Relaxed primer specificity associated with reverse transcriptases encoded by the pFOXC retroplasmids of Fusarium oxysporum

Abstract

The pFOXC mitochondrial retroplasmids are small, autonomously replicating linear DNAs that have a telomere-like repeat of a 5-bp sequence at their termini. The plasmids are possible evolutionary precursors of the ribonucleoprotein complex telomerase, as they encode an active reverse transcriptase (RT) that is involved in plasmid replication. Using an in vitro system to study reverse transcription, we show that the pFOXC RT is capable of copying in vitro-synthesized RNAs by use of cDNA primers or extension of snapped-back RNA templates. The ability of the pFOXC RT to use base-paired primers distinguishes it from the closely related RTs encoded by the Mauriceville and Varkud mitochondrial retroplasmids of Neurospora spp. Reaction products are similar, but not identical, to those obtained with conventional RTs, and differences reflect the ability of the pFOXC RT to initiate cDNA synthesis with loosely associated primers. The pFOXC RT can also copy DNA templates and extend 3' mismatched DNA oligonucleotide primers. Analysis of pFOXC in vivo replication intermediates suggests that telomeric repeats are added during reverse transcription, and the ability to extend loosely associated primers could play a role in repeat formation by mechanisms similar to those associated with telomerase and certain non-long-terminal-repeat retrotransposons.

FIG. 1.

Model for the replication of the pFOXC retroplasmids. (A) Plasmid replication cycle. The 1.9-kb double-stranded DNA contains a single open reading frame (ORF; box) encoding a 527-amino-acid RT. The open reading frames encoded by pFOXC2 and pFOXC3 are identical in length, and predicted polypeptides have 88.8% amino acid identity (24). Transcription initiates at a region downstream of the hairpin structure with the plus strand as template and proceeds around the hairpin to copy the entire minus strand. The resulting transcript (dashed gray line) serves both as an mRNA to produce the plasmid-encoded RT and as a template for cDNA synthesis. Reverse transcription begins at or near the 3′ end of the plasmid RNA and synthesizes a full-length minus-strand cDNA. The 3′ end of the resulting minus-strand cDNA likely contains an inverted repeat that could fold back upon itself and be elongated by the plasmid RT or an mtDNA polymerase. The RNA template may be displaced during plus-strand DNA synthesis or be degraded by a host RNase H. The mature plasmid DNA has a 5′ covalently linked protein (black circle), which appears to be added following reverse transcription. (B) Model for initiation of minus-strand cDNA synthesis and repeat addition. Plasmid transcripts have approximately three pentameric repeats and potentially snap back to prime minus-strand cDNA synthesis. Following initiation, nascent cDNAs may slide back on the template prior to elongation, resulting in minus-strand cDNAs having approximately four copies of the pentameric repeat. Additional details are described in the text. RNA sequences are in italics, and DNA sequences potentially added during the initiation process are underlined.

FIG. 2.

Pre- and posttreatment of endogenous reactions with DEAE-RT preparations. Reaction mixtures were preincubated in standard endogenous reaction buffer with or without actinomycin D (Ac) or RNase A (R) for 5 min prior to the addition of [³²P]dCTP. (Left panel) Reaction products were treated with RNase A (R) or DNase I (D) or left untreated prior to precipitation and electrophoresis in a 1.5% nondenaturing agarose gel. (Middle panel) Endogenous reaction mixtures were subjected to the following treatments: lane 6, mock incubation; lane 7, incubation in 0.1 M NaOH; lane 8, incubation with proteinase K; lane 9, incubation with proteinase K and then extraction with phenol-CIA; lane 10, mock incubation, followed by extraction with phenol-CIA. Reaction products were treated with glyoxal prior to electrophoresis. (Right panel) Endogenous reaction products were incubated with (lane 12) or without (lane 11) λ exonuclease. Reaction products were treated with glyoxal prior to electrophoresis. The filled arrows indicate the 1.95-kb cDNA reaction product, and the unfilled arrow indicates a band which is detected in nondenaturing gels and is more pronounced in reactions that retain endogenous RNA. The size of the products was determined by 5′-end-labeled λ-PstI and λ-HindIII/EcoRI restriction fragment markers (data not shown).

FIG. 3.

