TEFM is a potent stimulator of mitochondrial transcription elongation in vitro

Abstract

A single-subunit RNA polymerase, POLRMT, transcribes the mitochondrial genome in human cells. Recently, a factor termed as the mitochondrial transcription elongation factor, TEFM, was shown to stimulate transcription elongation in vivo, but its effect in vitro was relatively modest. In the current work, we have isolated active TEFM in recombinant form and used a reconstituted in vitro transcription system to characterize its activities. We show that TEFM strongly promotes POLRMT processivity as it dramatically stimulates the formation of longer transcripts. TEFM also abolishes premature transcription termination at conserved sequence block II, an event that has been linked to primer formation during initiation of mtDNA synthesis. We show that POLRMT pauses at a wide range of sites in a given DNA sequence. In the absence of TEFM, this leads to termination; however, the presence of TEFM abolishes this effect and aids POLRMT in continuation of transcription. Further, we show that TEFM substantially increases the POLRMT affinity to an elongation-like DNA:RNA template. In combination with previously published in vivo observations, our data establish TEFM as an essential component of the mitochondrial transcription machinery.

Figure 1.

Production of recombinant human TEFM and mouse Tefm. (A) Schematic representation of human TEFM and mouse Tefm. MTS (mitochondrial targeting signal), HhH (helix–hairpin–helix fold) and RnaseH fold (RuvC type) are indicated. (B) SDS-PAGE of recombinant human TEFM (amino acids 36–360) and mouse Tefm (amino acids 40–364). Both proteins run slightly below the expected size of 37.6 kDa. Molecular weight markers (kDa) are found in the first and last lanes. (C) Immunoblotting of recombinant human TEFM and endogenous TEFM from HeLa cells. Molecular weight marker is indicated.

Figure 2.

TEFM interacts directly with POLRMT to increase the polymerase processivity. (A) Microscale thermophoresis (MST) on labeled TEFM against POLRMT as described in materials and methods. The estimated fraction bound based on temperature jump data was plotted against POLRMT concentration giving a K_d value of 269 ± 14.4 nM. The experiment was performed in triplicate, and error bars show standard deviation. (B) Cartoon of the template used, the pUC18 vector with a mitochondrial LSP region insert including the CSB region, producing a run-off transcript of approximately 400 nts. (C) In vitro transcription with titration of TEFM. Lanes 1 and 2 contain DNA ladders (NEB 100 bp ladder and Affymetrix low molecular weight 10–100 nt ladder respectively; lane 3 is empty, and lanes 4–10 contain transcription reactions). Lane 4 has no TEFM added followed by lanes with TEFM in increasing concentrations (2.5, 5, 10, 20, 40, 80 nM in lanes 5–10, respectively). POLRMT to TEFM ratio, CSB region pre-terminations, CSB II, as well as run-off (RO) transcripts are indicated.

Figure 3.

TEFM strongly promotes POLRMT processivity on longer templates. (A) Cartoon of the templates used, produced from pUC18 containing an HSP or LSP insert (mtDNA insert). Note that only the LSP template contains the CSB region. To create run-off products of different lengths, the plasmids were cleaved with different restriction enzymes (RE) as specified in materials and methods. (B) In vitro transcription from LSP (lanes 1–4 and 9–10) or HSP (lanes 5–8) templates. These templates generate transcripts of about 400 nts (lanes 1–2 and 5–6), 3000 nts (lanes 3–4 and 7–8) as shown in panel (A), or 2000 nts (lanes 9–10, LSP PCR fragment template). TEFM was added at 40 nM where indicated. Note that only the LSP templates contain the CSB region, explaining the indicated CSB II-terminated bands present in LSP templates in the absence of TEFM.

Figure 4.

