New Results

Use of a ubiquitous gene-editing tool in budding yeast causes off-target repression of neighboring gene protein synthesis

Emily Nicole Powers, Lidia Llacsahuanga Allcca, Ella Doron-Mandel, Jenny Kim Kim, Marko Jovanovic, Gloria Ann Brar
doi: https://doi.org/10.1101/2022.06.27.497784

Abstract

Precision genome-editing approaches have long been available in budding yeast, enabling introduction of gene deletions, epitope tag fusions, and promoter swaps through a selection-based strategy. Such approaches allow loci to be modified without disruption of coding or regulatory sequences of neighboring genes. Use of this approach to delete DBP1 however, led to silencing of expression and the resultant loss of function for the neighboring gene MRP51. We found that insertion of a resistance cassette to delete DBP1, drove a 5’ extended alternative transcript for MRP51 which dampened Mrp51 protein synthesis. Misregulation of MRP51 occurred through an integrated transcriptional and translational repressive long undecoded transcript isoform (LUTI)-based mechanism that was recently shown to naturally regulate gene expression in yeast and other organisms. Cassette-induced MRP51 repression drove all mutant phenotypes we detected in cells deleted for DBP1. Selection cassette-mediated aberrant transcription events are not specific to this locus or a unique cassette but can be prevented by insertion of transcription insulators flanking the cassette. Our study suggests the existence of confounding off-target mutant phenotypes resulting from misregulated neighboring loci following genome edits in yeast. Furthermore, features of LUTI-based regulation are broadly conserved to eukaryotic organisms which indicates the potential that similar misregulation could be unnoticed in other edited organisms as well.

Introduction

Genome engineering to create targeted gene deletions, mutations, and epitope-tagged proteins is a key strategy for the mechanistic dissection of almost any biological process. Budding yeast was the first eukaryotic organism in which this became highly facile, thanks to the development of a one-step PCR-based editing strategy, frequently used with a shared toolkit of selection cassettes1–5. These tools accelerated the use of yeast as a model organism and have become common, routinely used by thousands of labs, as well as enabling large global endeavors, like creation of the yeast deletion collection6. The utility of this approach stems from the ease of creating targeted mutations, simply by designing primers that amplify a cassette of choice while adding homologous sequences that flank the genomic region to be edited1-5. After a cassette is amplified, the linear PCR product is transformed into yeast and incorporated into the targeted site within the genome by cellular homologous recombination machinery1–5. Finally, selection for a cassette-encoded resistance marker allows one to isolate a clonal population of edited cells. Cassette incorporation is highly specific to the targeted genomic location and mutant strains created in this manner are typically regarded with high confidence1–5. Because of the ease of use and rapid nature of this selection-mediated editing strategy, it has remained commonplace despite the development of CRISPR/Cas9-based editing methods which allow the creation of “seamless” non-selection-marked mutations7.

While selection-mediated editing strategies have been used to advance countless discoveries in yeast, their utility relies on the assumption that insertion of a selection cassette at one locus will not disrupt the expression of neighboring loci. However, a growing body of evidence suggests that yeast genomic loci are not the simple discrete units they were long thought to be, but instead can have complex and linked transcriptional outputs. For example, it is now understood that activity from one transcription start site (TSS) can interfere with the output of nearby TSSs in an adjacent sense or antisense configuration, and that many promoters are bidirectionally active8–19. Transcription interference is seen in diverse organisms, including yeast and human cells, and can occur between TSSs driving coding or noncoding RNAs and controlling both transcript isoform identity and transcript levels15,15,20–29. It has also been shown that more than one TSS is often present at a single locus, even in the simple budding yeast14.

Transcription interference between TSSs within the same genomic locus is an important feature of a naturally occurring type of regulation dependent on LUTIs (long undecoded transcript isoforms)20,30. For genes regulated by this strategy, 5’ extended LUTIs are transcribed in place of canonical mRNAs to temporally downregulate protein synthesis20,30–32. Here, use of an upstream alternate TSS represses transcription from the canonical TSS in cis-via transcriptional interference20,30. In concert, translation of the main ORF-encoded protein products from LUTIs are repressed by translation of competitive AUG-initiated upstream ORFs (uORFs)30,33–37. Natural LUTI-based regulation has been shown to modulate expression for many genes during meiosis and the unfolded protein response in yeast30–32. Features of LUTI-based regulation are conserved across eukaryotes, and evidence suggests this regulation may also occur in more complex eukaryotes such as during mammalian embryonic stem cell differentiation and larval development in Drosophila melanogaster38,39. Interestingly, we find that insertion of resistance cassettes commonly used to mark genomic edits in yeast, can induce the repression of neighboring genes though synthetic and constitutive LUTI-based regulation.

We report that standard selection-cassette replacement of the DBP1 ORF causes mis-regulation of the adjacent gene MRP51, by aberrant LUTI-based repression, highlighting an overlooked side-effect of this genome engineering strategy. Every phenotype we observed in cells with a cassette-mediated deletion of the DBP1 ORF was caused by cassette-induced mis-regulation of the neighboring gene MRP51, rather than loss of DBP1. Non-specific cassette-induced phenotypes were general to all selection cassettes that we tested to replace the DBP1 ORF. These data suggest the standard method for making selection-marked mutations in yeast may drive unexpected side effects that have thus far been overlooked. Consistent with this, we found that cassette driven aberrant transcripts are not unique to the DBP1 locus and can be detected in roughly 25% of loci edited in this manner. While this situation may suggest the use of “seamless” genome editing methods to avoid cassette-driven artifacts, we find that even CRISPR/Cas9 unmarked mutations can introduce surprising effects on regulation of mutated loci. This, as well as the increased ease of use of selection-marked mutations, motivated us to design and test a panel of “insulated” cassettes that prevent aberrant transcription from interfering with neighboring genes in vivo. Use of these cassettes in yeast should improve the specificity of engineered mutant strains for future studies. More broadly, our study highlights a long-overlooked and potent side effect that can result from genome-editing, and that may impact published studies performed both in yeast and more complex eukaryotes.

Results

Insertion of a resistance cassette at the DBP1 locus causes aberrant transcription and reduced protein production from the neighboring gene MRP51

We had previously observed upregulation of the Dbp1 helicase in meiotic yeast cells, coincident with downregulation of its paralog, Ded140. Ded1 is important for translation initiation, and has been well studied under conditions of mitotic exponential growth41,42. In contrast, the role of Dbp1 has remained less clear, partially a result of its absent or low expression under commonly studied laboratory conditions. Based on the identity of Dbp1 as a RNA helicase43, we predicted that Dbp1 could have a role in regulating gene expression during meiosis. To test this, we deleted DBP1 by replacing the ORF (from start to stop codon) with a standard cassette encoding Geneticin (G418)-resistance, amplified from the pFA6a-KANMX6 plasmid1, in the SK1 budding yeast strain background44. We then compared gene expression profiles of wild-type and dbp1Δ::KANMX6 cells during meiosis, using ribosome profiling and mRNA-sequencing (mRNA-seq). To our surprise, MRP51, the gene located directly adjacent to DBP1, exhibited a profound decrease in translation in the dbp1Δ::KANMX6 strain relative to wild-type controls (Figure 1A). Translation of MRP51 was 8.4-fold lower in dbp1Δ::KANMX6 cells despite a 1.8-fold increase in MRP51 mRNA abundance (Figure 1A-B). These changes reflected a 15-fold decrease in MRP51 translation efficiency (TE) in the dbp1Δ::KANMX6 cells compared to the wild-type control (Figure 1C; TE: ribosome footprint FP RPKM/ mRNA RPKM) and led to a ∼40% reduction in Mrp51 protein level compared to wild type, as assessed by western blot (Figure 1D). Because of the close genomic proximity of these genes, we hypothesized that this change in TE could be due to cis effects from the dbp1Δ::KANMX6 insertion rather than regulation of MRP51 by Dbp1 protein. Upon closer examination of the mRNA transcripts produced at this locus, we found that replacing the DBP1 ORF with the KANMX6 resistance cassette led to production of a 5’ extended MRP51 mRNA compared to that in wild-type cells. The extended transcript contained 3 AUG-initiated upstream ORFs (uORFs) not present on the wild-type transcript, which were translated at the expense of the MRP51 ORF in dbp1Δ::KANMX6 cells (Figure 1E).

Figure 1.
Figure 1. Insertion of a resistance cassette at the DBP1 locus causes aberrant transcription and reduced protein production from the neighboring MRP51 gene.

