regulation of two drug resistance factors under Mrr1 and that plasticity in drug resistance regulation could be instrumental in the development of multi-drug-resistant species.

We previously characterized the C. lusitaniae clinical isolate strain U04 and its mrr1∆ derivative complemented with either MRR1 ancestral , which confers Mrr1 activity typical of most C. lusitaniae isolates, or MRR1 Y813C , which confers constitutive Mrr1 activity that renders cells resistant to FLZ (21). Published transcriptomic comparisons of these strains revealed significantly higher levels of MDR1 (CLUG_01938_39 [18]), CDR1 (CLUG_03113 [34]), and CLUG_05825 (homolog of C. albicans FLU1), all of which encode drug efflux proteins, when Mrr1 was constitutively active (21) (Fig. 1A). The MDR1, CDR1, and FLU1 (CLUG_05825) transcripts were 8.2-fold, 2.7-fold, and 1.7-fold higher, respectively, in strains with constitutive Mrr1 activity when compared to strains with low Mrr1 activity (Table S1A) (21).

To investigate if C. lusitaniae Mrr1 regulation of these transporters was direct, we analyzed Mrr1-DNA interactions. We first generated N-terminal 6×His-3×FLAG (HF)-tag ged versions of different Mrr1 variants. HF-Mrr1-encoding alleles were expressed from the native MRR1 promoter after introduction into the U04 mrr1∆ mutant background. We found that HF-Mrr1 ancestral and HF-Mrr1 Y813C were stably produced, and both were detected at a slightly higher molecular weight (150 kDa) than the predicted ~140 kDa. This band was absent in the western blot of samples from the U04 mrr1∆ strain (Fig. 1B). We did not observe any significant differences in Mrr1 levels between strains expressing HF-Mrr1 ancestral and HF-Mrr1 Y813C (Fig. 1B). We compared the activities of the HF-Mrr1 variants to their untagged counterparts by evaluating the minimum inhibitory concen tration (MIC) of FLZ in strains with either HF-tagged or untagged Mrr1 variants (Fig. 1C). The U04 mrr1∆ strain with untagged MRR1 Y813C had a FLZ MIC that was 64-fold higher than the strain with untagged MRR1 ancestral (32 µg/mL vs 0.5 µg/mL) (21). The U04 mrr1∆ strain complemented with HF-MRR1 Y813C had a similarly high FLZ MIC relative to the strain with HF-MRR1 ancestral (Fig. 1C). Thus, the N-terminal HF-tagged Mrr1 is functional.

We evaluated genome-wide binding of HF-Mrr1 Y813C using cleavage under targets and release using nuclease (CUT&RUN) in two independent experiments (37). An α-FLAG antibody (Ab) was used for the enrichment of HF-Mrr1-bound DNA, and an IgG Ab was used to assess non-specific binding. The recovered DNA was sequenced and aligned to the genome of C. lusitaniae strain L17 (NCBI accession: ASM367555v2). Both U04 and L17 were isolated from the same clinical sample and differ by only ~108 single-nucleotide polymorphisms and ~130 insertions/deletions (18). The genome of L17 was utilized as it is a highly accurate genome produced by sequencing and assembly of reads obtained using Oxford Nanopore long-read and Illumina technologies. Genomic regions that showed significant fold enrichment in DNA recovered from the α-FLAG when compared to the IgG control of each sample are represented as peaks and indicate HF-Mrr1 Y813C interaction sites (Fig. 1D). The average enrichment of reads in α-FLAG relative to the IgG background within an identified peak region was quantified as peak signal (38). Peaks were filtered using a peak signal cutoff of 2, a false discovery rate (FDR) of <0.05, and a genomic position within 1 kb of an open reading frame (ORF). Approximately 329 CUT&RUN peaks were identified (File S1).

The upstream regions of MDR1, CDR1, and FLU1 all showed strong evidence for Mrr1 Y813C binding. The upstream region of the MDR1 ORF containing its promoter had a significant HF-Mrr1 Y813C peak with an average signal of 10.5 (Fig. 1D). The HF-Mrr1 Y813C peak associated with the MDR1 ORF spanned ~1.5 kb and extended into the neighboring coding regions of MDR1 (Fig. 1D). Thus, as in C. albicans (20,39), C. lusitaniae Mrr1 bound directly upstream of MDR1. The regions upstream of CDR1 also had a significantly enriched CUT&RUN peak with an average signal of 6.7 and a peak width of ~1.2 kb (Fig. 1E). An Mrr1 binding peak was similarly found upstream of the gene encoding Flu1 (Fig. 1F). The peaks associated with the FLU1 ORF had an average signal of 4.8 and covered a length of ~0.8 kb (Fig. 1F). The signal of the HF-Mrr1 Y813C peak upstream of the MDR1 ORF was 1.5-and 2.1-fold higher than upstream of the CDR1 and FLU1 ORFs. Together, these data are consistent with previous reports of Mrr1 regulation of MDR1 and provide evidence for direct regulation of CDR1 and FLU1 by Mrr1 in C. lusitaniae.

To investigate the phenotypic consequences of Mrr1 regulation of MDR1, CDR1, and FLU1, we determined the concentrations of various azoles required to inhibit 90% (MIC 90 ) of the growth of strain U04 with Mrr1 Y813C and its mdr1∆, cdr1∆, and flu1∆ derivatives. As the C. albicans homologs of Mdr1, Cdr1, and Flu1 were all capable of FLZ efflux (29,30,40), we first evaluated the FLZ MIC 90 . The U04 strain expressing Mrr1 Y813C had a 32-fold higher FLZ MIC 90 than the isogenic strain with Mrr1 ancestral (Fig. 2A; Table 1). The Mrr1 Y813C mdr1∆ mutant exhibited an eightfold lower FLZ MIC 90 than its parent mrr1∆+MRR1 Y813C strain (4 µg/mL vs 32 µg/mL; Fig. 2A; Table 1). Although the FLZ MIC 90 values were unchanged in a cdr1∆ and flu1∆ mutant, the flu1∆ mutant grew slightly less well than the parent strain across concentrations (Fig. 2A; Table 1).

