Full Paper
Mutations in the 5' untranslated region fine-tune translational control of heterologously expressed genes
2025 Volume 100 Article ID: 24-00188
Details
2025 Volume 100 Article ID: 24-00188
Strict control of the expression levels of heterologously introduced protein-coding genes is important for the functional analysis of the protein of interest and its effective use in new situations. For this purpose, various promoters with different expression strengths, codon optimization, and expression stimulation by low-molecular-weight compounds are commonly used. However, methods to control protein expression levels by combining regulation of translation efficiency have not been studied in detail. We previously observed relatively high basal expression of Cre when it was heterologously expressed in fission yeast. Here, we used a fission yeast strain that is susceptible to centromere disruption, and thus highly sensitive to Cre levels, and report successful fine-tuning of heterologous Cre expression by modulating the Cre translation efficiency. To inhibit Cre translation initiation, we generated two mutations in the 5' untranslated region of the Cre mRNAs, both of which interfered with the scanning process of start codon recognition, mediated by specialized ribosomal subunits. These mutations successfully reduced the levels of exogenously expressed Cre to different degrees in fission yeast. Combining them with promoters of different strengths allowed us to conduct centromere disruption experiments in fission yeast. Our data indicate that modification of translational control is an additional tool in heterologous gene expression.
Controlling gene expression levels is desirable for experimental studies that involve exogenous genes. According to natural genetic selection, gene expression levels are generally optimal when genes function in their endogenous environments (Dahan et al., 2011; Sorrells and Johnson 2015; Sztal and Stainier 2020; Parvathy et al., 2022). However, when they are artificially translocated to ectopic environments (e.g., as a result of experimental design), genetic selection is no longer valid, and appropriate expression levels must be artificially achieved.
We previously used the ectopic expression of Cre, a site-specific recombinase originating from bacteriophage P1 (Sternberg and Hamilton, 1981), in fission yeast, Schizosaccharomyces pombe, to conduct large-scale in vivo chromosome engineering called centromere disruption (Ishii et al., 2008). When Cre was conditionally expressed in fission yeast harboring two loxP sites at either end of a given centromere, the loxP-flanked centromere was detached from the chromosome by means of Cre-mediated recombination, resulting in the formation of a mitotically untransmittable acentric chromosome and cell lethality. This experimental design, combined with marker gene selection, allowed us to study the mechanisms of rare chromosomal events that rescue acentric chromosomes, such as neocentromere formation and telomere fusion (Ishii et al., 2008; Ogiyama et al., 2013; Ohno et al., 2016). The same approach has been extended to organisms other than fission yeast and has been shown to yield even more fruitful results, including the identification of essential structural cores within complex kinetochores in chickens and the discovery of new trade-off relationships between positive and negative effectors of telomeres in budding yeast (Hori et al., 2013; Shang et al., 2013; Pobiega et al., 2021).
However, finding a suitable Cre expression condition for centromere disruption in fission yeast has proven challenging. We established that the frequency of neocentromere formation and telomere fusion can be accurately measured if the centromere disruption design strain we used (designated as loxP-cen1 for chromosome I) is first transformed with Cre expression plasmid under Cre-repressed conditions. After establishing transformants that silently retained the plasmid, we induced uniform Cre expression from the plasmid, which resulted in the simultaneous detachment of the loxP-flanked centromeres in the transformants, and counted the number of surviving cells (Ishii et al., 2008). We used the thiamine-repressible promoter of the nmt1 (no message in thiamine) gene to conditionally express Cre, and transformants were obtained in the presence of thiamine (Maundrell, 1990). However, although the nmt1 promoter is frequently used as an inducible promoter in fission yeast with an up to 70-fold induction potential in the absence of thiamine, its basal expression in its putatively repressed state is relatively high (Maundrell, 1990; Forsburg, 1993). This non-negligible Cre expression in the presence of thiamine caused a reduction in the overall plasmid transformation efficiency and a potential accumulation of unknown mutations that could somehow eliminate Cre expression in the transformants. To overcome this limitation, we adopted the weakest TATA box-mutated derivative of the nmt1 promoter (nmt81) for Cre expression (Basi et al., 1993). However, even under these conditions, the transformation efficiency was poor, and in many of the obtained transformants the lethality expected on thiamine-free medium was lost. Consequently, despite the reduced transformation efficiency, transformant screening was additionally required to identify lethality on thiamine-free medium prior to screening for centromere disruption survivors. Because this isolation process was uncertain and laborious, we sought to develop alternative methodology.