(A) Translation (ribosome profiling FP) and (B) mRNA-seq reads per kilobase million mapped reads (RPKM) for every ORF expressed in wild-type and dbp1Δ::KANMX6 cells. (C) Fold-change of translation efficiency (TE: FP RPKM/mRNA RPKM) for all expressed genes. For RPKM for all quantified genes, see Figure 1 – Source Data 1. (D) Western blots of mitotic yeast demonstrate that Mrp51 protein level is decreased to 58% that of wild type in dbp1Δ::KANMX6 cells. Mrp51 levels were normalized to alpha tubulin. For full blot scans see Figure 1 – Source Data 1. (E) mRNA and FP reads mapped to the DBP1/MRP51 loci in wild-type (grey/black) and dbp1Δ::KANMX6 (purple) cells. Note that MRP51 encoding transcripts are 5’ extended in the dbp1Δ::KANMX6 cells compared to wild type (see mRNA tracks). The extended transcript contains 3 AUG initiated uORFs translated at the expense of the MRP51 ORF (see FP tracks and inset). (F) Model: replacement of DBP1 ORF with a resistance cassette causes aberrant expression of a long undecoded transcript isoform (LUTI) for MRP51, which results in lower Mrp51 protein expression as an off-target effect.

These data suggest that KANMX6 cassette insertion at the DBP1 locus drives the transcription of an aberrant 5’ extended MRP51 mRNA containing repressive uORFs and causing reduced Mrp51 protein production. Thus, the features of MRP51 mis-regulation in this strain, mimicked the natural LUTI-based mode of regulation that conditionally downregulates protein synthesis from many genes in yeast (Figure 1F)20,30–32. Here, the primary difference was that repression of MRP51 expression appears to occur constitutively as result of KANMX6 insertion at the neighboring DBP1 ORF rather than as a result of temporally regulated toggling between two TSSs.

Because of the widespread and long-term use of cassette-mediated gene deletion in yeast, we were surprised to observe this dramatic off-target effect. To confirm aberrant MRP51 regulation was not an artifact of our strain, experimental conditions, or selection cassette, we analyzed a published ribosome profiling and mRNA-seq dataset comparing wild type and dbp1Δ cells of the S288C budding yeast background. These dbp1Δ cells were generated by replacement of the DBP1 ORF with a cassette encoding resistance to Hygromycin (dbp1Δ::HYGMX4)45. Consistent with our data, insertion of HYGMX4 to replace the DBP1 ORF caused mis-regulation of the adjacent gene MRP51. Despite levels of the Mrp51-encoding transcript remaining similar to wild type in dbp1Δ::HYGMX4 cells, translation of MRP51 was decreased 4.4-fold in these conditions (Figure 1—figure supplement 1A-B). The 3.7-fold decrease in TE of MRP51 in dbp1Δ::HYGMX4 cells in this study led to the interpretation that the canonical MRP51 transcript is highly dependent on Dbp1 for its translation (Figure 1—figure supplement 1C). A closer look at the transcripts produced from this locus, however, revealed the presence of a 5’-extended MRP51 LUTI in the dbp1Δ::HYGMX4 cells, like the one we had observed. This 5’ extended MRP51 transcript contained competitive AUG-initiated uORFs that were translated in place of the MRP51 ORF, as in our experiments (Figure 1—figure supplement 1D). Together these data indicate that cassette replacement of the DBP1 ORF drives aberrant transcription independent of strain, cassette selection identity, and experimental conditions. Instead, the observed mis-regulation of MRP51 appears to be a general side-effect of the cassette-insertion.

Figure 1—figure supplement 1.
Figure 1—figure supplement 1. A published dataset confirms aberrant transcription and mis-regulation of MRP51 following cassette-mediated DBP1 replacement.

Ribosome profiling and mRNA-seq of dbp1Δ::HYGMX4 and wild-type S288C budding yeast during mitotic growth from Sen et al., 2019. (A) Translation (FP RPKM) and (B) mRNA RPKM counts of all genes expressed. (C) Fold-change TE for every gene expressed in dbp1Δ::HYGMX4 and wild-type cells. (D) mRNA and FP reads aligned to the DBP1/MRP51 loci for wild type (grey/black) and dbp1Δ::HYGMX4 (purple). Note the MRP51 mRNA is 5’ extended in the dbp1Δ::HYGMX4 cells (mRNA tracks) and the translation of uORFs not present on the wild-type mRNA (FP tracks).

Mis-regulation of MRP51 by insertion of a resistance cassette leads to systemic phenotypic consequences

We next sought to determine whether the mutant phenotypes that we observed in dbp1Δ::KANMX6 cells were due to loss of DBP1 function or resistance cassette-dependent mis-regulation of MRP51, a nuclear encoded mitochondrial small subunit ribosome protein (mt-SSU)46. Polysome analysis of dbp1Δ::KANMX6 cells during meiosis demonstrated that overall translation rates were diminished relative to wild-type cells, made evident by lower polysome peaks in the mutant (Figure 2A; right, fractions 5-6). Consistent with this finding, measurement of radioactive amino acid incorporation rates revealed a 24% decrease in bulk translation in dbp1Δ::KANMX6 cells (Figure 2A—figure supplement 1A). Polysome traces also indicated differences in the accumulation of a ribonucleoprotein (RNP) species roughly the size of the cytoplasmic large 60S subunit (Figure 2A; LSU, fraction 3) and a decrease in cytoplasmic small 40S subunit signal (Figure 2A; SSU, fraction 2). We next performed label-free mass spectrometry on fractionated wild-type and dbp1Δ::KANMX6 polysomes and used hierarchical clustering of the data to assess global differences in polysome composition. To our surprise, we identified a prominent cluster of proteins enriched in fraction 3 of polysomes from dbp1Δ::KANMX6 cells compared to wild type, that was highly enriched for the mitochondrial large ribosome subunit (mt-LSU). This cluster contained 34 of the 42 proteinaceous mt-LSU components quantified in our study (Figure 2A; left). Further analysis of fraction 3 revealed every mitochondrial 54S protein component quantified in this fraction was enriched in dbp1Δ::KANMX6 cells compared to the wild-type control (Figure 2—figure supplement 1B). These data suggest the accumulated RNP observed in dbp1Δ::KANMX6 cells is the mt-LSU. A previous study from our lab observed that when individual cytoplasmic SSU proteins are lost, free cytoplasmic 60S LSUs accumulate, presumably a result of their inability to find a SSU to complex with and form fully assembled 80S species.47 The build-up of free mt-LSU in dbp1Δ::KANMX6 cells, which express lower amounts of mitochondrial small subunit (mt-SSU) component Mrp51 is reminiscent of that effect. Furthermore, we identified a small cluster of cytoplasmically translated mitochondrial proteins that were decreased in dbp1Δ::KANMX6 high polysome fractions compared to wild type, hinting at effects of broader mitochondrial dysfunction. These data suggested the cassette-mediated mis-regulation of MRP51 in dbp1Δ::KANMX6 cells may have been responsible for the mutant phenotypes we observed, rather than loss of Dbp1 protein.

Figure 2.
Figure 2. Resistance cassette-driven mis-regulation of MRP51 causes widespread phenotypic consequences.

(A) Label free mass spectrometry of fractionated polysomes from wild-type and dbp1°::KANMX6 cells during meiosis. Note decreased polysomes, accumulation of mitochondrial 54S subunit (fraction 3), and a decrease in cytoplasmic 40S peak (fraction 2) for dbp1°::KANMX6 cells. Two biological replicates were analyzed, one is shown. Replicates and their agreement by spearman correlation are shown in figure supplement. For full data see Figure 2 – Source Data 1. (B-E). Analysis of wild type, dbp1°::KANMX6, and dbp1°::KANMX6 leu2::pDBP1-DBP1::LEU2 under conditions where dbp1°::KANMX6 phenotypes were observed. (B) Growth curves in the non-fermentable carbon source glycerol, and ethanol (C: post-diauxic). (D) 24h sporulation efficiency counts, data shown is the average of 3 biological replicates with error bars showing SEM. Statistical significance was determined by a two-way ANOVA adjusted for multiple comparisons with Tukey’s multiple comparison test. (E) Polysome profiles from meiotic cells. Two biological replicates were analyzed, one representative trace is shown. (F) Growth curves of wild type compared to mutant strains where DBP1 was replaced by varied resistance cassettes in the non-fermentable carbon source glycerol.

Disruption of mitochondrial translation, and ultimately function, would be expected to lead to global cellular fitness defects. Based on the apparent mitochondrial defects observed in the mass spectrometry data, we hypothesized that dbp1Δ::KANMX6 cells would grow poorly in conditions where elevated mitochondrial function is required. Consistently, dbp1Δ::KANMX6 cells exhibited severe growth defects in media containing only non-fermentable carbon sources, such as glycerol or acetate (Figure 2—figure supplement 1D). Additionally, growth defects were evident in dbp1Δ::KANMX6 cells grown in rich media (YEP dextrose or sucrose) during post-diauxic growth stages, after saturated cultures have exhausted fermentable carbon source availability and begin to utilize ethanol through respiration (Figure 2—figure supplement 1E). To determine whether these phenotypes were due to cassette-induced MRP51 mis-regulation, and not caused by loss of Dbp1 protein, we tested whether they could be rescued by exogenous expression of Dbp1. We expressed Dbp1 at endogenous levels in the dbp1Δ::KANMX6 cells by integrating a single copy of the DBP1 ORF and all its regulatory sequences at the LEU2 locus. All dbp1Δ::KANMX6 growth defects observed in conditions requiring respiratory growth remained similar with and without Dbp1 (Figure 2B-C), confirming the respiratory growth defects observed in dbp1Δ::KANMX6 cells result from cassette insertion rather than lack of Dbp1 protein.