Similar Mdr1-dependent resistance was observed for the other short-tailed azole voriconazole (VOR) in strains with constitutively active Mrr1; the mrr1∆+MRR1 Y813C strain had a 32-fold higher VOR MIC 90 than the mrr1∆+MRR1 ancestral strain (0.5 µg/mL vs 0.0156 µg/mL; Fig. 2B; Table 1). While the mdr1∆ mutation resulted in a fourfold lower VOR MIC 90 than the mrr1∆+MRR1 Y813C and the U04 WT (MRR1 Y813C ) strains, no difference in MIC 90 was observed for the cdr1∆ and flu1∆ mutants (Fig. 2B; Table 1). Overall, strains expressing constitutively active Mrr1 exhibited similar Mdr1-mediated resistance to the triazoles FLZ and VOR. Interestingly, while MDR1 was necessary for short-tailed azole resistance, MDR1 deletion alone was not sufficient to abrogate resistance as the mdr1∆ mutant still had a fourfold higher FLZ (4 µg/mL vs 1 µg/mL) and eightfold higher VOR (0.125 µg/mL vs 0.0156 µg/mL) MIC 90 values than the mrr1∆+MRR1 ancestral strain (Fig. 2A and B; Table 1). Hence, we tested the FLZ MIC 90 of an mdr1∆cdr1∆ double deletion derivative of the U04 WT (MRR1 Y813C ) strain. The mdr1∆cdr1∆ mutant had a 32-fold lower FLZ MIC 90 value than the mrr1∆+MRR1 Y813C mdr1∆ (0.125 µg/mL vs 4 µg/mL) strain, suggesting that Cdr1 also contributed to FLZ resistance in strains with constitutive Mrr1 activity (Fig. S1; Table 2). Absence of both MDR1 and CDR1 in U04 WT (MRR1 Y813C ) reduced the FLZ MIC 90 from 32 µg/mL to 0.125 µg/mL (Fig. S1; Table 2). Thus, our results suggest that both Mdr1 and Cdr1 mediate resistance to fluconazole in strains with constitutive Mrr1 activity.

We evaluated the susceptibility of the different strains to the long-tailed azoles: ketoconazole (KTZ), itraconazole (ITZ), and isavuconazole (ISA). The mrr1∆+MRR1 Y813C strain had a 32-, 8-, and >16-fold higher MIC 90 values for KTZ, ITZ, and ISA, respectively, than the mrr1∆+MRR1 ancestral strain (Fig. 2C through E; Table 1). Consistent with prior reports of Cdr1-mediated resistance to long-tailed azoles (34,41), the cdr1∆ strain had a >8-fold reduction in MIC 90 values for KTZ, ITZ, and ISA than the mrr1∆+MRR1 Y813C parental strain (Fig. 2C through E; Table 1). The mdr1∆ and flu1∆ strains were not more susceptible to the tested long-tailed azoles than the parent strain (Fig. 2C through E; Table 1). These data indicate that constitutively active Mrr1 confers resistance to long-tailed azoles via Cdr1.

Transporter-mediated efflux of other antifungal compounds of agricultural and clinical relevance has been demonstrated (42), and strains with constitutive Mrr1 activity exhibited broad-spectrum resistance against multiple toxic substrates in a Biolog Phenotype Microarray screen (21). Thus, we evaluated the MICs of 5-flucytosine (5-FC), cycloheximide, myclobutanil, terbinafine, and fluphenazine for mrr1∆+MRR1 ancestral and mrr1∆+MRR1 Y813C strains. Here, MIC was defined as the concentration at which no visible growth was observed. The mrr1∆+MRR1 Y813C strain had a 2-to 32-fold increase in the MIC values of the different tested antifungals compared to the mrr1∆+MRR1 ancestral strain (Fig. 3A). Furthermore, in the mrr1∆+MRR1 Y813C strain background, the mdr1∆ derivative

TABLE 1 Clinical azole MIC 90 values a Strain MIC 90 [µg/mL (fold difference)] FLZ VOR KTZ ITZ ISA U04 WT (MRR1 Y813C ) 32 (32) 0.5 (32) 0.25 (32) 0.05 (16) 0.0625 (>16) U04 mrr1∆+MRR1 ancestral 1 (1) 0.0156 (1) 0.0078 (1) 0.003125 (1) <0.0039 (1) U04 mrr1∆+MRR1 Y813C 32 (32) 0.5 (32) 0.25 (32) 0.025 (8) 0.0625 (>16) U04 mrr1∆+MRR1 Y813C mdr1∆ 4 (4) 0.125 (8) 0.5 (64) 0.05 (16) 0.125 (>32) U04 mrr1∆+MRR1 Y813C cdr1∆ 32 (32) 0.5 (32) 0.016 (2) 0.003125 (1) <0.0039 (1) U04 mrr1∆+MRR1 Y813C flu1∆ 32 (32) 0.5 (32) 0.25 (32) 0.05 (16) 0.25 (>64) a MIC 90 values the indicated strains were calculated using broth microdilution assays. MIC 90 was defined as the concentration at which 90% growth was inhibited. Fold difference in MIC 90 relative to the azole-sensitive U04 mrr1∆+MRR1 ancestral strain is presented within parentheses. FLZ, fluconazole; VOR, voriconazole; KTZ, ketoconazole; ITZ, itraconazole; and ISA, isavuconazole.

TABLE 2 FLZ MIC 90 values a Strain FLZ MIC 90 [µg/mL (fold reduction)] U04 WT (MRR1 Y813C ) 32 (1) U04 mrr1∆+MRR1 Y813C mdr1∆ 4 (8) U04 mrr1∆+MRR1 Y813C cdr1∆ 32 (1) U04 WT (MRR1 Y813C ) mdr1∆ cdr1∆ 0.125 (256)

a MIC 90 values of the indicated strains were calculated using broth microdilution assays. MIC 90 was defined as the concentration at which 90% growth was inhibited. Fold reduction in MIC 90 relative to the azole-resistant U04 (MRR1 Y813C ) strain is presented within parentheses.

resulted in increased susceptibility to 5-FC, cycloheximide, and myclobutanil (Fig. 3B).

The MIC values of cycloheximide and 5-FC decreased by fourfold in the mdr1∆ mutant (Fig. 3B); support for Mdr1-mediated resistance against the pyrimidine analog 5-FC has been previously shown in C. lusitaniae (12,25,34). While the protein synthesis inhibitor cycloheximide was a substrate of both Mdr1 and Cdr1 (29) in C. albicans, the cdr1∆ mutation did not alter the cycloheximide resistance of the mrr1∆+MRR1 Y813C strain. For the agricultural triazole myclobutanil, the mdr1∆ and cdr1∆ mutants had twofold to fourfold lower MIC values than the mrr1∆+MRR1 Y813C parental strain (2-4 µg/mL vs 8 µg/mL) (Fig. 3B). However, both still had eightfold higher MIC values than the mrr1∆+MRR1 ancestral strain (2-4 µg/mL vs 0.25 µg/mL), suggesting that other Mrr1 targets contributed to myclobutanil resistance. FIG 4 The Mrr1 regulon of C. lusitaniae. (A) Venn diagram shows the overlap between differentially expressed (DE) genes from RNA-seq in Demers et al. (21) and ORFs with HF-Mrr1 Y813C peaks in their intergenic regions from CUT&RUN. The 25 differentially regulated genes that have HF-Mrr1 Y813C peaks in either the 5′ or 3′ regions are listed in Table S1A. (B, C) HF-Mrr1 Y813C CUT&RUN read coverage plots normalized per 20 bp bin size. Chromosomal positions of regions containing MGD1 and MRR1 and adjacent genes are represented to scale with boxes and arrows. Peaks from HF-Mrr1 Y813C -bound DNA recovered by an α-FLAG antibody and for the non-specific binding control recovered via IgG are shown. The signal indicates the average read density in α-FLAG relative to IgG within the peak region. Two independent experiments were performed, and the results of both are shown.