Besides promoter selection, several ways to regulate heterologous gene expression levels exist. We sought to modulate the translation efficiency, as it could easily be combined with our centromere disruption experiments. The efficiency of eukaryotic mRNA translation is determined primarily by the frequency of ribosome assembly at the mRNA coding regions (Hinnebusch, 2014). Prior to ribosome assembly, the 43S preinitiation complex (PIC), which contains the 40S small ribosomal subunit, multiple initiation factors and initiator tRNAMet, is recruited to the mRNA 5' end and scans it downstream until it encounters the first AUG. Start codon recognition by the 43S PIC then elicits complete ribosome assembly and in-frame protein synthesis along the coding region (Hinnebusch, 2014). Recently, however, advanced genomic and proteomic approaches have revealed the presence of upstream open reading frames (uORFs) in 5' untranslated regions (UTRs) of many mRNAs in various organisms, including fission yeast (Ingolia et al., 2009, 2011; Hsu et al., 2016; Duncan et al., 2018; Chikashige et al., 2020; Chothani et al., 2022). These uORFs have been proposed to function in decreasing the translation efficiency of downstream main ORFs as a function of developmental stage and in response to external stress (Hinnebusch et al., 2016; Kearse and Wilusz, 2017). This reduction of main-ORF translation by uORFs could be used to suppress basal Cre expression under transcriptionally repressed conditions. Furthermore, translation of many uORFs was found to be initiated at non-AUG codons, suggesting that mRNA scanning by 43S PICs is not always stringent and can misrecognize non-AUG triplets (Kearse and Wilusz, 2017). Thus, AUG to non-AUG substitutions could be exploited to suppress basal Cre expression. In this study, we attempted to control gene expression at the translational level by modifying the Cre 5' UTR start codon recognition mechanism. Our results present a new means of controlling the expression levels of both exogenous and endogenous genes.
Cre mRNAs expressed from plasmids used in previous centromere disruption experiments contained a 72-nucleotide (nt) 5' UTR, derived from the nmt1+ gene (Maundrell, 1990, 1993; Ishii et al., 2008), which is considered to be an average length for eukaryotic 5' UTRs and satisfies scanning by 43S PICs (Leppek et al., 2018). We intentionally miscloned Cre downstream of the multi-cloning site of the nmt1 expression plasmid to create a 40-amino-acid (aa) ORF upstream of the Cre coding region that was not in-frame but rather overlapped with Cre by 62 nts (Fig. 1A and 1B). This synthetic mRNA (uORF-Cre) was expressed from the strongest available nmt1 promoter, a moderate nmt41 promoter derivative, and the weakest available nmt81 promoter derivative (Basi et al., 1993). We also generated a non-synonymous mutant clone in which the first codon of Cre was changed from AUG to AUA, encoding isoleucine (Fig. 1A and 1B). The first AUG from the 5' end in this clone is located 49 nts further downstream and belongs to a 10-aa ORF of a different reading frame. The first AUG in the same ORF as Cre is located 81 nts downstream from the original AUG (Fig. 1B). This non-AUG mutant mRNA (nAUG-Cre) was expressed similarly to uORF-Cre, and the effects of both mRNAs were compared with the original Cre mRNA.
