Key RNA-binding domains in the La protein establish tRNA modification levels in Trypanosoma brucei

Abstract

The RNA-binding protein La is found in most eukaryotes, and despite being essential in many organisms, its function is not completely clear. Trypanosoma brucei, the causative agent of human African trypanosomiasis, encodes a 'classical' La protein (TbLa) composed of a La-motif, two RNA recognition motifs (RRM1 and RRM2α), a C-terminal short basic motif (SBM), and a nuclear localization signal (NLS). In T. brucei, like in most eukaryotes, position 34 of tRNATyr, -Asp, -Asn and -His is modified with queuosine (Q34). The steady-state levels of queuosine-modified tRNA in the insect form (procyclic) of T. brucei can fluctuate dynamically depending on growth conditions, but the mechanism(s) controlling Q34 levels are not well understood. A well-established function of La is in precursor-tRNA 3'-end metabolism, but in this work, we demonstrate that La also controls Q34-tRNA levels. Individual domain deletions showed that while deletion of La motif or RRM1 causes dysregulation of Q34-tRNA levels, no other domain plays a similar role. We also show that La is important for the normal balance of several additional tRNA modifications. These findings are discussed in the context of substrate competition between La and modification enzymes, also highlighting subcellular localization as a key determinant of tRNA function.

Graphical Abstract

Figure 1.

Knockdown of TbLa increases Q₃₄ levels in tRNA. (A) Growth curves of T. brucei wild type (WT, triangles), RNAi uninduced (RNAi-, circles) and TbLa RNAi induced cells (RNAi+, squares). The inset shows RT-PCR using total RNA extracted on day 5 of the growth curve. RNAi was induced by addition of tetracycline to the media (as described in the “Materials and methods” section). (B) An APB-northern blot with radioactive oligonucleotide probes specific for tRNA^Tyr and tRNA^Glu and total RNA samples as above from WT, TGT knockout (TGTΔ), TbLa RNAi- and RNAi + cells, tRNA^Glu serves as an RNA input normalization control_. (C) Graphical representation of percent of Q modified tRNA^Tyr shown in from (B). (D) Fluorescence in situ hybridization of TbLa RNAi uninduced (RNAi-) and induced (RNAi+) cells, probed for mature tRNA^Tyr with a Cy3-labeled oligonucleotide probe specific for the spliced anticodon arm. DIC refers to the phase contrast image, DAPI stains nuclear and kinetoplastid (mitochondrial) DNA and merge shows the overlapped images of DAPI and Cy3 probes.

Figure 2.

Recoded La rescues the growth defect caused by the knockdown of endogenous TbLa protein. (A) Growth curves of wild type (WT), uninduced (RNAi-) and induced (RNAi+) of endogenous TbLa cells expressing a constitutive myc-tagged recoded version of TbLa (myc-TbLaRec). (B) Agarose gels showing RT-PCR of endogenous TbLa mRNA (WT La) and recoded myc-TbLa mRNA (La Rec) using total RNA extracted on day 5 of the growth curve (as in Fig. 1). The “WT” lane indicates the RT-PCR of T. brucei total RNA from WT (29–13) cells and the “C” lane indicates the PCR using genomic DNA as the positive control. In this experiment RNAi refers to downregulation of endogenous TbLa. (C) Western blot with anti-c-myc antibodies and total protein extracted from uninduced (RNAi-) and induced (RNAi+) cells expressing myc-TbLaRec. The blot was stripped and probed for ADAT2 using anti-ADAT2 polyclonal antibodies as the loading control. (D) Immunofluorescence microscopy of TbLa RNAi cells expressing myc-TbLaRec, where Alexa 488-conjugated antibodies were used to detect myc-TbLaRec, while Mitotracker stains the mitochondria and DAPI nuclear and kinetoplast (mitochondria) DNA, merge shows the overlapping stains and DIC refers to a phase contrast image.

Figure 3.