Susceptibility of other antifungals was dependent on Cdr1. The MICs for the allylamine antifungal terbinafine and the antipsychotic fluphenazine were lower in the cdr1∆ mutant. Since FLU1 deletion made drug-sensitive C. albicans hypersusceptible to the metabolic inhibitor mycophenolic acid (MPA) (40), we also investigated the MPA susceptibility of our strains. Despite the mrr1∆+MRR1 Y813C having a twofold to fourfold increase in MPA MIC relative to the mrr1∆+MRR1 ancestral strain, its MIC was not impacted by deletion of FLU1. The mdr1∆ and cdr1∆ mutants were also not more susceptible to MPA (Fig. 3B). Taken together, our results show that constitutive Mrr1 activity conferred resistance to a broad spectrum of antifungals, largely through its control of Mdr1 and Cdr1, with evidence for redundancy in Mrr1-regulated antifungal resistance mechanisms.

To examine other genes that were co-regulated with MDR1, CDR1, and FLU1, we identified additional genes that were differentially expressed due to a direct conse quence of constitutive Mrr1 activity. There were 25 genes that were differentially expressed when Mrr1 was constitutively active (Mrr1 Y813C ) compared to mrr1∆ and low-activity Mrr1 (21) (FDR < 0.05 and fold change ≥ 1.5) and had an HF-Mrr1 Y813C peak located within 1 kb from their ORF regions, including MDR1, CDR1, and FLU1 (Fig. 4A; Table S1A). These 25 genes will be referred to as the C. lusitaniae Mrr1 regulon (Table S1A). Slim Gene Ontology (GO) analysis of the C. albicans homologs of the C. lusitaniae Mrr1 regulon genes found transport, response to chemicals, response to stress, and cellular homeostasis as the most enriched biological process terms (Table S1B). The Mrr1 regulon included two putative peptide transporters (OPT1 and OPT5), two extracellular cell wall proteins (ECM33 and CSA1), two involved in metal homeostasis (CTR2 and CFL4), a putative glycerol transporter (HGT10/STL1), an alternative oxidase (AOX2), and multiple metabolic enzymes or putative oxidoreductases (Table S1A). Of note, the 77 indirect Mrr1 targets (Fig. 4A) were further enriched for transport, chemical, and stress response processes in a Slim GO analysis of their C. albicans homologs (File S2).

We previously showed that the C. lusitaniae Mrr1 induced MGD1 and MGD2 in the presence of exogenous MGO (24), a toxic 2-oxo-aldehyde released by metabolically dysregulated cells and activated macrophages at sites of infection (43). Furthermore, upregulation of MGD1 and MGD2 by constitutively active Mrr1 conferred a growth advantage in the presence of MGO (24). Methylglyoxal dehydrogenases are co-regulated with MDR1 in several other Candida spp., including C. albicans and C. auris (18,21,23,25,26,28). Interestingly, despite high expression of both MGD1 and MGD2 transcripts in strains with Mrr1 Y813C (21), only the promoter regions of MGD1 had a HF-Mrr1 Y813C CUT&RUN peak with an average signal of 6.5 (Fig. 4B). A HF-Mrr1 Y813C peak of average signal 2.8 was also present in the promoter regions of MRR1 (Fig. 4C), indicating a mechanism for potential positive self-regulation of MRR1 transcripts, which is consistent with previously published RNA-seq data (21). Three Mrr1 regulon genes (CLUG_04865, CLUG_01574, and CLUG_04429) had no clear homologs in C. albicans, but did have homologs in the more closely related C. auris. Although not differentially expressed in the U04 transcriptome, a putative alcohol dehydrogenase (CLUG_00171) and a putative phospholipase C (CLUG_01152) had HF-Mrr1 Y813C peaks in their promoter regions and were upregulated in the clinical C. lusitaniae P3 isolate with an MRR1 V668G GOF allele (25). Five genes were less abundant in strains with activated Mrr1 (Table S1A); one of these, CLUG_01020 (STL1), was the only locus with an HF-Mrr1 Y813C peak in its 1 kb downstream intergenic region with no peak in its upstream region (Table S1A).

To better understand direct Mrr1 regulation of targets, we used the STREME algorithm (44) to determine if specific motifs were enriched within sequences corresponding to 329 HF-Mrr1 Y813C CUT&RUN peaks (File S1). The 100 bp sequences upstream and downstream of peak summits (the most enriched point within an identified peak) were used as input for discriminative de novo motif discovery (44). A set of sequences chosen at random from the C. lusitaniae L17 genome and matched in length and number was used as background to identify enriched motifs in the input set. A 14-nucleotide (nt) consensus sequence RCGGAGWTARSVNN was the topmost motif predicted by STREME (Fig. 5A).

