ABSTRACT
Atmospheric and room-temperature plasma (ARTP) is a potent tool for microbial strain engineering; however, the mechanistic basis of its genome-wide mutagenic effects remains poorly understood. Utilizing Saccharomyces cerevisiae as a model, we performed a comprehensive characterization of the mutational landscape and DNA repair pathways elicited by ARTP. Under high-lethality conditions (~99% mortality), ARTP treatment induced a substantial mutational burden, with SNVs and InDels rates increasing 25.2- and 36.7-fold relative to untreated controls, respectively. These mutations exhibited a pronounced bias toward C-T/G-A and A-T/T-A substitutions and InDels within AT-rich regions. Remarkably, the frequency of loss of heterozygosity (LOH) surged 57.4-fold, suggesting a potent capacity for inducing DNA breaks. Through sectored colony assays, we demonstrate that ARTP primarily inflicts pre-replicative double-strand breaks (DSBs), with reciprocal crossovers serving as the predominant driver of terminal LOH. The distribution and pattern of LOH were also altered in ARTP-treated cells. Mechanistically, our study establishes that the translesion synthesis polymerases Rev1 and Pol ζ are indispensable for converting ARTP-induced lesions—likely pyrimidine dimers—into fixed point mutations. In contrast, Rad30 facilitates high-fidelity bypass across these dimers. The base excision repair pathway appears largely dispensable for the processing of these lesions. In contrast, the nucleotide excision repair pathway plays a predominant role in mitigating ARTP-induced mutations. Finally, the strategic application of iterative ARTP mutagenesis facilitated the rapid evolution of acetate-tolerant strains. Our findings provide fundamental insights into the molecular signatures of ARTP mutagenesis, offering a framework for its optimized application in directed evolution.
IMPORTANCE
While ARTP is a cornerstone of non-thermal mutagenesis in industrial biotechnology, its genetic impact has long been treated as a “black box.” This study provides a high-resolution molecular dissection of ARTP-induced genomic alterations using a diploid yeast model. We demonstrate that ARTP is not merely a source of point mutations but a potent catalyst for pre-replicative double-strand breaks and extensive mitotic recombination. By elucidating the pivotal roles of specific DNA repair pathways in mutation fixation, we transition ARTP from a purely empirical screening tool to a mechanistically grounded platform for directed evolution. Furthermore, we demonstrate the practical utility of this platform by evolving acetate-tolerant strains and identifying the underlying genetic drivers of their resilience.
KEYWORDS: ARTP, stress tolerance, DNA recombination, genomic alteration, mutation, yeast
INTRODUCTION
Atmospheric and room-temperature plasma (ARTP) is a powerful non-thermal mutagenesis platform based on radio-frequency glow discharge (1). By generating a helium-based plasma jet enriched in reactive oxygen and nitrogen species (ROS/RNS), charged particles, and UV radiation, ARTP inflicts diverse DNA damage that triggers cellular repair, thereby introducing genome-wide genetic diversity (1, 2). Owing to its operational simplicity, safety, and high mutagenic efficiency, ARTP has been widely used to improve traits in industrial microorganisms (1). Its utility spans from enhancing protein production in Corynebacterium glutamicum and L-cysteine tolerance in Escherichia coli (3, 4), to engineering yeasts for robust industrial fermentation—such as breeding Saccharomyces cerevisiae with improved low-pH tolerance and superior aroma production (5), and obtaining a genetically stable Candida tropicalis strain with markedly increased xylitol production (6). It has additionally been extended to germplasm improvement in higher aquatic organisms and environmental bioremediation (7, 8). However, despite these phenotypic successes, the genetic mechanistic framework for ARTP-mediated mutagenesis remains elusive. Elucidating the ARTP-induced mutational spectrum and its underlying genetic mechanisms will offer valuable insights for optimizing the use of ARTP technology.
In this study, we used S. cerevisiae as a model to systematically elucidate the mutational landscape and molecular mechanisms of ARTP-induced genomic instability. Leveraging a diploid yeast system characterized by high-density single nucleotide polymorphism (SNP) markers, we mapped diverse genetic alterations at high resolution. By applying ARTP to G1-synchronized cells, we determined the temporal origin of recombinational DNA lesions. Furthermore, we investigated the specific contributions of the translesion synthesis (TLS) polymerases (Rev1, Rev3, and Rad30), the base excision repair (BER) pathway, and nucleotide excision repair (NER) pathway to ARTP-induced mutation patterns.
In addition to elucidating the mutational mechanisms of ARTP, we also utilized ARTP to generate acetate-tolerant yeast strains and identified candidate mutations associated with the observed resistance phenotypes. In industrial settings, particularly when utilizing lignocellulosic hydrolysates (9) as the fermentation substrate, acetic acid acts as a ubiquitous and potent inhibitor (10, 11). The excessive intracellular accumulation of acetic acid triggers cytoplasmic acidification (12) and a surge in reactive oxygen species (ROS) (13), severely impairing the growth, metabolism, and fermentation efficiency of microorganisms such as S. cerevisiae (14, 15). Consequently, improving yeast acetic acid tolerance is a core bottleneck for cost-effective bioproduction. To date, researchers have employed various strategies to address this challenge, including rational genetic engineering (16–18) and nonrational adaptive laboratory evolution and whole-genome shuffling (19). Here, we demonstrate that rational and non-rational strategies are highly complementary rather than mutually exclusive. Specifically, key genomic variants identified through high-throughput sequencing of ARTP-generated mutants offer direct, actionable targets for subsequent rational engineering.
RESULTS
ARTP induces dose-dependent mutation rates and a distinct mutational spectrum
To evaluate the mutagenic impact of ARTP, we conducted a mutation accumulation (MA) experiment followed by whole-genome sequencing. We utilized the diploid S. cerevisiae strain WYspo11—generated by crossing the W303-1A and YJM789 haploid strains—which harbors approximately 50,000 SNP markers between homologs, enabling high-resolution mapping of loss-of-heterozygosity (LOH) events (20, 21). Multiple WYspo11 isolates were subjected to iterative ARTP exposure at three distinct treatment times: 30 s (3 rounds, n = 5), 70 s (3 rounds, n = 10), and 110 s (8 rounds, n = 12). Cell lethality ranged from 26% to 99% and increased in a dose-dependent manner with exposure time (Fig. 1A). During each round, log-phase cells were exposed to ARTP under standardized helium plasma conditions (10 SLM, 120 W, and 2 mm distance) and plated for recovery, with colonies randomly selected for the subsequent treatment cycle.
Fig 1.

ARTP induces dose-dependent single-nucleotide variations (SNVs). (A) Cell lethality after ARTP treatment at the indicated exposure times. (B) SNV rates in WYspo11 isolates treated at the indicated ARTP exposure times. (C) Mutational spectra of SNVs in untreated and 110-s ARTP-treated isolates. (D) Trinucleotide sequence contexts of ARTP-induced SNVs. SNVs are positioned in the middle of each three-nucleotide sequence. Data in B are shown as means ± SDs. P values were determined using the Wilcoxon rank-sum test for panel B and the χ2 test for panels C and D. *P < 0.05; ***P < 0.001.
Whole-genome sequencing of the 12 isolates treated with ARTP for 110 s identified 290 single-nucleotide variations (SNVs) (Data Set S1). Given that a single cell undergoes approximately 25 divisions to form a visible colony per round (20), and considering the 23 Mb diploid genome size, the mutation rate was calculated as 5.3 × 10−9 SNVs/bp/division (Fig. 1B). Using the same parameters, the mutation rates for the 30 s and 70 s conditions were determined to be 8.1 × 10−10 and 2.5 × 10−9 SNVs/bp/division, respectively (Fig. 1B). Compared to the spontaneous rate of 2.1 × 10−10 SNVs/bp/division previously established for the WYspo11 strain (20), ARTP treatment at 30 s, 70 s, and 110 s resulted in roughly 3.9-, 12-, and 25.2-fold increases in mutagenic frequency. These results demonstrate that ARTP is a potent mutagen whose impact is strictly dose-dependent.
Among the 290 SNVs identified in the 110 s ARTP-treated isolates, approximately 70% were localized within coding regions, comprising 134 missense, 57 synonymous, and 10 stop-gain/loss variants (Data Set S1). ARTP treatment elicited a distinct shift in the mutational spectrum (Fig. 1C). Specifically, we observed a significant elevation in the relative proportions of C-T/G-A transitions (increased by 39%) and A-T/T-A transversions (increased by 100%), whereas the frequencies of C-A/G-T and A-C/G-T transversions were markedly reduced compared to the untreated control (Fig. 1C).
