Establishment of an “in saccharo” experimental system

ABSTRACT

Many proteins form complexes that function in reaction pathways. Therefore, to understand protein function, it is necessary to reconstitute complexes and pathways in vitro. However, it is not straightforward to achieve full activity in reconstituted systems. To address this problem, we present a yeast system named “in saccharo” analysis, which uses budding yeast for simultaneous expression and analysis of many kinds of non-host proteins, such as human proteins. For this purpose, vectors that can accommodate many genes are required. Here, we describe the construction of a chromosome vector by insertion of unique barcode sequences (BCs) into the ribosomal RNA gene repeat (rDNA). Each unit of the rDNA has a BC that is used as the target for integration of an external gene. Because rDNA is naturally capable of maintaining many repetitive copies, the vector is expected to retain the numerous external genes that may be required for reconstitution of functional protein complexes and reaction pathways. Consistent with this prediction, we were able to clone three human genes that form the RNA silencing pathway, which has no functional equivalent in budding yeast, and to demonstrate functionality in this in saccharo analysis system.

INTRODUCTION

To understand functions of proteins, or genes, two main methods are used, namely in vivo (e.g., using genetics) and in vitro (e.g., using biochemistry) analyses. In in vivo analysis, for example, by using mutants, we are able to identify genes linked to the altered phenotype and analyze the relationship between genes and their roles. In in vitro analysis, we test enzymatic activities using purified proteins. By these methodologies, the fundamental physical reactions in cells have been revealed. However, there are limits to these methodologies. For example, in biochemistry in a tube, it is not easy to reconstitute complicated reactions, such as DNA repair, because stability and activity of the enzymes are usually reduced outside the cell. Furthermore, some proteins need structural bases (scaffolds) for their activity, such as membranes or the cytoskeleton, and these are difficult to reconstitute in a tube. Thus, only factors that are soluble and central to reactions have been tested, and our biochemical knowledge is therefore limited. In addition, it has become clear that many proteins function in complexes (Clancy and Hovig, 2014). Consequently, authentic in vitro analysis is not easy for some reactions because it is difficult to reconstitute an intact protein complex in which all of the subunits are active. To solve the problems presented by such limited methodology, a next-generation reconstitution system that goes beyond the current in vitro systems is required.

One approach to accomplishing this goal is to create an artificial or synthetic cell. If a cell-like structure could be constructed in which the correct physical conditions were present, a protein complex could be stabilized and its activity tested. Although the field is developing (Stano, 2018), it takes time to construct a cell-like structure because of technical difficulties and a lack of knowledge about what constitutes the essential components of the structure. Furthermore, to test the activity of some protein complexes, customization of the cell will likely be required for that particular reaction and this is unlikely to be straightforward. Thus, before addressing the challenge of artificial cell construction, it is more reasonable to collect further information about individual reactions in cells. Here, we present a system using yeast as a host for reconstituted reactions in lieu of a tube.

Budding yeast, Saccharomyces cerevisiae, is one of the best studied eukaryotic cells. In addition to the wealth of knowledge of the organism, the huge evolutionary distance between yeast and mammals and plants (~1 billion years) is advantageous for reconstituting pathways and complexes from more complex organisms that are likely to be lacking in the yeast. For example, in budding yeast the chromosome does not fully condense in the G2 phase of the cell cycle (Guacci et al., 1994), genes with introns are rare (Stajich et al., 2007), and there is no DNA methylation (Proffitt et al., 1984), no centrosome (Kilmartin, 2014) and no gene silencing by RNAi (Drinnenberg et al., 2009). It may therefore be possible to reconstitute these missing activities and thus determine their component factors. Furthermore, powerful yeast genetics will facilitate analysis of the reconstituted systems. Such a combination may allow the use of the yeast system in drug screening to find candidates that can affect the reconstituted reaction. This approach, named “in saccharo” analysis, could contribute to reducing the number of experimental animals and also shorten the time for drug development.