We next assessed whether our other observed phenotypes in dbp1Δ::KANMX6 cells were also due to off-target effects, rather than loss of Dbp1 protein. Indeed, the sporulation defect observed in dbp1Δ::KANMX6 cells was not rescued by Dbp1 (Figure 2D), which we now believe to result from the dependency of meiotic cells on respiratory function48. Finally, the reduced cytoplasmic 40S signal and diminished polysome levels representing lower translation rates in meiotic dbp1Δ::KANMX6 cells was also not rescued by Dbp1 (Figure 2A, Figure 2—figure supplement 1A). Instead, we believe the bulk cytoplasmic translation phenotypes may reflect a cellular response to mitochondrial dysfunction or reduced mitochondrial translation rates, both of which have been shown to interact with the TOR pathway thus influencing global translation49–52. We conclude that resistance cassette insertion at the DBP1 locus is sufficient to cause widespread cellular phenotypes not associated with loss of DBP1 function. These data raised concerns as they confirm that resistance cassette-driven cis effects can confound the interpretation of phenotypes in mutants created in this manner. While these results were disappointing regarding our study of the function of Dbp1 helicase, they highlighted this locus as a useful context to dissect the mechanism behind, and ideally fix, resistance-cassette-related side effects.

To assess whether cassette-driven off-target effects were common to a specific feature of the KANMX6 and HYGMX4 selection cassettes, we replaced DBP1 ORF in wild-type cells with three additional cassettes. Among the cassettes tested, we varied the promoter and terminator sequences used to express the resistance ORF, as well as the identity of the resistance ORF. Insertion of every resistance cassette tested lead to a similar growth defect in a non-fermentable carbon source (Figure 2F), a phenotype which was not dependent on Dbp1 (Figure 2B). We concluded that the cis effects exhibited with DBP1 replacement were not specific to any one cassette and instead are a consequence of the resistance-mediated gene replacement approach.

Figure 2—figure supplement 1.
Figure 2—figure supplement 1. DBP1 ORF replacement with KANMX6 causes diverse phenotypes resulting from mitochondrial dysfunction.

(A) S35 amino acid incorporation of dbp1°::KANMX6 and wild-type cells during meiosis. Statistical significance determined by an ordinary one-way ANOVA (p value <0.05). Data is average of three biological replicates; error bars show SEM. (B) Spearman correlation of biological replicates from label free quantification (LFQ) of fractionated polysome mass spectrometry experiment. (C) Normalized LFQ of all proteins quantified in fraction 3 shown as average of both replicates. (D-E) Growth curves of wild-type and dbp1°::KANMX6 cells in non-fermentable carbon sources (D) and fermentable (E: before diauxic shift) and non-fermentable (E: post diauxic shift) carbon sources.

Resistance cassette-induced aberrant transcription events are not specific to the DBP1 locus

Because of the dramatic off-target effects observed with cassette replacement of DBP1, we wanted to determine if cassette insertion could drive aberrant transcription and neighboring gene misregulation at other loci. To assess this, we observed mRNA-seq reads over genomic intervals surrounding replaced ORFs, using previously published mRNA-seq and ribosome profiling data from our lab47 as well as an unpublished mRNA-seq dataset, also from our lab. All strains were made by replacing ORFs with either the KANMX6 or NATMX6 resistance cassettes and samples were collected during mitotic exponential growth in rich media. We found aberrant transcripts stemming from cassette-mediated gene replacements at 5 of these 19 additional observed loci (Figure 3 A-E). For each of these 5 new examples, the aberrant transcript appeared to be driven from the 5’ end of the resistance cassette, like in the case of dbp1Δ::KANMX6 (Figure 1E, Figure 3 A-E). For 4 of these new cases in which aberrant transcription resulted from a cassette-based deletion, the adjacent gene was positioned in the same orientation as the aberrant TSS thus it would be possible for synthetic LUTI-based repression to occur as was observed for MRP51. However, unlike the strong mis-regulation of MRP51 seen with DBP1 ORF replacement, these aberrant transcription events had no obvious effect on the translation of the adjacent genes, consistent with the lack of apparent uORF translation in the extended 5’ mRNA regions (Figure 3A-D). Although, we also note that for 3 of these 4 cases, expression of the adjacent ORF was low, even in wild-type cells, and thus misregulation of these genes may be difficult to detect under these conditions compared to what we observed for MRP51 (Figure 3B-D). For the last case, in which UPF1 was replaced with the NATMX6 cassette, an abundant aberrant transcript was observed. However, the adjacent ISF1 gene was oriented antisense to this transcript and there was no evidence of its transcription with or without NATMX6 insertion at UPF1 (Figure 3E); hence, there was no baseline expression to disrupt. Together, these data suggest that cassette-induced-aberrant transcription events are not a phenomenon unique to the DBP1/MRP51 genomic context, but that such severe phenotypic consequences may be more rare.

Figure 3.
Figure 3. Resistance cassette-induced aberrant transcripts are not specific to the DBP1 locus.

(A-D) 18 ribosomal protein deletions generated by ORF replacement with KANMX6 from Cheng et al., 201847, were observed for aberrant transcription events. Left, ribosome profiling footprint (FP) and mRNA-seq reads aligned to the ORF-replaced locus for each strain. Right, FP RPKM for every ORF expressed in the mutant and wild-type strains. Data was collected from vegetative exponentially growing cells. (D) Note: RPL41B is not quantified in translation graph because the ORF is only 25 amino acids long. (E) mRNA-seq reads aligned to the replaced locus in a upf1°::NATMX6 mutant. Data was collected from vegetative exponentially growing cells. No matched ribosome profiling data was available for this study. See Figure 3 – Source Data 1.

The finding that deleting an ORF by replacement with a resistance cassette can cause mis-regulation of adjacent genes and potent off-target effects may seem to motivate the use of marker-less gene editing strategies for future studies. However, when we created a marker-less deletion of the DBP1 ORF using CRISPR/Cas9 editing7 we observed surprising effects on the expression of surrounding genes as well. “Seamless” deletion of the DBP1 ORF by this strategy led to the production of a transcript containing only the 5’ and 3’ UTRs of DBP1. Expression of this mutant transcript led to the aberrant translation of a dubious ORF contained within the DBP1 3’ UTR, not translated in wild type (Figure 3—figure supplement 1). These data, combined with the greater versatility of selection-marked mutations, highlight the need for improved modes of generating resistance marked mutations.

Figure 3—figure supplement 1.
Figure 3—figure supplement 1. Cas9-mediated “seamless” deletion of DBP1 ORF causes aberrant translation of an ORF within the dbp1° 3’UTR.

(A) Ribosome profiling footprint (FP) and mRNA-seq reads mapped to the DBP1 locus in wild-type and dbp1° (ORF removed with no resistance marker inserted) cells. Samples were collected during meiosis. (B) Translation (FP RPKM) for all genes expressed in dbp1° and wild-type cells during meiosis. See Figure 3—figure supplement 1-Source Data 1 for all quantified genes.

Insulated resistance cassettes prevent the mis-regulation of insertion adjacent genes

Because bi-directional transcription has been shown to occur from many promoters8,11,53, and all observed aberrant transcripts came from the end of the cassette containing the resistance ORF promoter, we thought this promoter was the most likely cause of the aberrant transcription events. Thus, to circumvent these events from disrupting nearby genes, we placed strong transcription terminator sequences flanking both ends of the resistance cassette54. We placed a copy of the DEG1 terminator 5’ to the TEF1 promoter, and either the same DEG1 terminator or the CYC1 terminator 3’ to the TEF1 terminator within the KANMX6 cassette (Figure 4A). We named these plasmids “KANMX6-ins” and used them to replace the DBP1 ORF using the same primers and insertion location as before. Both dbp1Δ::KANMX6-ins strains grew at wild-type rates in the non-fermentable carbon source glycerol, in contrast to cells housing the original cassette replacement (Figure 4B). Consistent with this finding, levels of Mrp51 protein in the dbp1Δ::KANMX6-ins strains were similar to those in wild-type cells (Figure 4C, D). Additionally, 5’ rapid amplification of cDNA ends (5’RACE) confirmed that the MRP51 transcript produced in the dbp1Δ::KANMX6-ins strains was the same as that seen in wild-type cells, as opposed to the synthetic MRP51 LUTI in the initial cassette-inserted strains (Figure 4E—figure supplement 1A-B). We were pleased that this simple change to the resistance cassette structure prevented the neighboring gene mis-regulation and resulting systemic phenotypes seen previously with DBP1 ORF replacement. These data suggest that this strategy may prove useful to prevent aberrant transcription and neighboring gene misregulation in future studies. We thus created insulated versions of cassettes housing three additional commonly used selection markers TRP1, HIS5 (S. Pombe HIS5; which complements his3), and NAT (nourseothricin resistance) into the classic “MX6” backbone1-3. We believe that use of these insulated cassettes in place of the traditional set should drastically reduce off-target effects from future genetic studies.

