Alignment of vertebrate pol γB subunits with T.Thermophilus GRS. The program Clustal (PCGene) was used to align the sequences of human (HSGB; GenBank accession number AF177201), mouse (MMGB; accession number AF177202) and Xenopus (XLGB; accession number AF124606) pol γB subunits with T.thermophilus GRS (TTGRS; accession number P56206). Identical amino acids are indicated with asterisks; similar ones, with dots. The first amino acid of the mature Xenopus protein and the predicted ones for the human and mouse proteins are indicated in bold. The consensus motifs in class II aaRS sequences are underlined in the TTGRS sequence (31). The DNA sequence we determined for human pol γB includes a few changes with respect to the original sequence reported in GenBank U94703, as follows: silent T to C changes at positions 371 and 998; C to G at 448, resulting in a T122 to R change in the protein sequence; A to G at 488, resulting in S136 changed to G; G to A in 587, resulting in A169 changed to T; and a short frame shift caused by an insertion of an A in position 940 and a deletion of a C in position 961, resulting in a change of sequence TNFTTI to NKLYYN.

All three vertebrate pol γB subunits stimulate processive DNA synthesis by human pol γA. (A) Coomassie blue stained SDS gel showing recombinant Xenopus (X), mouse (M) and human (H) pol γB subunits. 4.7 pmol of each protein was loaded on the gel, corresponding to ~230 ng of each polypeptide. The sizes in kDa of molecular mass markers run in parallel are indicated on the left. (B) Primer extension assay on oligo(dT)-primed poly(dA). Reactions were carried out as described in Materials and Methods using 40 fmol of pol γA and 240 fmol of pol γB per reaction. Lane 1, no protein; lanes 2–5 correspond to reactions carried out with pol γA alone, A + Xenopus B, A + mouse B and A + human B, respectively, incubated for 1 min; lanes 6–9 are as 2–5 but incubated for 2 min; lanes 10–13 correspond to reactions incubated for 5 min. The sizes in nucleotides of MspI restriction fragments of pUC 18 are indicated on the left.

Copurification of human γA and B subunits with pol γ activity. Mitochondrial DNA polymerase was purified from HeLa cells as described in Materials and Methods. Fractions from the SP Sepharose and Heparin Hi Trap chromatography steps were assayed for reverse transcriptase activity (indicated as c.p.m. × 1000 on the right of each chart) and also analyzed by western blot. The upper part of each membrane (containing proteins above 81 kDa) was incubated with an anti-human pol γA antibody (A). The lower part was incubated with an anti-human pol γB antibody (B). Fraction numbers indicated at the bottom of the histogram also refer to the lanes in the associated western blot; L, loaded fraction; FT, flow through; AB, recombinant pol γA plus γB used as positive control. The sizes in kDa of protein molecular mass markers run in parallel are indicated on the left.

Deletion analysis of human pol γ B. (A) Diagram showing the locations of deletions within human pol γB in comparison to the full-length mature protein (HGB). Black rectangles indicate the regions that align with conserved motifs 1–3 (from left to right, respectively) of class II aminoacyl-tRNA synthetases; gray rectangles represent the two conserved blocks close to the C-terminus. Amino acid sequences at the end points of the deletion sites are indicated. ΔI1 is depicted with the two N-terminal deletions due to the proximity of the deleted region to the N-terminus. For ΔI3, two extra amino acids, E and F, were inserted in the deletion site. The size of each protein (including the his-tag) is indicated on the right. (B) Coomassie blue-stained gel showing the mutant proteins purified by Ni-NTA affinity chromatography together with recombinant wild-type pol γB (GB). The pol γB derived proteins are indicated with asterisks. Note that protein samples contained a variable contamination with bacterial proteins in the 70 to 80 kDa range. The sizes in kDa of molecular mass markers run in parallel are indicated on the left. (C) Activity of wild-type human pol γB and deletion mutants on poly(dA):oligo(dT). (D) EMSA assays to test the ability of pol γB derivatives to bind with pol γA to a primer-template oligonucleotide. 0 indicates no enzyme. A, human pol γA alone; B, human pol γA plus pol γB; N1 to I3, human pol γA plus each of the deletion mutants. Binding reactions contained 40 fmol of pol γA and 40 or 400 fmol (triangles) of pol γB derived protein with 40 fmol of a 26mer:45mer primer:template.

Model for the recognition of the origin of replication by mitochondrial DNA polymerase. (A) Structures present at the heavy strand (O_H) and light strand (O_L) origins of replication of human mitochondrial DNA [adapted from Brown et al. (27) and Wong and Clayton (30)]. RNA primers are indicated by the gray outline, sites of DNA synthesis are indicated by black arrows. Two alternative start sites are shown at O_H. (B) mtRNA polymerase initiates transcription at the light strand promoter (LSP) and proceeds through conserved sequence block 2 (CSB2). The GC-rich sequence at CSB2 has been proposed to stabilize the RNA–DNA hybrid. (C) The RNA adopts a secondary structure that can be recognized by MRP endonuclease. The double headed arrow at the bottom of the panel indicates that single-stranded DNA in the origin region may fold in an RNA-like secondary structure. We suggest that DNA pol γB may interact with the nascent RNA, with the folded single-stranded DNA, or with both.