Recoded La protein restores Q₃₄ levels in tRNA. (A) Levels of Q₃₄ in tRNA^Tyr determined by APB-northern blot. ³²P-oligonucleotide probes specific for tRNA^Tyr or tRNA^Glu were incubated with membranes from APB-PAGE separated total RNA isolated from WT, TbTGT knockout (TGTΔ), TbLa RNAi uninduced (−) and induced cells (+) (lanes 1–4), as well as the endogenous TbLa RNAi cell line constitutively expressing the recoded version of TbLa (LaRec) with RNAi uninduced (−) or induced (+) (lanes 5 and 6). (B) Graphical representation of the percentage of Q₃₄ modified tRNA^Tyr as shown in (A). Error bars obtained by measurement from three independent biological samples.

Figure 4.

Only LaM and RRM1 domains of TbLa are important for normal cell growth. (A) Schematic representation of each domain of myc-TbLaRec, showing the La motif (LaM), RNA recognition motif 1 (RRM1), RNA recognition motif 2, which includes the α3 helix (RRM2α), Short basic motif (SBM), and the nuclear localization signal (NLS). (B) Growth curves of TbLa RNAi uninduced (RNAi -) and induced (RNAi +) cells constitutively expressing TbLaRec domain deletion mutants as indicated. (C) Agarose gels showing RT-PCR of endogenous TbLa mRNA (WT La) and recoded myc-TbLa mRNA (La Rec) using total RNA extracted on day 5 of the growth curves shown in (B). (D) Western blot of total protein extracted from RNAi uninduced (-) and induced (+) cells expressing each myc-TbLaRec mutant. The blot was stripped and probed for ADAT2 using anti-ADAT2 polyclonal antibodies as the loading control.

Figure 5.

The key RNA-binding domains of TbLa (LaM and RRM1) are important for maintaining normal Q₃₄ levels. (A) APB-northern blot with radioactive oligonucleotide probes specific for tRNA^Tyr and tRNA^Glu and total RNA from WT, TGT knockout (TGTΔ) (lanes 1 and 2), and the endogenous TbLa RNAi cell line constitutively expressing deletion mutants of each domain of TbLaRec as indicated: La motif (LaM) (lanes 3 and 4), RNA recognition motif 1 (RRM1) (lanes 5 and 6), RNA recognition motif 2, which includes the α3 helix (RRM2α) (lanes 7 and 8), Short basic motif (SBM) (lanes 9 and 10) when TbLa RNAi is uninduced (−) and RNAi induced (+). The signal from tRNA^Glu served as a normalization control. (B) Graphical representation of the percentage of Q₃₄ modified tRNA^Tyr as shown in (A). Error bars obtained by measurement from three independent biological samples.

Figure 6.

LaM and RRM1 are important for tRNA binding in vitro. Electrophoretic mobility shift assay (EMSA) performed on native, mature tRNA^Tyr isolated from T. brucei, which was T4 PNK 5′ ³²P-end labeled prior to incubation with increasing concentrations of TbLa variants (A) wild type (B) LaM domain mutant (LaMΔ), or (C) RRM1 domain mutant (RRM1Δ). Protein-tRNA complexes were separated from unbound tRNA (free tRNA) on non-denaturing polyacrylamide gels. In all cases, lane 1 contains a no protein control, and lanes 2–7 contain TbLa, or domain deletion mutants at 40, 80, 160, 320, 640, and 800 nM, respectively. (D) Plotts of fraction bound tRNA graphed against increasing concentration of TbLa (circles), LaMΔ (squares), or RRM1Δ (triangles). (E) The apparent dissociation constants (K_Dapp) expressed in nM, and hill coefficients for TbLa and mutants derived from graph (D). All the EMSAs, plots, and calculations are representative of three independent replicates.

Figure 7.