For the establishment of an in saccharo system, it is necessary to have a vector that can accommodate the many genes required for complex reconstitution. Usual plasmid vectors cannot accommodate multiple genes because of their limited number of selection markers. However, YAC (yeast artificial chromosome) vectors can accommodate large DNA inserts that can include many genes (Murray and Szostak, 1983). Furthermore, artificially synthesized yeast chromosomes can also be used to clone external genes (Richardson et al., 2017). However, the weak points of cloning many genes into YACs and synthetic chromosomes are that construction is not easy and homologous recombination may occur if the same promoters are used to control the expression of multiple external genes. It is noteworthy that chromosomal integration methods, such as the EasyClone method, are relatively simple means of cloning genes. Such methods have the potential to reduce the risk of recombination and subsequent loss of the external genes so long as the integration sites are kept distant from each other (Jensen et al., 2014). However, the stability problem caused by recombination between homologous sequences remains if the integrated genes share the same promoter/terminator sequences.

Here, we describe a new type of “chromosome vector” that can be used to clone up to ~100 genes in yeast (Fig. 1A). The vector was constructed using the ribosomal RNA genes (rDNA), which have a highly repetitive structure in eukaryotic cells. In general, repetitive sequences are unstable because homologous recombination occurs between the repeats. However, for rDNA, cells have systems for maintenance of the high-copy-number situation because of the huge demand for ribosomes and, in fact, each organism maintains its rDNA copy number (for a review, see Kobayashi, 2006). One of the systems used for this maintenance is gene amplification. When rDNA copies are lost by homologous recombination, the amplification system compensates for the shortage. After the copy number reaches the wild type level, recombination is repressed (Kobayashi et al., 1998, 2019). Our studies of this amplification identified the key protein Fob1, which is associated with the replication fork barrier sequence (RFB) and arrests the fork against transcription (Kobayashi et al., 1998; Kobayashi, 2003). The blockage triggers a DNA double-strand break at the RFB and induces recombination and amplification (Supplementary Fig. S1). The recombination also contributes to gene conversion that homogenizes and maintains the quality of the sequence (Ganley and Kobayashi, 2011). Therefore, in a fob1 mutant, recombination is severely repressed and rDNA is highly stabilized (Kobayashi et al., 1998). Of course, in long-term culture without rDNA recombination, the copy number may accidentally decline and mutations may accumulate. However, during the period of assay, the rDNA is quite stable (Kobayashi et al., 1998). This indicates that rDNA has a system for avoiding homologous recombination in nature. Indeed, we identified several genes that affect rDNA stability (Saka et al., 2016). For example, SIR2 overexpression and EAF3 deletion enhance rDNA stability (Ganley et al., 2009; Wakatsuki et al., 2019). By using such a strain as the genetic background, we may be able to accommodate more genes in a more stabilized rDNA array.

Fig. 1.

Construction of rDNA vectors (rVECs). (A) Structure of rVECs. Unique barcode sequences (BCs) integrated into the intergenic spacers of rDNA are the targets for external gene insertion. (B) Construction of BCs. A total of 143 BCs (~600 bp) were constructed using long random-DNA oligos, PCR and ligation. The BCs were cloned into a pUC vector (pBCs). See Materials & Methods for details. (C) Plasmid structure for integration of BCs. p1^stBC, p2^ndBC and p3^rdBC were used for the first, second and third integrations of BCs (BC1, BC2 and BC3), respectively. (D) Integration of BCs (BC1 to BC3). BC (n) (where n = 1, 2 or 3) was integrated stepwise by homologous recombination with the previously integrated BC (n – 1) and non-rDNA region. (E) Confirmation of an rDNA chromosome vector with three BCs (mini rVEC). Integration of three BCs into the rDNA was tested by PCR; the leftmost lanes contain size markers. Two mating type (MAT) strains (a and α) were used.

In the fob1 mutant, we inserted a unique barcode sequence in each rDNA copy that is used as an integration target for an external gene, allowing many genes to be accommodated and stably maintained. Using this “rDNA chromosome vector (rVEC)” in a trial experiment, we succeeded in stably cloning three human genes that are required for gene silencing by destabilizing mRNA. In these yeast cells, the human system was functional and repressed the expression of a reporter gene. This suggests that the “in saccharo” system supported by a rVEC can be used to analyze other non-yeast reactions.