RT activity and specificity in reactions having exogenous RNA templates. The pFOXC RT was liberated from mtRNP particles by treatment with MN, and reverse transcription was assayed using exogenous template-primer substrates or RNAs alone. (A) Reactions with MN-treated pFOXC3 mtRNPs with poly(rC)-oligo(dG) template-primer substrates. The ability to incorporate [³²P]dGTP into high-molecular-weight products was measured as described in Materials and Methods, and counts per minute represent an average of two or more reactions having specified amounts of MgCl₂. (B) Products of MN-treated mtRNP reactions containing total mtRNA isolated from a pFOXC3-containing strain used as a probe for a Southern blot containing HindIII-digested mtDNA from the same strain. The ethidium bromide (EtBr)-stained gel is shown on the left, and the autoradiogram is on the right. An arrow identifies the band that corresponds to the 1.9-kb pFOXC2 plasmid DNA. The size of DNA markers (lambda DNA digested with PstI) is indicated to the left in kilobases. (C) Partial restriction map of pFOXC2 and pFOXC3. Restriction sites for BglII (Bg), EcoRV (E5), XmaI (X), and EcoRI (E) of pFOXC2 are shown above the “clothespin” DNA, and the sites that are also found in pFOXC3 are indicated by a vertical line below the hairpin structure. The long open reading frame is indicated by diagonal lines, and the location of highly conserved amino acids characteristic of RTs is shown (boxes 1, 2, 2a, and A to E). DNA constructs used to generate clones for the synthesis of in vitro RNAs (dashed horizontal lines) are shown below the hairpin DNA. The terminal EcoRI fragment of pFOXC3 was cloned into pBluescribe and engineered to have two or three copies of the 3′ pentameric repeat (small boxes). DNA templates were digested with NlaIII and used in in vitro transcription reactions with T7 RNA polymerase to synthesize 93-nt (C3-2R) and 98-nt (C3-3R) RNAs having two or three copies of the 5-bp repeat, respectively. Additional RNAs were synthesized from DNAs that were digested with BglII (Bg) and BamHI (B). (D) Determination of optimal MgCl₂ and KCl concentrations for cDNA synthesis with C3-2R RNAs. Reactions were carried out with MN-treated pFOXC3 RT and 0.5 μg of RNA in standard exogenous reaction conditions having variable concentrations of MgCl₂ or KCl (in millimolar concentrations), as indicated. The prominent cDNA product of 173 nt is shown.

FIG. 4.

Exogenous reverse transcription reactions with MN-treated pFOXC RT and MMLV RT. Reactions with mtRNP particles isolated from pFOXC3-containing strains digested with MN (lanes 1 to 5) or MMLV RT (lanes 7 to 10). Lane 1, no exogenous RNA. Lanes 2 to 5 and 7 to 10, reaction mixtures containing 93-nt C3-2R RNA that corresponds to the 3′ terminus of the pFOXC3 plasmid transcript. Reactions were carried out with (lanes 4, 5, 9, and 10) or without (lanes 2, 3, 7, and 8) a 34-nt oligonucleotide that is complementary to a 25-nt region at the 3′ end of the RNA template. Following cDNA synthesis, products were incubated with RNase A (lanes 3, 5, 8, and 10) or left untreated (lanes 1, 2, 4, 7, and 9), prior to electrophoresis in a 6% polyacrylamide gel containing 8 M urea. Numbers on the left indicate the sizes (nucleotides) of a 100-bp marker and Sau3AI fragments of pBS(−) molecular weight markers (M, lane 6). Numbers on the right indicate the sizes (in nucleotides) of the ³²P-labeled DNA products as well as a schematic drawing of the most favorable base-pairing interactions of snapped-back C3-2R RNAs that may correspond to the products observed (dashed line, C3-2R RNA; solid line, cDNA; open box, 2R oligonucleotide). Potential base-pairing interactions of the 2R oligonucleotide self-dimer that is extended by the pFOXC RT are shown at the bottom, with the incorporated nucleotides indicated in lowercase letters.

FIG. 5.

Extension of 3R oligonucleotide in reactions lacking RNA templates. (A) Unlabeled 3R oligonucleotide was used in reverse transcription reactions with MN-pFOXC RT containing 0.33 μM [³²P]dATP and either 20 μM dCTP, dGTP, and TTP (lane 1); no additional nucleotides (lane 2); [³²P]dATP plus 100 μM dideoxy-TTP (lane 3); or [³²P]dATP plus 20 μM dCTP or TTP and 100 μM dideoxy-GTP (lane 4). Lane 5 contains 5′-end-labeled oligonucleotide 3R. Sizes are indicated on the left (in nucleotides) and are based on DNA size standards (not shown). (B) The sequence and length of predicted extension products for reactions with the three most favorable base pairings. Nucleotides added during polymerization are lowercase and are indicated only for the oligonucleotides on the bottom strand. Vertical lines indicate Watson-Crick base pairing, and a colon indicates potential G-T pairing.

FIG. 6.

Comparison of cDNA synthesis with internally and terminally annealed DNA primers with AMV, MMLV, or pFOXC RT. All reactions contain the 98-nt C3-3R RNA (dashed line) having three copies of the terminal 5-bp repeat. (Left panel) Lanes 1, 4, and 7 lack DNA oligonucleotides; lanes 2, 5, and 8 contain the 23-nt INT (I) primer (open box) that is complementary to an internal region of the template; and lanes 3, 6, and 9 contain the 2Rc (2c) primer having 10 nt of complementarity to the 3′ end of the template. A band of the same size as the RNA template was detected in lanes 1 to 6 and may represent partial end-labeling activity in reactions having AMV and MMLV RT. Reaction products were not posttreated with alkali, and the ∼175-nt band in lanes 7 and 8 is RNA-cDNA hybrids formed by extension of snapped-back RNA templates. (Right panel) Primer extension reactions with three oligonucleotides having diminishing amounts of complementarity to the 3′-terminal sequences of the C3-3R RNA template. Primer 1R+ATc (1⁺) has 7 nt of complementarity, and primer 1Rc (1c) has 5 nt of complementarity with the template. Reactions were carried out under standard reaction conditions with 10 mM MgCl₂, and products were subjected to alkaline hydrolysis. The sizes (in nucleotides) of products that derive from primers bound to the 3′ end of the RNAs are shown below the gel. The size (in nucleotides) of the DNA markers is shown on the left of each panel, and they are end-labeled Sau3AI and AluI pBS restriction fragments.