Analysis of the ±1 bp sequence context showed that ARTP-induced SNVs exhibited clear trinucleotide-context preferences rather than a random distribution. Significant enrichments were detected primarily for C-T, T-A, and T-C substitutions (Fig. 1D). Among these, T-A substitutions were enriched in the sequences 5′-ATA-3′, 5′-TTA-3′, and 5′-TTT-3′, indicating a strong preference for A/T-rich sequence contexts (Fig. 1D). T-C substitution showed a significant enrichment in the sequences 5′-TTA-3′ and 5′-TTT-3′ (Fig. 1D), reflecting a pronounced bias toward T-rich sequence contexts. In addition, C-T substitutions were preferentially associated with the sequences 5′-CCT-3′, 5′-TCC-3′, and 5′-TCT-3′ (Fig. 1D). Although the T-C substitution within the 5′-TTA-3′ motif was the most prevalent individual event (Fig. 1D), C-T substitutions were broadly enriched across multiple sequence contexts, driving a significant overall increase in the relative proportion of C-T mutations (Fig. 1C). Together, these data indicate that ARTP drives SNV formation through the fixation of specific base lesions in preferred local sequence contexts.
Beyond SNVs, ARTP treatment significantly elevated the frequency of small insertions and deletions (InDels). We identified 24 InDels across the 12 sequenced isolates (Data Set S1), corresponding to a frequency of 4.3 × 10−10 events/bp/division—a 36.7-fold increase relative to the untreated control (1.2 × 10−11 events/bp/division). While the majority of these events (67%; 16/24) were situated in intergenic regions, a substantial proportion (33%; 8/24) occurred within coding sequences, all of which resulted in frameshift mutations (Data Set S1). Sequence context analysis further demonstrated that InDels preferentially localized to A/T-rich regions, such as (A/T)n or (AT)n motifs (Data Set S1), indicating that most InDels likely resulted from replication slippage.
BER pathway is not involved in repairing ARTP-induced DNA lesions
BER is a primary pathway for the removal of damaged bases and abasic (AP) sites in organisms (22, 23). Since ARTP exposure is expected to induce diverse oxidative and deaminative base damages, we investigated whether the BER pathway influences the processing and mutational outcomes of ARTP-induced lesions. To address this, we sequenced the genomes of diploid ung1Δ, ogg1Δ, mag1Δ, and apn1Δ strains treated with ARTP (110 s for 3 rounds) and compared their genomic alteration profiles directly with those of the ARTP-treated wild-type strain. The overall SNV rates in the ung1Δ, ogg1Δ, mag1Δ, and apn1Δ mutants were determined to be 5.0 × 10−9, 7.1 × 10−9, 4.6 × 10−9, and 5.8 × 10−9 events/bp/division, respectively. Notably, these values did not differ significantly from the rate observed in the wild-type strain (Fig. 2A). These results suggest that, in contrast to canonical lesions such as 8-oxo-G or uracil, the specific base damages induced by ARTP are not effectively recognized or processed by the intrinsic BER machinery.
Fig 2.

Contribution of BER and TLS pathways to ARTP-induced point mutations. (A) SNV rate in ARTP-treated wild-type, ung1, and apn1 strains. (B) SNV rate in ARTP-treated wild-type and translesion synthesis-deficient mutants rev1, rev3, and rad30. (C) Base-substitution spectra in wild-type, rev1Δ, rev3Δ, and rad30Δ backgrounds after ARTP treatment. (D) Absolute rate of individual base-substitution classes. (E) InDel rate in wild-type and TLS-deficient strains after ARTP treatment. P values were determined using the Wilcoxon rank-sum test for rate comparisons and the χ2 test for distribution or enrichment analyses. *P < 0.05; **P < 0.01. BER, base excision repair; TLS, translesion synthesis.
Rev1-Pol ζ complex contributes to SNV and InDel formation
TLS allows replication to proceed across damaged DNA templates and can therefore contribute to damage-induced mutagenesis (24–26). In S. cerevisiae, the TLS machinery comprises Rev1, a deoxycytidyl transferase that functions within a complex with Pol ζ; Pol ζ, which utilizes Rev3 as its catalytic subunit; and Pol η, encoded by RAD30 (27, 28). Because ARTP generates base lesions, we asked whether TLS is required for the formation of ARTP-induced point mutations. To address this question, we treated diploid rev1Δ, rev3Δ, and rad30Δ strains with ARTP (110 s) and sequenced six independent clones per genotype. Following a 110-s ARTP treatment, cell viability in the rev1Δ and rev3Δ deletion strains reached only ~65% of that in the wild-type control (P < 0.05), whereas viability in the rad30Δ mutant showed no statistically significant difference from the wild-type. Relative to the ARTP-treated wild type, the SNV rates were markedly reduced in rev1Δ and rev3Δ, but not in rad30Δ (Fig. 2B). The SNV rates in rev1Δ and rev3Δ strains were 1.6 × 10−9 and 1.2 × 10−9 events/bp/division, respectively, corresponding to 30.9% and 23.2% of the wild-type ARTP level (Fig. 2B). In contrast, the rad30Δ strain exhibited an SNV rate of 5.2 × 10−9 events/bp/division, corresponding to 1.1-fold of the wild-type ARTP level (Fig. 2B). These results indicate that the Rev1-Pol ζ complex is required for efficient conversion of ARTP-induced base lesions into heritable point mutations.
The substitution spectrum was significantly altered in TLS-deficient backgrounds compared to the wild type. In ARTP-treated WT cells, C-T/G-A transitions were the predominant class (36%), followed by A-G/T-C transitions (25%) and A-T/T-A transversions (20%). In the rev1Δ and rev3Δ backgrounds (lacking Pol ζ activity), the proportion of A-T/T-A substitutions dropped precipitously to approximately 5%, while C-G/G-C transversions became undetectable (Fig. 2C). Consequently, the relative fraction of C-T/G-A transitions increased to >50% (Fig. 2C). This shift indicates that the Rev1-Pol ζ complex is specifically responsible for the generation of A-T and C-G transversions during the bypass of ARTP-induced lesions. Conversely, the rad30Δ strain maintained high proportions of A-T/T-A (23%), C-A/G-T (17%), and A-C/T-G (12%) substitutions, exhibiting a mutational spectrum distinct from that of the Pol ζ mutants (Fig. 2C). Notably, the absolute rates of A-T/T-A (1.2 × 10−9 events/bp/division), C-A/G-T (8.7 × 10−10 events/bp/division), and A-C/T-G (6.4 × 10−10 events/bp/division) in rad30Δ were approximately 1.2-, 1.9-, and 2.2-fold those in WT cells, respectively (Fig. 2D). Although these differences did not reach statistical significance, Pol η (Rad30) is well established to facilitate high-fidelity lesion bypass—such as across thymine–thymine dimers—by incorporating correctly matched nucleotides (28). We therefore speculate that in the absence of Rad30, ARTP-induced lesions are more likely routed to alternative, error-prone translesion synthesis polymerases, thereby increasing the frequency of transversion events. Together, these results demonstrate that individual TLS polymerases make distinct and non-redundant contributions to lesion bypass, thereby shaping the final SNV spectrum in a polymerase-specific manner.
Relative to the ARTP-treated wild-type isolates, the InDel rate was reduced to 40% in rev1Δ and 26.7% in rev3Δ, respectively (Fig. 2E; P < 0.05). By contrast, rad30Δ retained 66.6% of the wild-type ARTP InDel level (Fig. 2E). These results confirm that the Rev1-Pol ζ complex is the principal driver of replication slippage and the subsequent formation of InDel mutations in ARTP-treated cells, whereas Pol η contributes minimally to these specific mutagenic events.
NER pathway is required to avoid ARTP-induced mutations
The NER pathway is a primary cellular defense mechanism for removing bulky DNA adducts, playing a critical role in protecting cells against UV-induced mutagenesis (29). To investigate the role of the NER pathway in ARTP-induced mutagenesis, we subcultured isolates of rad1Δ and rad2Δ strains under both untreated and ARTP-treated (3 rounds, 70 s) conditions. Phenotypically, both rad1Δ and rad2Δ mutants displayed significantly increased sensitivity to ARTP (Fig. 3A). Because a 110-s exposure resulted in complete lethality (no viable cells remained), an ARTP treatment duration of 70 s was selected for subsequent experiments. Wild-type isolates were subjected to the identical regime (3 rounds, 70 s) as controls. Genomic analysis revealed that under untreated conditions, the SNV rate of the rad1Δ mutant was 2.3-fold higher than that of the WT, whereas ARTP treatment further increased its SNV rate to 13.4-fold that of the WT (Fig. 3B). In contrast, while the untreated rad2Δ mutant showed no noticeable difference in SNV rate compared to the WT, ARTP exposure triggered a striking 21.5-fold increase in SNV rate (Fig. 3B). These findings demonstrate that the NER pathway serves as the primary mechanism for repairing ARTP-induced nucleotide damage. Furthermore, mutational spectrum analysis showed a marked elevation in the relative proportion of C-T/G-A substitutions in both rad1Δ and rad2Δ mutants (Fig. 3C). This specific substitution pattern is a canonical hallmark of UV-induced mutagenesis, primarily reflecting the misincorporation of adenine opposite damaged cytosine residues within pyrimidine dimers (30, 31).
