Fig 2: Model Illustrating the Possible Role of TGS1 in hTR BiogenesisIn TGS1-proficient cells (TGS1+), TGS1 restricts hTR interaction with the CBC and Sm complexes. hTR is enriched in CBs and excluded from nucleoli. In TGS1 mutant cells (TGS1-), hTR is monomethylated and acquires higher affinity for the CBC and Sm complexes, it is no longer enriched in CBs, relocates to nucleoli, and a fraction is exported into the cytoplasm. Total hTR levels increase, possibly because mislocalization sequesters hTR from the RNA degradation machinery, including the RNA exosome. A fraction of hTR may be reimported back into the nucleus, as it occurs for snRNAs and may be incorporated into telomerase. In the absence of TGS1, hTR is incorporated into functional telomerase complexes and telomerase activity is increased, proportional to hTR increase.

Fig 3: Mutations in TGS1 Induce Formation of Extended hTR Molecules but Do Not Affect 3' Processing(A) Diagram indicating the localization of the primers used to detect the extended hTR species (EXT1 and EXT2) generated by transcription readthrough beyond the 3' end of mature transcript (nt 451).(B) qRT-PCR showing increased expression of EXT1 and EXT2 in TGS1-depleted cells compared to TGS1-expressing cells: TGS1 M versus TGS1 MR, average of three different mutant clones compared to their rescued counterparts; siRNA for TGS1 (siTGS1) or non-targeting CTR (siCTR) in UMUC3 cells. Data are normalized to GAPDH. TGS1 M1 and M2 data are relative to parental cells (set to 1); UMUC3 siTGS1 data are relative to siCTR cells (set to 1). TGS1 M and TGS1 MR bars are means of three clones. UMUC3 siTGS1 are means of two treatments (*p < 0.05; **p < 0.01; one-way ANOVA).(C and D) 3'RACE-seq was performed to determine the abundance of hTR precursors terminating 1–10 nt beyond the 3' end of the hTR mature transcript (>451)(C) and oligoadenylated hTR molecules (OligoA) (D). No significant differences in these two types of precursors were observed in TGS1 M cells compared to CTR or TGS1 M1R cells. PAPD5-KO cells display an 80% reduction in OligoA species and a 75% reduction in >451 nt hTR molecules (C), as previously observed (Roake et al., 2019). In PARN-KO cells there is a 3-fold increase in both OligoA reads and >451 nt hTR reads (D). >35,000 reads per sample were analyzed.

Fig 4: Human TGS1 rescues the lethality induced by RNAi-mediated depletion of dTgs1.(A) Top panel: Representative western blot showing the dTgs1 levels in brains from larvae bearing the UAS-dTgs1 RNAi construct (dTgs1RNAi) and no driver (lane1), dTgs1RNAi and the Actin-Gal4 driver (Act>dTgs1RNAi) in a dTgs1+ background (lane 2) or in a dTgs1m1/+ background (lane 3), or homozygous for dTgs1m1 (lane 4). Bottom panel: Quantification of the dTgs1 protein levels relative to the Actin loading control in brains from larvae of the indicated genotypes. Data are from three experiments, 10 larval brains per sample. *, p< 0.05; ** p< 0,01; one-way ANOVA. (B) Western blot showing the reduced level of dTgs1 in brains from larvae heterozygous for the dTgs1R1 mutation, bearing the dTgs1RNAi construct driven by Actin-GAL4. In lanes 2, 3 and 4, the amount of extract loaded is 1/27, 1/9 and 1/3 of that of lane 5, respectively. (C) Phenotypes induced by the expression of UAS-dTgs1 RNAi driven by the indicated Gal4 drivers, in the presence or absence of a UAS-hTGS1 construct. (D) Schematic representation of catalytic dead versions of the Drosophila Tgs1 (dTgs1CD) and human TGS1 (hTGS1CD) proteins, carrying the indicated amino acid substitutions targeting the conserved amino acid stretch (grey), within the methyltransferase domain (MT, yellow). (E) Representative western blot showing that hTGS1 and hTGS1CD are expressed at similar levels in brains from larvae bearing the Actin-Gal4 driver and either UAS-hTGS1 (Act>hTGS1) or UAS-hTGS1CD (Act>hTGS1CD). Protein extracts are from 10 larval brains; Actin was used as the loading control.

Fig 5: Physical interactions between dTgs1 and the SMN complex.Proteins co-purifying with dTgs1 or Smn were identified by affinity purification (AP) using GFP-TRAP beads, followed by mass spectrometry (MS) and stringent filtering (see Materials and Methods for details). AP/MS was carried out using 0–3 hr embryos from mothers expressing UAS-GFP-dTgs1 driven by Actin-Gal4 or Smn-GFP under the control of the tubulin promoter. (A) Efficiency of GFP-TRAP-mediated AP assayed using an anti-GFP antibody (T, total protein extract; I, input (10%); S supernatant; IP, immunoprecipitate). (B) Schematic representations of the snRNP maturation pathway; note that in Drosophila there are only four Gemin proteins. See Introduction for a detailed description of this pathway. (C, D) dTgs1 (C) and Smn (D) interacting proteins. All protein IDs conforming to stringent filtering (see Materials and Methods) are shown in the Tables. Mean area corresponds to top 3 protein quantification (T3PQ), the mean of the three highest abundance peptides identified for each protein. The complete lists of the Tgs1- and Smn-interacting proteins are shown in S1 and S2 Tables.