Global tRNA modification changes upon TbLa downregulation are not limited to Q₃₄. Total tRNA samples from uninduced (RNAi-) and induced (RNAi+) were digested to nucleosides and analyzed by LC-MS/MS. Measurements were performed on triplicate biological samples. To calculate the fold difference, the raw values of each sample, were internally normalized to the values of Cytidine in each sample. Each sample quotient was then divided by the triplicate average of RNAi- samples. Asterisks (*) denote the P-value thresholds calculated using two-tailed unpaired t-test; *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001. (ct⁶A: Cyclic N⁶‐threonylcarbamoyladenosine, ac⁴C: N⁴‐acetylcytidine, hm⁵C: 5‐Hydroxymethylcytidine, t⁶A: N⁶‐threonylcarbamoyladenosine, ψ: pseudouridine, i⁶A: N⁶‐isopentenyladenosine, ncm⁵U: 5‐Carbamoylmethyluridine, m⁶t⁶A: N⁶‐methyl‐N⁶‐threonylcarbamoyladenosine, mnm⁵U: 5‐Methylaminomethyluridine, m³U: 3‐Methyluridine, mcm⁵Um: 5‐Methoxycarbonylmethyl‐2′‐O‐methyluridine, m⁶A: N⁶‐methyladenosine or 6‐methyladenosine, acp³U: 3‐(3‐Amino‐3‐carboxypropyl)uridine, m¹G: 1‐Methylguanosine, acp³D: 3(3‐Amino‐3‐carboxypropyl)‐5,6‐dihydrouridine, m¹I: N¹‐inosine or 1‐methylinosine, D: dihydrouridine, Ψm: 2′‐O‐methylpseudouridine, m⁵C: 5‐Methylcytidine, m^2,2G: N²,N²‐dimethylguanosine, m¹A: N¹‐methyladenosine or 1‐methyladenosine, mchm⁵U: 5‐Carboxyhydroxymethyluridine methyl ester, mcm⁵U: 5‐Methoxycarbonylmethyluridine, m¹ψ: 1‐Methylpseudouridine , ho⁵U: 5‐Hydroxyuridine, ncm⁵Um: 5‐Carbamoylmethyl‐2′‐O‐methyluridine, I: inosine, Cm: 2′‐O‐methylcytidine, m³C: 3‐Methylcytidine, Gm: 2′‐O‐methylguanosine, yW-58: 7‐Aminocarboxypropylwyosine methyl ester, Q: Queuosine, Am: 2′‐O‐methyladenosine, cm⁵U: 5‐Carboxymethyluridine, mimG: Methylwyosine, OHyW: Hydroxywybutosine, m¹Am: N¹,2′‐O‐dimethyladenosine, imG: wyosine, ms²t⁶A: 2‐Methylthio‐N⁶‐threonylcarbamoyl‐adenosine)

Figure 8.

OTTR sequencing reveals m¹A₅₈ modification is decreased in a subset of tRNA upon TbLa downregulation. (A) Reverse Transcription (RT) % misincorporation across all T. brucei tRNA isodecoders in tRNA samples from TbLa RNAi uninduced (RNAi-) or (B) TbLa RNAi induced (RNAi+) cells. (C) Fold-change difference for % RT misincorporation for a subset of tRNA positions on select tRNAs (color spectrum log₂ scale). TGT substrate tRNAs, which can be modified with Q₃₄, are indicated in bold text. Circle areas indicate adjusted P-value thresholds as indicated. (D) The volcano plot comparing the log₂ fold change of tRNA species between RNAi induced (RNAi+) and uninduced (RNAi-) conditions. The red dots indicate the tRNA species with a significant increase (P< 0.05) above a fold change threshold of 1.5 (log₂(0.59)) and green dots indicate the tRNA species with a significant decrease (P< 0.05) below a fold change threshold of 1.5 (log₂(0.59)) when TbLa is downregulated (RNAi+). Pre-tRNA species are labeled in smaller text while mature tRNA species are labeled in bigger text.