Figure 4.
Figure 4. Insulated resistance cassettes prevent cassette-induced mis-regulation of adjacent genes.

(A) Design of “insulated” KANMX6-ins cassettes and proposed model. The DEG1 transcription terminator sequence was inserted 5’ of the TEF1 promoter and either the DEG1 or CYC1 terminator was inserted 3’ of the TEF1 terminator in pFA6a-KANMX61. (B) Growth of dbp1°::KANMX6 and dbp1°::KANMX6-ins strains compared to wild type in YEP-glycerol. (C) Western blots for Mrp51-3v5 levels in dbp1°::KANMX6 and dbp1°::KANMX6-ins cells compared to wild type during mitotic exponential growth in rich media. For full blot scans see Figure 4 – Source Data 1. (D) Quantification of western blots in (C) with Mrp51 normalized to Tub1 (alpha tubulin). Data represent average of two biological replicates with error bars showing range. Statistical significance was determined by a one-way ANOVA adjusted for multiple comparisons using Tukey’s multiple comparisons test, P values less than .005 are shown. (E) 5’ rapid amplification of cDNA ends (RACE) demonstrates the 5’ ends of transcripts produced in wild type, dbp1°::KANMX6, and dbp1°::KANMX6-ins strains. Exact 5’ mRNA end sequences listed in figure supplement.

Figure 4—figure supplement 1.
Figure 4—figure supplement 1. 5’ RACE confirms the presence of wild-type 5’ end of MRP51 transcript when DBP1 ORF is replaced with KANMX6-ins cassettes.

(A) PCR amplified 5’ ends of cDNA from wild type, dbp1°::KANMX6, dbp1°::NATMX6, dbp1°::tDEG1-KANMX6-tDEG1, and dbp1°::tDEG1-KANMX6-tCYC1. Colored dots represent the color each sample is depicted with in Figure 4E. Bottom image is darker exposure of the same gel to demonstrate lack of MRP51 LUTI in insulated cassette samples. See Figure 4 – figure supplement 1 – Source Data 1 for uncropped images of gels. (B) The exact 5’ ends of transcripts sequenced to identify the start site of each transcript as well as the number of reads sequenced at each position.

Discussion

In this study, we provide evidence that resistance-cassette-mediated genome engineering in budding yeast can result in overlooked and potent off-target effects. Dissection of the mechanism behind these off-target effects at the DBP1/MRP51 loci demonstrate their potential to ablate expression of neighboring genes through use of cassette-induced alternative TSSs and production of repressive LUTIs. This regulation, integrates transcription interference and translation control, and occurs despite the ORF and normal regulatory sequences for the affected adjacent gene remaining intact at the genomic level20,30,31. This, combined with condition-specific manifestation of phenotypes caused by off-target effects, as was seen at the DBP1/MRP51 locus, can make identification of the off-target effects difficult. Had we not generated mRNA-seq and ribosome profiling data and been aware of the features of naturally occurring LUTI regulation, we would have misinterpreted mutant phenotypes in the dbp1Δ::KANMX6 cells as relevant to Dbp1 function. Additionally, the well characterized function of Mrp51 allowed us to test whether the cassette-mediated mis-regulation of MRP51 was sufficient to cause cellular phenotypes. We believe cassette-induced mis-regulation of adjacent genes may be a more elusive problem in studies where genome-wide gene expression information is not generated and that this side effect can cause misinterpretation of mutant phenotypes.

How common is this unexpected effect? Confidence in the specificity of yeast genome-editing cassettes has motivated their widespread use and many landmark studies defining the functions of conserved proteins have been performed using this approach. It is routine for multiple loci to be disrupted within the same strain to enable analysis of genetic interactions, and global yeast deletion, hypomorphic, and epitope tag collections were created by use of this exact insertion cassette55–61. Our ability to identify cassette-driven aberrant transcripts in previous studies from ourselves and others demonstrate that this side effect can be induced with a variety of cassettes and at multiple loci. These data suggest this type of genome engineering artifact may be an overlooked and relevant factor in the interpretation of mutants from many published studies. Of the 20 selection-cassette-mediated loss-of-function mutants for which our lab had generated mRNA-seq data, 6 produced a clear aberrant transcript (Figure 1E, Figure 3 A-E). However, none of the 5 additional cases resulted in severe effects on translation of neighboring genes, as was seen with DBP1/MRP51.

We believe the dramatic shut down of Mrp51 synthesis resulted from the convergence of multiple genomic features allowing for maximal mis-regulation via the LUTI-based mechanism. These features included a gene antisense and 5’ to the cassette insertion, translated competitive AUG uORFs, and lack of an intervening transcriptional terminator between the replaced and adjacent ORF. The combination of all these features is not common to most loci but would be expected to be seen in many cases genome-wide. It is also worth noting that phenotypic consequences resulting from cassette mis-regulation of adjacent genes may be condition-specific, as was the case with MRP51. Exponentially growing dbp1Δ::KANMX6 cells in rich media grow at wild-type rates despite having lower levels of Mrp51 protein, presumably a result of the reduced need for mitochondrial function under these conditions (Figure 2C: prior to the diauxic growth phase). Under conditions requiring elevated mitochondrial function, the MRP51 mis-regulation becomes limiting and widespread phenotypes occur (Figure 2A-F, Figure 2— figure supplement F-G). The cassette-induced-aberrant transcription events reported here that did not have evident mis-regulation of adjacent genes under the single condition we assayed may cause further adjacent gene disruption and phenotypic consequences under conditions not tested in this study.

While we focused on the potential for integrated transcription and translation LUTI-based neighboring-gene disruption in this study, mis-regulation of adjacent genes by cassette-driven transcription interference alone may happen more frequently. This mis-regulation can occur regardless of the adjacent gene’s direction in relation to the cassette and does not require the transcript to be continuous over the adjacent TSS. For example, in haploid cells induction into meiosis is blocked by the expression of two separate non-coding RNAs that repress an antisense mRNA (RME2/IME4) or a downstream sense RNA (IRT1/IME1) via transcription interference. Transcription interference between loci can be effective over a large distance, like the ∼1800 nt span over which transcription of the repressive IRT1 long noncoding RNA represses IME122. Considering the number of factors that can affect whether a gene adjacent to a resistance cassette could be mis-regulated, we believe it is hard to predict when and where the side effect reported here would result in misregulation severe enough to mislead studies. Two recent studies analyzing the non-essential yeast deletion collection show clear signs of the cassette driven off-target effects reported here and indicate that this effect may be more common than our analyses suggest. One study found that one of the most significant factors determining whether a specific protein’s abundance would change following deletion of a gene encoding another protein was whether the genes encoding the protein pair were genomic neighbors62. Another study, which analyzed genetic interaction data, determined that neighbor gene effects of unknown mechanism appeared to account for non-specific phenotypes in ∼ 10% of deletion strains63.

What is driving aberrant transcription from selection cassette insertion? It has been recently shown that most promoters in yeast are bidirectional and that resulting antisense transcripts are often not evident due to their targeting by mRNA decay mechanisms8,12,53. We believe the aberrant transcripts in our study arise from bidirectional transcription stemming from promoters within the cassette used to express resistance ORFs. Instability of such antisense transcripts in RNA-decay-proficient backgrounds may explain why we do not observe aberrant transcripts driven by cassettes in 14 of the 20 cases we analyzed. Thus, we may be underestimating the frequency by which adjacent genes are affected by cassette-mediated transcription interference. Alternatively, it is possible that the chromatin context of cassette insertion controls the bidirectional activity of the cassette promoters. Studies using resistance cassette generated mutants in strain backgrounds incompetent for mRNA decay will be informative in deciphering between these possibilities. Interestingly, the upf1Δ::NATMX6 antisense transcript produced appeared to be more abundant than many of our other examples (Figure 3E compare to A-D). Upf1’s role as an RNA helicase that drives degradation of mRNAs through the non-sense mediated decay pathway64,65 could be a relevant factor to investigate in future studies.

