1

Lesion-induced Mutation in the Hyperthemophilic Archaeon Sulfolobus acidocaldarius and

its Avoidance by the Y-family DNA Polymerase Dbh

Cynthia J. Sakofsky*1 and Dennis W. Grogan§

Department of Biological Sciences

University of Cincinnati

1Current address:

Department of Biology

University of Iowa

143 Biology Building

Iowa City, IA 52242-1324

Genetics: Early Online, published on July 29, 2015 as 10.1534/genetics.115.178566

Copyright 2015.

2

Running title: TLS in hyperthermophilic archaea

Keywords:

trans-lesion DNA synthesis (TLS) / abasic site / 7,8-dihydro-8-oxo-deoxyguanosine /

oligonucleotide-mediated transformation

§Corresponding author:

Dennis Grogan

Department of Biological Sciences

614 Rieveschl Hall, Clifton Court

Cincinnati Ohio 45221-0006

513-556-9748

grogandw@ucmail.uc.edu

3

ABSTRACT

Hyperthermophilic archaea offer certain advantages as models of genome replication, and

Sulfolobus Y-family polymerases Dpo4 (S. solfataricus) and Dbh (S. acidocaldarius) have been

studied intensively in vitro as biochemical and structural models of trans-lesion DNA synthesis

(TLS). However, the genetic functions of these enzymes have not been determined in the native

context of living cells. We developed the first quantitative genetic assays of replication past

defined DNA lesions and error-prone motifs in Sulfolobus chromosomes, and used them to

measure the efficiency and accuracy of bypass in normal and dbh-minus strains of Sulfolobus

acidocaldarius. Oligonucleotide-mediated transformation allowed low levels of abasic-site

bypass to be observed in S. acidocaldarius, and demonstrated that the local sequence context

affected bypass specificity; in addition, most erroneous TLS did not require Dbh function.

Applying the technique to another common lesion, 8-oxo-dG, revealed an anti-mutagenic role of

Dbh. The efficiency and accuracy of replication past 8-oxo-dG was higher in the presence of

Dbh, and up to 90% of the Dbh-dependent events inserted dC. A third set of assays, based on

phenotypic reversion, showed no effect of Dbh function on spontaneous -1 frameshifts in

mononucleotide tracts in vivo, despite the extremely frequent slippage at these motifs

documented in vitro. Taken together, the results indicate that a primary genetic role of Dbh is to

avoid mutations at 8-oxo-dG that occur when other Sulfolobus enzymes replicate past this lesion.

The genetic evidence that Dbh is recruited to 8-oxo-dG raises questions regarding the

mechanism of recruitment, since Sulfolobus spp. have eukaryotic-like replisomes but no

ubiquitin.

4

INTRODUCTION

The fact that DNA damage occurs in all living cells poses a threat to the accurate replication and

partitioning of their genomes. Cells counter this threat with an array of diverse damage-coping

systems, some of which repair the DNA, whereas others enable replication to continue past

persistent lesions. Lesion bypass, also termed damage tolerance, can follow various alternatives,

which are broadly distinguished as error-free vs. error-prone (Lehman et al. 2007). The error-

free pathways generally use recombination-associated functions to pair a strand, newly

synthesized on intact template, to the damaged DNA strand; the mechanisms for this include

transient reversal of the replication fork and strand exchange at gaps left behind the fork (Sale

2012). These mechanisms bypass lesions accurately, although they can promote other forms of

genetic instability, such as ectopic recombination between dispersed repeats (Izhar et al. 2013).

Error-prone mechanisms of lesion bypass, in contrast, use specialized, trans-lesion synthesis

(TLS) polymerases to continue strand elongation using the damaged template; most of these

enzymes belong to the Y family of DNA polymerases (Sale et al. 2012). Large catalytic sites,

combined with the absence of proofreading activity, allow these polymerases to insert

nucleotides opposite diverse DNA lesions, but also limit the accuracy with which they replicate

intact template (Yang & Woodgate 2007; Pata 2010). This latter property creates a risk of

mutagenesis, which is mitigated somewhat by other features of these enzymes, including low

processivity and a bias toward insertion of the correct nucleotide opposite certain DNA lesions

(Jarosz et al. 2007). TLS polymerases are nevertheless subject to mechanisms which limit their

access to template and co-ordinate TLS with other lesion-bypass strategies. The potential

5

complexity of this co-ordination, and its importance for achieving genetic stability, is illustrated

in human cells, which have an extensive repertoire of TLS polymerases (Waters et al. 2009).

Y-family DNA polymerases of hyperthermophilic archaea have become valuable models for

elucidating mechanisms of TLS, due in part to their favorable crystallization properties. As the

third, deeply diverging, lineage of cellular organisms distinct from bacteria and eukaryotes,

archaea provide a unique perspective on the diversity and evolution of DNA replication and

damage-coping mechanisms. The scheme of DNA replication in archaea incorporates many

eukaryotic features but is simpler than that of eukaryotes (Ishino et al. 2013). Archaeal

replication origins are recognized by Cdc6/Orc1 homologues, and the replicative helicase is an

MCM homologue that moves 3’ to 5’ on the leading-strand template (Barry & Bell 2006; Li et

al. 2013). Two molecules of a B-family polymerase, one for each arm of the fork, are

complexed to corresponding rings of PCNA, which attach to RFC and four GINS-like proteins to

form a replisome (Li et al. 2013). Archaeal Okazaki fragments are short, and are processed for

ligation by a Fen1 homologue (Barry & Bell 2006; Li et al. 2013; O’Donnell et al. 2013;

Raymann et al. 2014).

The systems by which archaea cope with DNA damage and replication errors are less clear,

however. Although several hyperthermophilic archaea encode the components necessary for

complete systems of base excision repair (BER), homologous recombination, and TLS, all

hyperthermophilic archaea lack certain nucleotide excision repair (NER) and mismatch repair

(MMR) proteins that are broadly conserved in other organisms, including mesophilic archaea

(White & Grogan 2008). For a few hyperthermophilic archaea, including Sulfolobus spp. (which

6

grow optimally at about 80° C and pH 3), the functional properties of genome-maintenance

systems have been investigated by genetic analyses. Homologous recombination in S.

acidocaldarius appears to rely heavily on a short-patch, gene-conversion process which is not

affected by mismatches in the substrates, and thus seems capable of promoting ectopic events

(Grogan & Stengel 2008; Grogan & Rockwood 2010; Rockwood et al. 2013).

Diverse DNA-damaging treatments induce mutation in Sulfolobus cells, providing functional

evidence of error-prone TLS (Reilly & Grogan 2002). Sulfolobus species each have a single Y-

family polymerase; the enzyme from S. solfataricus (Dpo4) and its counterpart from S.

acidocaldarius (Dbh), exhibit the low fidelity typical of these enzymes, making them potential

sources of mutations. The S. acidocaldarius polymerase Dbh, appears to be particularly

mutagenic; when Dbh replicated short homopyrimidine tracts followed by G (i.e., 5’G(Y)n

template) in vitro, up to 50% of the extension products had single-nt deletions within the tract

(Potapova et al. 2002).

Despite the multiple threats to genome stability posed by growth temperatures that accelerate

DNA damage, the presence of an error-prone DNA polymerase, the lack of MutS and MutL

proteins, and promiscuous recombination, Sulfolobus cells achieve remarkably accurate genome

replication. S. acidocaldarius, for example, exhibits an overall error rate per genome replication

below that of most MMR-proficient mesophilic micro-organisms (Grogan et al. 2001), and an

intra-specific sequence divergence far below that of Sulfolobus islandicus (Mao & Grogan

2012b). The mechanistic basis of the observed replication fidelity remains unclear. In addition

to the inability to identify certain repair genes in Sulfolobus genomes, the genetic consequences

7

of unrepaired lesions and the specific roles played by TLS polymerases in determining these

consequences have not been defined experimentally.