MATERIALS AND METHODS

Yeast strains and medium

BY4743 background yeast strains with low-copy rDNA used in this study are listed in Supplementary Table S1. The TRP1 gene in the reporter strains was replaced with the natMX gene (trp1∆::natMX) to construct mating partners of transgene-expressing strains (YTT440 and YTT441). For silencing analysis of human RNAi components, rDNA-barcoded strains with or without human RNAi genes and strains harboring GFP reporter genes were mated to construct diploid strains. For yeast culture, yeast media for marker selection were based on Hartwell’s complete (HC) medium containing 2% D-glucose or D-galactose without ammonium sulfate, selected amino acids or uracil. URA⁺ counter-selection was performed on HC plates with 0.1% 5-fluoroorotic acid (5-FOA). Standard yeast culture in rich medium was in yeast extract peptone (YP) medium containing 2% D-glucose or D-galactose.

Construction of unique barcode sequence (BC) and plasmids

Unique BC fragments were assembled from double-stranded DNA (dsDNA) fragments with random sequences (Fig. 1B). The first dsDNA fragments (W1, X1, X2, Y1, Y2 and Z1) were synthesized from long random-DNA oligos (OLI002, 004, 006, 008, 010 and 012) and their complementary primers (OLI003, 005, 007, 009, 011 and 013) using KOD FX Neo DNA polymerase (Toyobo). The 100-bp dsDNA fragment sets (W1-X1, X2-Y1 and Y2-Z1) were ligated with quick-DNA ligase (New England BioLabs) and amplified by PCR with KOD FX Neo DNA polymerase as WX, XY1 and YZ1 fragments. The WX1 fragment was generated by PCR from the WX fragment with primers OLI001, 005 and 014. The XY1 and YZ1 fragments were assembled as the XZ1 fragment by PCR with primers OLI007 and 0013. The WX1 and XZ1 fragments were further assembled as BC fragments by PCR with primers OLI013 and 0014. The BC fragments were cloned into the pUC19 vector fragment amplified by PCR with primers OLI015 and 0016 using NEBuilder HiFi DNA assembly (New England BioLabs). The sequences of the cloned BC fragments were checked by Sanger sequencing. Inserts in the BC plasmids (pBCs) without long G or C stretches (< 6) or restriction enzyme target sites (AvrII, BamHI, Eco47III and PstI) were selected as the BCs. The sequences of BC1–3 are listed in Supplementary Table S2. All the primers used in this study are listed in Supplementary Table S3.

To produce BC integration plasmids, two plasmids with a rDNA-associated sequence and marker genes, pUC_RDN1L_LYS2 and pUC_RDN1L_URA3, were constructed. The rDNA-associated sequence (chromosomal position chrXII: 450760–451418) was amplified by PCR with primers OLI020 and 025 from the BY4743 genome. The LYS2 marker fragments were amplified by PCR with primer sets OLI022–023 and OLI021–024 from the BY4743 genome. The URA3 marker fragments were amplified by PCR with primers OLI033 and 034 from YIplac211. The vector fragment was amplified by PCR with primers OLI 017 and 026 from pUC19. The rDNA-associated sequence and marker gene fragments were cloned with the pUC19 vector fragment using NEBuilder HiFi DNA assembly as pUC_RDN1L_LYS2 and pUC_RDN1L_URA3. To supply rDNA sequence to the BC integration plasmids, a hygromycin-resistant rDNA gene (RDN-Hyg1) was cloned with the pUC19 vector fragment as the pUCEco47RDNhyg1BamHI plasmid. Two rDNA fragments were amplified by PCR with primer sets OLI029–030 and OLI027–028 from the BY4743 genome and cloned with the pUC19 vector fragment to generate pUCEco47RDNheadBamHI. The PaeI-RDN-Hyg1-NheI and SmaI-RDN-Hyg1-PaeI fragments from pRDN-Hyg1 were cloned into the SmaI-NheI site of pUCEco47RDNheadBamHI to construct pUCEco47RDNhyg1BamHI.

The BC integration plasmids, p1^stBC1, p2^ndBC and p3^rdBC, were assembled from BC fragments, RDN-Hyg1 fragments and pUC_RDN1L_marker vector fragments. The marker-proximal and -distal BCs were amplified by PCR from pBCs with primer sets OLI040–041 and OLI044–045, respectively. The RDN1-end fragment for p1^stBC1 was amplified by PCR from pUCEco47RDNhyg1BamHI with primers OLI042 and 043. The pUCEco47RDNhyg1BamHI plasmid was digested by AfeI and BamHI, and the 9.2-kb fragment was purified as full-length RDN-Hyg1 fragment. The p1^stBC1 plasmid was assembled from the marker-proximal BC1, RDN1-end, and SalI-PstI-digested pUC_RDN1L_LYS2 fragments. The others were assembled from the marker-distal BCx-1, the marker-proximal BCx, RDN-Hyg1 and SalI-PstI-digested pUC_RDN1L_marker fragments (x = 2 or 3).

