The fission yeast methyl phosphate capping enzyme Bmc1 guides 2'-O-methylation of the U6 snRNA

Abstract

Splicing requires the tight coordination of dynamic spliceosomal RNAs and proteins. U6 is the only spliceosomal RNA transcribed by RNA Polymerase III and undergoes an extensive maturation process. In humans and fission yeast, this includes addition of a 5' γ-monomethyl phosphate cap by members of the Bin3/MePCE family as well as snoRNA guided 2'-O-methylation. Previously, we have shown that the Bin3/MePCE homolog Bmc1 is recruited to the S. pombe telomerase holoenzyme by the LARP7 family protein Pof8, where it acts in a catalytic-independent manner to protect the telomerase RNA and facilitate holoenzyme assembly. Here, we show that Bmc1 and Pof8 are required for the formation of a distinct U6 snRNP that promotes 2'-O-methylation of U6, and identify a non-canonical snoRNA that guides this methylation. We also show that the 5' γ-monomethyl phosphate capping activity of Bmc1 is not required for its role in promoting snoRNA guided 2'-O-methylation, and that this role relies on different regions of Pof8 from those required for Pof8 function in telomerase. Our results are consistent with a novel role for Bmc1/MePCE family members in stimulating 2'-O-methylation and a more general role for Bmc1 and Pof8 in guiding noncoding RNP assembly beyond the telomerase RNP.

Graphical Abstract

Figure 1.

Bmc1, Pof8, and Thc1 guide 2′-O-methylation of U6. (A) qRT-PCR of U6 and sno530 in Bmc1 PrA immunoprecipitates, normalized to immunoprecipitation from an untagged strain (mean ± standard error, two-tailed unpaired t test) (n = 3 biological replicates). (B) Glycerol gradient sedimentation of myc-tagged Pof8, U4 and U6, and U3 from wild type (Pof8 myc) and bmc1Δ strains. U4 and U6 signals were normalized to U3 for calculating relative migration in the gradient. (C) Quantification of relative 2′-O-methylation-induced reverse transcriptase stops at A64, compared to a wild type strain (mean ± standard error, two-tailed paired t test) (n = 3 biological replicates). (D) 2′-O-methylation primer extension of U6 at high (1.5 mM) and limiting (0.1 mM) dNTP concentrations. 2′-O-methylated sites are indicated.

Figure 2.

Bmc1, Pof8, and Thc1 promote U6 snRNP assembly. (A) Native northern blot analysis of spliceosomal and non-spliceosomal (U3) snRNPs from native yeast cell extracts. (B) Quantification of U6-containing snRNPs from wild type and knockout yeast cell extracts (mean ± standard error, two-tailed paired t test) (n = 4 biological replicates). (C) Native northern blot analysis of total and Bmc1-immunoprecipitated U6. (D) Solution hybridization of U4/U6 pairing in wild type and knockout yeast strains using radiolabeled probes targeting the 5′ end of U4 and 3′ end of U6. (E) Quantification of U4/U6 pairing from solution hybridization assay, expressed as the fraction of non-duplexed U4 and U6 (‘free RNA’) (mean ± standard error, two-tailed paired t test) (n = 3 biological replicates). (F) T_m values from UV melt curve analysis of U4/U6 pairing with unmodified and A64-2′-O-methylated U6 oligos (mean ± standard error, two-tailed paired t test) (n = 6 technical replicates). (G) Northern and western blot analysis of U4, U6 and myc-tagged Prp24 from total cell extracts and myc-immunoprecipitates. (H) Quantification of Prp24-immunoprecipitated U4 and U6, relative to Prp24 myc (mean ± standard error, two-tailed paired t test) (n = 4 biological replicates).

Figure 3.