Two widely occurring DNA lesions, abasic, or ‘AP’, sites and 7,8-dihydro-8-oxo-

deoxyguanosine (8-oxo-dG), are expected to arise frequently in Sulfolobus chromosomes. The

extremely high temperatures and mildly acidic conditions within actively dividing Sulfolobus

cells (Meyer & Schäfer 1992) create abasic sites at high rates both directly, via spontaneous

depurination (Lindahl & Nyberg 1972), and indirectly via enzymatic removal of deaminated

bases, which also form spontaneously at high temperatures (Lindahl & Nyberg 1974). In vitro,

chemically stabilized abasic sites impede Dbh progression (Gruz et al. 2001), and the residual

bypass that does occur tends to skip over the lesion, generating a -1 bp deletion (Potapova et al.

2002). The catalytic properties of Dbh thus do not seem to implicate it specifically in coping

with unrepaired abasic sites in the S. acidocaldarius genome. Similarly, 8-oxo-dG forms in the

DNA of all aerobes by spontaneous oxidation of guanine, which is predicted to be accelerated

under aerobic conditions by high temperature (Cadet et al. 1997). Unlike an abasic site, 8-oxo-

dG does not significantly alter the helical structure of DNA, and does not block most

polymerases (Freisinger et al. 2002), but its structure encourages insertion of dA, generating G:C

to T:A transversions (Hsu et al. 2004). In vitro, Dbh and Dpo4 preferentially insert dC opposite

8-oxo-dG (Gruz et al. 2001; Rechkoblit et al. 2006; Zang et al. 2006; Maxwell & Suo 2012); this

predicts that these enzymes would suppress transversion in vivo if recruited to 8-oxo-dG during

replication of Sulfolobus chromosomes. Consistent with this prediction, disrupting the dbh gene

(Saci_0554) resulted in increased rates of spontaneous G:C to T:A transversions in a

chromosomal gene (pyrE) where forward mutations can be selected (Sakofsky et al. 2012).

8

The present study developed new genetic assays, based on existing selections for loss or

restoration of pyrE gene function in S. acidocaldarius (Grogan and Stengel 2008), and used them

to address two related questions: i) what genetic consequences result from placing abasic sites

and 8-oxo-dG in the chromosome of an hyperthermophilic archaeon, and ii) what roles do the Y-

family polymerase and other DNA polymerases of the organism play in determining these

outcomes? The results provide evidence for lesion-specific recruitment of a Y-family

polymerase in archaea, mutagenesis by B-family Sulfolobus polymerases, an influence of

sequence context on the accuracy of bypass, and avoidance of 8-oxo-dG-promoted mutation

mediated by Dbh.

MATERIALS AND METHODS

Strains and growth conditions: The strains and oligonucleotides used in this study are

described in Supporting Information (Tables S1 - S3). S. acidocaldarius pyrE mutations were

selected in wild-type and dbh- backgrounds via spontaneous 5-fluoro-orotic acid (FOA)

resistance (Grogan & Gunsalus 1993), which is similar to the selection of Saccharomyces

cerevisiae ura3 mutants (Boeke et al. 1984). Mutants were confirmed phenotypically by

pyrimidine auxotrophy, and the identity of the mutation was confirmed by dye-terminator

sequencing of PCR products of the pyrE gene. Pyr- strains were grown on xylose-tryptone (XT)

medium supplemented with uracil (20 mg/L) (Grogan and Gunsalus 1993). Pyr+ revertants or

transformants were selected by plating on solid XT medium lacking uracil. For genotyping, Pyr+

transformants were clonally purified on selective plates, followed by growth in liquid culture and

extraction of genomic DNA.

9

Genetic assays: Rates of reversion were measured by fluctuation assays of 16-20 independent

cultures (0.2 mL each), grown from inocula of about 104 cells to a final size of about 108 cells.

The average number of viable cells per culture was determined by dilution and plating on non-

selective medium, and the number of Pyr+ revertants was determined individually for each

culture by harvesting the entire culture and plating on selective medium (xylose-tryptone without

uracil). The rate of mutation was calculated using the Ma-Sandri-Sarkar maximum-likelihood

estimator via the FALCOR web interface (Hall et al. 2009).

To analyze TLS, isogenic pairs of dbh+ and dbh- strains were prepared in which each strain

contained a non-synonymous BPS inactivating the pyrE gene next to a synonymous site in the

coding sequence. Three alleles in the S. acidocaldarius pyrE gene satisfied these criteria and

were used in the study (Fig. 1 (e-g)); a fourth site in a heterologous pyrE gene (Sulfolobus

solfataricus) was used to evaluate 8-oxo-dG bypass only (Fig. 1(h, i)). Corresponding sets of

commercially synthesized, single stranded (ss) oligonucleotides (Table S2) were prepared as

10μM or 20μM stock solutions and were electroporated into recipient cells as previously

described (Grogan & Stengel 2008). All four bases at the query sites were confirmed to yield

transformants; a strand bias favoring transformation by anti-sense oligonucleotides,

corresponding to that observed for other pyrE alleles (Grogan & Stengel 2008), was also

confirmed (Table S4).

Except for the assays depicted in Fig. 1 (h, i), lesion-bearing oligonucleotides electroporated into

pyrE recipients carried a centrally located abasic site or 8-oxo-dG residue directly adjacent to the

10

nucleotide used for selection (Table S2). The abasic site was a tetrahydrofuran (THF) spacer

incorporated by synthesis (Integrated DNA Technologies, Coralville, Iowa USA).

Oligonucleotides containing 8-oxo-dG lesions were synthesized and purified by HPLC at the

University of Utah DNA/Peptide Core Facility.

To identify the outcome of bypass events, genomic DNA was extracted from clonally purified

transformants, and the S. acidocaldarius pyrE region was amplified by PCR (Table S4). The

appropriate synonymous site within pyrE was then scored by ligase chain reaction (LCR;

Wiedmann et al. 1994) using the pyrE PCR product as template, as follows. A set of four

upstream oligonucleotides probed for each of the four nucleotides at the site of the lesion, and

four corresponding LCR reactions were completed for each recombinant (Table S3). The

ligation product was detected by analyzing a melting curve from 65C to 95C of the DNA

upon completion of the LCR cycles. All transformants were confirmed to yield product in one,

but only one, of the four nucleotide-specific LCR reactions.

AP Endonuclease assays: S. acidocaldarius cells were suspended in 50 mM MES buffer,

potassium salt (pH 6.3), and disrupted by sonication; the resulting extracts were clarified by

centrifugation (10 min, 13000 xg), and stored frozen until use. Each assay contained 5 μL

extract (27 μg protein), 40 pmol fluorescein-labelled oligonucleotide (5'-FAM), 50 mM MES

buffer (as above), and 2 mM MgSO4. Sequences of the substrate oligonucleotides are given in

Table S3. Duplex substrates were formed by annealing a fluorescently labeled ssDNA with a

20% excess of unlabeled complement.

11

Reactions were started by adding cell extract to the chilled mixture, followed by mixing and

immediate incubation at 70° C. Reactions were stopped by rapid cooling to 25° C and adding an

equal volume of 90% formamide-50 mM EDTA, followed by heating 5 min at 100° C. To detect

strand scission products, the entire assay mixture was electrophoresed in a 15% polyacrylamide

gel containing 7M urea in TE buffer. Electrophoresis standards consisted of unmodified

fluorescent oligonucleotide, combined with the cleavage product generated by treatment of the

corresponding dU-containing oligonucleotide with E. coli Endonuclease III (New England

Biolabs) at 37° C for 30 min, followed by chemical cleavage with piperidine (90° C, 30 min).

Quantitative Analyses: Transformation efficiencies of oligonucleotides carrying abasic or 8-

oxo-dG lesions are reported as medians in Dbh+ and Dbh- backgrounds, and were calculated

based on a minimum of six independent experiments. Statistical comparisons between

transformation efficiencies of damaged DNA into Dbh+ and Dbh- at each nt position were

performed using the Mann-Whitney U test. Chi-square analyses were used to determine

statistical differences between the distributions of nucleotides placed; Yate’s correction was used

where appropriate. Frequencies of nucleotides at the query positions of the S. acidocaldarius

pyrE gene were corrected by subtracting the background reversion frequencies for the

corresponding recipient strain.