When this consensus sequence was scanned for in the upstream intergenic regions of MDR1 from C. lusitaniae L17 and ATCC 42720 (ASM383v1), the motif was observed seven times in the C. lusitaniae L17 MDR1 promoter region, and six of these were conserved in C. lusitaniae strain ATCC 42720 (Fig. S2A). The 14-nt consensus sequence had substantial nucleotide ambiguity at both ends (positions 1 and 11-14) (Fig. 5A). Therefore, for subsequent motif analyses, we focused on the internal 9-nt CGGAGWTAR motif (Fig. 5A, boxed). The 9-nt CGGAGWTAR motif and the 14-nt RCGGAGWTARSVNN motif were similarly detected in the promoter regions of MDR1 in both strains (Fig. S2A). Henceforth, we refer to the 9-nt CGGAGWTAR motif as the consensus Mrr1-binding DNA motif (cMBM) (Fig. 5A). We determined if the cMBM sites in the MDR1 promoter were conserved in other C. lusitaniae strains; all six cMBM sites were found in the clinical isolate AR0398 (GCA_032599225.1) and distinct environmental isolates 79-1 (GCA_032599145.1) and 76-31 (GCA_032599085.1) (45). Each of the cMBMs was at identical positions and orientations relative to the MDR1 translational start sites across the different strains (Fig. 5B). cMBMs were also found upstream of CDR1 and FLU1 (Fig. S2B; File S3). The cMBMs upstream of CDR1 were conserved in position in both L17 and ATCC 42720 strains despite differences in the length of the CDR1 adjacent intergenic regions (Fig. S2B). At least one cMBM, and often multiple cMBMs, was found within the peak spanning regions associated with all but two of the genes in the Mrr1-regulon (File S3). Furthermore, the upstream intergenic regions of the 25 direct Mrr1 targets (Fig. 4A) had a significant enrichment of cMBMs when compared to the intergenic regions of genes that were indirect targets (CUT&RUN peak absent but differentially expressed in RNA-seq) (File S2), and negative control genes that were not present in either the Mrr1 CUT&RUN or RNA-seq data sets (Fig. S2C).

Mrr1 binding to the predicted cMBM was evaluated in vitro using analytical size-exclu sion chromatography (SEC) and electrophoretic mobility shift assays (EMSA). For this, we heterologously expressed and purified the N-terminal 1-196 amino acid region, which encompassed the Zn 2 -Cys 6 motif capable of DNA binding. We chose a 50 bp region, from -734 to -684, upstream of the translational start site of the C. lusitaniae L17 MDR1, with two predicted cMBMs, as the DNA probe (Fig. 5C). Three samples-Mrr1 1-196 , 50 bp DNA probe, and the Mrr1 1-196 and 50 bp probe mixture were individually evaluated for their size/shape-based separation in SEC with detection of DNA and/or protein by monitoring absorbance at 280 nm (A 280 ). Protein standards of different molecular weights were also analyzed (Fig. 5D). The A 280 peak of the protein-DNA mixture appeared earlier than the A 280 peaks of the DNA-only and Mrr1 1-196 -only samples (Fig. 5D). SDS-PAGE analysis confirmed the presence of Mrr1 1-196 in the earlier eluted fractions when DNA was present, suggesting the formation of a higher molecular weight Mrr1-DNA complex (Fig. S3).

We used 30 bp DNA probes that contained either wild-type (cMBM probe) or mutated cMBM (mut-cMBM probe) for EMSA (Fig. 5C). When the cMBM probe was titrated with increasing concentrations of Mrr1 1-196 , a shift in the mobility of the probe was detected which corresponded to the Mrr1 1-196 -cMBM probe complex (Fig. 5E). Furthermore, at higher concentrations of Mrr1 1-196 , a decrease in the signal of the unbound cMBM probe was observed (Fig. 5E). The mut-cMBM probe was mutated in five of the nine cMBM nucleotides, including the highly conserved CGG (positions 1-3) and T (position 7) (Fig. 5C). Mrr1 1-196 did not induce a shift of the mut-cMBM probe, suggesting that the mutations in the cMBM eliminated formation of a Mrr1 1-196 -DNA complex (Fig. 5E). Thus, our data suggest that cMBM is sufficient for Mrr1 binding to the DNA, and that the highly conserved residues within the cMBM are necessary for this interaction.

We also scanned for the cMBM in the promoter sequences of the MDR1 and CDR1 homologs in Candida spp. At least three copies of cMBM were found in the MDR1 and CDR1 promoter sequences of C. albicans and C. parapsilosis, and one cMBM in C. auris (Fig. 5F; Fig. S2D). In the case of C. albicans, two cMBMs occurred in locations previously annotated to be important for MDR1 transcriptional regulation. These cMBMs were discovered between the -200 to -400 regions, which encompassed the benomyl response element (-260 and -296) (46) and the Mrr1-binding region that contained the C. albicans Mrr1-binding DNA motif DCSGHD (-342 to -492) (39). In a chromatin immunoprecipitation-quantitative reverse transcription PCR (ChIP-qRT) analysis of Mrr1 binding to the C. albicans MDR1 promoter, DNA recovery was highest at these cMBMcontaining regions relative to the rest of the MDR1 promoter sequence (20). Together, these data strongly suggest that the consensus Mrr1-binding DNA motif discovered in C. lusitaniae is conserved in other Candida species.

Previous studies on C. lusitaniae Mrr1 suggested that expression at some loci (e.g., MDR1 and MGD1) (21, 24, 25) was repressed by low-activity Mrr1 variants and induced in the presence of benomyl and MGO inducers of Mrr1 or by constitutively active Mrr1 variants. Thus, we compared the DNA localization of the HF-Mrr1 Y813C to the genomewide binding of low activity HF-Mrr1 ancestral in the absence of Mrr1-inducing stimuli. Using the same parameters as for the analysis of HF-Mrr1 Y813C , we found around 1,276 peaks associated with HF-Mrr1 ancestral -bound DNA (File S4). The MDR1 intergenic region revealed a significant HF-Mrr1 ancestral peak that spanned a region of ~1.6 kb and had a signal of 15.1 (Fig. 6A). HF-Mrr1 ancestral peaks were also found upstream of CDR1 and FLU1 (1.6 and 2.2 peak signals, respectively; Fig. 6B and C). Comparison of HF-Mrr1 ancestral and HF-Mrr1 Y813C -bound sites upstream of MDR1, CDR1, and FLU1 exhibited a striking similarity in their peak profiles (Fig. S4A through C). The remarkable overlap of HF-Mrr1 ancestral and HF-Mrr1 Y813C CUT&RUN peaks present in over 930 genomic locations (Fig. 6D) suggests that Mrr1-mediated repression and induction are not due to differences in Mrr1 localization to the DNA.

