Mammalian transcription factor A is a core component of the mitochondrial transcription machinery

Abstract

Transcription factor A (TFAM) functions as a DNA packaging factor in mammalian mitochondria. TFAM also binds sequence-specifically to sites immediately upstream of mitochondrial promoters, but there are conflicting data regarding its role as a core component of the mitochondrial transcription machinery. We here demonstrate that TFAM is required for transcription in mitochondrial extracts as well as in a reconstituted in vitro transcription system. The absolute requirement of TFAM can be relaxed by conditions that allow DNA breathing, i.e., low salt concentrations or negatively supercoiled DNA templates. The situation is thus very similar to that described in nuclear RNA polymerase II-dependent transcription, in which the free energy of supercoiling can circumvent the need for a subset of basal transcription factors at specific promoters. In agreement with these observations, we demonstrate that TFAM has the capacity to induce negative supercoils in DNA, and, using the recently developed nucleobase analog FRET-pair tC(O)-tC(nitro), we find that TFAM distorts significantly the DNA structure. Our findings differ from recent observations reporting that TFAM is not a core component of the mitochondrial transcription machinery. Instead, our findings support a model in which TFAM is absolutely required to recruit the transcription machinery during initiation of transcription.

Fig. 1.

TFAM is required for HSP1 and LSP transcription in mitochondrial extracts. (A) Sequences of the HSP1 and LSP promoters. The TFAM binding sites (13, 14) are underlined. The boxed sequences were mutated to inactivate TFAM-dependent transcription. (B) Immunodepletion of TFAM from mitochondrial extracts. Immunoblotting against the indicated proteins demonstrates that POLRMT and TFB2M levels remained unaffected. (C) Run-off transcription in mitochondrial extracts was monitored using run-off transcription on linearized templates containing HSP1 or LSP. Experiments were performed as described in Materials and Methods. TFAM (1 pmol) was added when indicated. As a control, in vitro transcription was performed with recombinant TFAM, TFB2M, and POLRMT (lane 1). (D) Run-off transcription was monitored on linearized DNA templates containing (WT) or lacking (MUT) the high-affinity TFAM binding site. Experiments were performed as described in Materials and Methods. TFAM (2.5 pmol) was added when indicated. The TFAM binding sites were mutated by changing A to C, C to A, T to G, and G to T at positions −26 to −35 relative to the transcription start site. LSP PT corresponds to a shorter LSP product caused by premature transcription termination at CSBII (3).

Fig. 2.

The requirement of TFAM can be relaxed at low ionic strength. (A) The effect of increasing ionic strength on run-off transcription was monitored on linearized DNA templates containing LSP (Middle), HSP1 (Bottom), or both HSP1 and LSP (Top). Transcription reactions contained the indicated template, POLRMT (400 fmol) and TFB2M (400 fmol). Transcription was monitored in the absence (Left) and presence (Right) of TFAM (2.5 pmol). (B) Promoter-specific transcription is less sensitive to salt concentration than promoter-independent transcription. The effect of ionic strength on HSP1 transcription was monitored in the presence of TFAM as in A. (C) Bacterially expressed POLRMT and TFB2M (indicated with a B) require TFAM for HSP1 transcription at 40 mM NaCl. Identical results were obtained with TFB2M expressed in insect cells (SF9).

Fig. 3.

Negative supercoiling can relax the requirement of TFAM at HSP1, but not at LSP. TFAM dependence was monitored on negatively supercoiled or linearized templates containing LSP or HSP1. To produce transcripts of a defined length, CTP was omitted from the LSP transcription reaction, generating an 18-nt–long product. Similarly, UTP was omitted from the HSP1 reaction, leading to the formation of a 17-nt transcript. Transcription reactions contained the indicated template, and POLRMT (400 fmol), TFB2M (400 fmol), and TFAM (2.5 pmol) were added when indicated.

Fig. 4.

Topoisomerase I restores the requirement of TFAM for initiation of transcription at HSP1. (A) Transcription reactions were performed as in Fig. 3. When indicated, the template was incubated with TOP1mt (280 fmol) before transcription, as described in Materials and Methods. (B) TFAM induces negative supercoils in plasmid DNA. The input DNA (lane 2) was relaxed by incubation with TOP1mt for 15 min (lane 3) before increasing amounts of TFAM (0, 5.5, 11, 22, or 45 pmol) were added to the DNA and incubated together with TOP1mt for another 20 min (lanes 4–8).