RESULTS

Molecular assays of TLS in the Sulfolobus chromosome: Oligonucleotide-mediated

transformation (OMT) methods have been used to analyze the fate of DNA lesions at specific

12

sites in yeast and bacterial chromosomes. In this approach, incorporation of a lesion-containing

DNA is selected by restoration or alteration of a chromosomal gene, and individual bypass

events are documented by sequencing the region (Otsuka et al. 2002; Kroeger et al. 2004; Weiss

2008; Rodriguez et al. 2013). The properties of OMT in S. acidocaldarius (Kurosawa & Grogan

2005; Grogan & Stengel 2008; Mao & Grogan 2012a) suggested the feasibility of corresponding

genetic assays that would enable both accurate and erroneous bypass events to be analyzed in

vivo. In the method we developed for Sulfolobus (Fig. 1), a synthetic ssDNA restores the wild-

type sequence at the site of a chromosomal pyrE mutation which inactivates the encoded UMP-

biosynthetic enzyme (orotate:phosphoribosyl transferase). If the oligonucleotide is incorporated

into the chromosome and replicated, the recipient cell gains the ability to grow in medium

lacking uracil. A synonymous site adjacent to the selected nucleotide in the transforming DNA

contains the DNA lesion to be evaluated (Fig. 1a, b). Any base-pair substitution at this “query

position” yields the same amino acid as in the wild-type, so that pyrE gene function is restored,

regardless of base inserted opposite the lesion at the first replication (Fig. 1c, d).

Eight S. acidocaldarius recipients, representing four alleles in a chromosomal pyrE gene (Fig. 1

(e – i)), provided query sites in different positions and sequence contexts, thus allowing the

generality of bypass patterns to be evaluated. Each query and selection site was also provided in

both a Dbh+ and Dbh- background, allowing the function of Dbh to be evaluated independently at

each site (Table S1). Control oligonucleotides confirmed that any base placed at the query sites

yielded transformants with high efficiency (Table S4). In addition, phenotypic reversion

(whether spontaneous or induced by the electroporation procedure) (Berkner & Lipps 2008) was

13

confirmed to be rare (Table 1), and the distribution of nucleotides provided corrections for

corresponding analyses of transformants (Table 2).

To enforce stringency in analyzing lesion bypass, the OMT assay (Fig. 1) was designed to make

transformation sensitive both to repair activities and polymerase blockage. Even precise removal

of the lesion from the annealed oligonucleotide by an endonuclease was expected to negate

transformation, because a 1-nt, 5’ flap on the 3’ side of the repair gap provides a favorable

substrate for removal by the Sulfolobus Fen1/XPG homolog (Horie et al. 2007; Hutton et al.

2008) but not for the highly selective Sulfolobus ligase (Lai et al. 2002). We evaluated the

feasibility of OMT as a probe of lesion bypass in Sulfolobus experimentally by placing a

chemically stabilized abasic site (THF spacer) in three transforming oligonucleotides. Prior in

vitro studies have shown this lesion to impede progression of the Y-family polymerase Dbh of S.

acidocaldarius and its S. solfataricus counterpart Dpo4 (Boudsocq et al. 2001, Gruz et al. 2001;

Potapova et al. 2002; Kokoska et al. 2003; Boudsocq et al. 2004; Ling et al. 2004). In vivo,

BER enzymes that act on ssDNA can also block OMT before the recombination step. Although

THF spacers are not repaired by AP lyases (Kroeger et al. 2004), they are substrates for AP

endonucleases, which have been detected in archaea (Grasso and Tell 2014). To determine the

relevance of these enzymes to OMT in S. acidocaldarius, we incubated fluorescently labelled

oligonucleotides with cell extracts and monitored backbone cleavage. The results confirmed that

S. acidocaldarius has significant endonuclease activity on oligonucleotides containing a THF

spacer (Fig. S1). Under the assay conditions, the DNA backbone was cleaved more rapidly in

duplex than in single-stranded form, and the cleavage of dsDNA was not strongly influenced by

the identity of the opposing base (Fig. S1). In vivo, therefore, the observed repair activity could

14

be expected to preclude OMT either before or after a transforming oligonucleotide has annealed

to its complement in the recipient chromosome.

Genetic consequences of abasic sites. Consistent with all these properties, a THF spacer

decreased the average transformant yield of the three oligonucleotides by a factor of 200 to 700,

yet the lowest observed frequencies of transformation still remained nearly 10-fold higher than

phenotypic reversion (Table 1), suggesting that most of the selected clones were generated by

some form of lesion bypass. Identifying the nucleotide placed in the query site in transformants

via ligase chain reaction (LCR) assays (see Methods) showed that the distributions of nucleotides

placed opposite the DNA lesion were non-random (P< 0.05 for expected values of equal

distribution of bases), and differed dramatically among the three sites tested (Table 2). The

dominant outcome at nt 189 and 219 (A in the sense strand) maintained the bp found at the query

site in the recipient (Table 2). The pattern at these two sites thus remained formally consistent

with three mechanistically distinct alternatives, singly or in combination: i) single-nt repair of the

lesion that preserved the selected nucleotide, ii) error-free damage tolerance (typically due to

strand-exchange processes), or iii) TLS by one or more non-Dbh polymerases, with a strong

intrinsic bias toward insertion of dA. At the third site (bp 348), neither (i) nor (ii) applied to the

most common outcome, however, making the results more informative (Table 2). At this

position, fewer transformants retained the original bp (reflecting either BER or error-free

tolerance), and the primary effect of Dbh was to decrease two erroneous insertions (dA and dC;

Table 3). The consistent recovery of a non-templated dA at position 348 is consistent with

erroneous TLS but not error-free tolerance (Table 2, Table 3). Furthermore, because this result

15

was seen in Δdbh recipients, it could be attributed to one or more of the B-family enzymes of S.

acidocaldarius.

Genetic impact of 8-oxo-dG: Although replicative DNA polymerases can proceed past 8-oxo-

dG, they are prone to insert dA, which yields G:C to T:A transversions. In contrast, certain Y-

family polymerases, including E. coli DinB and human Pol κ, insert dC opposite 8-oxo-dG

(Waters et al. 2009), and thus avoid this mutagenesis. To investigate whether Dbh plays the

latter role in S. acidocaldarius, we repeated the OMT assay with oligonucleotides that placed 8-

oxo-dG at the three pyrE locations used previously. This lesion decreased genetic transformation

relative to undamaged DNAs, but the decrease (15 to 100-fold) was smaller than that observed

for abasic sites. Transformation by DNAs containing 8-oxo-dG at pyrE positions 219 and 348

was significantly higher for Dbh+ recipients than for Dbh-, indicating that Dbh increased the

overall success of strand incorporation and replication (Table 1). In addition, the pattern of

nucleotide insertion opposite 8-oxo-dG was distinct from that observed with abasic

oligonucleotides, indicating that these transformants did not simply result from TLS past

intermediates formed during BER of 8-oxo-dG. In particular, insertion of dC was the most

frequent outcome under all but one condition in Dbh+ cells (Table 2). Since none of the

recipients have C in the top strand at these sites, this result is consistent neither with repair nor

error-free damage tolerance. The specificity of nucleotide insertion opposite 8-oxo-dG was

strongly influenced by the functional status of Dbh at all target positions within the pyrE gene

(nt189, P <0.0001; nt219, P <0.0001; nt348, P =0.0032). The primary impact of functional Dbh

was to increase the frequency of dC insertion opposite 8-oxo-dG. This was accompanied by a

decrease in insertion of dA (the native nucleotide) at positions 189 and 219 (Table 3).

16

The strong bias toward insertion of dC opposite 8-oxo-dG correlated with Dbh function in three

different contexts and conformed to the behavior predicted for accurate TLS. To test the

generality of this result, we evaluated a fourth set of isogenic (Δdbh vs. dbh+) S. acidocaldarius

recipients, which placed the selected nt (defined by the pyrE mutation) on the 5’ side of the

lesion (Fig 1h), so that polymerases encountered 8-oxo-dG before the selected nucleotide. In

further contrast to the previous assays, one bp separated the selected and damaged sites, and the

transforming oligonucleotide differed from the recipient at this intermediate position. This latter

feature was designed to limit the efficiency of single-nt repair of the lesion via BER and strand-

exchange processes, while providing a marker for any such events that did occur (Fig. 1h).