Reduction in Cre expression caused by AUG-associated 5' UTR mutations
First, a series of Cre plasmids with the strong nmt1 promoter were transformed into wild-type fission yeast, and expression levels of each construct were examined by Western blotting. Under both nmt1 promoter-repressed (+Thi) and nmt1 promoter-induced (-Thi) conditions, a prominent decrease in Cre expression levels was observed in the mutants of 5' UTR (Fig. 2A). The decrease was higher in the non-AUG substitution mutant than in the uORF creation mutant. Here, the C-terminal Flag-tagged variants of the same constructs were co-examined, which generated almost the same results as those using their untagged counterparts (Fig. 2A). Western blotting using an antibody against the C-terminal Flag tag further highlighted a nAUG-Cre-specific N-terminal truncation under +Thi and -Thi, which was also detectable in the untagged version, whose size was consistent with Cre having been translated from its own second AUG (Fig. 1B and Supplementary Fig. S1). Note that such protein truncation was partial and many products derived from nAUG-Cre migrated at the full-length position, suggesting that the AUA of nAUG-Cre is also recognized as a start codon. Cre expression from the nmt1 promoter has been reported to be toxic in wild-type fission yeast even under repressed conditions (Iwaki and Takegawa, 2004). In agreement with this finding, we found it challenging to obtain and maintain normal Cre transformants (Supplementary Fig. S2A). However, C-terminal Flag-tagging improved our ability to maintain these strains, presumably owing to reduced recombinase activity, and relatively high amounts of Flag-tagged Cre transformants were obtained and maintained fairly normally (Supplementary Fig. S2B). Therefore, further examination of Cre expression levels was performed using only the Flag-tagged variants.
Next, the induction time course of the mutant constructs was examined. Consistent with previous observations (Maundrell, 1990), Cre expression was induced around 10–12 h and peaked around 14–16 h after thiamine removal from the medium, which was similar to that of the 5' UTR mutant constructs despite an overall decrease in expression levels (Fig. 2B). This implies that the reduction of gene expression may occur posttranscriptionally through changes in translation efficiency, without excluding alternative mechanisms.
Gene expression reduction was also examined in the presence of different promoter combinations. In addition to the potent nmt1 promoter, a 5' UTR mutation-dependent reduction in Cre expression was observed for the moderate-strength nmt41 and the weak nmt81 promoters (Fig. 2C). Quantitative reverse transcription PCR (RT-PCR) revealed that the Cre mRNA levels in each condition were virtually the same with and without the mutations, further supporting the notion that the reduction occurs at the protein level (Supplementary Fig. S3). Interestingly, the reduction of Cre expression in the presence of a uORF was more effective in combination with the nmt41 and nmt81 promoters than with the nmt1 promoter. This suggests that although Cre expression from the nmt1 promoter may be occurring at higher levels, its detection is hampered by cytotoxicity. Nevertheless, the expression levels of uORF-Cre and nAUG-Cre from the nmt41 and nmt81 promoters under repressive conditions were similar to or lower than that of the original Cre from the nmt81 promoter (Fig. 2C). We conclude that these constructs are promising candidates for improving our centromere disruption experiment.
Optimization of transformant isolation for centromere disruptionPlasmids carrying nmt41- and nmt81-regulated variants of uORF-Cre and nAUG-Cre were transformed into the centromere disruption strain loxP-cen1. Contrary to the poor transformation efficiency of the plasmid bearing Cre, no obvious decrease in transformation efficiency was observed with plasmids bearing uORF-Cre or nAUG-Cre (Fig. 3A). Furthermore, the colony size of the obtained transformants was relatively uniform (Supplementary Fig. S4). The uniformity of growth among the individual transformants of uORF-Cre and nAUG-Cre and their lower degree of growth inhibition compared with that of the original Cre were also evident in the doubling time of each transformant (Fig. 3B). Taken together, our findings indicate that the toxic effects of Cre on the loxP-cen1 strain are largely inhibited in the uORF-Cre and nAUG-Cre transformants under nmt1 promoter-repressive conditions.