A major challenge presented by the cassette-mediated off-target effects we found in this study is that they are difficult to detect. PCR checking, sequencing, and backcrossing of cassette-inserted strains are common approaches to identify or prevent potential non-specific mutations and phenotypes, but none of these strategies can prevent or detect cassette-driven aberrant transcription. No unexpected mutations exist in these cases and even mRNA measurements of neighboring genes are not diagnostic unless an approach is designed to assess all transcript isoforms from a given locus. We find that measurement of translation or protein levels from neighboring genes can be diagnostic of this undesirable side-effect, but ribosome profiling is not a routine experiment for analysis of most mutants, and tagging of many proteins does not retain their normal function and regulation. Additionally, as many genes are expressed in a condition-specific manner, comparison of the adjacent gene’s protein levels in a mutant as compared to wild type may give false confidence unless the proper conditions are chosen. Here, for 3 of the 6 mutants with observable aberrant transcripts that could disrupt adjacent genes via LUTI-based regulation, those genes were lowly expressed, even in wild-type cells, under the conditions surveyed.

Despite this complexity, use of complementation-based rescue experiments— namely single copy insertion of the deleted gene with its entire endogenous regulatory region at another locus— allowed us to clearly identify the unexpected regulation in the case of DBP1 disruption. This type of approach should be more commonly used to confirm mutant phenotypes, even in “seamless” deletes, but may be complex or unfeasible in cases in which cassette insertion is used for epitope tagging or experiments in which genetic interactions between several genes are being investigated.

We suggest the future use of “insulated” resistance cassettes to preventing off-target effects and misinterpretation of mutant phenotypes in future genome editing experiments. Precision Cas9-based ORF deletion is a also powerful approach to genome editing, but can also result in changes to the chromatin context of an adjacent gene, potentially resulting in LUTI-based repression13,7. For the case of a Cas9-generated marker-less deletion of the DBP1 ORF, we observed translation of a previously untranslated ORF within the mutant transcript containing only the DBP1 5’ and 3’ UTR (Figure 3—figure supplement 1). One could imagine that in cases like these, translation of novel ORFs could also cause misleading neomorphic mutant phenotypes not linked to the deleted ORFs loss of function. While our study was limited to yeast, all factors behind the cassette-induced off-target effects identified here are conserved to higher eukaryotes and we believe genome-editing-driven LUTI-based regulation is likely to occur in other organisms. Our data highlight this as a previously overlooked side-effect of genome-editing which should be considered in the analysis of edited cells. They also emphasize the need for proper controls to establish causality of observed phenotypes to the intended mutation, even when using precision genome engineering approaches in yeast and other organisms.

Materials and Methods

Strains

All strains were derived from the Saccharomyces cerevisiae SK1 background44. Detailed strain genotypes can be found in supplemental table 1. Strains were diploid MATa/alpha unless otherwise specified as haploid (MATa). Selection-mediated loss of function mutants were created by transforming wild-type haploids with a linear PCR fragment encoding a selection cassette that would replace the ORF of interest via homologous DNA repair mechanisms(1-5). Sequences directly upstream and downstream of each replaced ORF were always left intact. Genotypes for all strains were confirmed using PCR and mutants were backcrossed to a wild-type strain of the opposite mating type to ensure clearance of possible background mutations from transformation. The marker-less dbp1Δ was created by transforming a wild-type haploid with a plasmid encoding Cas9 and a sgRNA targeting the DBP1 ORF, alongside a linear repair fragment that removes the DBP1 ORF while preserving the upstream and downstream adjacent sequence. Strains containing the Mrp51-3v5 allele were created by transforming wild type or the selection replaced dbp1Δ strain of interest with a plasmid encoding Cas9 and a sgRNA targeting MRP51, alongside a linear repair template that c terminally tags Mrp51 with the 3v5 sequence and introduces a synonomous mutation in the guide target sequence to prevent re-editing. Strains expressing exogenous Dbp1 from the LEU2 locus were created by cloning the DBP1 ORF as well as ∼1kb of upstream and downstream regulatory sequence into a single integration plasmid targeted to LEU2 This plasmid was linearized by digestion with SwaI nuclease (NEB, Ipswich, MA) and transformed into the dbp1Δ::KANMX6 strain.

Yeast Growth Conditions

Yeast were grown for experiments as in (Powers et al., 2021). Briefly, strains were thawed on YPG plates from glycerol stocks overnight, then grown on YPD plates for a day before use in experiments. All growth experiments were carried out at 30°C using YEP media supplemented with 2% w/v of the carbon source, except for glycerol which was used at 3% v/v. All growth experiments shown were repeated at least twice with one representative growth curve shown for each. Synchronous sporulating cells were prepared by first growing cells in YEP 2% dextrose for 24 hours at room temperature then diluted into BYTA at OD600 = .25 and grown overnight at 30°C. Next, cells were pelleted, washed with water, and resuspended in sporulation (SPO) medium: either (0.5% KAc [pH = 7.0] supplemented with 0.02% raffinose) for Figure 1 ribosome profiling and mRNA-seq experiments, or rich SPO medium (2% KAc [pH = 7.0] supplemented with 40 mg/L adenine, 40 mg/L uracil, 20 mg/L histidine, 20 mg/L leucine, 20 mg/L tryptophan for sporulation efficiency, polysome experiments, and ribosome profiling in Figure 3—figure supplement 1. Sporulation efficiency was counted at 24 hours after induction into SPO media and 200 cells were counted per sample for 3 independent biological replicates. Samples were blinded prior to counting.

Ribosome profiling and mRNA-seq

Meiotic yeast cells were harvested after 4 or 6 hours in sporulation medium (Figure 1 data: 4 hours; Figure 3—figure supplement 1: 6 hours), by brief treatment with 100uM then flash freezing. They were subjected to ribosome profiling and mRNA-seq as described in37. Reads per ORF were determined following mapping of all reads to the SK1 genome. RPKM values for ribosome footprints (FP) and mRNA reads were calculated by dividing the number of raw reads per gene by the total number of million mapped reads per sample and by the length in kilobases of each gene. Translation efficiency (TE) was calculated by dividing the FP RPKM / mRNA RPKM for each gene. Data visualization of genome tracts was made using MochiView66.

Immunoblotting

Trichloroacetic acid (TCA) extractions were performed to collect total protein from samples as described in (Chen 2017)30. Briefly, ∼2.5OD600 of yeast were treated with 5% TCA at 4°C overnight, washed with acetone, dried then lysed in lysis buffer (50 mM Tris–HCl [pH 7.5], 1 mM EDTA, 2.75 mM DTT, protease inhibitor cocktail (cOmplete EDTA-free, Roche) for 5 minutes on a Mini-Beadbeater-96 (Biospec Products). Next, 3x SDS sample buffer (187.5 mM Tris [pH 6.8], 6% ß-mercaptoethanol, 30% glycerol, 9% SDS, 0.05% bromophenol blue) was added to 1X and the cell lysate was boiled for 5 min. Proteins were run on 4-12% Bis-Tris Bolt gels (Thermo Fischer) then transferred to .45uM nitrocellulose membranes using the 30 minute mixed molecular weight protocol on the Trans-Blot Turbo System from Bio-rad. Membranes were blocked with Intercept (PBS) blocking buffer (LI-COR Biosciences, Lincoln, NE) at room temperature then incubated overnight at 4°C with mouse anti-v5 (1:1000, Invitrogen, RRID:AB_2556564), and rat anti-tubulin alpha (1:10,000, Serotec, RRID:AB_325005). Membranes were washed in PBS-.08% tween then incubated with an anti-mouse secondary antibody conjugated to IR Dye 800 (RRID:AB_621842) at a 1:15,000 dilution, and either an anti-rat secondary antibody conjugated to IR Dye 680 (RRID:AB_10956590) at a 1:15,000 dilution at room temperature for 1-2h (LI-COR Biosciences). Immunoblot images were generated and quantified using the Odyssey system (LI-COR Biosciences).

Polysome Analysis

Cells were treated with 100uM cycloheximide for 30s then filter collected, flash frozen, lysed, and prepared for sucrose gradient centrifugation of as in67, with the following exceptions. Samples were thawed just prior to their loading on sucrose gradients, and SUPERase·In was added as for “mock digested samples”. Also, the polysome lysis buffer used was modified to contain protease and phosphatase inhibitors and is as follows 20mM Tris pH8, 140mM KCl, 1.5mM MgCl2, 100ug/ml cycloheximide, 1% Triton X-100, 2ug/ml Aprotinin, 10ug/ml Leupeptin, 1mM PMSF, 1:100 PIC2, 1:100 PIC3. Lysates were loaded onto 7-47% sucrose gradients and spun in a SW 41 Ti rotor for 3 hours at 35,000rcf at 4°C. Following the spin, samples were kept at 4°C and immediately collected using the Gradient Master gradient station from BioComp while monitoring 260nm absorbance with the BioRad EM-1 Economonitor. Samples were fractionated as shown in Figure 2, flash frozen, and submitted for mass spectrometry analysis.

