of m⁶A levels in 10-day-old AmiR-mta transgenic plants and fip37–4 LEC1:FIP37. (D) m⁶A levels of selected genes are reduced in AmiR-mta. Shoot apices of 10-day-old wild-type and AmiR-mta seedlings were harvested for m⁶A-IP-qPCR. Error bars, mean ± SD; *p < 0.05, two-tailed paired Student’s t test; n = 3 biological replicates. (E) qPCR analysis of expression of MTA, FIP37, WUS, and STM in 10-day-old wild-type (WT) and AmiR-mta with strong or weak phenotypes. The gene expression level in wild-type seedlings is set as 1. Error bars, mean ± SD; n = 3 biological replicates.

Figure 6.. SAM Overproliferation in fip37–4 LEC1:FIP37 Is Attributable to WUS and STM Activity

(A) SAM overproliferation in fip37–4 LEC1:FIP37 is partially suppressed by wus-8 or stm-10, and fully suppressed by stm-10 wus-8. Upper panels show top views of 18-day-old seedlings with various genetic backgrounds, while lower panels show close-up views of their SAM regions. Scale bars, 0.5 mm. (B) SEM analysis of the SAMs of 18-day-old fip37–4 LEC1:FIP37, stm-10 wus-8, and stm-10 wus-8 fip37–4 LEC1:FIP37 seedlings. Scale bars, 100 μm. See also Figure S7.

Figure 7.. FIP37-Mediated m⁶A Methylation Accelerates Decay of STM and WUS Transcripts

(A) qPCR analyses of transcripts levels of STM, WUS, TUB8, and WRKY6 in 5-day-old wild-type and fip37–4 LEC:FIP37 seedlings treated by actinomycin D versus mock treated for 0, 8, and 24 hr. The relative expression of each gene at different time points was calculated by normalizing the gene expression in actinomycin D-treated samples against that in mock-treated samples. Error bars, mean ± SD; *p < 0.001, two-tailed paired Student’s t test; n = 3 biological replicates. (B) A model describing regulation of shoot meristem proliferation by FIP37-mediated m⁶A modification in Arabidopsis. FIP37 determines the m⁶A mRNA modification landscape and acts as a core component of the plant methyltransferase complex including MTA. In the wild-type SAM (left panel), FIP37-mediated m⁶A modification confines WUS expression in the organizing center (blue dotted circle) and STM expression in the undifferentiated region of the SAM (green dotted lines). Loss of FIP37 abolishes m⁶A modification of WUS and STM (right panel). This delays decay of WUS and STM mRNAs, causing their expanded expression and a resulting overproliferated SAM. SAM, shoot apical meristem; LP, leaf primordium. See also Table S4.