Upon induction of the nmt1 promoters by thiamine removal, many of the uORF-Cre transformants bearing the nmt41 and nmt81 promoters showed lethality (Fig. 3C and Supplementary Fig. S2). In the case of nAUG-Cre, however, a high degree of lethality was observed only in transformants bearing the nmt41 promoter and not those bearing the nmt81 promoter (Fig. 3C and Supplementary Fig. S2). This suggests that the Cre expression level from the nmt81 promoter in the nAUG-Cre transformants is too low to result in lethality due to centromere disruption. Given the above behavior of the transformants, we decided to perform centromere disruption experiments using the loxP-cen1 strain with nmt41-nAUG-Cre, nmt41-uORF-Cre and nmt81-uORF-Cre plasmids.
Centromere disruption experiments using 5' UTR mutant CreWe observed that the cell growth profiles of the different loxP-cen1 transformants exhibited Cre construct-dependent differences after nmt1-mediated induction of centromere disruption (Fig. 4A and Supplementary Fig. S5A). nmt41-nAUG-Cre and nmt41-uORF-Cre transformants exhibited an induction-dependent growth retardation phenotype at about 12–16 h after thiamine removal. nmt41-uORF-Cre transformants also showed a basal growth retardation phenotype, but to a lesser extent than the nmt81-Cre transformants, which exhibited virtually no induction-dependent growth retardation. Induction-dependent growth retardation was also not prominent in nmt81-uORF-Cre transformants, but unlike nmt81-Cre, their basal growth was normal and individual growth profiles appeared somewhat promising (Fig. 4A and Supplementary Fig. S5A). Whatever the growth profile, all transformants in which Cre was induced for 18 h were processed downstream for the isolation of survivors.
Centromere disruption survivors are isolated based on centromere disruption-dependent acquisition of marker genes that confer drug resistance (Ishii et al., 2008). Drug selection eliminates cells bearing intact centromeres as well as those that do not initiate a survival response upon centromere disruption, allowing only centromere-disrupted cells that undergo survival responses such as neocentromere formation and telomere fusion to proliferate. The average survival rate after 18 h of Cre induction, as determined by drug resistance, was highest for the nmt81-Cre construct (Fig. 4B and 4C). However, the variation in survival rates was also the highest among the Cre constructs examined; the remaining Cre constructs bearing 5' UTR mutations produced fewer survivors, but the rates were relatively uniform among the transformants (Fig. 4B). Interestingly, the resulting average survival values were proportional to the Cre expression levels we previously determined by Western blotting, except for nmt81-Cre, where the diversity among transformants was extreme (Fig. 4C). These data suggested that Cre expression levels play an important role in determining the rate of survival.
To establish whether the obtained survivors had undergone chromosome reorganizations such as neocentromere formation and telomere fusion, nmt41-nAUG-Cre transformant survivors were further subjected to pulsed-field gel electrophoresis (PFGE) to determine their chromosomal configurations. The PFGE data indicated that the survivors had evenly acquired both types of chromosome reorganization (neocentromere formation and telomere fusion) without bias toward either type per transformant (Supplementary Fig. S5B). The overall ratio between neocentromere formation and telomere fusion was comparable to that observed previously for the nmt81-Cre construct (Fig. 4D). Therefore, we conclude that the generated 5' UTR mutations reduce the overall survivor frequency after centromere disruption, but the event itself remains unchanged from the original experimental design.
Transcription of mRNA and its subsequent translation into protein are fundamental cellular processes. However, attempts to artificially modify these processes have so far focused predominantly on mRNA transcription. This has been more obvious in the introduction of exogenous genes, since a great challenge is to get foreign DNA transcribed in an ectopic environment, whereas translation mechanisms are remarkably conserved among organisms. Here, using ectopic Cre expression in fission yeast, we showed that controlling translation efficiency is as effective as promoter selection in the modulation of the levels of exogenous gene expression. Introduction of 5' UTR mutations that reduce Cre expression levels to varying degrees by interfering with translation initiation provided a solution to an experimental challenge, namely, uncontrollably high basal transcription of Cre.