Polysome fraction processing for LC-MS/MS measurements

Proteins were precipitated and desalted using the SP3 method, as described in (Hughes et al., 2018)68. 50% from each polysome fraction (volume) were processed. Disulfide bonds were reduced with 5mM dithiothreitol and cysteines were subsequently alkylated with 10mM iodoacetamide. Proteins were precipitated on 0.5 µg/µL speedBead magnetic carboxylated modified beads (1:1 mix of hydrophobic and hydrophilic beads, cat# 6515215050250, 45152105050250, GE) by addition of 100% ethanol in a 1:1 (vol:vol) sample:ethanol ratio followed by 15 min incubation at 25°C, 1000 rpm. Protein-bound beads were washed in 80% ethanol and proteins were digested off the beads by addition of 0.8ug sequencing grade modified trypsin (Promega) in 100mM Ammonium bicarbonate, incubated 16 hrs at 25°C, 600 rpm. Beads were removed and the resulting tryptic peptides evaporated to dryness in a vacuum concentrator. Dried peptides were further desalted by another round of SP3 precipitation; Peptides were reconstituted in 200ul of 95% Acetonitrile (ACN), followed by addition of bead-mix to a final concentration of 0.5 µg/µL and incubated 15 min at 25°C, 1000 rpm. Beads were subsequently washed in 80% ethanol, peptides were eluted off the beads in 50 µL 2% DMSO and samples were evaporated to dryness in a vacuum concentrator. Dried peptides were then reconstituted in 3% ACN / 0.2% Formic acid to a final concentration of 0.5 µg/µL.

LC-MS/MS analysis on a Q-Exactive HF

About 1 μg of total peptides were analyzed on a Waters M-Class UPLC using a 25cm Ionopticks Aurora column coupled to a benchtop Thermo Fisher Scientific Orbitrap Q Exactive HF mass spectrometer. Peptides were separated at a flow rate of 400 nL/min with a 190 min gradient, including sample loading and column equilibration times. Data was acquired in data dependent mode using Xcalibur software. MS1 Spectra were measured with a resolution of 120,000, an AGC target of 3e6 and a mass range from 300 to 1800m/z. Up to 12 MS2 spectra per duty cycle were triggered at a resolution of 15,000, an AGC target of 1e5, an isolation window of 1.6 m/z and a normalized collision energy of 27.

All raw data were analyzed with MaxQuant software version 1.6.10.43 69 using a UniProt yeast database (release 2014_09, strain ATCC 204508 / S288c), and MS/MS searches were performed with the following parameters: Oxidation of methionine and protein N-terminal acetylation as variable modifications; carbamidomethylation as fixed modification; Trypsin/P as the digestion enzyme; precursor ion mass tolerances of 20 p.p.m. for the first search (used for nonlinear mass re-calibration) and 4.5 p.p.m. for the main search, and a fragment ion mass tolerance of 20 p.p.m. For identification, we applied a maximum FDR of 1% separately on protein and peptide level. “Match between the runs” was activated, as well as the “LFQ” normalization (at least two ratio counts were necessary to get an LFQ value). We required 1 or more unique/razor peptides for protein identification and a ratio count of 2 or more for label free protein quantification in each sample. This gave intensity values for a total of 2143 protein groups across both replicates. “LFQ” normalized values were used for all subsequent analyses.

LFQ normalized values were then clustered with Cluster 3.070 and visualized with Java TreeView71. Proteins not quantified in 2 or more samples were excluded in clustering. GO enrichment was assessed on the clusters specified (June 2022) and compared to the background population of all proteins quantified in the experiment (Boyle EI, et al. 2004).

Radioactive amino acid incorporation assays

Cells were transferred to sporulation (SPO) media and incubated at 30°C for 4 hours. To metabolically label the cells 5uL of EasyTag EXPRESS 35S protein labeling mix (PerkinElmer), was added to 10mL of SPO cultures and incubated with shaking for 10 minutes at 30°C. Protein was precipitated by addition of 100uL 100% trichloroacetic acid to 900uL SPO culture and incubated at 95°C with shaking. Samples were chilled on ice then pelleted and washed with cold 10% trichloroacetic acid, followed by a wash in cold 100% ethanol. Samples were resuspended in 5mL of Econo-Safe scintillation fluid (RPI). Scintillation was counted for 2 minutes and the 35S incorporation rates were derived from counts per minute normalized to cell density between samples and normalized to wild type.

Design and testing of insulated selection cassettes

Based on the previous report that the Deg1 and Cyc1 terminators can be used as insulator sequences72, we amplified these terminators from wild-type yeast and placed them into the 5’ and 3’ ends of previously designed selection cassettes using gibson assembly just inside the primer amplification sites. We chose to conserve the primer amplification sites of these plasmids so that they would be useable with any primers designed to the previous system (pFA6a)1. Insulated plasmids were fully sequenced over cloning junctions and the entire region to be amplified and integrated into yeast during transformation. To test the ability of these insulators to prevent aberrant transcription at the DBP1 locus, the KANMX6-ins cassettes were amplified and transformed into yeast exactly as had been done for the previous KANMX6 selection cassettes used in this study. When used to replace DBP1 ORF, Mrp51-3v5 levels and 5’ RACE confirmed the efficacy of the KANMX6-ins cassettes in preventing aberrant LUTI regulation of MRP51. We then cloned 3 additional cassettes using varying selection (including Saccharomces pombe HIS5: complements S.cerevisiae HIS3, TRP1, and NATr) such that they can be used conveniently in combination as was done for the original toolkit. Descriptions of these plasmids can be found in supplemental table 1.

5’ RACE

5’ RACE was performed with total RNA extracted from samples using phenol chloroform precipitation and the GeneRacer™ Kit with SuperScript III™ RT and Topo TA Cloning™ Kit for Sequencing. For each sample, 4.8ug of total RNA was used. Random primers were used to reverse transcribe the de-capped cDNA library. The 5’ ends of the MRP51 cDNA were then amplified using the GeneRacer™ 5’ primer and a reverse gene specific primer (5’ ACCGTCCAAGCAACTCTGCCAATGTC 3’) in a reaction with Platinum™ Taq High Fidelity DNA Polymerase. SnapGene was used to visualize and align sequenced cDNA ends to MRP51 sequence.

Materials availability

All newly created materials are available upon request. Newly created “Insulated” selection cassettes will also be deposited to Addgene and made publicly available.

Data Availability

We are currently in the process of uploading all relevant newly created datasets to public databases. Repository information will be shared when available. Source data for all figures generated from newly created datasets is currently shared within the manuscript.

Competing interests

The authors declare no competing interests.

Source Data Information

Figure 1 – Source Data 1:

RPKM values for mRNA-seq and ribosome profiling data for all genes quantified in the experiment shown in Figure 1 plots.

Figure 1 – Source Data 2:

Zipped folder containing uncropped tif image of western blot shown in Figure 1 with and without labels on the relevant samples and bands.

Figure 2 – Source Data 1:

Label free quantification (LFQ) of all proteins quantified in fractionated polysome experiment shown in Figure 2 and Figure 2 – figure supplement 1.

Figure 3 – Source Data 1:

Zipped folder containing .txt (wiggle files) used to generate Figure 3E genome browser track figure. UPF1 ORF coordinates 415257 – 418173 on chr 13. 2KB upstream of UPF1 and 1 KB downstream of UPF1 are included.

Figure 3 – figure supplement 1 – Source Data 1:

RPKM values for all genes quantified in ribosome profiling experiment shown in Figure 3 – figure supplement 1.

Figure 4 – Source Data 1:

Zipped folder containing uncropped tif image of western blot shown in Figure 4 with and without labels on the relevant samples and bands.

Figure 4 – figure supplement 1 – Source Data 1:

Zipped folder containing uncropped tif image of agarose gel showing the amplified 5’ RACE products sequenced in Figure 4 and shown in Figure 4 – figure supplement 1.

Acknowledgments

We thank N. Ingolia, E. Ünal, and current members of the Brar and Ünal labs for their comments on this manuscript. This work was supported by National Institutes of Health [1R35GM134886] funding to G.A.B. E.N.P was funded by an NSF predoctoral Fellowship.