TEFM prevents termination at transcription pause sites and increases POLRMT affinity to DNA. (A) In vitro transcription from LSP on the 3000 nts run-off template (Figure 3A) at different time points (0, 3, 6, 9, 12, 15 and 30 min) in the absence (lanes 1–7) or presence (lanes 8–14) of 40 nM TEFM. The upper part of the figure is underexposed, and the bottom part is overexposed to compensate for a difference in labeling due to transcript length. For comparison, the pre-terminated transcript labeled PT* is shown in both parts. Full figures of both exposures are found in Supplementary Figure S3A and B. CSB region transcripts as well as run-off transcripts are indicated. (B) Quantification of run-off transcripts in panel (A). White squares with full black lines indicate samples in the absence of TEFM, and black squares with dotted lines indicate samples in the presence of TEFM. The transcript levels are measured in photostimulated luminescence per area (PSL/mm²) and the time in minutes. (C) The same quantification as in panel (B) but for the CSB II transcript. (D) Pulse-chase experiment on the 400 nts run-off LSP template. Transcription was initiated in the absence (lanes 1 and 6) or presence of 40 nM of TEFM (lane 11). After 3 min incubation, an excess of cold UTP was added to stop labeling. At this time point (0*), one of the reaction mixtures lacking TEFM was supplemented with 40 nM of TEFM (lanes 6–10). The reactions were then allowed to progress and samples were taken for analysis after 2.5, 5, 10 and 30 min. Transcripts prematurely terminated in the CSB region (PT and CSB II) as well as run-off transcripts are indicated. LMW marker (New England Biolabs) is indicated. (E) Microscale thermophoresis on an 18-mer 5′ Alexa488 labeled DNA hybridized to 8 nts of a 12-mer of RNA against POLRMT in the presence of ATP and in the presence or absence of TEFM. In the presence of ATP, the template allows for one nt incorporation before pausing. The estimated fraction bound based on combined thermophoresis and temperature jump data was plotted against POLRMT in the absence, white squares, or in the presence, filled black squares, of 2000 nM TEFM. The K_d of the interactions were determined as 47.8 ± 2.88 nM in the absence and 11.2 ± 1.00 nM in the presence of TEFM. The experiments were performed in triplicate and error bars show standard deviation. (F) DNase I footprinting on an LSP PCR template. Lane 1 is a no-protein control whereas lane 2 contains the initiation machinery (POLRMT, TFAM and TFB2M) and lane 3 contains the initiation machinery complemented with TEFM. Arrows indicate differences between samples in the absence or presence of TEFM (lanes 2 and 3, respectively). Relative positions to LSP +1 as well as the TFAM binding site are indicated.

Figure 5.

TEFM is a transcription elongation factor with cross-species function between human and mouse that do not require the NTE of POLRMT and aid the polymerase in bypass of DNA damage. (A) Cartoon of the template used for transcription with the mouse system, linearized to create a template with a run-off transcription product of 4500 nts. (B) In vitro transcription with the template described in panel (A) in the presence of mouse Tfam and mouse Tfb2m. Mouse Polrmt (lanes 1–3), mouse Δ320-Polrmt (lanes 4–6), or human POLRMT (lanes 7–9) was added at 20 nM. Mouse Tefm or human TEFM was added at 40 nM where indicated. Positions of run-off and CSB II-terminated transcripts are indicated. In the exposure used here, CSB II transcripts are difficult to visualize, since they are much shorter than the run-off transcription products and therefore less labeled by incorporation of radioactive UTP. (C) LCR-produced templates used for transcription. The templates include the TFAM binding site and the LSP promoter and generates a run-off product of 65 nts. The +50 nt, denoted as X in the figure, is either dG, 8-Oxo-dG, or an AP site in the experiment. (D) In vitro transcription on the templates described above with dG in lanes 1–2, AP in lanes 3–4 and 8-Oxo-dG in lanes 5–6. TEFM was added where indicated. The site of the 50 nts pre-termination (PT) and the 65 nt run-off (RO) are indicated. Molecular weight is indicated (Affymetrix low molecular weight 10–100 nt ladder). (E) Quantification of run-off and pre-terminated transcripts from panel (D). Run-off and pre-terminated transcripts are displayed as a ratio of total transcription (RO+PT). The experiment was performed in triplicate and the error bars show standard deviation.