Although we demonstrated a 5' UTR mutation-dependent decrease in Cre expression (Fig. 2), the actual mechanism by which expression levels are decreased remains to be elucidated. The nature of both introduced mutations (Fig. 1) and their minimal effect on mRNA levels (Supplementary Fig. S3) suggest that the reduction in Cre levels is due to a reduction in Cre mRNA translation efficiency, but to ascertain this, translation efficiency must be directly examined by ribosome profiling or other methods. These experiments remain to be conducted in future studies.
The mutation scheme we employed took advantage of a leaky mRNA scan by the 43S PIC to reduce the efficiency of canonical Cre translation initiation (Hinnebusch et al., 2016; Kearse and Wilusz, 2017). For the 43S PIC to optimally recognize the AUG of the mRNA as a start codon, base pair interaction between the anticodon in the initiator tRNAMet within the 43S PIC and the AUG of the mRNA is not sufficient; the 43S PIC also requires the presence of a favorable sequence context around AUG, typically provided by the Kozak sequence (Hernández et al., 2019). Therefore, both AUGs in unfavorable sequence contexts and non-AUGs in favorable sequence contexts undergo leaky scanning by the 43S PIC, which results in unrecognized AUGs and misrecognized non-AUGs, respectively. The uORF-Cre mutant was intended to generate the former type of aberration, and nAUG-Cre the latter. In this sense, it is intriguing that the degrees of reduction in Cre expression in uORF-Cre and nAUG-Cre appear to be anticorrelated (Fig. 2A). However, the degree to which the 43S PIC preferentially recognizes its cognate sequences is highly context-dependent and difficult to discuss rigorously (Diaz de Arce et al., 2018; Hernández et al., 2019). Our current data are not suitable to shed light on this matter.
We detected a substantial amount of Cre enzyme produced in the start codon-mutated nAUG-Cre clone (Fig. 2–4). Unlike other missense mutations in the coding region, start codon mutations are considered equivalent to gene knockout and are often used for studying loss of gene function (Wang et al., 2020). However, our results show that this stereotypical view of start codon mutations is not always correct and calls for caution. AUA was chosen as the substitution codon in this study, based on several reports suggesting that it may function as a start codon in vivo in eukaryotes including fission yeast, albeit with low efficiency in most cases (Hernandez et al., 2002; Chikashige et al., 2017; Diaz de Arce et al., 2018; Eisenberg et al., 2020). Codon discrimination between AUA and AUG is controlled in eukaryotes by base modifications introduced in the tRNAIle anticodon nucleosides (Senger et al., 1997), although how the modified anticodon prevents AUG reading is not entirely clear (Mandal et al., 2010; Suzuki and Numata, 2014). AUA is also known to encode methionine instead of isoleucine in the mitochondria of many organisms (Osawa et al., 1992). It remains unclear whether the AUA substitution we employed has a particular function or is just one of the possible start codon substitutions (Diaz de Arce et al., 2018; Eisenberg et al., 2020).
In this study, we sought to simplify the procedure of the Cre-mediated centromere disruption experiments we had previously established in fission yeast (Ishii et al., 2008). The key for the success of this modification lies in the proper control of Cre expression levels. Excessive amounts of Cre are toxic to wild-type fission yeast (Iwaki and Takegawa, 2004), and even more so to our centromere disruption design strain (Supplementary Fig. S2). Combined with the regulation of translation efficiency, we succeeded in controlling the basal expression of Cre to non-toxic levels and eliminating the pre-selection step in centromere disruption experiments (Fig. 4 and Supplementary Fig. S5). Since the modifications also affected the levels of Cre induction, which in turn affected the efficiency of eventual survivor acquisition (Fig. 4C), it is not possible at present to conclude which Cre constructs are best suited for centromere disruption. However, with the choice of promoters increasing in fission yeast (Erler et al., 2006; Zilio et al., 2012; Kjaerulff and Nielsen, 2015; Ohira et al., 2017), the success of Cre repression at the translational level is valuable in that such modifications can be further combined with promoters of different expression strengths to synergistically expand the options for gene expression levels. Additionally, translational control of heterologous gene expression commonly benefits from codon optimization of the transgene (Parvathy et al., 2022). That said, codon optimization is unidirectional and not helpful in suppressing gene expression. The new approach presented in this study may now expand the possibilities for the regulation of heterologous gene expression in diverse organisms.