References

  1. 1.↵
    Longtine, M. S. et al. Additional modules for versatile and economical PCR-based gene deletion and modification in Saccharomyces cerevisiae. Yeast 14, 953–961 (1998).
    OpenUrlCrossRefPubMedWeb of Science
  2. 2.
    Wach, A., Brachat, A., Alberti-Segui, C., Rebischung, C. & Philippsen, P. HeterologousHIS3 Marker and GFP Reporter Modules for PCR-Targeting inSaccharomyces cerevisiae. Yeast 13, 1065–1075 (1997).
    OpenUrlCrossRefPubMedWeb of Science
  3. 3.↵
    Wach, A., Brachat, A., Pöhlmann, R. & Philippsen, P. New heterologous modules for classical or PCR-based gene disruptions inSaccharomyces cerevisiae. Yeast 10, 1793–1808 (1994).
    OpenUrlCrossRefPubMedWeb of Science
  4. 4.
    Baudin, A., Ozier-Kalogeropoulos, O., Denouel, A., Lacroute, F. & Cullin, C. A simple and efficient method for direct gene deletion in Saccharomyces cerevisiae. Nucl Acids Res 21, 3329–3330 (1993).
    OpenUrlCrossRefPubMedWeb of Science
  5. 5.↵
    Lorenz, M. C. et al. Gene disruption with PCR products in Saccharomyces cerevisiae. Gene 158, 113–117 (1995).
    OpenUrlCrossRefPubMedWeb of Science
  6. 6.↵
    Giaever, G. et al. Functional profiling of the Saccharomyces cerevisiae genome. Nature 418, 387–391 (2002).
    OpenUrlCrossRefPubMedWeb of Science
  7. 7.↵
    Jinek, M. et al. A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity. Science 337, 816–821 (2012).
    OpenUrlAbstract/FREE Full Text
  8. 8.↵
    Xu, Z. et al. Bidirectional promoters generate pervasive transcription in yeast. Nature 457, 1033–1037 (2009).
    OpenUrlCrossRefPubMedWeb of Science
  9. 9.
    Churchman, L. S. & Weissman, J. S. Native elongating transcript sequencing (NET-seq). Curr Protoc Mol Biol Chapter 4, Unit 4.14.1-17 (2012).
  10. 10.
    Wery, M. et al. Native elongating transcript sequencing reveals global anti-correlation between sense and antisense nascent transcription in fission yeast. RNA 24, 196–208 (2018).
    OpenUrlAbstract/FREE Full Text
  11. 11.↵
    Neil, H. et al. Widespread bidirectional promoters are the major source of cryptic transcripts in yeast. Nature 457, 1038–1042 (2009).
    OpenUrlCrossRefPubMedWeb of Science
  12. 12.↵
    Nevers, A. et al. Antisense transcriptional interference mediates condition-specific gene repression in budding yeast. Nucleic Acids Res 46, 6009–6025 (2018).
    OpenUrlCrossRef
  13. 13.↵
    Brooks, A. N. et al. Transcriptional neighborhoods regulate transcript isoform lengths and expression levels. Science 375, 1000–1005 (2022).
    OpenUrl
  14. 14.↵
    Pelechano, V., Wei, W. & Steinmetz, L. M. Extensive transcriptional heterogeneity revealed by isoform profiling. Nature 497, 127–131 (2013).
    OpenUrlCrossRefPubMedWeb of Science
  15. 15.↵
    Martens, J. A., Laprade, L. & Winston, F. Intergenic transcription is required to repress the Saccharomyces cerevisiae SER3 gene. Nature 429, 571–4 (2004).
    OpenUrlCrossRefPubMedWeb of Science
  16. 16.
    van Werven, F. J. et al. Transcription of two long noncoding RNAs mediates mating-type control of gametogenesis in budding yeast. Cell 150, 1170–81 (2012).
    OpenUrlCrossRefPubMedWeb of Science
  17. 17.
    Nadal-Ribelles, M. et al. Control of Cdc28 CDK1 by a stress-induced lncRNA. Mol Cell 53, 549–561 (2014).
    OpenUrlCrossRefPubMedWeb of Science
  18. 18.
    Greger, I. H. & Proudfoot, N. J. Poly(A) signals control both transcriptional termination and initiation between the tandem GAL10 and GAL7 genes of Saccharomyces cerevisiae. EMBO J 17, 4771–4779 (1998).
    OpenUrlAbstract/FREE Full Text
  19. 19.↵
    Castelnuovo, M. et al. Bimodal expression of PHO84 is modulated by early termination of antisense transcription. Nat Struct Mol Biol 20, 851–858 (2013).
    OpenUrlCrossRefPubMed
  20. 20.↵
    Chia, M. et al. Transcription of a 5’ extended mRNA isoform directs dynamic chromatin changes and interference of a downstream promoter. Elife 6, (2017).
  21. 21.
    Hirschman, J. E., Durbin, K. J. & Winston, F. Genetic evidence for promoter competition in Saccharomyces cerevisiae. Mol Cell Biol 8, 4608–15 (1988).
    OpenUrlAbstract/FREE Full Text
  22. 22.↵
    van Werven, F. J. et al. Transcription of two long noncoding RNAs mediates mating-type control of gametogenesis in budding yeast. Cell 150, 1170–1181 (2012).
    OpenUrlCrossRefPubMedWeb of Science
  23. 23.
    Hongay, C. F., Grisafi, P. L., Galitski, T. & Fink, G. R. Antisense Transcription Controls Cell Fate in Saccharomyces cerevisiae. Cell 127, 735–745 (2006).
    OpenUrlCrossRefPubMedWeb of Science
  24. 24.
    Hausler, B. & Somerville, R. L. Interaction in vivo between strong closely spaced constitutive promoters. J Mol Biol 127, 353–6 (1979).
    OpenUrlCrossRefPubMed
  25. 25.
    Adhya, S. & Gottesman, M. Promoter occlusion: transcription through a promoter may inhibit its activity. Cell 29, 939–44 (1982).
    OpenUrlCrossRefPubMedWeb of Science
  26. 26.
    Proudfoot, N. J. Transcriptional interference and termination between duplicated alpha-globin gene constructs suggests a novel mechanism for gene regulation. Nature 322, 562–5 (1986).
    OpenUrlCrossRefPubMedWeb of Science
  27. 27.
    Boussadia, O. et al. Transcription of unr (upstream of N-ras) down-modulates N-ras expression in vivo. FEBS Lett 420, 20–4 (1997).
    OpenUrlCrossRefPubMedWeb of Science
  28. 28.
    Emerman, M. & Temin, H. M. Genes with promoters in retrovirus vectors can be independently suppressed by an epigenetic mechanism. Cell 39, 449–67 (1984).
    OpenUrlCrossRefPubMed
  29. 29.↵
    Corbin, V. & Maniatis, T. Role of transcriptional interference in the Drosophila melanogaster Adh promoter switch. Nature 337, 279–82 (1989).
    OpenUrlCrossRefPubMed
  30. 30.↵
    Chen, J. et al. Kinetochore inactivation by expression of a repressive mRNA. Elife 6, (2017).
  31. 31.↵
    Cheng, Z. et al. Pervasive, coordinated protein level changes driven by transcript isoform switching during meiosis. Cell 172, 910–923 (2018).
    OpenUrlCrossRefPubMed
  32. 32.↵
    Van Dalfsen, K. M. et al. Global Proteome Remodeling during ER Stress Involves Hac1-Driven Expression of Long Undecoded Transcript Isoforms. Developmental Cell 46, 219-235.e8 (2018).
    OpenUrlCrossRef
  33. 33.↵
    Wethmar, K. The regulatory potential of upstream open reading frames in eukaryotic gene expression. Wiley Interdiscip Rev RNA 5, 765–778 (2014).
    OpenUrl
  34. 34.
    Barbosa, C., Peixeiro, I. & Romao, L. Gene expression regulation by upstream open reading frames and human disease. PLoS Genet 9, e1003529 (2013).
    OpenUrlCrossRefPubMed
  35. 35.↵
    Hinnebusch, A. G., Ivanov, I. P. & Sonenberg, N. Translational control by 5’-untranslated regions of eukaryotic mRNAs. Science 352, 1413–6 (2016).
    OpenUrlAbstract/FREE Full Text
  36. 36.
    Law, G. L., Bickel, K. S., MacKay, V. L. & Morris, D. R. The undertranslated transcriptome reveals widespread translational silencing by alternative 5’ transcript leaders. Genome Biol. 6, R111 (2005).
    OpenUrlCrossRefPubMed
  37. 37.↵
    Brar, G. A. et al. High-resolution view of the yeast meiotic program revealed by ribosome profiling. Science 335, 552–557 (2012).
    OpenUrlAbstract/FREE Full Text
  38. 38.↵
    Jorgensen, V. et al. Tunable Transcriptional Interference at the Endogenous Alcohol Dehydrogenase Gene Locus in Drosophila melanogaster. G3 Genes|Genomes|Genetics 10, 1575–1583 (2020).
    OpenUrl
  39. 39.↵