Fig 3.

The role of the NER pathway in ARTP-induced mutagenesis. (A) Cell viability of wild-type (WT) and NER-deficient strains (rad1Δ and rad2Δ) following 70 s of ARTP treatment. P values were determined using the t test. ***P < 0.001. (B) SNV rates of WT and NER-deficient strains under untreated (left) and ARTP-treated (right) conditions. P values were determined using the Wilcoxon rank-sum test. **P < 0.01; ***P < 0.001. (C) Mutational spectra of SNVs in WT and NER-deficient strains.
ARTP stimulates mitotic recombination and LOH events
LOH is the most prevalent genetic event in heterozygous diploid yeast (32–34), occurring at a spontaneous frequency of 4.6 × 10−3 per cell division in the wild-type strain WYspo11 (20). LOH are generally categorized by their chromosomal architecture: interstitial (I-LOH), where the event is restricted to the interior of the chromosome, or terminal (T-LOH), which extends from the breakpoint to the chromosome terminus (20, 33). In the 12 isolates subjected to a 110-s ARTP treatment, we identified a total of 496 I-LOH and 138 T-LOH events (examples are provided in Fig. 4A and the coordinates of all LOH are listed in Data Set S2). The frequencies of I-LOH and T-LOH rose to 2.1 × 10−1 and 5.8 × 10−2 events per division, representing a 64-fold and 43-fold elevation over spontaneous background levels, respectively (Fig. 4B). The cumulative LOH region per strain ranged from 0.94 to 6.79 Mb. On average, a single ARTP treatment event induced LOH across approximately 0.469 Mb of the genome—equivalent to 4.1% of the genome size. Notably, despite this massive induction in overall LOH frequency, the relative proportion of I-LOH to T-LOH remained remarkably stable, with I-LOH accounting for approximately 78% of total events. In comparison with the high LOH frequency observed following 110 s of ARTP treatment (2.6 × 10−1 events/division), the frequencies for the 70 s and 30 s treatments were substantially lower, at 9.9 × 10−2 and 1.3 × 10−2 events/division, respectively (Fig. 4C). This decline mirrors the trend observed for SNVs, indicating that ARTP-induced DNA recombination occurs in a clear dose-dependent manner.
Fig 4.

ARTP stimulates genome-wide LOH and alters LOH tract complexity. (A) Representative LOH events detected by genome sequencing. The left panel shows an interstitial LOH (I-LOH) event, and the right panel shows a terminal LOH (T-LOH) event. (B) Frequencies of I-LOH and T-LOH events in untreated and 110-s ARTP-treated isolates. (C) Total LOH frequencies after different ARTP exposure durations. (D) Representative complex gene conversion tract associated with T-LOH in an ARTP-treated isolate. More than two transitions (A, B, C) between W303- and YJM789-derived. haplotypes within the conversion region define a complex tract. (E) Distribution of gene conversion tract lengths in untreated and ARTP-treated isolates. Horizontal lines indicate median tract sizes. (F) Genome-wide distribution of LOH and chromosomal rearrangement events across the 16 yeast chromosomes. Horizontal colored segments indicate the genomic spans of I-LOH, T-LOH, and I-DUP/DEL events.
We further investigated whether ARTP treatment altered the patterns of LOH events. Regarding T-LOH, we found that 72 out of 152 events (47.4%) were associated with detectable gene conversion tracts. Notably, 40.1% of all ARTP-induced T-LOH events (61 out of 152) featured complex tracts—defined by more than two sequence transitions between homologs (as exemplified in Fig. 4D)—representing a 2.4-fold increase over the 16.6% (59 out of 356) observed in untreated isolates. Similarly, the ratio of complex to simple tracts for non-crossover gene conversion events surged from 11% to 30% following ARTP treatment. The median size of I-LOH (gene conversion) tracts was 5,889 bp, which is not significantly different from that of untreated isolates (4,079 bp) (Wilcoxon rank-sum test, P > 0.05; Fig. 4E) (20). The median size of T-LOH-associated tracts (homologous region) was 201,189 bp, which was significantly smaller than that of untreated isolates (391,252 bp) (Wilcoxon rank-sum test, P = 2.62 × 10⁻⁸). This result suggests that ARTP-induced breakpoints tend to occur closer to the telomeres compared to spontaneous ones.
To assess the spatial distribution of LOH breakpoints, we mapped the “windows” that are likely to contain the breakpoints for each event across the 16 chromosomes (Fig. 4F). Our analysis revealed no striking hotspots for LOH induction. A χ2-based enrichment analysis was subsequently performed to determine whether specific chromosomal features—such as centromeres, replication termination sites, highly transcribed genes, high-GC content, and low-GC content—were preferentially associated with LOH breakpoints (35). We found that low-GC content regions were significantly enriched, while high-GC content regions were underrepresented near the breakpoints (Data Set S3). Later we will discuss the possible explanation of this phenomenon.
The origins of the DNA lesions leading to recombination and mutation
In the established framework of yeast genetics, mitotic homologous recombination primarily serves to repair DNA double-strand breaks (DSBs) (36–38). DSBs occurring in the G1 phase are particularly significant; as the cell progresses into the S/G2 phase, both sister chromatids will inherit the lesion, necessitating repair using the homologous chromosome as a template. As illustrated in Fig. 5A, the repair of G1-induced DSBs yields a 4:0 gene conversion pattern at the recombination breakpoint (where “4” indicates that both daughter cells inherit the template chromosome’s sequence). Conversely, DSBs initiated during the S/G2 phase—often resulting from the conversion of single-strand breaks into DSBs during DNA replication—typically manifest as a 3:1 segregation pattern (Fig. 5A) (39, 40). To determine the temporal origin of ARTP-induced recombinogenic lesions, we employed a yeast red/white sectored colony assay (Fig. 5A). Diploid JSC25 cells (isogenic to WYspo11) were synchronized in the G1 phase using alpha-factor prior to ARTP treatment. JSC25 is homozygous for the ade2-1 ochre mutation and carries a single hemizygous SUP4-o ochre suppressor inserted near the right telomere of chromosome IV (Fig. 5A) (39). A reciprocal crossover between the centromere and the SUP4-o locus alters the suppressor copy number in the resulting daughter cells, manifesting as red/white sectored colonies on solid medium (Fig. 5A). This phenotypic readout allows for the isolation and molecular analysis of the two distinct daughter lineages derived from a single, defined mitotic recombination event.
Fig 5.

ARTP-induced recombination and mutations frequently originate before DNA replication. (A) LOH patterns used to infer the timing of ARTP-induced recombination in red/white sectored colonies. Pre-replicative DSBs produce 4:0 conversion tracts, whereas S/G2 DSBs produce 3:1 conversion tracts. (B) Representative selected crossover on chromosome IV containing a 4:0 gene conversion region. The same converted interval is detected in both reciprocal sectors, consistent with repair of a pre-replicative DSB. (C) Classification of ARTP-induced one-strand mutations (OSMs) and two-strand mutations (TSMs) in paired sectors. OSMs arise when a lesion is fixed as a mutation in only one replicated chromatid and are therefore detected in one sector only. TSMs arise when a lesion is fixed before DNA replication and are detected as the same SNV in both sectors. Red and blue chromatids indicate W303- and YJM789-derived homologs, respectively.
We isolated and purified cell populations from the red and white sectors of 19 colonies for whole-genome sequencing. Detailed analysis of the gene conversion tracts associated with the 19 selected crossovers on the right arm of chromosome IV revealed that the gene conversion tracts of 12 events contained a 4:0 region (Data Set S4; one example is shown in Fig. 5B), confirming that the initiating DSBs occurred during the G1 phase. The remaining seven events were identified as being initiated by DSBs during the S/G2 phase. Beyond the targeted crossovers on chromosome IV, we identified 63 additional I-LOH and 24 T-LOH events throughout the genome (Data Set S4).
Among I-LOH events without crossovers, complex patterns (Class 1C; mixed 3:1/4:0 or more complex tracts) were common (n = 27) and outnumbered both simple 3:1 events (Class 1A; n = 10) and simple 4:0 events (Class 1B; n = 18) (Data Set S4). A smaller number of other complex patterns (Class 1D; n = 8) was also observed. Among the gene conversion events that were informative for timing (Classes 1A–1C), 78.2% (43/55) contained 4:0 or mixed 3:1/4:0 tracts, indicating that most recombinogenic lesions were initiated before replication.