In Demers et al. (21), we characterized MRR1 alleles with GOF mutations that resulted in constitutive activity and Mdr1-dependent FLZ resistance (Fig. 7A), as well as alleles with both GOF mutations and secondary suppressor mutations that restored the inducible low activity state, such as MRR1 L1191H+Q1197*(L1Q1*) (Fig. 7A). The mrr1∆+MRR1 L1Q1* strain had more than a 32-fold lower FLZ MIC value (0.125 µg/mL vs 32 µg/mL) than strains with MRR1 GOF alleles (MRR1 Y813C and MRR1 L1191H ) (Fig. 7B). Since GOF muta tions in Mrr1 did not affect DNA localization, we evaluated whether secondary sup pressor mutation(s) altered these interactions by performing CUT&RUN on U04 mrr1∆ strains expressing HF-Mrr1 L1Q1* from its endogenous promoter. Western blot confirmed that the truncated HF-Mrr1 L1Q1* was present at levels similar to that of the full-length HF-Mrr1 ancestral and HF-Mrr1 Y813C (Fig. S5A). The HF-tag did not affect Mrr1 L1Q1* activity as strains expressing tagged Mrr1 L1Q1* exhibited similar 32-to 64-fold lower FLZ MIC as untagged Mrr1 L1Q1* when compared to strains expressing the constitutively active Mrr1 Y813C variant (Fig. S5B). Our CUT&RUN analysis found HF-Mrr1 L1Q1* -bound DNA to be significantly enriched in the upstream intergenic regions of MDR1, CDR1, and FLU1 ORFs (Fig. 7C through E) with a peak profile identical to HF-Mrr1 ancestral and HF-Mrr1 Y813C . The HF-Mrr1 L1Q1* peak recapitulated the 1.5-and 2-fold higher signal upstream of MDR1 relative to CDR1 and FLU1. Across the entire C. lusitaniae genome, the HF-Mrr1 L1Q1* -bound genomic sites (see File S5 for peaks) were strikingly similar to the HF-Mrr1 ancestral and HF-Mrr1 Y813C -bound sites, suggesting that secondary suppressor mutation(s) do not likely impact Mrr1 localization to the DNA (Fig. S6). Hence, our results illustrate that Mrr1 localization at the C. lusitaniae DNA is unaltered by the tested mutations and is independent of Mrr1 activation state.

In this study, we demonstrated that constitutively active C. lusitaniae Mrr1 directly upregulates several multi-drug transporter-encoding genes, including MDR1 and CDR1, leading to reduced susceptibility to both short-tailed and long-tailed azoles and other antifungals (Fig. 1 to 3). The coordinated regulation of both MDR1 and CDR1 by Mrr1 in C. luistaniae differs from their regulation in the well-studied species C. albicans, wherein Mrr1 is the primary regulator of MDR1 and Tac1 is the main CDR1 transcriptional activator (23,33). We identified a consensus Mrr1-binding motif (cMBM; CGGAGWTAR) that colocalized with Mrr1 CUT&RUN peaks and that was present in multiple positions within the peaks found in regions adjacent to C. lusitaniae MDR1, CDR1, and in almost all other Mrr1-regulated genes (Fig. 5A; File S3). The cMBM sequences in the Mrr1 peak regions upstream of MDR1 and CDR1 were conserved in other C. lusitaniae strains (Fig. 5B; Fig. S2B). Furthermore, the cMBM was also enriched in the regions upstream of MDR1 homologs in C. albicans, C. auris, and C. parapsilosis, and in C. albicans, the cMBM was present in regions shown to bind C. albicans Mrr1 (Fig. 5F) (20). The cMBM was also upstream of C. lusitaniae and C. parapsilosis CDR1 (Fig. S2D), which is consistent with reports that constitutive Mrr1 activity also induces expression of CDR1 in these species. Moreover, we noted the presence of cMBMs in regions upstream of CDR1 in species that The shapes indicate the type of experimental data used to support the model, including protein-DNA studies (this study; 20,38,49), expression and phenotypic studies (18-21, 23-28, 32-35, 41, 48, 49, 51-53). (B) Possible mechanism(s) for gene induction by Mrr1 based on published studies (20,39,47,55,(57)(58)(59). The mechanisms that impact Mrr1-mediated gene expression may vary between promoters and conditions within a strain, and there may be differences across strains and species.

have no reports for Mrr1 regulation of CDR1, including C. albicans and C. auris (Fig. S2D) (33,47,48). Consistent with the potential for Mrr1 regulation of CDR1 in C. albicans, a ChIP-ChIP analysis detected Mrr1 in the upstream regions of CDR1 (49), though CDR1 was not reported as a Mrr1 target because its expression was not increased by constitutively active Mrr1.

Studies in C. albicans and recent work in C. auris have shown that Tac1 with an activating mutation upregulates CDR1 expression, and the C. albicans Tac1 regulates CDR1 by binding a consensus CGGN 4 CGG motif in the promoter region (49). Though C. lusitaniae has a Tac1 homolog (Clug_02369) (34) and a CGGN 4 CGG motif at -761 in the CDR1 promoter region (data not shown), strains with low Mrr1 activity and a cdr1∆ mutant had similar susceptibilities (MIC 12.5-25 µg/mL) to the Cdr1 substrate fluphenazine (Fig. 3) (22,50). In fact, while activating mutations in TAC1 have been character ized in FLZ-resistant C. parapsilosis (51) and C. auris (36), to our knowledge, there are no reports on activating mutations in the TAC1 gene leading to FLZ resistance in C. lusitaniae. While we (21) and others (34) have shown that Mrr1 is sufficient to upregulate C. lusitaniae CDR1, Tac1 may induce CDR1 under conditions not tested in this study. For instance, estradiol is an inducer of Tac1-mediated CDR1 expression in C. albicans (52,53). Together, these data underscore the evolutionary plasticity in transporter regulation in Candida spp. through the adoption of targets from one transcriptional circuit to another (Fig. 8A; 54-56). In the case of C. lusitaniae, the coordinated regulation of drug efflux proteins may be a mechanism for cross-resistance to multiple antifungals and may promote the development of other resistance mutations through a reduction in drug susceptibility.

Our data on Mrr1 levels and Mrr1 variants binding to upstream, or in some cases downstream, regions of Mrr1-regulated genes provide insight into Mrr1 regulation. First, we found that Mrr1 variants with differing activities did not have differences in total protein levels (Fig. S5A). Second, activated Mrr1 and inducible but inactive Mrr1 had indistinguishable localization at all cMBMs (Fig. S6), which is consistent with ChIP-qRT analysis of Mrr1 interactions with the MDR1 promoter region in C. albicans (20). Third, the subset of genes repressed by inducible but inactive Mrr1 had similar Mrr1 localization in their promoter regions as those genes that were not repressed by Mrr1 (Fig. S4A). Thus, Mrr1 is likely regulated through mechanisms, such as induced conformational change by ligand, co-factor binding (60), phosphorylation (61), or differential activity of co-regulatory proteins. These mechanisms are not mutually exclusive (57). In C. albicans, changes in activity of coregulatory proteins (Cap1 or Mcm1 [39,58]), mediator (20,53) or chromatin remodeling complexes like the Swi/Snf complex influence Mrr1 induction of MDR1 and other genes (20) (Fig. 8B). The >1 kb width of our CUT&RUN peaks is consistent with the presence of multiple cMBMs in regions adjacent to Mrr1-regulated genes and may also reflect the presence of co-regulators or chromatin remodeling complexes that could influence micrococcal nuclease access to DNA. The involvement of multiple mechanisms allows for the controlled and differential expression of unique gene subsets in different strain backgrounds (12) in response to environmental cues that may be present in an infection environment (e.g., decreased nutrient availability or metabolites like methylglyoxal or inflammatory molecules). The presence of diverse regulatory mechanisms may promote survival under diverse conditions and may also promote the evolution of novel regulatory circuits across species and even strains (Fig. 8A).