Normalized efficiencies of transformation were similar to those of the previous TLS assays, and

were significantly elevated in the Dbh+ recipient (Table 1, entries for position 443). Correct

insertion (dC opposite 8-oxo-dG) was about 9-fold more frequent in Dbh+ transformants than in

Dbh- transformants (Table 3).

No detectable contribution of Dbh to spontaneous frameshift mutation: In vitro, Dbh has a

characteristic signature on undamaged template, dominated by single-nt deletions which are

especially frequent opposite 5’G(Y)n motifs (Potapova et al. 2002). Crystallography of putative

intermediates, combined with single-nucleotide addition kinetics, indicate that these deletions

occur when the nascent-strand terminus has extended into the homopyrimidine tract of a G(Y)n

template (Wilson & Pata 2008). The nascent strand re-aligns to exclude one of the template

pyrimidines from the helix, and Dbh extends these intermediates efficiently, particularly the one

in which the extra-helical template pyrimidine is three bp back from the nascent-strand terminus

17

(Wilson & Pata 2008). The catalytic and structural data therefore predict that about half of the

Dbh molecules which gain access to G(Y)n in vivo would relinquish back to the replicative

polymerase a substrate with a bulge several bp back from the 3’ end. Such sub-terminal

mismatches evade correction by polymerase proofreading, which accounts for the pronounced

length-dependence of mononucleotide-repeat instability seen in eukaryotic and bacterial cells,

and its further enhancement by inactivation of post-replicational MMR (Tran et al. 1997; Harfe

& Jinks-Robertson 1999).

S. acidocaldarius appears to lack post-replicational MMR naturally (White & Grogan 2008; Mao

& Grogan 2012), and G(Y)n motifs are abundant in its chromosome. These two properties

combine to argue that the S. acidocaldarius chromosome should be a particularly sensitive

detector of Dbh-mediated mutagenesis. At least two hot-spots of spontaneous frameshifts occur

within the S. acidocaldarius pyrE gene (Grogan et al. 2001; Sakofsky et al. 2012). One hotspot

matches the Dbh deletogenic motif closely (5’GCCCCC on the minus strand), whereas the other

lacks the characteristic G (minus-strand 5’ATTTTTT). Both sites generate approximately equal

proportions of +1 and -1 frameshifts in vivo (Grogan et al. 2001; Sakofsky et al. 2012), whereas

purified Dbh makes -1 frameshifts almost exclusively (Potapova et al. 2002). Thus the

deletogenic activity exhibited by Dbh on undamaged DNA in vitro is consistent with some, but

not all, of the observed properties of spontaneous mutation at these sites.

In order to resolve the contribution of Dbh to this prominent form of spontaneous mutation in S.

acidocaldarius, we assembled a set of S. acidocaldarius pyrE mutants that allowed four

frameshifts on homopyrimidine templates to be detected individually in both Dbh- and Dbh+

18

backgrounds: i) C added to GCCCC, ii) T added to ATTTTT, iii) C deleted from GCCCCCC, or

iv) T deleted from ATTTTTTT (Table S1). Each of these events restored pyrE gene function,

and could therefore be measured by selecting revertants. This was done in side-by-side

fluctuation assays which included multiple strains with the same mutations, to ensure that the

only difference between groups of cultures was the dbh genotype of the strains. Differences in

the pyrE reversion rates for the corresponding Dbh- and Dbh+ strains could thus reveal the

contribution, if any, of the Dbh polymerase to the corresponding frameshift event.

Similar to the pattern observed in other systems (Tran et al. 1997; Harfe & Jinks-Robertson

1999), the frameshift rates generally increased with tract length (Table 4), but in every case the

rate measured in the dbh+ strain could not be distinguished from that measured in the isogenic

Δdbh strain. These assays therefore detected no contribution of Dbh to spontaneous frameshifts

at motifs similar to those which it replicates with extremely high frameshift rates in vitro. This

result provided evidence that Dbh either does not have deletogenic properties in vivo, or

normally does not gain access to this template during replication of the S. acidocaldarius

chromosome.

A second test of Dbh-dependent -1 events at G(Y)n motifs was designed to exploit the apparent

specificity of Dbh for 8-oxo-dG in vivo (Tables 1 and 2). We modified one of the OMT assays

by designing two ssDNAs that contained an extra nucleotide and thus required a -1bp bypass

event near the selected and query sites in order to generate transformants. One DNA contained

undamaged G at the query site, the other contained 8-oxo-dG (Fig. 1(i)), and each was

electroporated into corresponding isogenic Dbh+ and Dbh- recipients. Consistent with the assay

19

design, the undamaged oligonucleotide yielded Pyr+ clones at low efficiencies (Table 1), and all

the transformants examined contained intragenic suppressor mutations in the form of 1-bp

deletions uniformly distributed within about 5 bp of the selected nt. In all cases, these deletions

occurred outside of the short homopyrimidine tract, and thus did not represent the type of

slipped-strand events mediated by Dbh in vitro (Potapova et al. 2002). Replacing the G at the

query site with 8-oxo-dG increased the yield of transformants about 10-fold, yielding a relative

efficiency similar to that of in-frame OMT in both Dbh- and Dbh+ recipients (Table 1). In

addition, 15 of 16 transformants scored showed precise deletion of the query position, not of a

subsequent G:C bp in the homopyrimidine tract. Thus, this assay detected 8-oxo-dG-promoted

1-bp deletions in S. acidocaldarius that were both frequent and localized at a particular site, but

not mediated by Dbh.

DISCUSSION

In Archaea, DNA replication and responses to DNA damage occur in the context of a relatively

simple prokaryotic cell, yet these processes involve proteins and other molecular features

homologous to those of eukaryotic cells, rather than bacteria (Ishino et al. 2013). This situation

makes archaea strategic for understanding the diversity and evolution of mechanisms that

preserve genome integrity, but the archaeal mechanisms remain largely unexplored in their

cellular context. In particular, despite extensive studies of Sulfolobus DNA polymerases in vitro,

the contributions of Sulfolobus Y-family vs. B-family DNA polymerases to mutagenesis by

particular DNA lesions have not been defined in molecular terms. The present study is the first

to investigate this question experimentally in vivo.

20

The genetic consequences of DNA lesions transferred to the S. acidocaldarius chromosome

differed as a function of DNA lesion, location within the gene sequence, and Dbh status (Table

2). Differences between the abasic-site and 8-oxo-dG patterns include a significant level of

accurate bypass with little input from Dbh in the case of abasic sites vs. a strong dependence on

Dbh for accurate bypass in the case of 8-oxo-dG. Both for abasic sites and 8-oxo-dG, the pattern

of outcomes varied with the location of the damage within the target gene. Although the

accuracy of oligonucleotide synthesis may have contributed to the site-specific differences, it

does not account for the similarity seen for abasic and 8-oxo-dG oligonucleotides at different

sites, and the differential impact of Dbh status across the sites (Table 2). The results thus suggest

that S. acidocaldarius, and presumably other hyptherthermophilic archaea, can use multiple

mechanisms for lesion bypass, and that the identity of the lesion and its local sequence context

determines which mechanism predominates.