The origin of T-LOH events was further examined. Events with reciprocal T-LOH in both sectors, as expected for a crossover, were scored as CO. Events with nonreciprocal outcomes, including unilateral T-LOH in one sector, as expected for break-induced replication (BIR) (41). Among the T-LOH events analyzed, CO was the predominant outcome. Thirty-two events showed reciprocal T-LOH patterns consistent with CO, whereas only seven were BIR-associated. Thus, in this system, T-LOH predominantly arises from reciprocal crossovers rather than BIR. Among events with assignable donor usage, the W303-1A and YJM789 homologs served as repair templates at similar frequencies (52 and 43 times, respectively) (Data Set S4), indicating that there is no fixed template bias in the homologous repair of ARTP-induced breaks.
Across the 19 sectored colonies, we also identified a total of 55 SNVs. Assuming these mutations occurred within a single cell division, the estimated maximum SNV rate was 1.2 × 10−7 events/bp/division, approximately 600 times that under spontaneous conditions. Notably, identical mutations were frequently detected in both sectors of the same colony, accounting for 36.4% (20/55) of the identified SNVs. We defined these shared mutations as two-strand mutations (TSMs) and those present in only one sector as one-strand mutations (OSMs) (Fig. 5C). Mechanistically, TSMs and OSMs are indicative of DNA lesion processing during the G1 and S phases, respectively (Fig. 5C) (42). A TSM arises when a G1-phase lesion is repaired in an error-prone manner on both strands of the DNA duplex prior to replication, leading to the mutation being inherited by both sister chromatids (Fig. 5C) (42). Conversely, an OSM typically results from the replication of a damaged template, where the mutation is fixed in only one of the two daughter DNA molecules (Fig. 5C) (42). The observed TSM:OSM ratio of approximately 1:1.75 indicates that a substantial fraction of ARTP-induced mutations is fixed on a homolog in G1-phase cells before the onset of chromosome replication.
ARTP effectively improves acetic acid resistance in yeast
To assess the efficacy of ARTP mutagenesis in generating acetic acid-tolerant yeast, the haploid strain 0912-1α (derived from the bioethanol strain YJS329 [43]) underwent 3 rounds of ARTP exposure (70 s each) and subsequent selection under increasing acetic acid stress. This procedure yielded 14 0912-1α-derived isolates with improved acetic acid tolerance to 5.5 g/L acetic acid (Fig. 6A). As a control, several isolates were serially passaged on acetic acid-containing plates for three rounds without ARTP treatment; however, no mutants with stably enhanced tolerance were obtained. Although we cannot exclude the possibility that prolonged cultivation under acetic acid stress might eventually yield tolerant strains, ARTP treatment significantly increases the probability of acquiring such phenotypes by elevating the overall mutation rate.
Fig 6.

ARTP-mediated evolution reveals OTU2 as a determinant of acetic acid tolerance. (A) Acetic acid tolerance of ARTP-evolved isolates. The parental strain 0912-1α (WT) and 14 ARTP-evolved isolates were spotted on YPD plates with or without 5.5 g/L acetic acid. (B) Candidate coding mutations in acetic acid-tolerant isolates identified by whole-genome sequencing. (C) Comparative growth of the parental strain and three independent otu2Δ transformants (1–3) on YPD and acetic acid-containing plates. All otu2Δ transformants showed improved acetic acid tolerance.
To identify the genetic basis of the improved acetic acid resistance, we performed whole-genome sequencing on all 14 evolved isolates. Analysis of SNVs and small InDels identified a set of candidate genes carrying coding changes—specifically frameshift mutations—that were potentially associated with enhanced acetic acid tolerance (Fig. 6B). To validate these findings, we performed targeted gene deletions of five candidates: DTR1, OTU2, PPH21, PPR1, and ASF2. A literature survey of functional annotations indicated that these five genes are more strongly linked to acetic acid tolerance than the other identified candidates. Chronologically, mutations in OTU2 and PPH21 were present across all isolates from the second and third rounds of treatment, indicating that they arose during the first round of ARTP mutagenesis. Mutations in PPR1 and ASF2 emerged during the second round, whereas the DTR1 mutation occurred in the third round. Among the tested mutants, the otu2Δ strain displayed significantly improved growth on 4.5 g/L and 5.0 g/L acetic acid plates relative to the wild-type control (Fig. 6C). Notably, the deletion of OTU2 effectively phenocopied the robust tolerance observed in the isolates recovered from the initial ARTP screen, confirming its role as a causal factor in the evolved stress-response phenotype. These results show that ARTP treatment, combined with genome sequencing and reverse validation, can efficiently convert selected mutants into tractable engineering targets.
DISCUSSION
This study establishes ARTP as a composite mutagen that generates a diverse spectrum of localized base lesions and large-scale recombinogenic DNA damage. Within this framework, the final genomic landscape is jointly determined by the duration of plasma exposure and the cell’s hierarchical choice of DNA repair pathways. Consequently, SNVs, InDels, and LOH are not merely parallel outputs of chemical damage but represent distinct biological endpoints shaped by the processing of specific lesions. Furthermore, we demonstrate the high efficacy of ARTP-mediated directed evolution in facilitating the rapid generation of acetate-tolerant yeast strains. Below, we discuss the implications of these findings for microbial strain engineering and our understanding of ARTP-induced genomic instability.
Unique mutational signatures induced by ARTP and underlying genetic model
The mutational spectrum elicited by ARTP is characterized by a prominent enrichment of C-T/G-A transitions and A-T/T-A transversions. This signature diverges sharply from the highly specific chemical footprints of conventional agents, such as ethyl methanesulfonate (predominantly alkylation-driven C-T transitions) (31), bleomycin (T-G transversions and small InDels) (25), or H2O2 (oxidative G-T transversions) (21). While the ARTP spectrum exhibits a partial overlap with the UV-induced signature (42), this similarity is intrinsically linked to the physical properties of the plasma discharge. During operation, ARTP utilizes helium as a carrier gas to generate a uniform, high-activity plasma at atmospheric pressure. The formation and relaxation of this high-energy plasma state inherently emit a substantial flux of UV radiation (1).
Under spontaneous conditions, C-T/G-A transitions constitute the most prevalent SNV class, primarily reflecting the hydrolytic deamination of cytosine to uracil (44). Within the canonical BER framework, the uracil DNA glycosylase Ung1 specifically recognizes and excises these uracil residues to prevent mutation fixation (44). If ARTP-induced mutations were driven by simple cytosine deamination, a dramatic surge in C-T rates would be expected in an ung1Δ background. However, our results demonstrate that the loss of UNG1 does not significantly amplify the ARTP-induced SNV rate, suggesting that ARTP does not primarily induce simple cytosine deamination. Beyond Ung1, cells employ specialized glycosylases such as Ogg1 (45) and Mag1 (46) to remove specific damaged bases, creating AP sites that are subsequently processed by the AP endonuclease Apn1 (47). Notably, the deletion of OGG1, MAG1, and APN1 also failed to elicit a significant shift in ARTP-induced mutagenic potency, indicating that the BER pathway is largely dispensable for processing ARTP-induced lesions. Instead, these lesions are fixed into permanent mutations primarily via the Rev1-Pol ζ-mediated TLS pathway.
A plausible explanation for this phenomenon is that ARTP generates complex or clustered DNA lesions, such as pyrimidine dimers, rather than isolated, simple chemical modifications. These bulky lesions impose significant steric hindrance that prevents canonical BER glycosylases from successfully performing “base flipping”—a necessary step to rotate the damaged base into the enzyme’s active site. Instead, such bulky adducts are primarily targeted by the NER pathway. This mechanism is directly corroborated by our observation that deletion of key NER genes (RAD1 or RAD2) markedly increases cellular sensitivity to ARTP and dramatically surges the SNV rate (Fig. 3), accompanied by a pronounced elevation in C-T substitutions—a canonical signature of unrepaired pyrimidine dimers. When these lesions escape NER repair and stall replication forks, the cell recruits TLS polymerases. Specifically, the Rev1-Pol ζ complex is engaged to extend DNA strands from distorted primer termini (Fig. 7A), as evidenced by the significantly reduced mutation rates in rev1Δ and rev3Δ mutants compared to wild-type levels. Furthermore, sequence context analysis reveals a strong bias toward mutations localized at dipyrimidine motifs (5′-CT-3′, 5′-TT-3′, and 5′-CC-3′; Fig. 1D). Together, these findings indicate that the characteristic C-T/G-A and A-T/T-A substitutions largely arise from error-prone bypass of damaged C or T residues—primarily CT, TT, and CC pyrimidine dimers—by Pol ζ, whose specialized extender activity facilitates incorporation opposite distorted templates (Fig. 7A). In contrast to the error-prone nature of Rev1-Pol ζ (48), Rad30 functions as a high-fidelity polymerase specifically adapted to bypass TT dimers accurately (28). By correctly inserting AA opposite the dimer, Rad30 facilitates error-free bypass, consistent with our observation that the rates of certain base substitutions increase upon RAD30 deletion.