For many azoles, there was an 8-to 16-fold increase in the MIC values of strains with activated Mrr1 variants compared to strains with low-activity Mrr1, and these differences were primarily dependent on either Mdr1 (FLZ and VOR) or Cdr1 (KTZ, ITZ, and ISA) (Fig. 2 and Table 1). The contribution of Cdr1 toward fluconazole resistance became evident only in the absence of Mdr1 (Fig. S1; Table 2). These data are consistent with published data in another C. lusitaniae strain with an activated Mrr1 (P3), which showed that the mdr1∆cdr1∆ double mutant was much more susceptible to FLZ, VOR, and ITZ than the single mdr1∆ and cdr1∆ deletion mutants (34). Redundancy in transporter efflux was also observed in the case of other broad-spectrum antifungals. In addition to Mdr1 and Cdr1, susceptibility to the tested antifungals could also be mediated by other efflux pumps in the Mrr1 regulon, including the MFS family transporter Flu1. While FLU1 is a conserved Mrr1 target in other Candida spp., such as C. albicans (17) and C. parapsilo sis (35), the promiscuity for substrates between transporters may have concealed any apparent contribution of Flu1 to efflux in an MDR1/CDR1 overexpression strain (Fig. 2 and 3; Table 1). Beyond drug efflux, the C. lusitaniae Mrr1 regulon (Table S1A) included genes involved in other transporter activities like oligopeptide transport (OPT1), and chemical and stress response, which is consistent with published Mrr1 regulons of C. albicans (23) and C. parapsilosis (28,35). In C. auris, the homolog is upregulated in response to stress, such as antifungal exposure (59) or macrophage phagocytosis (62). While OPT1 may be involved in nutrient uptake under stress conditions (63), other metabolic factors, including aldehyde and methylglyoxal dehydrogenases and aldo-keto reducta ses, are speculated to protect cells from reactive molecules generated by azole stress (29). Thus, the Mrr1-regulated metabolic and stress response genes may be important for the persistence of the MRR1 GOF mutants in vivo or could lead to the selection for MRR1 GOF mutants in drugless conditions (18). Understanding Mrr1 regulation of these additional targets across Candida spp. can provide insights into the mechanisms that change multi-drug transporter regulation in Candida.

Strains used in this study are listed in Table S2. All strains were stored as frozen stocks with 25% glycerol at -80°C and maintained regularly on YPD (1% yeast extract, 2% peptone, 2% glucose, 1.5% agar) plates incubated at 30°C, then stored at room temperature. Strains were grown in YPD liquid medium (5 mL) at 30°C on a roller drum for ~16 h prior to inoculation into specified culture conditions. For drug susceptibility assays, cells were grown in RPMI-1640 (Sigma, containing L-glutamine, 165 mM MOPS, 2% glucose, pH 7) liquid, as noted. Escherichia coli strains were grown in LB with either 150 µg/mL carbenicillin or 15 µg/mL gentamicin as necessary to maintain plasmids.

Gene replacement constructs for knocking out MRR1 (CLUG_00542, as annotated in reference 18) and MDR1 (CLUG_01938/9 [18]) were generated by fusion PCR, as described in Grahl et al. (64). All primers (IDT) used are listed in Table S3. Briefly, 0.5 to 1.0 kb of the 5′ and 3′ regions flanking the gene was amplified from U04 DNA, isolated using the MasterPure Yeast DNA Purification Kit (epiCentre). The nourseothricin (NAT1) or hygromycin B (HygB) resistance cassette was amplified from plasmids pNAT (65) and pYM70 (66), respectively. Nested primers within the amplified flanking regions were used to stitch the flanks and resistance cassette together. Gene replacement constructs for knocking out CDR1 (CLUG_03113) and FLU1 (CLUG_05825) were generated by introducing 30-to 50-bp of the 5′ and 3′ regions flanking the gene of interest into the replacement NAT1 cassette using PCR. PCR products for transformation were purified and concentra ted with the Zymo DNA Clean & Concentrator kit (Zymo Research) with a final elution in molecular biology grade water (Corning).

Plasmids for complementing untagged MRR1 were created as described in Biermann et al. (24). Plasmids for complementing N-terminal HF-tagged MRR1 were made as follows. We amplified (i) the 6×His-3×FLAG-tag from an HF-MRR1 tagged C. albicans strain DH2561 using primers ED207 and ED208, (ii) the ~1,150 bp upstream region of the MRR1 gene for homology, from the respective MRR1 allele complementation plasmids, using primers ED103 and ED206, and (iii) ~1,500 bp of the MRR1 gene using primers ED209 and ED132. The 6×His-3×FLAG-tag is placed after the first codon of MRR1. PCR products were cleaned up using the Zymo DNA Clean & Concentrator kit (Zymo Research). The amplified PCR products were assembled into a pMQ30 vector using the Saccharomyces cerevisiae recombination technique described in Shanks et al. (67). Plasmids created in S. cerevisiae were isolated using a yeast plasmid miniprep kit (Zymo Research) and transformed into High-Efficiency NEB5-alpha competent E. coli (New England BioLabs). E. coli containing pMQ30-derived plasmids were selected for on LB containing 15 µg/mL gentamicin. Plasmids from E. coli were isolated using a Zyppy Plasmid Miniprep kit (Zymo Research) and subsequently verified by Sanger sequencing. MRR1 complementation plasmids were linearized with the NotI-HF restriction enzyme (New England BioLabs), cleaned up using the Zymo DNA Clean & Concentrator kit (Zymo Research), and eluted in molecular biology grade water (Corning) before transformation of 2 µg into C. lusitaniae strain U04 mrr1Δ as described below.

Mutants were constructed as previously described in Grahl et al. using an expression-free ribonucleoprotein CRISPR-Cas9 method (64). One to 2 µg of DNA for gene knockout constructs generated by PCR or 2 µg of digested plasmid, purified and concentrated with a final elution in molecular biology grade water (Corning), was used per transformation. E. coli strains containing the complementation and knockout constructs and crRNAs are listed in Tables S2 and S3, respectively. Transformants were selected on YPD agar containing 200 µg/mL nourseothricin or 600 µg/mL hygromycin B.