As depicted in the scheme of Figure 2, the most consistent and biologically significant function

of the Y-family DNA polymerase Dbh evident in this study was the avoidance of base-pair

substitutions during replication past 8-oxo-dG. Dbh seems to be specialized in this respect; it did

not contribute to TLS past abasic sites nor to spontaneous -1 events in frameshift hotspots. Both

mutation avoidance at 8-oxo-dG and poor bypass of abasic sites are consistent with known

enzymological properties of Dbh (Gruz et al. 2001). In contrast, the failure to promote -1 events

at homo-pyrimidine tracts, which we documented in two distinct experimental contexts (normal

growth and OMT), is not consistent with the strong deletogenic propensity of Dbh demonstrated

repeatedly in vitro (Potapova et al. 2002; Wilson and Pata 2008). The formal possibility remains

that deletogenesis on intact template reflects some artifact of in vitro assays, but this seems

21

unlikely, as this property has been linked to specific structural features of the polymerase

(Boudocq et al. 2004; Pata 2010) and conformational dynamics of the template in the enzyme

active site (Wilson & Pata 2008; Manjari et al. 2014). Conversely, deletogenesis seems unlikely

to be masked in vivo by non-specific factors, such as cytoplasmic solutes or elevated

temperature. Accordingly, we interpret the irrelevance of Dbh to spontaneous frameshifts in the

pyrE gene as genetic evidence that Dbh is normally prevented, with some specificity, from

replicating undamaged DNA in vivo. The extent to which Dbh promotes 1-nt deletions at

persistent lesions in vivo remains unresolved, but the fact that Dbh function increases the rate of

transformation by 8-oxo-dG-containing oligonucleotides argues that such lesion-triggered

deletions would be relatively rare. Since a frameshift nullifies detection of the TLS event in

most of the assays used, any such -1 events appear to be significantly outnumbered by in-frame

TLS, for example.

Ironically, Dbh has been characterized biochemically as an extremely error-prone DNA

polymerase, yet most of the erroneous bypass we observed in S. acidocaldarius did not respond

to Dbh function. Sulfolobus spp. generally encode four DNA polymerases: a replicative enzyme

responsible for the bulk of DNA synthesis (B-family), a Y-family polymerase implicated in TLS,

and two additional B-family enzymes which remain relatively obscure. Biochemical studies of

the latter enzymes of S. solfataricus indicate limited affinity for DNA substrate, limited

processivity, and relatively low rates of polymerization (Choi et al. 2011). More recently, the

DNA primase of a distantly related hyperthermophilic archaeon was reported to catalyze TLS in

vitro (Jozwiakowski et al. 2015). In vivo, we observed several distinct types of mutagenesis by

non-Dbh enzymes, including OMT-associated base substitution and single-bp deletion, and

22

spontaneous frameshift mutation. The mechanism of the second mutagenic process has yet to be

defined, but its efficiency and specificity seems to pose an additional mutagenic threat to

Sulfolobus genomes from 8-oxo-dG which, to our knowledge, has not been predicted by other

studies.

At the four sites we evaluated, Dbh mediated 75% to 90% of the observed TLS past 8-oxo-dG,

despite the ability of other enzymes to perform bypass in its absence. This observation

complements the evidence of Dbh exclusion from undamaged template, indicating that some

property favors Dbh over other alternative pathways in replicating 8-oxo-dG. Recruitment of

TLS polymerases to unrepaired lesions is commonly associated with polymerase switching

(Lehman et al. 2007). Although this process occurs both in bacteria and in eukaryotic cells, and

involves corresponding replisome components (i.e., the sliding clamp), the bacterial and

eukaryotic mechanisms differ dramatically. Bacterial polymerases appear to compete for

preferred binding sites on the dimeric β clamp, where binding appears to be largely reversible,

i.e., determined by the relative concentrations of these enzymes (Friedberg et al. 2006; Lovett

2007; Andersson et al. 2010; Kath et al. 2014). In eukaryotes, however, different patterns of

ubiquitination of the trimeric PCNA dictate which damage-tolerance pathways will be activated.

A diversity of ubiquitination patterns allows a range of distinct responses to be specified; the

possible choices encompass several TLS polymerases and multiple error-free processes,

including replication-fork regression or Rad51-dependent strand exchange (Nick McElhinny et

al. 2008; Yang et al. 2013).

23

The eukaryotic features of archaeal replisomes include trimeric PCNAs instead of the dimeric β

clamps found in bacteria. Since archaea encode no ubiquitin, however, it remains unclear

whether archaea switch DNA polymerases in response to covalent modification of their PCNA.

Ubiquitin alternatives, with corresponding ligation systems, do occur in archaea, but they have

been implicated so far only in other processes, such as cofactor synthesis, tRNA modification,

and proteolysis (Makarova & Koonin 2010). Conversely, properties observed in vitro raise the

possibility that biologically relevant switching could occur via passive exchange among archaeal

DNA polymerases, as has been proposed for bacteria. For example, Dpo1, the replicative

polymerase of S. solfataricus, disengages from PCNA frequently (Bauer et al. 2013), and 8-oxo-

dG in the template blocks its progress, but increases the catalytic efficiency of the TLS

polymerase Dpo4 (Maxwell & Suo 2012). In this context, the present study demonstrates that

genetic analysis of TLS in S. acidocaldarius can provide important functional insight into

mutation avoidance, polymerase recruitment and other aspects of DNA damage tolerance

modeled by the enzymes of hyperthermophilic archaea.

Acknowledgements: CJS acknowledges Wieman-Wendel-Benedict awards from the

Department of Biological Sciences, University of Cincinnati. Initial stages of this work were

supported by grant MCB 0543910 from the National Science Foundation.

24

LITERATURE CITED

Barry, E. R., and S. D. Bell, 2006 DNA replication in the archaea. Microbiol. Mol. Biol. Rev.

70: 876-887.

Bauer, R. J., I. D. Wolff, X. Zuo, H. K. Lin and M. A. Trakselis, 2013 Assembly and distributive

action of an archaeal DNA polymerase holoenzyme. J. Mol. Biol. 425: 4820-4836.

Berkner, S., and G. Lipps, 2008 Mutation and reversion frequencies of different Sulfolobus

species and strains. Extremophiles 12: 263-270.

Boeke, J. D., F. LaCroute and G. R. Fink, 1984 A positive selection for mutants lacking

orotidine-5'-phosphate decarboxylase activity in yeast: 5-fluoro-orotic acid resistance. Mol. Gen.

Genet. 197: 345-346.

Boudsocq, F., S. Iwai, F. Hanaoka and R. Woodgate, 2001 Sulfolobus solfataricus P2 DNA

polymerase IV (Dpo4): An archaeal DinB-like DNA polymerase with lesion-bypass properties

akin to eukaryotic Pol eta. Nucleic Acids Res. 29: 4607-4616.

Boudsocq, F., R. J. Kokoska, B. S. Plosky, A. Vaisman, H. Ling et al, 2004 Investigating the role

of the little finger domain of Y-family DNA polymerases in low fidelity synthesis and

translesion replication. J. Biol. Chem. 279: 32932-32940.

Cadet, J., M. Berger, T. Douki and J. L. Ravanat, 1997 Oxidative damage to DNA: Formation,

measurement, and biological significance. Rev. Physiol. Biochem. Pharmacol. 131: 1-87.

Choi, J. Y., R. L. Eoff, M. G. Pence, J. Wang, M. V. Martin et al, 2011 Roles of the four DNA

polymerases of the crenarchaeon Sulfolobus solfataricus and accessory proteins in DNA

replication. J. Biol. Chem. 286: 31180-31193.

25

Freisinger, E., A. P. Grollman, H. Miller and C. Kisker, 2004 Lesion (in)tolerance reveals

insights into DNA replication fidelity. EMBO J. 23: 1494-1505.

Grasso, S., and G. Tell, 2014 Base excision repair in archaea: Back to the future in DNA repair.

DNA Repair (Amst) 21: 148-157.

Grogan, D. W., and J. Rockwood, 2010 Discontinuity and limited linkage in the homologous

recombination system of a hyperthermophilic archaeon. J. Bacteriol. 192: 4660-4668.

Grogan, D. W., and K. R. Stengel, 2008 Recombination of synthetic oligonucleotides with

prokaryotic chromosomes: Substrate requirements of the Escherichia coli/lambdaRed and

Sulfolobus acidocaldarius recombination systems. Mol. Microbiol. 69: 1255-1265.

Grogan, D. W., and R. P. Gunsalus, 1993 Sulfolobus acidocaldarius synthesizes UMP via a

standard de novo pathway: Results of biochemical-genetic study. J. Bacteriol. 175: 1500-1507.