Fig 7.

Genetic models of ARTP-induced genome instability. (A) Mutagenic and error-free bypass of ARTP-induced lesions. ARTP-induced cyclobutane pyrimidine dimers (CPDs; indicated by asterisks) are bypassed by the Rev1-Pol ζ complex via error-prone translesion synthesis, leading to the fixation of point mutations. In contrast, Pol η (Rad30) facilitates high-fidelity bypass by inserting the correct matching bases opposite the dimers, thereby mitigating the mutational burden. (B) Mechanism of double-strand break (DSB) formation. Chromosomal DNA is represented as unreplicated double-stranded molecules, with newly synthesized DNA shown as dashed lines. (i) The reactive species and high-energy particles generated by ARTP can strike the genome to induce immediate, direct double-strand DNA cleavage. (ii) Simultaneous excision of two closely opposed dimers (triangles) on opposite strands generates a short, unstable double-stranded region. The collapse of this region leads to a functional DSB, which, upon replication, produces two broken sister chromatids. (C) Homologous recombination (HR) pathways leading to loss of heterozygosity (LOH). The repair process is initiated by the recognition of a DSB, followed by 5'−3' resection to generate long 3' single-stranded DNA overhangs. The extent of this resection is positively correlated with the length of the resulting gene conversion tract. These 3' overhangs invade the homologous template to initiate repair synthesis. Synthesis-dependent strand annealing (SDSA): After limited repair synthesis, the extended invading strand dissociates from the template and anneals back to the complementary ssDNA on the other side of the original DSB, resulting in non-crossover (NCO) gene conversion and interstitial LOH (I-LOH). Double-strand break repair (DSBR): Alternatively, the second end of the broken DNA is captured by the displaced strand of the homologous template, forming a double Holliday junction (dHJ) intermediate. Resolution of the dHJ structure can yield either NCO or reciprocal crossovers (CO), the latter being a primary driver of terminal LOH (T-LOH). Break-induced replication (BIR): A broken DNA end may also invade the homologous chromosome to initiate a unidirectional replication process that can extend for hundreds of kilobases to the chromosome terminus, leading to non-reciprocal T-LOH.
Pre-replicative DSBs and recombination dynamics
The capacity of ARTP to induce high-frequent recombination is a defining feature of its mutagenic profile. Our sector analysis demonstrated that approximately 74% of recombination events in ARTP-treated G1 cells originate from lesions initiated prior to the S phase, as evidenced by 4:0 or more complex gene conversion patterns (Data Set S4). This ability to induce direct DNA cleavage independent of the replication fork suggests that ARTP behaves similarly to radiomimetic agents like bleomycin or H2O2 (21, 25). We propose two non-exclusive mechanisms to explain this high pre-replicative DSB frequency. The first explanation posits that the high-activity plasma jet possesses the intrinsic energy to directly sever the DNA phosphodiester backbone (Fig. 7B). ARTP generates a potent “cocktail” of high-energy electrons, ions, and highly reactive oxygen and nitrogen species. When these species strike the genome, they may cause immediate, direct double-strand cleavage through high-energy impacts or concentrated oxidative attacks, bypassing the need for biological processing to create a break. Alternatively, the elevated DSB frequency may arise from the stochastic and dense distribution of ARTP-induced lesions (Fig. 7B). Such “clustered DNA damage,” characterized by proximal lesions (e.g., pyrimidine dimers) on opposite strands, can be inadvertently converted into a DSB by the cellular repair machinery (Fig. 7B). A prime example occurs when the NER pathway concurrently processes two proximal lesions on opposing strands, leading to closely opposed nicks that collapse into a double-strand break (40). Our finding that LOH breakpoints are significantly enriched in low-GC (A/T-rich) regions (Data Set S3) strongly supports the prevalence of this mechanism. This enrichment suggests that ARTP-induced lesions are predominantly pyrimidine dimers formed at 5′-TT-3′ motifs, which are more abundant in these genomic contexts. Beyond repair-mediated mechanisms, the proximal collapse of replication forks on both strands provides an additional pathway for generating DSBs that affect both sister chromatids.
However, a distinction must be made regarding the absolute quantity of damage across the cell cycle. While ARTP is highly effective in G1, the total number of DSBs is numerically higher in the S/G2 phases. This is largely because DSBs occurring after replication are preferentially repaired using the sister chromatid as a template (39). Consequently, the recombination events we observe in G1 represent a unique window where the cell is forced to utilize homologous chromosomes, leading to significant genomic alterations. Crucially, we identified reciprocal crossovers, rather than BIR, as the primary driver of terminal LOH in ARTP-treated cells. This is accompanied by a significant surge in complex gene conversion tracts. As illustrated in Fig. 7C, the initial step of homologous recombination involves the 5′−3′ resection of DNA ends at the DSB site, which generates long 3′ single-stranded DNA overhangs. The extent of this resection is positively correlated with the resulting gene conversion tract length. These 3′ overhangs subsequently invade the homologous template to initiate repair synthesis. During this elongation phase, the repair machinery may undergo frequent template switching, especially when encountering obstacles on the template strand. The observed increase in complexity of gene conversion tracts in ARTP-treated cells suggests that ARTP-induced DNA damage likely forces the invading strand to switch templates multiple times during repair synthesis.
Our quantitative analysis reveals that a single round of ARTP treatment induces LOH across approximately 4% of the yeast S. cerevisiae genome. In heterozygous industrial yeast strains, beneficial recessive alleles are frequently masked by dominant counterparts. ARTP-induced LOH facilitates “phenotypic unmasking” by rapidly converting these hidden alleles to homozygosity across the entire genome, thereby accelerating the evolution of superior industrial phenotypes.
Implications for industrial strain development
Our successful engineering of acetate-tolerant strains, leading to the identification of Otu2 as a pivotal causal factor, validates the efficacy of this framework. While the precise adaptive landscape remains to be fully elucidated, we hypothesize that the deletion of OTU2—a deubiquitinase associated with the 40S ribosomal subunit (49)—enhances stress resilience by modulating translational reset and ribosome recycling. This may facilitate more efficient protein synthesis or altered mRNA selectivity under the physiological constraints of acetic acid stress. Furthermore, our study reveals that ARTP generates a distinct mutational profile that is different from traditional chemical or physical mutagens. Leveraging the synergy between ARTP and other mutagenic agents will further expand the accessible sequence space, accelerating the discovery of novel genetic drivers for increasingly complex industrial phenotypes.
In conclusion, this study delineates the genomic signature and underlying mechanisms of ARTP-induced mutagenesis in S. cerevisiae. ARTP primarily generates pyrimidine dimers and DNA double-strand breaks, whose mutagenic outcomes depend on the Rev1-Pol ζ pathway but not the BER system. Our findings establish a mechanistic framework for ARTP and validate its practical use for evolving robust industrial strains.
MATERIALS AND METHODS
Strains and medium
The diploid strain WYspo11 (MATa/α leu2-3,112/LEU his3-11,15/HIS ura3-1/ura3 ade2-1/ade2-1 trp1-1/TRP can1-100/CAN1 RAD5/RAD5 GAL2/gal2 ho::hisG IV1510386/IV1510386::SUP4-o spo11::loxP/spo11::loxP) served as the parental background for exploring ARTP-induced mutagenesis (20). To evaluate the specific contributions of the BER and TLS pathways, a series of gene-deletion strains isogenic to the WYspo11 background was used (Table 1). These included BER-deficient mutants (apn1Δ, mag1Δ, ogg1Δ, and ung1Δ), NER-deficient mutants (rad1Δ and rad2Δ), and TLS-deficient mutants (rev1Δ, rev3Δ, and rad30Δ). For directed evolution experiments regarding acetic acid tolerance, the haploid strain 0912-1α derived from strain YJS329 was utilized (43). Yeast strains were routinely cultured at 30°C in YPD medium (1% yeast extract, 2% peptone, and 2% glucose). To prepare the selective medium for acetic acid tolerance, the base medium was first sterilized by autoclaving. Once the medium had cooled to 55°C, glacial acetic acid was added aseptically to the desired final concentration.
TABLE 1.