Mutants for CDR1 and FLU1 were generated using a microhomology-mediated end-joining repair method as described in Al Abdallah et al. (68). One to 2 µg of DNA for gene knockout constructs generated by PCR were used for transformation. crRNAs (IDT) used to target the 5′ and 3′ end of the gene of interest are listed in Table S3. CDR1 and FLU1 knockout transformants were selected on YPD agar containing 200 µg/mL nourseothricin.

Overnight cultures were back-diluted into 50 mL YPD and grown to the exponential phase (~5 h) at 30°C. Harvested cells were snap-frozen using ethanol and dry ice and stored at -80°C. Thawed cell pellets were resuspended in a homogenization buffer (10 mM Tris-HCl, 150 mM NaCl, and 5 mM EDTA, adjusted to pH 7.4 and 10% sucrose) with protease inhibitor (2× Halt protease, Thermo Scientific) and mixed with an equal volume of 1:1 of 0.5-and 1-mm silica bead mix in a bead beating tube (VWR). Bead beating was carried out for six cycles at a speed of 5.65 for 20 s, with a 1-min rest on ice between each cycle. The top liquid phase was collected and centrifuged to remove any cell debris. Supernatants were transferred to new tubes and stored at -80°C. Protein concentrations were determined using the Bradford assay (Quick Start Bio-Rad) with a standard curve generated using serial dilutions of 2 mg/mL bovine serum albumin (BSA).

Samples were diluted to equal concentrations in sample buffer (3.78% Tris, 5% SDS, 25% sucrose at pH 6.8, and 0.04% bromophenol blue prepared as a 5× stock solution. β-mercaptoethanol [0.05%] was freshly added). Samples were heated for 10 min at 95°C and loaded into 6.5% SDS page gels along with a BioRad All Blue Precision Plus MW marker. The gel was run for ~40 min at 180 V. The BioRad Turboblot semi-dry transfer system with custom settings (1.3 A constant and 25 V for 15 min) was used to transfer the protein bands to an LF-PVDF membrane (Immobilon Product IPFL00010). The blots were processed using the standard LICOR protocol for western blotting, including the optional drying step after transfer and REVERT total protein staining. A milk-based blocking buffer was used instead of the Odyssey blocking buffer. The α-FLAG monoclonal antibody (1 mg/mL) (Sigma-Aldrich M2 or ThermoFisher FG4R) was diluted 3,000-fold in blocking buffer with 0.1% Tween-20. The goat α-mouse secondary antibody (1 mg/mL) labeled with IRDye 700CW was diluted 15,000-fold in blocking buffer with 0.1% Tween-20. Blots were imaged using an Odyssey CLX scanner (LICOR) and analyzed using the Empiria software (LICOR).

Overnight cultures were back-diluted into 50 mL YPD and grown to the exponential phase (~5 h) at 30°C. Samples were processed using the Epicypher CUT&RUN kit (Epicypher) as per the protocol described in Qasim et al. (69). Briefly, yeast nuclei were isolated from the thawed cell pellets using Zymolase 100T (Zymoresearch). Digitonin (0.01%) was added to all buffers used hereafter to permeabilize nuclei and prevent bead clumping. The isolated nuclei were bound to activated concanavalin A (ConA)-coated magnetic beads. The nuclei-bound ConA beads were then split and incubated overnight at 4°C with either 1:100 IgG or α-FLAG primary antibody (Sigma-Aldrich M2 for experi ment-1 and ThermoFisher FG4R for experiment-2). After washing to remove unbound primary Ab, pAG-MNase was added to the nuclei and incubated for an hour. Targeted chromatin digestion by pAG-MNase was initiated by adding CaCl 2 and stopped after 30 min with the stop buffer spiked with 50 ng of E. coli DNA. The supernatant with the pAG-MNase-digested DNA was then collected and purified using an Epicypher DNA cleanup column. DNA libraries were prepared using the NEB Ultra II protocol kit, with slight modifications as recommended in the Epicypher CUTNRUN kit.

Our pilot experiment (experiment-1) was set up with U04 strains expressing one of the three alleles (HF-MRR1 ancestral , HF-MRR1 Y813C , and HF-MRR1 L1Q1* ) and sequenced using paired-end 150 bp reads on the Illumina Nextseq 2000 platform to achieve a sequencing depth of 10 M per sample. Based on the pilot study results, the sequencing depth was adjusted to 5-6 M per sample for the subsequent experiment (experiment-2), including two biological replicates of the U04 strain expressing HF-MRR1 Y813C , which were sequenced using paired-end 50 bp reads on the Illumina Nextseq 2000 platform.

Raw read quality was evaluated using FastQC (v0.12.1) prior to read trimming with Cutadapt (v.4.4) for adapter sequences with additional parameters "--nextseq-trim 20 --max-n 0.8 --trim-n -m 1. " Reads were mapped to Clavispora (Candida) lusitaniae strain L17 (NCBI accession: ASM367555v2) with Bowtie2 (v2.4.2) using parameters "--local --no-mixed --no-discordant. " Alignments were sorted coordinate with Samtools (v1.11), filtered for unmapped or multi-mapping reads using sambamba (v0.8.0), and downsam pled to 3 million reads per sample to ensure equal sensitivity for peak calling across samples. MarkDuplicates (Picard Tools) was used to identify and remove duplicate reads. Fragment size distributions of individual samples were visualized using deep Tools (v3.5.1) command "bamPEFragmentSize. " Peaks were called using the MACS2 (v2.2.7.1) command "callpeak" in narrowpeak mode using IgG IP samples as controls with parameters "-f BAMPE --keep-dup all -g 11999093 -q 0.05. " Significant peaks were further filtered to keep only those with twofold or greater signal increase relative to control (IgG) samples. The fraction of reads in peaks was calculated for each sample to assess individual quality. The BEDTools (v 2.31.1) command "merge, " with the parameter "-c" for averaging peak signal value, was used to merge peaks with a twofold or greater signal from all replicates of experiment-2. BEDTools (v 2.31.1) command "intersect" with the parameter "-a" was used to identify a set of reproducible overlapping peaks between HF-Mrr1 Y813C from experiment-1 and experiment-2. Since the nuclei isolation step was not controlled in our CUT&RUN experiments (37,69), it limited our ability to perform differential peak analysis between replicates and across strains expressing different MRR1 alleles.