Bmc1 catalytic activity is not a requirement for 2′-O-methylation of U6. (A) AlphaFold (84) structure prediction of Bmc1 aligned to the SAH-bound (yellow) catalytic domain of MePCE (PDB 6DCB) (83) with mutations indicated in red. Inset: side chain interactions with SAH. (B) U6 2′-O-methylation primer extension in bmc1Δ cells transformed with the indicated plasmid. 2′-O-methylated sites are indicated. Western blots for Bmc1-HA expression and b-actin are indicated below. (C) Quantification of relative 2′-O-methylation-induced reverse transcriptase stops at A64, compared to wild type Bmc1-HA (mean ± standard error, two-tailed paired t test) (n = 4 biological replicates). (D) Quantification of relative 2′-O-methylation-induced reverse transcriptase stops at A64, compared to wild type Bmc1-HA, normalized to average Bmc1-HA expression relative to b-actin (mean ± standard error, two-tailed paired t test) (n = 4 biological replicates). (E) Western blot and northern blot analysis of co-immunoprecipitation of HA-tagged Bmc1, myc-tagged Pof8 and U6.

Figure 4.

The xRRM and Lsm2-8-binding surface are important for Pof8-mediated 2′-O-methylation of U6. (A) Schematic of Pof8 domains and mutants used in this study. N = N-terminal domain, LaM = La motif, RRM1 = RNA Recognition Motif 1, xRRM = extended RNA Recognition Motif. (B) U6 2′-O-methylation primer extension in pof8Δ cells transformed with the indicated plasmid. 2′-O-methylated sites are indicated. Western blots for Pof8-HA expression and b-actin are indicated below. (C) Quantification of relative 2′-O-methylation-induced reverse transcriptase stops at A64, compared to wild type Pof8-HA (mean ± standard error, two-tailed paired t test) (n = 3 biological replicates). (D) Northern and western blot analysis of U6, PrA-tagged Bmc1 and HA-tagged Pof8 from total cell extracts and PrA-immunoprecipitates. *Bmc1 PrA cleavage products, **an additional band cross-reacting with the antibody, arrows indicate Pof8 HA. (E) Quantification of Bmc1-immunoprecipitated U6, relative to the Pof8-HA-expressing strain (mean ± standard error, two-tailed paired t test) (n = 3 biological replicates). (F) qRT-PCR of sno530 in Bmc1 PrA pof8Δ immunoprecipitates, normalized to immunoprecipitation from a strain transformed with an empty vector (mean ± standard error, two-tailed unpaired t test) (n = 3 biological replicates).

Figure 5.

Bmc1 contributes to splicing robustness at elevated temperature. (A) Average intron retention from three biological replicates for wild type and bmc1Δ strains grown at 32°C or heat shocked at 42°C (two-tailed unpaired t-test with Welch's correlation). (B) Semi-quantitative RT-PCR in wild type and bmc1Δ strains grown at 32°C or heat shocked at 42°C for intron 1 of pud1, intron 2 of bor1, intron 3 of alp41, and intron 1 of rpl1603 (mean ± standard error, two-tailed paired t test) (n = 3 biological replicates). Representative gels are provided in Figure S7. (C–E) Comparison of 5′ splice site scores (C), 3′ splice site scores (D) and intron minimum free energy (kcal/mol) (E) for heat shock-sensitive and insensitive introns in wild type and bmc1Δ cells. Heat shock-sensitive introns were classified as introns that exhibited a greater than 2-fold increase in intron retention following heat shock and a false discovery rate <0.05. Remaining introns are classified as heat shock insensitive. Only introns with greater than 4 reads supporting splicing in all biological replicates were included (two-tailed unpaired t-test with Welch's correlation).

Figure 6.

Evolutionary convergence and divergence of Bmc1/MePCE and Pof8/LARP7 in noncoding RNA processing. (A) Summary of Bmc1/MePCE and Pof8/LARP7/p65 functions in the 7SK snRNP, telomerase holoenzyme, and U6 snRNP. LaM = La motif, RRM1 = RNA Recognition Motif 1, xRRM = extended RNA Recognition Motif, NTD = N-terminal domain (Lsm2-8-interacting region), MID = MePCE-Interacting Domain. The existence and composition of a U6 snRNP in T. thermophila is currently unknown. (B) Schematic of the U6 biogenesis pathway in fission yeast. NTC = NineTeen Complex.