Yeast strains used in this study
| Strain | Genotype | Resource |
|---|---|---|
| WYspo11 | MATa/α leu2-3,112/LEU his3-11,15/HIS ura3-1/ura3 ade2-1/ade2-1 trp1-1/TRP can1-100/CAN1 GAL2/gal2 ho::hisG IV1510386/IV1510386::SUP4-o spo11::loxP/spo11::loxP | (20) |
| WY105 (apn1Δ) |
MATa/α leu2-3,112/LEU his3-11,15/HIS ura3-1/ura3 ade2-1/ade2-1 trp1-1/TRP can1-100/CAN1 GAL2/gal2 ho::hisG IV1510386/IV1510386::SUP4-o spo11::loxP/spo11::loxP ung1::URA3/ung1/URA3 | Lab stock |
| WY1 (ung1Δ) |
MATa/α leu2-3,112/LEU his3-11,15/HIS ura3-1/ura3 ade2-1/ade2-1 trp1-1/TRP can1-100/CAN1 GAL2/gal2 ho::hisG IV1510386/IV1510386::SUP4-o spo11::loxP/spo11::loxP apn1::URA3/apn1::URA3 | Lab stock |
| WY145 (mag1Δ) |
MATa/α leu2-3,112/LEU his3-11,15/HIS ura3-1/ura3 ade2-1/ade2-1 trp1-1/TRP can1-100/CAN1 GAL2/gal2 ho::hisG IV1510386/IV1510386::SUP4-o spo11::loxP/spo11::loxP mag1::URA3/mag1::URA3 | Lab stock |
| WY39 (ogg1Δ) |
MATa/α leu2-3,112/LEU his3-11,15/HIS ura3-1/ura3 ade2-1/ade2-1 trp1-1/TRP can1-100/CAN1 GAL2/gal2 ho::hisG IV1510386/IV1510386::SUP4-o spo11::loxP/spo11::loxP ogg1::URA3/ogg1::URA3 | Lab stock |
| WY50 (rad1Δ) |
MATa/α leu2-3,112/LEU his3-11,15/HIS ura3-1/ura3 ade2-1/ade2-1 trp1-1/TRP can1-100/CAN1 GAL2/gal2 ho::hisG IV1510386/IV1510386::SUP4-o spo11::loxP/spo11::loxP rad1::URA3/rad1::URA3 | Lab stock |
| WY51 (rad2Δ) |
MATa/α leu2-3,112/LEU his3-11,15/HIS ura3-1/ura3 ade2-1/ade2-1 trp1-1/TRP can1-100/CAN1 GAL2/gal2 ho::hisG IV1510386/IV1510386::SUP4-o spo11::loxP/spo11::loxP rad2::URA3/rad2::URA3 | Lab stock |
| WY74 (rev1Δ) |
MATa/α leu2-3,112/LEU his3-11,15/HIS ura3-1/ura3 ade2-1/ade2-1 trp1-1/TRP can1-100/CAN1 GAL2/gal2 ho::hisG IV1510386/IV1510386::SUP4-o spo11::loxP/spo11::loxP rev1::URA3/rev1::URA3 | Lab stock |
| WY75 (rev3Δ) |
MATa/α leu2-3,112/LEU his3-11,15/HIS ura3-1/ura3 ade2-1/ade2-1 trp1-1/TRP can1-100/CAN1 GAL2/gal2 ho::hisG IV1510386/IV1510386::SUP4-o spo11::loxP/spo11::loxP rev3::URA3/rev3::URA3 | Lab stock |
| WY63 (rad30Δ) |
MATa/α leu2-3,112/LEU his3-11,15/HIS ura3-1/ura3 ade2-1/ade2-1 trp1-1/TRP can1-100/CAN1 GAL2/gal2 ho::hisG IV1510386/IV1510386::SUP4-o spo11::loxP/spo11::loxP rad30::URA3/rad30::URA3 | Lab stock |
| 0912-1α | MATα | (43) |
| SY165 | MATa leu2-3,112 his3-11,15 ura3-1 ade2-1 trp1-one can1-100 spo11::loxP | (20) |
| otu2Δ | MATa leu2-3,112 his3-11,15 ura3-1 ade2-1 trp1-one can1-100 spo11::loxP otu2::KanMX6 | This study |
Gene deletion
Gene deletions were executed using a PCR-mediated gene replacement strategy. The specific primers utilized for strain construction are listed in Data Set S5. PCR was performed with Phanta Max Super-Fidelity DNA Polymerase (P505-d1; Vazyme Biotechnology Co. LTD., China) over 30 cycles. Yeast transformations were performed with 5 µg of the gene deletion cassette using the lithium acetate/single-stranded carrier DNA/polyethylene glycol (LiAc/ssDNA/PEG) protocol (50). The G418 resistance gene (KanMX6) from plasmid pUG6 was used for target gene disruption and selection (51). Successful locus-specific genomic integrations were verified via colony PCR using primers flanking the deletion sites.
ARTP treatment
ARTP mutagenesis was performed using an ARTP-II mutation system (Tmaxtree, Wuxi, China). Log-phase yeast cells (OD600 ≈ 1) were harvested by centrifugation at 6,000 rpm at 30°C, washed once, and resuspended in 1 mL sterile water to a concentration of 107 cells/mL. A 10 µL cell suspension was spread onto a sterile stainless-steel plate and exposed to a helium plasma jet (10 SLM, 120 W) at a 2 mm distance. Exposure time–lethality curves were generated (0–150 s) by plating on YPD to determine survival.
G1 phase synchronization and sectoring assay
The colony-sectoring assay was used to detect recombination events as previously described (21, 52). The strain JSC25 used for this assay carries a heterozygous SUP4-o marker inserted near the right telomere of chromosome IV (39). JSC25 cells were cultured in 10 mL of YPD medium to log phase (OD600 ≈ 1) and arrested with 5 μg/mL α-factor for 2.5 h. G1 arrest was confirmed by microscopy (>90% unbudded cells). Synchronized cells were treated with ARTP for 110 s and plated onto solid YPD medium. Red/white sectored colonies were identified after incubation at 30°C for 3 to 5 days. Sectors from the same colony were purified and enriched (cultured in 10 mL YPD for 24 h) for genomic DNA extraction and genome sequencing.
Genome sequencing and analysis
Genomic DNA of yeast cells (OD600 ≈ 5) was extracted from overnight cultures using a commercial yeast DNA extraction kit (Omega Bio-Tek EZNA). DNA library construction and sequencing were performed as described in our previous studies (24, 31, 53). Whole-genome sequencing was performed on the BGI T7 platform to generate 150-bp paired-end reads. The average sequencing depth was approximately 100×. Raw reads were processed with Trimmomatic to remove adaptors and low-quality bases (54). The clean reads were aligned to the S. cerevisiae reference genome (S288C, R64-1-1) using the Burrows-Wheeler Aligner software (55). PCR duplicates were marked and removed using SAMtools (56). SNVs and InDels were identified using VarScan 2.3.9 (57). High-confidence variants were filtered based on read depth (>20× ) and allele frequency (>0.25). Background variants inherently present in the parental strain were identified and removed. To evaluate loss of heterozygosity (LOH), we monitored 50,000 high-quality, pre-validated heterozygous SNP markers spanning the genome that distinguish the W303-1A and YJM789 parental backgrounds. An LOH event was defined as a shift in a contiguous block of markers from a heterozygous state (0.25 < allele ratio < 0.7) to a homozygous state (allele ratio <0.1 or >0.9). Interstitial LOH events were defined as homozygous tracts flanked on both sides by heterozygous markers, whereas terminal LOH events extended continuously to the chromosome end.
Statistical analysis
Differences in mutation rates were evaluated using the Wilcoxon rank-sum test. The relative distribution of genetic events was analyzed using Fisher’s exact test (two-tailed). The enrichment or underrepresentation of chromosomal elements associated with LOH breakpoints was determined by the χ2 test. For all statistical analyses, a P-value < 0.05 was considered to indicate statistical significance.
ACKNOWLEDGMENTS
This study was supported by the National Key R&D Program of China (2023YFC2813000 and 2023YFE0124700), the National Natural Science Foundation of Zhejiang Province (LZ24C010002), and the National Natural Science Foundation of China, China (32270086).
Contributor Information
Ke Zhang, Email: zhangke726@zju.edu.cn.
Yvonne Nygård, Chalmers tekniska hogskola AB, Gothenburg, Sweden.
DATA AVAILABILITY
The raw whole-genome sequencing data of yeast isolates were deposited in the SRA database with BioProject accession number PRJNA1463135.
SUPPLEMENTAL MATERIAL
The following material is available online at https://doi.org/10.1128/aem.00955-26.
Data Sets S1 to S5.
ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.