    Hollerer, I. et al. Evidence for an Integrated Gene Repression Mechanism based on mRNA Isoform Switching in Human Cells. G3 (Bethesda, Md.) 9, 1045–1053 (2019).
    OpenUrl
  40. 40.↵
    Brar, G. A. et al. High-resolution view of the yeast meiotic program revealed by ribosome profiling. Science 335, 552–557 (2012).
    OpenUrlAbstract/FREE Full Text
  41. 41.↵
    de la Cruz, J., Iost, I., Kressler, D. & Linder, P. The p20 and Ded1 proteins have antagonistic roles in eIF4E-dependent translation in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. U.S.A. 94, 5201–5206 (1997).
    OpenUrlAbstract/FREE Full Text
  42. 42.↵
    Sen, N. D., Zhou, F., Ingolia, N. T. & Hinnebusch, A. G. Genome-wide analysis of translational efficiency reveals distinct but overlapping functions of yeast DEAD-box RNA helicases Ded1 and eIF4A. Genome Res. 25, 1196–1205 (2015).
    OpenUrlAbstract/FREE Full Text
  43. 43.↵
    Jamieson, D. J. & Beggs, J. D. A suppressor of yeast spp81/ded1 mutations encodes a very similar putative ATP-dependent RNA helicase. Molecular Microbiology 5, 805–812 (1991).
    OpenUrlCrossRefPubMedWeb of Science
  44. 44.↵
    Padmore, R., Cao, L. & Kleckner, N. Temporal comparison of recombination and synaptonemal complex formation during meiosis in S. cerevisiae. Cell 66, 1239–1256 (1991).
    OpenUrlCrossRefPubMedWeb of Science
  45. 45.↵
    Sen, N. D. et al. Functional interplay between DEAD-box RNA helicases Ded1 and Dbp1 in preinitiation complex attachment and scanning on structured mRNAs in vivo. Nucleic Acids Research gkz595 (2019) doi:10.1093/nar/gkz595.
    OpenUrlCrossRef
  46. 46.↵
    Green-Willms, N. S., Fox, T. D. & Costanzo, M. C. Functional interactions between yeast mitochondrial ribosomes and mRNA 5’ untranslated leaders. Mol Cell Biol 18, 1826–1834 (1998).
    OpenUrlAbstract/FREE Full Text
  47. 47.↵
    Cheng, Z. et al. Small and Large Ribosomal Subunit Deficiencies Lead to Distinct Gene Expression Signatures that Reflect Cellular Growth Rate. Molecular Cell (2018) doi:10.1016/j.molcel.2018.10.032.
    OpenUrlCrossRef
  48. 48.↵
    Jambhekar, A. & Amon, A. Control of Meiosis by Respiration. Current Biology 18, 969–975 (2008).
    OpenUrlCrossRefPubMedWeb of Science
  49. 49.↵
    Gao, Y. et al. Mammalian elongation factor 4 regulates mitochondrial translation essential for spermatogenesis. Nature Structural & Molecular Biology 23, 441–449 (2016).
    OpenUrl
  50. 50.
    Jazwinski, S. M. The retrograde response: When mitochondrial quality control is not enough. Biochimica et Biophysica Acta (BBA) - Molecular Cell Research 1833, 400–409 (2013).
    OpenUrl
  51. 51.
    Raught Brian, Gingras Anne-Claude, & Sonenberg Nahum. The target of rapamycin (TOR) proteins. Proceedings of the National Academy of Sciences 98, 7037–7044 (2001).
    OpenUrlAbstract/FREE Full Text
  52. 52.↵
    Topf, U., Uszczynska-Ratajczak, B. & Chacinska, A. Mitochondrial stress-dependent regulation of cellular protein synthesis. Journal of Cell Science 132, jcs226258 (2019).
    OpenUrlAbstract/FREE Full Text
  53. 53.↵
    Wei, W., Pelechano, V., Järvelin, A. I. & Steinmetz, L. M. Functional consequences of bidirectional promoters. Trends Genet. 27, 267–276 (2011).
    OpenUrlCrossRefPubMedWeb of Science
  54. 54.↵
    Song, W., Li, J., Liang, Q. & Marchisio, M. A. Can terminators be used as insulators into yeast synthetic gene circuits? J Biol Eng 10, 19 (2016).
    OpenUrl
  55. 55.↵
    Giaever, G. et al. Functional profiling of the Saccharomyces cerevisiae genome. Nature 418, 387–391 (2002).
    OpenUrlCrossRefPubMedWeb of Science
  56. 56.
    Winzeler, E. A. et al. Functional Characterization of the S. cerevisiae Genome by Gene Deletion and Parallel Analysis. Science 285, 901–906 (1999).
    OpenUrlAbstract/FREE Full Text
  57. 57.
    Giaever, G. & Nislow, C. The Yeast Deletion Collection: A Decade of Functional Genomics. Genetics 197, 451–465 (2014).
    OpenUrlAbstract/FREE Full Text
  58. 58.
    1. Osterman, A. L. &
    2. Gerdes, S. Y.
    Chu, A. M. & Davis, R. W. High-Throughput Creation of a Whole-Genome Collection of Yeast Knockout Strains. in Microbial Gene Essentiality: Protocols and Bioinformatics (eds. Osterman, A. L. & Gerdes, S. Y.) vol. 416 205–220 (Humana Press, 2008).
    OpenUrl
  59. 59.
    Ghaemmaghami, S. et al. Global analysis of protein expression in yeast. Nature 425, 737–741 (2003).
    OpenUrlCrossRefPubMedWeb of Science
  60. 60.
    Huh, W.-K. et al. Global analysis of protein localization in budding yeast. Nature 425, 686–691 (2003).
    OpenUrlCrossRefPubMedWeb of Science
  61. 61.↵
    Breslow, D. K. et al. A comprehensive strategy enabling high-resolution functional analysis of the yeast genome. Nature Methods 5, 711–718 (2008).
    OpenUrl
  62. 62.↵
    Messner, C. B. et al. The Proteomic Landscape of Genome-Wide Genetic Perturbations. bioRxiv 2022.05.17.492318 (2022) doi:10.1101/2022.05.17.492318.
    OpenUrlAbstract/FREE Full Text
  63. 63.↵
    Atias, N., Kupiec, M. & Sharan, R. Systematic identification and correction of annotation errors in the genetic interaction map of Saccharomyces cerevisiae. Nucleic Acids Research 44, e50–e50 (2016).
    OpenUrlCrossRefPubMed
  64. 64.↵
    Leeds, P., Peltz, S. W., Jacobson, A. & Culbertson, M. R. The product of the yeast UPF1 gene is required for rapid turnover of mRNAs containing a premature translational termination codon. Genes & Development 5, 2303–2314 (1991).
    OpenUrlAbstract/FREE Full Text
  65. 65.↵
    Rebbapragada, I. & Lykke-Andersen, J. Execution of nonsense-mediated mRNA decay: what defines a substrate? Curr. Opin. Cell Biol. 21, 394–402 (2009).
    OpenUrlCrossRefPubMedWeb of Science
  66. 66.↵
    Homann, O. R. & Johnson, A. D. MochiView: versatile software for genome browsing and DNA motif analysis. BMC Biol. 8, 49 (2010).
    OpenUrlCrossRefPubMed
  67. 67.↵
    1. Labunskyy, V. M.
    Powers, E. N. & Brar, G. A. Performing Ribosome Profiling to Assess Translation in Vegetative and Meiotic Yeast Cells. in Ribosome Profiling (ed. Labunskyy, V. M.) vol. 2252 89–125 (Springer US, 2021).
    OpenUrl
  68. 68.↵
    Hughes, C. S. et al. Single-pot, solid-phase-enhanced sample preparation for proteomics experiments. Nature Protocols 14, 68–85 (2019).
    OpenUrl
  69. 69.↵
    Cox, J. & Mann, M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nat. Biotechnol. 26, 1367–1372 (2008).
    OpenUrlCrossRefPubMedWeb of Science
  70. 70.↵
    de Hoon, M. J. L., Imoto, S., Nolan, J. & Miyano, S. Open source clustering software. Bioinformatics 20, 1453–1454 (2004).
    OpenUrlCrossRefPubMedWeb of Science
  71. 71.↵
    Saldanha, A. J. Java Treeview--extensible visualization of microarray data. Bioinformatics 20, 3246–3248 (2004).
    OpenUrlCrossRefPubMedWeb of Science
  72. 72.↵
    Song, W., Li, J., Liang, Q. & Marchisio, M. A. Can terminators be used as insulators into yeast synthetic gene circuits? J Biol Eng 10, 19 (2016).
    OpenUrl
Back to top
Posted June 29, 2022.
Download PDF
Email
Use of a ubiquitous gene-editing tool in budding yeast causes off-target repression of neighboring gene protein synthesis
Emily Nicole Powers, Lidia Llacsahuanga Allcca, Ella Doron-Mandel, Jenny Kim Kim, Marko Jovanovic, Gloria Ann Brar
bioRxiv 2022.06.27.497784; doi: https://doi.org/10.1101/2022.06.27.497784
Twitter logo Facebook logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area