Manipulation of fission yeast was performed according to standard procedures (Moreno et al., 1991). Cells were grown at 33 °C. Standard rich medium (YES) and synthetic minimal medium (EMM2) for fission yeast were used as described (Moreno et al., 1991). Antibiotics and chemicals were added at the following concentrations: G418 (Nacalai Tesque), 50 mg/l; 5-fluoroorotic acid (5-FOA, FUJIFILM Wako Pure Chemical), 1 g/l. Fission yeast cells were transformed using the lithium chloride method as described (Gietz, 2014). Genotypes of the fission yeast strains used in this study are as follows: wild type (KYP88), h- leu1-32; loxP-cen1 (KYP378), h- leu1-32 ura4-D18 cen1L::Padh1-loxP cen1R::ura4+-loxP-kanRORF (Ishii et al., 2008).
Plasmid constructionpREP1, pREP41 and pREP81 were used to express Cre from the nmt1, nmt41 and nmt81 promoters, respectively (Basi et al., 1993; Maundrell, 1993). 5' UTR mutations were generated through the PCR amplification of Cre coding sequence using the following primers: uORF creation, Cre-NotI-F (5'-ATAAGACCGCGGCCGCTATGTCCAATTTACTGACCG-3') and Cre-NotI-R (5'-ATAGTTGGGCGGCCGCCTAATCGCCATCTTCCAGC-3'); non-AUG substitution, InFus_CreATA5 (5'-CTTTGTTAAATCATATATCCAATTTACTGACCGTACACC-3') and InFus_CreATA3 (5'-TCTAGAGTCGACATACTAATCGCCATCTTCCAGCAG-3'). A NotI site was created in the middle of the SmaI site in the multi-cloning sites of pREP1, pREP41 and pREP81, generating pREP1N, pREP41N and pREP81N, respectively, and the mutation-harboring PCR products were ligated into these NotI sites for uORF creation and In-Fusion HD cloning (Takara Bio) for non-AUG substitution. Nucleotide sequences of the resulting plasmids were confirmed by sequencing. For C-terminal Flag-tagging of Cre in the above plasmids, a fragment encompassing 3 × Flag and part of the nmt1 terminator sequence up to the PacI site was PCR-amplified using the primers PAPP-IF5 (5'-TTAATTAAGATGGATTATAAAGATG-3') and PAPP-IF3 (5'-TTTCATCGTTTTTTAATTAATATTC-3') from pFA6a-3Flag-PA-hphMX6 in which the budding yeast ADH1 terminator sequence in pFA6a-3Flag-hphMX6 (Bähler et al., 1998) had been replaced with the nmt1 terminator sequence. The resulting PCR product was cloned into the above-mentioned Cre expression plasmids digested by BssHII and PacI via In-Fusion HD cloning together with another PCR product covering part of the Cre coding region downstream of BssHII, which was amplified using the primers CreBP-IF5 (5'-ATATCTTCAGGCGCGCGG-3') and CreBP-IF3 (5'-ATCCATCTTAATTAAATCGCCATCTTCCAGCAGGCGC-3').