Grogan, D. W., G. T. Carver and J. W. Drake, 2001 Genetic fidelity under harsh conditions:

Analysis of spontaneous mutation in the thermoacidophilic archaeon Sulfolobus acidocaldarius.

Proc. Natl. Acad. Sci. U. S. A. 98: 7928-7933.

Gruz, P., M. Shimizu, F. M. Pisani, M. De Felice, Y. Kanke et al, 2003 Processing of DNA

lesions by archaeal DNA polymerases from Sulfolobus solfataricus. Nucleic Acids Res. 31:

4024-4030.

Gruz, P., F. M. Pisani, M. Shimizu, M. Yamada, I. Hayashi et al, 2001 Synthetic activity of

Sulfolobus solfataricus DNA polymerase Y1, an archaeal DinB-like DNA polymerase, is

stimulated by processivity factors proliferating cell nuclear antigen and replication factor C. J.

Biol. Chem. 276: 47394-47401.

26

Hall, B. M., C. X. Ma, P. Liang and K. K. Singh, 2009 Fluctuation analysis CalculatOR: A web

tool for the determination of mutation rate using Luria-Delbruck fluctuation analysis.

Bioinformatics 25: 1564-1565.

Harfe, B. D., and S. Jinks-Robertson, 1999 Removal of frameshift intermediates by mismatch

repair proteins in Saccharomyces cerevisiae. Mol. Cell. Biol. 19: 4766-4773.

Horie, M., K. Fukui, M. Xie, Y. Kageyama, K. Hamada et al, 2007 The N-terminal region is

important for the nuclease activity and thermostability of the flap endonuclease-1 from

Sulfolobus tokodaii. Biosci. Biotechnol. Biochem. 71: 855-865.

Hsu, G. W., M. Ober, T. Carell and L. S. Beese, 2004 Error-prone replication of oxidatively

damaged DNA by a high-fidelity DNA polymerase. Nature 431: 217-221.

Huber, H., and D. Prangishvili, 2006 Chapter 3. Sulfolobales, pp. 28-51 in The Prokaryotes: A

Handbook on the Biology of Bacteria, edited by M. Dworkin, S. Falkow, E. Rosenberg, K. -.

Schleifer and E. Stackebrandt. 2006 Springer Science+Business Media, LLC, New York, N.Y.

Hutton, R. D., J. A. Roberts, J. C. Penedo and M. F. White, 2008 PCNA stimulates catalysis by

structure-specific nucleases using two distinct mechanisms: Substrate targeting and catalytic

step. Nucleic Acids Res. 36: 6720-6727.

Ishino, S., L. M. Kelman, Z. Kelman and Y. Ishino, 2013 The archaeal DNA replication

machinery: Past, present and future. Genes Genet. Syst. 88: 315-319.

Izhar, L., O. Ziv, I. S. Cohen, N. E. Geacintov and Z. Livneh, 2013 Genomic assay reveals

tolerance of DNA damage by both translesion DNA synthesis and homology-dependent repair in

mammalian cells. Proc. Natl. Acad. Sci. U. S. A. 110: E1462-9.

Jarosz, D. F., V. G. Godoy and G. C. Walker, 2007 Proficient and accurate bypass of persistent

DNA lesions by DinB DNA polymerases. Cell. Cycle 6: 817-822.

27

Jozwiakowski, S. K., F. Borazjani Gholami and A. J. Doherty, 2015 Archaeal replicative

primases can perform translesion DNA synthesis. Proc. Natl. Acad. Sci. U. S. A. 112: E633-8.

Kath, J. E., S. Jergic, J. M. Heltzel, D. T. Jacob, N. E. Dixon et al, 2014 Polymerase exchange on

single DNA molecules reveals processivity clamp control of translesion synthesis. Proc. Natl.

Acad. Sci. U. S. A. 111: 7647-7652.

Kokoska, R. J., S. D. McCulloch and T. A. Kunkel, 2003 The efficiency and specificity of

apurinic/apyrimidinic site bypass by human DNA polymerase eta and Sulfolobus solfataricus

Dpo4. J. Biol. Chem. 278: 50537-50545.

Kroeger, K. M., M. F. Goodman and M. M. Greenberg, 2004 A comprehensive comparison of

DNA replication past 2-deoxyribose and its tetrahydrofuran analog in Escherichia coli. Nucleic

Acids Res. 32: 5480-5485.

Kurosawa, N., and D. W. Grogan, 2005 Homologous recombination of exogenous DNA with the

Sulfolobus acidocaldarius genome: Properties and uses. FEMS Microbiol. Lett. 253: 141-149.

Lai, X., H. Shao, F. Hao and L. Huang, 2002 Biochemical characterization of an ATP-dependent

DNA ligase from the hyperthermophilic crenarchaeon Sulfolobus shibatae. Extremophiles 6:

469-477.

Lasken, R. S., D. M. Schuster and A. Rashtchian, 1996 Archaebacterial DNA polymerases

tightly bind uracil-containing DNA. J. Biol. Chem. 271: 17692-17696.

Lehmann, A. R., A. Niimi, T. Ogi, S. Brown, S. Sabbioneda et al, 2007 Translesion synthesis: Y-

family polymerases and the polymerase switch. DNA Repair (Amst) 6: 891-899.

Li, X. T., N. Costantino, L. Y. Lu, D. P. Liu, R. M. Watt et al, 2003 Identification of factors

influencing strand bias in oligonucleotide-mediated recombination in Escherichia coli. Nucleic

Acids Res. 31: 6674-6687.

28

Li, Z., L. M. Kelman and Z. Kelman, 2013 Thermococcus kodakarensis DNA replication.

Biochem. Soc. Trans. 41: 332-338.

Lim, S., I. Song, F. P. Guengerich and J. Y. Choi, 2012 Effects of N(2)-alkylguanine, O(6)-

alkylguanine, and abasic lesions on DNA binding and bypass synthesis by the euryarchaeal B-

family DNA polymerase vent (exo(-)). Chem. Res. Toxicol. 25: 1699-1707.

Lin, H. K., S. F. Chase, T. M. Laue, L. Jen-Jacobson and M. A. Trakselis, 2012 Differential

temperature-dependent multimeric assemblies of replication and repair polymerases on DNA

increase processivity. Biochemistry 51: 7367-7382.

Lindahl, T., and B. Nyberg, 1974 Heat-induced deamination of cytosine residues in

deoxyribonucleic acid. Biochemistry 13: 3405-3410.

Lindahl, T., and B. Nyberg, 1972 Rate of depurination of native deoxyribonucleic acid.

Biochemistry 11: 3610-3618.

Ling, H., F. Boudsocq, R. Woodgate and W. Yang, 2004 Snapshots of replication through an

abasic lesion; structural basis for base substitutions and frameshifts. Mol. Cell 13: 751-762.

Lovett, S. T., 2007 Polymerase switching in DNA replication. Mol. Cell 27: 523-526.

Makarova, K. S., and E. V. Koonin, 2010 Archaeal ubiquitin-like proteins: Functional versatility

and putative ancestral involvement in tRNA modification revealed by comparative genomic

analysis. Archaea 2010: 10.1155/2010/710303.

Manjari, S. R., J. D. Pata and N. K. Banavali, 2014 Cytosine unstacking and strand slippage at an

insertion-deletion mutation sequence in an overhang-containing DNA duplex. Biochemistry 53:

3807-3816.

Mao, D., and D. Grogan, 2012 Genomic evidence of rapid, global-scale gene flow in a

Sulfolobus species. ISME J. 6: 1613-1616.

29

Mao, D., and D. W. Grogan, 2012 Heteroduplex formation, mismatch resolution, and genetic

sectoring during homologous recombination in the hyperthermophilic archaeon Sulfolobus

acidocaldarius. Front. Microbiol. 3: 192.

Maxwell, B. A., and Z. Suo, 2012 Kinetic basis for the differing response to an oxidative lesion

by a replicative and a lesion bypass DNA polymerase from Solfolobus solfataricus.