REFERENCES
- 1. Li Y-H, Juo J-J, Ng I-S. 2025. Current breakthroughs and advances in atmospheric room temperature plasma (ARTP) technology for biomanufacturing. Bioresour Bioprocess 12:63. doi: 10.1186/s40643-025-00907-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Zhang X, Zhang X-F, Li H-P, Wang L-Y, Zhang C, Xing X-H, Bao C-Y. 2014. Atmospheric and room temperature plasma (ARTP) as a new powerful mutagenesis tool. Appl Microbiol Biotechnol 98:5387–5396. doi: 10.1007/s00253-014-5755-y [DOI] [PubMed] [Google Scholar]
- 3. Yang H, Zhang B, Wu Z, Pan J, Chen L, Xiu X, Cai X, Liu Z, Zheng Y. 2024. Synergistic application of atmospheric and room temperature plasma mutagenesis and adaptive laboratory evolution improves the tolerance of Escherichia coli to L-cysteine. Biotechnol J 19:e2300648. doi: 10.1002/biot.202300648 [DOI] [PubMed] [Google Scholar]
- 4. Meng L, Gao X, Liu X, Sun M, Yan H, Li A, Yang Y, Bai Z. 2021. Enhancement of heterologous protein production in Corynebacterium glutamicum via atmospheric and room temperature plasma mutagenesis and high-throughput screening. J Biotechnol 339:22–31. doi: 10.1016/j.jbiotec.2021.07.010 [DOI] [PubMed] [Google Scholar]
- 5. Tian T, Wu D, Ng C-T, Yang H, Sun J, Liu J, Lu J. 2020. A multiple-step strategy for screening Saccharomyces cerevisiae strains with improved acid tolerance and aroma profiles. Appl Microbiol Biotechnol 104:3097–3107. doi: 10.1007/s00253-020-10451-z [DOI] [PubMed] [Google Scholar]
- 6. Zhang C, Qin J, Dai Y, Mu W, Zhang T. 2019. Atmospheric and room temperature plasma (ARTP) mutagenesis enables xylitol over-production with yeast Candida tropicalis. J Biotechnol 296:7–13. doi: 10.1016/j.jbiotec.2019.01.026 [DOI] [PubMed] [Google Scholar]
- 7. Ji-Lun H, Xiao-Yan Z, Gui-Xing W, Zhao-Hui S, Wei D, Ya-Xian Z, Fei S, Li-Yan W, Xin-Hui X, Yu-Fen W. 2019. Novel breeding approach for Japanese flounder using atmosphere and room temperature plasma mutagenesis tool. BMC Genomics 20:323. doi: 10.1186/s12864-019-5681-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Qian X, Xin K, Zhang L, Zhou J, Xu A, Dong W, Jiang M. 2023. Integration of ARTP mutation and adaptive laboratory evolution to reveal 1,4‑butanediol degradation in Pseudomonas putida KT2440. Microbiol Spectr 11:e0498822. doi: 10.1128/spectrum.04988-22 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Jönsson LJ, Martín C. 2016. Pretreatment of lignocellulose: formation of inhibitory by-products and strategies for minimizing their effects. Bioresour Technol 199:103–112. doi: 10.1016/j.biortech.2015.10.009 [DOI] [PubMed] [Google Scholar]
- 10. Palma M, Guerreiro JF, Sá-Correia I. 2018. Adaptive response and tolerance to acetic acid in Saccharomyces cerevisiae and Zygosaccharomyces bailii: a physiological genomics perspective. Front Microbiol 9:274. doi: 10.3389/fmicb.2018.00274 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Geng P, Zhang L, Shi G-Y. 2017. Omics analysis of acetic acid tolerance in Saccharomyces cerevisiae. World J Microbiol Biotechnol 33:94. doi: 10.1007/s11274-017-2259-9 [DOI] [PubMed] [Google Scholar]
- 12. Guerreiro JF, Mira NP, Santos AXS, Riezman H, Sá-Correia I. 2017. Membrane phosphoproteomics of yeast early response to acetic acid: role of Hrk1 kinase and lipid biosynthetic pathways, in particular sphingolipids. Front Microbiol 8:1302. doi: 10.3389/fmicb.2017.01302 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Chaves SR, Rego A, Martins VM, Santos-Pereira C, Sousa MJ, Côrte-Real M. 2021. Regulation of cell death induced by acetic acid in yeasts. Front Cell Dev Biol 9:642375. doi: 10.3389/fcell.2021.642375 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Fletcher E, Feizi A, Bisschops MMM, Hallström BM, Khoomrung S, Siewers V, Nielsen J. 2017. Evolutionary engineering reveals divergent paths when yeast is adapted to different acidic environments. Metab Eng 39:19–28. doi: 10.1016/j.ymben.2016.10.010 [DOI] [PubMed] [Google Scholar]
- 15. Della-Bianca BE, de Hulster E, Pronk JT, van Maris AJA, Gombert AK. 2014. Physiology of the fuel ethanol strain Saccharomyces cerevisiae PE-2 at low pH indicates a context-dependent performance relevant for industrial applications. FEMS Yeast Res 14:1196–1205. doi: 10.1111/1567-1364.12217 [DOI] [PubMed] [Google Scholar]
- 16. Zhang M, Zhang K, Mehmood MA, Zhao ZK, Bai F, Zhao X. 2017. Deletion of acetate transporter gene ADY2 improved tolerance of Saccharomyces cerevisiae against multiple stresses and enhanced ethanol production in the presence of acetic acid. Bioresour Technol 245:1461–1468. doi: 10.1016/j.biortech.2017.05.191 [DOI] [PubMed] [Google Scholar]
- 17. Swinnen S, Henriques SF, Shrestha R, Ho P-W, Sá-Correia I, Nevoigt E. 2017. Improvement of yeast tolerance to acetic acid through Haa1 transcription factor engineering: towards the underlying mechanisms. Microb Cell Fact 16:7. doi: 10.1186/s12934-016-0621-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Meng J, Qiu Y, Zhang Y, Zhao H, Shi S. 2023. CMI: CRISPR/Cas9 based efficient multiplexed integration in Saccharomyces cerevisiae. ACS Synth Biol 12:1408–1414. doi: 10.1021/acssynbio.2c00591 [DOI] [PubMed] [Google Scholar]
- 19. Gurdo N, Novelli Poisson GF, Juárez ÁB, Rios de Molina MC, Galvagno MA. 2018. Improved robustness of an ethanologenic yeast strain through adaptive evolution in acetic acid is associated with its enzymatic antioxidant ability. J Appl Microbiol 125:766–776. doi: 10.1111/jam.13917 [DOI] [PubMed] [Google Scholar]
- 20. Sui Y, Qi L, Wu J-K, Wen X-P, Tang X-X, Ma Z-J, Wu X-C, Zhang K, Kokoska RJ, Zheng D-Q, Petes TD. 2020. Genome-wide mapping of spontaneous genetic alterations in diploid yeast cells. Proc Natl Acad Sci USA 117:28191–28200. doi: 10.1073/pnas.2018633117 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Zhang K, Zheng D-Q, Sui Y, Qi L, Petes TD. 2019. Genome-wide analysis of genomic alterations induced by oxidative DNA damage in yeast. Nucleic Acids Res 47:3521–3535. doi: 10.1093/nar/gkz027 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Krokan H-E, Bjørås M. 2013. Base excision repair. Cold Spring Harb Perspect Biol 5:a012583. doi: 10.1101/cshperspect.a012583 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Mitra S, Boldogh I, Izumi T, Hazra TK. 2001. Complexities of the DNA base excision repair pathway for repair of oxidative DNA damage. Environ Mol Mutagen 38:180–190. doi: 10.1002/em.1070 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Yan C, Wang Y-K, Xiong Y-R, Zhou X-Q, Fang Y-C, Nie R-T, Ye C, Zhang K, Zheng D-Q. 2025. The roles of NHEJ and TLS pathways in genomic alterations and phenotypic evolution in the yeast Yarrowia lipolytica. Appl Microbiol Biotechnol 109:183. doi: 10.1007/s00253-025-13575-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Zheng D-Q, Wang Y-T, Zhu Y-X, Sheng H, Li K-J, Sui Y, Zhang K. 2022. Uncovering bleomycin‑induced genomic alterations and underlying mechanisms in the yeast Saccharomyces cerevisiae. Appl Environ Microbiol 88:e0170321. doi: 10.1128/AEM.01703-21 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Kiktev DA, Sheng Z, Lobachev KS, Petes TD. 2018. GC content elevates mutation and recombination rates in the yeast Saccharomyces cerevisiae. Proc Natl Acad Sci USA 115:E7109–E7118. doi: 10.1073/pnas.1807334115 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Nelson J-R, Lawrence C-W, Hinkle D-C. 1996. Deoxycytidyl transferase activity of yeast REV1 protein. Nature 382:729–731. doi: 10.1038/382729a0 [DOI] [PubMed] [Google Scholar]
- 28. Johnson R-E, Prakash S, Prakash L. 1999. Efficient bypass of a thymine-thymine dimer by yeast DNA polymerase, Poleta. Science 283:1001–1004. doi: 10.1126/science.283.5404.1001 [DOI] [PubMed] [Google Scholar]