Western blottingLogarithmically growing fission yeast cells harboring the Cre expression plasmids cultured in EMM2 with or without 2 μM thiamine were harvested. Except for induction time-course experiments, the condition of 18 h after thiamine removal was used for thiamine-deprived cell cultures. Cells (5 × 107) were lysed in TEG150 buffer (50 mM Tris-HCl [pH 7.5], 5 mM EDTA, 150 mM NaCl, 10% glycerol, 1 mM PMSF) supplemented with cOmplete mini EDTA-free (Sigma-Aldrich). The supernatant of centrifugation at 2,400 × g for 5 min at 4 °C was applied for Western blotting as whole cell extract. Anti-Cre (Novagen) and anti-Flag M2 (Sigma-Aldrich) were used as primary antibodies. Band intensity was measured using ImageJ software (ver. 1.52a).
RNA preparation and quantitative RT-PCRCell cultures were prepared as in Western blotting. Cells (5 × 107) were resuspended in TES buffer (10 mM Tris-HCl [pH 7.5], 10 mM EDTA, 0.5% SDS), mixed with an equal volume of acid phenol:chloroform:isoamyl alcohol (25:24:1) [pH 5.2] (Nacalai Tesque), and extracted at 65 °C for 60 min. Total RNAs in the aqueous fraction were ethanol-precipitated and dissolved in RNase-free water. After digesting the contaminating DNA in the solution by TURBO DNA-free (Invitrogen), RNAs were reverse transcribed using the ReverTra Ace qPCR RT kit (TOYOBO). Relative cDNA amounts of Cre and LEU2 in the RT reaction were determined by quantitative PCR (qPCR) using the StepOnePlus Real-Time PCR System (Applied Biosystems) with the following primer sets: Cre, 5'-TTGCCGCGCCATCTG-3' and 5'-TTGCTTCAAAAATCCCTTCCA-3'; LEU2, 5'-TGTTGCCATCTGCGTCCTT-3' and 5'-CGTGGCATGGTTCGTACAAA-3'. qPCR of each target in each sample was repeated at least three times, and the absence of DNA contamination in the samples was always confirmed by examining the equivalent samples without the RT reaction. Ct values were averaged and used to calculate ΔCt values. Standard deviation was calculated as the square root of the sum of the squares of the standard deviations for each target.
Colony size measurementLogarithmically growing loxP-cen1 cells were transformed with 1 μg of the plasmids. They were plated onto EMM2 containing 2 μM thiamine and incubated at 33 °C for four days. Each plate was imaged and colony size of the obtained transformants on the plate was measured using the ROI tool in ImageJ software (ver. 1.52a).
Centromere disruptionThe centromere disruption assay was performed as described previously (Ishii et al., 2008). The genetic selection scheme of the assay was briefly as follows: the loxP-cen1 strain harbors an adh1 promoter fragment upstream of one loxP site, a promoterless kanR gene downstream of the other loxP site, and a ura4+ marker gene in addition to the centromere of chromosome I between the two loxP sites. Conditionally induced Cre excises the centromere and ura4+ from chromosome I and the kanR gene is expressed through the adjacent adh1 promoter at the recombination site, which confers resistance to both G418 and 5-FOA.
Author contributions: K. I. and R. K. conceived and designed the study; R. K. performed the experiments; K. I. and R. K. analyzed the data; K. I. wrote the paper.
Conflicts of interest: The authors declare no conflict of interest.
Supporting information: Supplementary Figure S1 shows the original Western blotting data in different exposures. Supplementary Figure S2 shows the growth retardation observed in transformants bearing different Cre expression plasmids under Cre-repressed or Cre-induced conditions. Supplementary Figure S3 shows the RT-PCR data detecting Cre mRNA levels in different conditions. Supplementary Figure S4 shows the colony size variation seen after transformation of loxP-cen1 strain with different Cre expression plasmids. Supplementary Figure S5 shows the behavior of individual loxP-cen1 transformants in the centromere disruption experiments.
We thank our laboratory members for valuable discussions. This work was supported by JSPS KAKENHI (JP21K06022 to K. I.), Joint Research of the Exploratory Research Center on Life and Living Systems (ExCELLS program No. 24EXC202 to K. I.), the Ohsumi Frontier Science Foundation (to K. I.) and JST SPRING (JPMJSP2141 to R. K.).