Biochemistry.

McDonald, J. P., A. Hall, D. Gasparutto, J. Cadet, J. Ballantyne et al, 2006 Novel thermostable

Y-family polymerases: Applications for the PCR amplification of damaged or ancient DNAs.

Nucleic Acids Res. 34: 1102-1111.

Meyer, W., and G. Schafer, 1992 Characterization and purification of a membrane-bound

archaebacterial pyrophosphatase from Sulfolobus acidocaldarius. Eur. J. Biochem. 207: 741-

746.

Nick McElhinny, S. A., D. A. Gordenin, C. M. Stith, P. M. Burgers and T. A. Kunkel, 2008

Division of labor at the eukaryotic replication fork. Mol. Cell 30: 137-144.

O'Donnell, M., L. Langston and B. Stillman, 2013 Principles and concepts of DNA replication in

bacteria, archaea, and eukarya. Cold Spring Harb Perspect. Biol. 5:

10.1101/cshperspect.a010108.

Otsuka, C., N. Kunitomi, S. Iwai, D. Loakes and K. Negishi, 2005 Roles of the polymerase and

BRCT domains of Rev1 protein in translesion DNA synthesis in yeast in vivo. Mutat. Res. 578:

79-87.

Otsuka, C., K. Kobayashi, N. Kawaguchi, N. Kunitomi, K. Moriyama et al, 2002 Use of yeast

transformation by oligonucleotides to study DNA lesion bypass in vivo. Mutat. Res. 502: 53-60.

30

Pages, V., G. Mazon, K. Naiman, G. Philippin and R. P. Fuchs, 2012 Monitoring bypass of

single replication-blocking lesions by damage avoidance in the Escherichia coli chromosome.

Nucleic Acids Res. 40: 9036-9043.

Palud, A., G. Villani, S. L'Haridon, J. Querellou, J. P. Raffin et al, 2008 Intrinsic properties of

the two replicative DNA polymerases of Pyrococcus abyssi in replicating abasic sites: Possible

role in DNA damage tolerance? Mol. Microbiol. 70: 746-761.

Pata, J. D., 2010 Structural diversity of the Y-family DNA polymerases. Biochim. Biophys. Acta

1804: 1124-1135.

Potapova, O., N. D. Grindley and C. M. Joyce, 2002 The mutational specificity of the dbh lesion

bypass polymerase and its implications. J. Biol. Chem. 277: 28157-28166.

Raymann, K., P. Forterre, C. Brochier-Armanet and S. Gribaldo, 2014 Global phylogenomic

analysis disentangles the complex evolutionary history of DNA replication in archaea. Genome

Biol. Evol. 6: 192-212.

Rechkoblit, O., L. Malinina, Y. Cheng, V. Kuryavyi, S. Broyde et al, 2006 Stepwise

translocation of Dpo4 polymerase during error-free bypass of an oxoG lesion. PLoS Biol. 4: e11.

Reilly, M. S., and D. W. Grogan, 2002 Biological effects of DNA damage in the

hyperthermophilic archaeon Sulfolobus acidocaldarius. FEMS Microbiol. Lett. 208: 29-34.

Rockwood, J., D. Mao and D. W. Grogan, 2013 Homologous recombination in the archaeon

Sulfolobus acidocaldarius: Effects of DNA substrates and mechanistic implications.

Microbiology 159: 1888-1899.

Rodriguez, G. P., J. B. Song and G. F. Crouse, 2013 In vivo bypass of 8-oxodG. PLoS Genet. 9:

e1003682.

31

Sakofsky, C. J., P. L. Foster and D. W. Grogan, 2012 Roles of the Y-family DNA polymerase

Dbh in accurate replication of the Sulfolobus genome at high temperature. DNA Repair (Amst) .

Sakofsky, C. J., L. A. Runck and D. W. Grogan, 2011 Sulfolobus mutants, generated via PCR

products, which lack putative enzymes of UV photoproduct repair. Archaea 2011: 864015.

Sale, J. E., 2012 Competition, collaboration and coordination--determining how cells bypass

DNA damage. J. Cell. Sci. 125: 1633-1643.

Sale, J. E., A. R. Lehmann and R. Woodgate, 2012 Y-family DNA polymerases and their role in

tolerance of cellular DNA damage. Nat. Rev. Mol. Cell Biol. 13: 141-152.

Tran, H. T., J. D. Keen, M. Kricker, M. A. Resnick and D. A. Gordenin, 1997 Hypermutability

of homonucleotide runs in mismatch repair and DNA polymerase proofreading yeast mutants.

Mol. Cell. Biol. 17: 2859-2865.

Vidal, A. E., and R. Woodgate, 2009 Insights into the cellular role of enigmatic DNA

polymerase iota. DNA Repair (Amst) 8: 420-423.

Waters, L. S., B. K. Minesinger, M. E. Wiltrout, S. D'Souza, R. V. Woodruff et al, 2009

Eukaryotic translesion polymerases and their roles and regulation in DNA damage tolerance.

Microbiol. Mol. Biol. Rev. 73: 134-154.

Weiss, B., 2008 Removal of deoxyinosine from the Escherichia coli chromosome as studied by

oligonucleotide transformation. DNA Repair 7: 205-212.

White, M. F., and D. W. Grogan, 2008 DNA stability and repair, pp. 179-188 in Thermophiles:

Biology and Technology at High Temperatures, edited by F. T. Robb, G. Antranikian, D. W.

Grogan and A. J. Driessen. CRC Press, Boca Raton, Florida.

Wiedmann, M., W. J. Wilson, J. Czajka, J. Luo, F. Barany et al, 1994 Ligase chain reaction

(LCR)--overview and applications. PCR Methods Appl. 3: S51-64.

32

Wilson, R. C., and J. D. Pata, 2008 Structural insights into the generation of single-base deletions

by the Y family DNA polymerase Dbh. Mol. Cell 29: 767-779.

Yang, K., C. P. Weinacht and Z. Zhuang, 2013 Regulatory role of ubiquitin in eukaryotic DNA

translesion synthesis. Biochemistry 52: 3217-3228.

Yang, W., and R. Woodgate, 2007 What a difference a decade makes: Insights into translesion

DNA synthesis. Proc. Natl. Acad. Sci. U. S. A. 104: 15591-15598.

Zang, H., A. Irimia, J. Y. Choi, K. C. Angel, L. V. Loukachevitch et al, 2006 Efficient and high

fidelity incorporation of dCTP opposite 7,8-dihydro-8-oxodeoxyguanosine by Sulfolobus

solfataricus DNA polymerase Dpo4. J. Biol. Chem. 281: 2358-2372.

33

Figure 1. Transformation assays of TLS in the S. acidocaldarius chromosome. Diagrams (a)

through (d) depict the transfer of the lesion to the Sulfolobus chromosome by OMT, and its

subsequent bypass. The transforming ssDNA (a) represents the anti-sense (minus) strand of the

S. acidocaldarius pyrE gene. Near its midpoint, one nucleotide (filled symbol) corrects a point

mutation in the recipient cell, while a DNA lesion replaces the adjacent nucleotide (x). This

ssDNA is electroporated into recipient cells (b), where it anneals to a transient gap (Li et al.

34

2003). The resulting heteroduplex (b) is then replicated (c), which requires some form of bypass

of the DNA lesion. Replication produces two completed chromosomes (i.e., sister chromatids)

(d), but only daughter cells that retain the marker provided by the transforming ssDNA generate

a colony.

The remaining panels show the central section of each transforming ssDNA (bottom strand)

opposite the recipient chromosome (top strand); numbers indicate positions within the pyrE

coding sequence. The mutant codon of the recipient is underlined, with the corresponding amino

acid shown to the right in bracketed italic type; the corresponding WT codon provided by the

transforming DNA is given for the bottom strand. The selected nucleotide and query site (x) are

shown in boldface type. Panels (e) – (g) show sites in the pyrE gene of S. acidocaldarius; (h)

and (i) represent the pyrE gene of S. solfataricus inserted into the S. acidocaldarius chromosome

(see Table S1).