- 29. Prakash S, Prakash L. 2000. Nucleotide excision repair in yeast. Mutat Res 451:13–24. doi: 10.1016/s0027-5107(00)00037-3 [DOI] [PubMed] [Google Scholar]
- 30. Bai Y-T, Sharma A, Xiang Q, Tian L-Y, Li K-J, Guo B-Y, Qi L, Zheng D-Q. 2025. Genomic alterations of marine yeast Scheffersomyces spartinae under spontaneous and mutagenic conditions. BMC Genomics 26:297. doi: 10.1186/s12864-025-11479-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Xiong Y-R, Fang Y-C, He M, Li K-J, Qi L, Sui Y, Zhang K, Wu X-C, Meng L, Li O, Zheng D-Q. 2025. Patterns of spontaneous and induced genomic alterations in Yarrowia lipolytica. Appl Environ Microbiol 91:e0167824. doi: 10.1128/aem.01678-24 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Vijayan N, Joshi S, Sarath P, Nishant KT. 2025. Loss of Heterozygosity associated with ubiquitous environments in yeast. PLoS Genet 21:e1011692. doi: 10.1371/journal.pgen.1011692 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Sampaio NMV, Ajith VP, Watson RA, Heasley LR, Chakraborty P, Rodrigues-Prause A, Malc EP, Mieczkowski PA, Nishant KT, Argueso JL. 2020. Characterization of systemic genomic instability in budding yeast. Proc Natl Acad Sci USA 117:28221–28231. doi: 10.1073/pnas.2010303117 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. James TY, Michelotti LA, Glasco AD, Clemons RA, Powers RA, James ES, Simmons DR, Bai F, Ge S. 2019. Adaptation by loss of heterozygosity in Saccharomyces cerevisiae clones under divergent selection. Genetics 213:665–683. doi: 10.1534/genetics.119.302411 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Zhang K, Sui Y, Li W-L, Chen G, Wu X-C, Kokoska RJ, Petes TD, Zheng D-Q. 2022. Global genomic instability caused by reduced expression of DNA polymerase ε in yeast. Proc Natl Acad Sci USA 119. doi: 10.1073/pnas.2119588119 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Symington L-S, Rothstein R, Lisby M. 2014. Mechanisms and regulation of mitotic recombination in Saccharomyces cerevisiae. Genetics 198:795–835. doi: 10.1534/genetics.114.166140 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Jinks-Robertson S, Petes T-D. 2021. Mitotic recombination in yeast: what we know and what we don’t know. Curr Opin Genet Dev 71:78–85. doi: 10.1016/j.gde.2021.07.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Lisby M, Rothstein R. 2015. Cell biology of mitotic recombination. Cold Spring Harb Perspect Biol 7:a016535. doi: 10.1101/cshperspect.a016535 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. St Charles J, Petes TD. 2013. High-resolution mapping of spontaneous mitotic recombination hotspots on the 1.1 Mb arm of yeast chromosome IV. PLoS Genet 9:e1003434. doi: 10.1371/journal.pgen.1003434 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Yin Y, Petes T-D. 2013. Genome-wide high-resolution mapping of UV-induced mitotic recombination events in Saccharomyces cerevisiae. PLoS Genet 9:e1003894. doi: 10.1371/journal.pgen.1003894 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Liu L, Malkova A. 2022. Break-induced replication: unraveling each step. Trends Genet 38:752–765. doi: 10.1016/j.tig.2022.03.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Zhu Y-X, Li K-J, He M, Zhang K, Zheng D-Q, Petes TD. 2025. Mitotic recombination events and single-base mutations induced by ultraviolet light in G1-arrested yeast cells. Proc Natl Acad Sci USA 122:e2518046122. doi: 10.1073/pnas.2518046122 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Zheng D-Q, Wang P-M, Chen J, Zhang K, Liu T-Z, Wu X-C, Li Y-D, Zhao Y-H. 2012. Genome sequencing and genetic breeding of a bioethanol Saccharomyces cerevisiae strain YJS329. BMC Genomics 13:479. doi: 10.1186/1471-2164-13-479 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Impellizzeri KJ, Anderson B, Burgers PM. 1991. The spectrum of spontaneous mutations in a Saccharomyces cerevisiae uracil-DNA-glycosylase mutant limits the function of this enzyme to cytosine deamination repair. J Bacteriol 173:6807–6810. doi: 10.1128/jb.173.21.6807-6810.1991 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Nash HM, Bruner SD, Schärer OD, Kawate T, Addona TA, Spooner E, Lane WS, Verdine GL. 1996. Cloning of a yeast 8-oxoguanine DNA glycosylase reveals the existence of a base-excision DNA-repair protein superfamily. Curr Biol 6:968–980. doi: 10.1016/s0960-9822(02)00641-3 [DOI] [PubMed] [Google Scholar]
- 46. Chen J, Derfler B, Maskati A, Samson L. 1989. Cloning a eukaryotic DNA glycosylase repair gene by the suppression of a DNA repair defect in Escherichia coli. Proc Natl Acad Sci USA 86:7961–7965. doi: 10.1073/pnas.86.20.7961 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Ramotar D, Popoff S-C, Gralla E-B, et al. 1991. Cellular role of yeast Apn1 apurinic endonuclease/3′-diesterase: repair of oxidative and alkylation DNA damage and control of spontaneous mutation. Mol Cell Biol 11:4537–4544. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Gibbs P-E, McDonald J, Woodgate R, et al. 2005. The relative roles in vivo of Saccharomyces cerevisiae Pol η, Pol ζ, Rev1 protein and Pol32 in the bypass and mutation induction of an abasic site, TT (6-4) photoadduct and TT cis-syn cyclobutane dimer. Genetics 169:575–582. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Ikeuchi K, Ivic N, Buschauer R, Cheng J, Fröhlich T, Matsuo Y, Berninghausen O, Inada T, Becker T, Beckmann R. 2023. Molecular basis for recognition and deubiquitination of 40S ribosomes by Otu2. Nat Commun 14:2730. doi: 10.1038/s41467-023-38161-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Gietz R-D, Schiestl R-H. 2007. High-efficiency yeast transformation using the LiAc/SS carrier DNA/PEG method. Nat Protoc 2:31–34. doi: 10.1038/nprot.2007.13 [DOI] [PubMed] [Google Scholar]
- 51. Gueldener U, Heinisch J, Koehler GJ, Voss D, Hegemann JH. 2002. A second set of loxP marker cassettes for Cre-mediated multiple gene knockouts in budding yeast. Nucleic Acids Res 30:e23. doi: 10.1093/nar/30.6.e23 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52. Zheng D-Q, Zhang K, Wu X-C, Mieczkowski PA, Petes TD. 2016. Global analysis of genomic instability caused by DNA replication stress in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 113:E8114–E8121. doi: 10.1073/pnas.1618129113 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53. Qi L, Li K-J, Tang X-X, Tian L-Y, Wang Y-K, He M, Zhang K, Petes TD, Zheng D-Q. 2026. Loss of Pol32, a subunit of DNA polymerases δ and ζ, leads to different patterns of genome stability than direct impairment of these individual polymerases. mBio 17:e0053126. doi: 10.1128/mbio.00531-26 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54. Bolger A-M, Lohse M, Usadel B. 2014. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30:2114–2120. doi: 10.1093/bioinformatics/btu170 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55. Li H, Durbin R. 2009. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25:1754–1760. doi: 10.1093/bioinformatics/btp324 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56. Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, Marth G, Abecasis G, Durbin R, 1000 Genome Project Data Processing Subgroup . 2009. The sequence Alignment/Map format and SAMtools. Bioinformatics 25:2078–2079. doi: 10.1093/bioinformatics/btp352 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57. Koboldt DC, Zhang Q, Larson DE, Shen D, McLellan MD, Lin L, Miller CA, Mardis ER, Ding L, Wilson RK. 2012. VarScan 2: somatic mutation and copy number alteration discovery in cancer by exome sequencing. Genome Res 22:568–576. doi: 10.1101/gr.129684.111 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Sets S1 to S5.
Data Availability Statement
The raw whole-genome sequencing data of yeast isolates were deposited in the SRA database with BioProject accession number PRJNA1463135.