MIC was determined using a broth microdilution method as previously described (70). Briefly, 2 × 10 3 cells were added to a twofold dilution series of the drug prepared in RPMI-1640, then incubated at 37°C. The MIC was defined as the minimum drug concentration that abolished visible growth compared to a drug-free control. The MIC 90 was defined as the minimum drug concentration that led to a 90% or greater decrease in growth relative to a drug-free control. No more than a twofold difference was observed between MICs recorded at 24 and 48 h; data from the 24 h time point were reported unless otherwise noted. The concentration range used for azoles was FLZ: 64 to 0.0625 µg/mL, VOR: 4 to 0.004 µg/mL, KTZ: 1 to 0.004 µg/mL, ITZ: 0.4 to 0.001 µg/mL, and ISA: 1 to 0.04 µg/mL. For the broad-spectrum antifungals, the following concentration ranges were used: myclobutanil: 32 to 0.0625 µg/mL, terbinafine: 64 to 0.125 µg/mL, cycloheximide: 32 to 0.0625 µg/mL, 5-FC: 4 to 0.008 µg/mL, fluphenazine: 200 to 0.39 µg/mL, and mycophenolic acid: 256 to 0.5 µg/mL.

Sequences spanning ±100 bp around the peak summits identified from CUT&RUN data were extracted from the L17 genome (NCBI accession: ASM367555v2) using BEDTools v2.30.0 (71). To establish a background control, we used BEDTools random to retrieve randomly selected 200 bp sequences from the genome of L17. STREME (44), part of the MEME Suite (streme --verbosity 1 --oc streme_results --dna --totallength 4000000 --time 14400 --minw 6 --maxw 20 --thresh 0.05 --align center --p around_peaks.fasta --n random_sequences.fasta) was employed for motif discovery and enrichment analysis. and C. lusitaniae ATCC 42720) was conducted using FIMO (72), part of the MEME Suite (fimo --oc fimo_results --verbosity 1 --bgfile --nrdb --thresh 1.0E-3 motif1.meme target_seqs.fasta). MDR1 and CDR1 gene IDs and their translational start site coordinates used for the sequence retrieval of the upstream regions are listed in Table S4.

For the phylogenetic gene trees, nucleotide sequences of the respective genes were extracted and aligned using MAFFT v7 (https://mafft.cbrc.jp/alignment/server/ index.html) with default parameters. A neighbor-joining tree was then constructed based on the aligned DNA sequences, with 1,000 bootstrap replicates to assess phylogenetic relationships. The results were visualized using the ggmotif v0.2.0 R package (73) and FigTree v1.4.4 (http://tree.bio.ed.ac.uk/software/figtree/).

Mrr1 1-196 was expressed from pET51b-MRR1 ) in E. coli BL21(DE3) cells grown at 37°C. Protein expression was induced with 1 mM IPTG, which was added to cultures in mid-log phase; subsequent incubation was either for 4 h at 37°C or 16 h at 16°C. Cell pellets were stored at -20°C. Lysis buffer (20 mM sodium phosphate at pH 7.4, 200 mM NaCl, 20 mM imidazole, and 10% glycerol) supplemented with EDTA-free protease inhibitors and 0.01 mg/mL lysozyme was used to resuspend the cell pellets. Cells were lysed using an LM10 microfluidizer processor at 18,000 psi for three cycles. The cell debris was removed by ultracentrifugation. The clarified lysate was loaded onto a 5 mL HisTrap HP column (Cytiva) pre-equilibrated with binding buffer (20 mM sodium phosphate at pH 7.4, 400 mM NaCl, 20 mM imidazole, and 10% glycerol) using the AKTA Pure25 fast liquid protein chromatography system. The His-tagged Mrr1 1-196 was eluted using 10% of elution buffer (20 mM sodium phosphate at pH 7.4, 400 mM NaCl, 500 mM imidazole, and 10% glycerol), followed by 100% of elution buffer. Eluted fractions were evaluated by SDS-PAGE. Pooled fractions containing the Mrr1 1-196 (~25 kDa) were loaded onto HiLoad Superdex 200 26/600 columns (Cytiva) for further purification by size exclusion chromatography. Gel filtration buffer (20 mM sodium phosphate at pH 7.4, 150 mM NaCl, and 10% glycerol) was used for column calibration and sample elution. Protein concentrations were determined using Bradford assays as described above.

One hundred microliters of Mrr1 1-196 protein (180 µM), DNA (25 µM), or protein-DNA mix (containing 100 µM protein and 25 µM DNA) were injected into a Superose 6 Increase 10/300 GL column and eluted with SEC buffer containing 25 mM HEPES, pH 7.5, 150 mM NaCl, 10% glycerol, and 1 mM DTT. The elution was performed at room temperature. A 280 chromatogram was recorded for all samples. Protein standards, including thyroglo bulin (GE Healthcare, #28-4038-42), aldolase (GE Healthcare, #28-4038-42), bovine serum albumin (Sigma-Aldrich, #MWGF70), and carbonic anhydrase (Sigma-Aldrich, #MWGF70), were used to calibrate the column. Eluted samples were collected as 0.5 mL/fraction. Peak fractions containing protein samples were subjected to SDS-PAGE on 10% NuPAGE Bis-Tris gels (Invitrogen) with MOPS buffer (Invitrogen) and followed by staining with InstantBlue Coomassie Protein Stain (Abcam).

Cy5.5-labeled 30 bp DNA probes were resuspended in 1× DNA annealing buffer (10 mM Tris, pH 7.5, and 50 mM NaCl). Recombinant Mrr1 1-196 was diluted to final concentra tions of 3, 2, 1, 0.75, 0.5, 0.25, 0.1, and 0 µM in gel filtration buffer. Mrr1 1-196 at each concentration was incubated with 0.5 µM Cy5.5-labeled 30 bp DNA probes, 0.025 mg/mL poly(dI:dC), and 0.002 mg/mL BSA at room temperature for 30 min. Glycerol at a final concentration of 22% was added to the samples. The samples were loaded into 12% tris-borate native acrylamide gels. The gel was run for ~160 min at 80 V and imaged using an Odyssey CLX scanner (LICOR). For the experiments with mutated Cy5.5-labeled 30 bp DNA probes, poly(dI:dC) was used at a final concentration of 0.05 mg/mL. DNA probe concentration was optimized to 0.5 µM since the unbound probe signal was eliminated by non-specific protein binding at nanomolar concentrations of the probe.

Ordinary one-way analysis of variance (ANOVA) and Dunnett's multiple comparisons testing, with a single pooled variance, were used for statistical evaluation. P values <0.05 were considered significant for all analyses performed and are indicated with asterisks: *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. Figures 7A and 8 were created in BioRender (https://BioRender.com/39wn5ht).