35

Figure 2. Dbh and bypass of 8-oxo-dG

The diagram depicts the most common responses to 8-oxo-dG ("oG" above) represented by

bypass events in normal and Dbh-deficient mutants of S. acidocaldarius. Individual processes

contributing to the bypass seen in dbh mutants were not resolved in this study, and may include

repair, error-free (recombinational) damage tolerance, or polymerase switching among B-family

enzymes.

36

Table 1. Effects of DNA lesions and Dbh function on overall efficiency of OMT (oligonucleotide-mediated transformation)

  Position  Transformation Efficiency  [median (range)]1  Rel. Efficiency2  Effect of Dbh3 

DNA lesion  in pyrE  Δdbh  dbh+  Δdbh   dbh+ Ratio P value 

none  189  1720  (196  ‐ 2436) 1012 (227  ‐ 1722)  100  100     

none  219  2280  (2120  ‐ 2808) 1927 (1600  ‐ 2790)   100  100     

none  348  547  (451  ‐ 817)  692 (510  ‐ 1918)  100  100     

none  4434  750  (317  ‐ 1390) 484.5 (191  ‐ 1200)  100  100     

abasic site  189  9  (6  ‐ 12)  5  (7  ‐ 17)  0.52  0.49 0.94   

abasic site  219  3  (1  ‐ 5)  2  (1  ‐ 5)  0.13  0.10 0.77   

abasic site  348  3  (1  ‐ 12)  1  (1  ‐ 5)  0.55  0.14 0.25   

C duplication5  4434  20  (7  ‐ 41) 9 (6  ‐ 34) 2.67  1.86 0.70 0.71

8‐oxo‐dG  189  38.5  (6  ‐ 58) 45 (11  ‐ 64) 2.24  4.45 1.99 0.093

8‐oxo‐dG  219  23.5  (18  ‐ 49) 133.5 (95  ‐ 171) 1.03  6.93 6.73 0.0022*

8‐oxo‐dG  348  14.5  (9  ‐ 21) 29.5 (17  ‐ 63)    2.65  4.26 1.61 0.045*

8‐oxo‐dG  4434  6  (2  ‐ 46) 18 (2  ‐ 53) 0.80  3.72 4.64 0.0008*

8‐oxo‐dG +C5  4434  159  (62  ‐ 376)  144 (41  ‐ 389)  21.2    29.7 1.40  0.25 

37

1 Transformants per electroporation, based on at least 6 independent experiments. Because the number of transformants did not

correlate with amount of DNA per electroporation over the ranges tested (20 to 100 pmol for normal ssDNAs and 50 to 200 pmol for

lesion-containing ssDNAs), results were pooled for each combination of DNA and recipient strain.

2 Calculated as 100 times the median transformation efficiency divided by the corresponding median for undamaged DNA in the same

strain.

3 P values for variation with Dbh status were from Mann-Whitney U tests comparing transformant yields in Dbh- and Dbh+ recipients,

as corrected for differences in transformation efficiencies (by normalizing to the same cells transformed with undamaged DNA).

Asterisks (*) denote statistical difference at the P< 0.05 level.

4 This query position was in the pyrE sequence of S. solfataricus, used as a selectable cassette to construct S. acidocaldarius strains

DG250 and DG251 (see Table S1).

5 The transforming DNAs contained an extra C between the query site and the selected nucleotide (see Results), thus requiring a -1

replication error for successful transformation.

38

Table 2. Nucleotide specificity of lesion bypassa

               Δdbh    dbh+ 

  Pos.b  (n)c  A  T  C G (n)c A T C G 

Abasic  189  54  98%  2%  0% 0% 60 100% 0% 0% 0% 

  219  60  98%  2%  0% 0% 54 96% 2% 2% 0% 

  348  49  32%d  38%  5% 25% 49 22% 30% 2% 46% 

           

oxo‐dG  189  66  48%  0%  52% 0% 61 3% 0% 97% 0% 

  219  69  36%  0%  63% 1% 63 2% 0% 98% 0% 

  348  59  48%d  26%  24% 2% 62 29% 13% 56% 2% 

  443e  15  0%e  27%  27% 27% 13 0%e 0% 69% 15% 

a The nucleotide found and the indicated position was identified by LCR (see Methods), except as noted below. Values for each base are %

occurrence among the transformants. Boldface numerals indicate the native nucleotide at that position, i.e., the one which could be templated.

b Nucleotide in the pyrE coding sequence

39

c Number of transformants analyzedd Values were corrected for the predicted contribution from spontaneous revertants in the indicated strain, as

determined from DNA-less electroporation controls. This correction changed most entries about 1% or less.

e The site and transforming oligonucleotides correspond to the pyrE gene of S. solfataricus (see Fig. 1h). For these transformations, insertion of

dA was precluded by the selection, whereas additional outcomes not listed in the table were possible. As a result, the transformants were scored

by dideoxy sequencing of PCR products. The most frequent of the additional events not depicted in the table accounted for about 13% of Δdbh

and 8% of dbh+ transformants; this was a complex event yielding CTTCACAAA in the coding strand, where the query site is underlined (see Fig.

1h).

40

Table 3. Normalized frequencies of bypass eventsa

dA insertion dT insertion dC insertion dG insertion .

pyrE

position Δdbh dbh+ ratiob Δdbh dbh+ ratiob Δdbh dbh+ ratiob Δdbh dbh+ ratiob

Abasic  189  0.533  0.592  1.11  0.01  und.c   ‐ ‐  und.c  und.c   ‐ ‐  und.c  und.c   ‐ ‐ 

site  219  0.180  0.070  0.39  0.006  0.001  0.3  und.c  0.001   ‐ ‐  und.c  und.c   ‐ ‐ 

  348  0.20  0.06  0.28  0.074  0.057  0.8  0.032  0.005  0.15  0.15  0.12  0.79 

                           

8‐oxo‐dG  189  1.0  0.1  0.12  und.c  und.c   ‐ ‐  1.11  3.91  3.52  und.c  und.c   ‐ ‐ 

  219  0.29  0.07  0.22  und.c  und.c   ‐ ‐  0.90  6.47  7.19  und.c  und.c   ‐ ‐ 

  348  1.4  1.3  0.97  0.18  0.34  1.8  0.71  2.5  3.5  0.050  0.08  2 

  443d  ‐ ‐  ‐ ‐  ‐ ‐  0.351  und.c   ‐ ‐  0.35  3.05  8.68  0.35  0.0068  1.9 

a Unless noted, numerical values are frequencies of the corresponding outcome, expressed as a percentage of the overall

transformation efficiency of the corresponding undamaged oligonucleotide.

41

b Transformant yield in dbh+/yield in Δdbh, calculated to indicate the effect of Dbh function.

c Events that were allowed by the selection but not detected (see Table 2).

d Position in the pyrE gene of S. solfataricus (see Fig. 1(h)).

42

Table 4. Rates of mononucleotide-repeat expansion and contraction at pyrE frameshift hotspotsa

  (G:C)4 expansionb  (A:T)5 expansionb  (G:C)6 contractionb  (A:T)7 contractionb 

  Rate  (95% ci) c  Rate (95% ci) c  Rate (95% ci) c  Rate (95% ci) c 

Dbh‐  2.6  (3.7   ‐   1.7)  1.5 (2.2   ‐  0.95) 33 (39   ‐  27)  43 (50   ‐  36)

Dbh+  2.3  (3.2   ‐   1.5)  1.8 (2.6   ‐  1.1) 32 (38   ‐  27  38 (45   ‐  32)

a Values represent the number of the indicated mutational (i.e., pyrE reversion) events per 108 cell divisions. Each rate was measured

using four sets of 15-20 independent cultures each (see Methods). Some of the pyrE mutations were represented by multiple strains

(CS2.3 and CS2.5, for example; see Methods), as a means to control for any genetic differences that might have arisen during initial

isolation. No differences in reversion rates between corresponding mutants were observed, however, so the data were pooled to yield

the analyses shown above.

b The base in the top (sense) strand is indicated first in each basepair.

c “ci” indicates confidence interval.