The GAL1 promoter (GAL1p), CYC1 terminator (CYC1t) and rS-URA3-Rs fragments were assembled in pUC19 vector to construct a galactose-inducible cassette with a pop-out URA3 marker as pUC_AsiSI-GAL1pCYC1t-rSURA3Rs-NotI. GAL1p and CYC1t were amplified by PCR from pAG413Gal-Ago2. The rS-URA3-Rs fragment was amplified by PCR from YIplac211 with primers OLI063 and 064. AsiSI-GAL1pCYC1t-rSURA3Rs-NotI was cloned into the AsiSI-NotI site of pBCs for human RNAi gene expression (pBCxGAL). To induce R-recombinase from the GAL1 promoter for URA3 marker pop-out, the SalI-PstI GAL1p-R fragment of pRINT was cloned into the SalI-PstI site of the YEplac181 vector (YEp181GALpR).

To construct YIp128PDA1pGFP, a GFP reporter plasmid, the PDA1 promoter and NLS-GFP-adh1t fragments were cloned into the EcoRI-HindIII site of YIplac128. The PDA1 promoter fragment was amplified by PCR with primers OLI073 and 074 from the BY4743 genome. The NLS-GFP-adh1t fragment was amplified by PCR with primers OLI075 and 076 from pKT127. YIp128PDA1pGFP was integrated into the ADH1 locus of YTT430 by AccI digestion.

The GFP hairpin plasmid, YIp211TEF1p-GFPhairpin, was constructed from the TEF promoter, the CYC1 terminator, two GFP fragments (pfg1 and gfp2) and EcoRI-HindIII-digested YIplac211 vector. The TEF promoter and CYC1 terminator fragments were amplified by PCR from the BY4743 genome with primer sets OLI065–066 and OLI071–072, respectively. The pfg1 and gfp2 fragments were amplified by PCR from pKT127 with primer sets OLI067–068 and OLI069–070, respectively. YIp211TEF1p-GFPhairpin was integrated into the CYC1 locus of YTT430 by HindIII digestion.

Integration of BCs

The first, second and third integration fragments were prepared from p1^stBC, p2^ndBC and p3^rdBC, respectively by PstI-AvrII digestion. The first BC fragment was transformed into yeast strains YTT399 and YTT401 with LYS⁺ selection. The second and third BCs were selected by URA⁺ and LYS⁺ to generate YTT430 and YTT431.

Integration and detection of human silencing genes

For marker pop-out by R recombinase, YEp181GALpR was introduced into the rDNA vector strain YTT431 (YTT435). HsTRBP, HsAgo2 and HsDicer genes were integrated into BC1, BC2 and BC3 regions, respectively. 1/2BCx-GAL and 1/2BCx-URA3 fragments were amplified by PCR from pBCxGAL plasmids (x = 1, 2 or 3). To integrate HsTRBP into YTT435, 1/2BC1-GAL (primers OLI013 and 086), 1/2BC1-URA3 (primers OLI014 and 087) and BamHI-TRBP-Eco32I fragments from pAG415Gal-TRBP were co-transformed, with URA⁺ selection. For HsAgo2 integration, 1/2BC2-GAL (primers OLI013 and 084), 1/2BC2-URA3 (primers OLI014 and 085) and BamHI-Ago2-EcoRI fragments from pAG413Gal-Ago2 were used. HsDicer was integrated into BC3 with 1/2BC3-GAL (primers OLI013 and 082), 1/2BC3-URA3 (primers OLI014 and 083) and HindIII-HsDicer-BcuI fragments. To reuse the URA3 gene for other BC integrations, transgene-integrated strains with YEp181GALpR were cultured in HC-Leu medium with galactose overnight at 30 ℃, and URA3 pop-out clones were selected on HC plates containing 5-FOA. For construction of strains with multiple RNAi genes, integration cycles were repeated (YTT443–YTT449).

Expression of human silencing genes was confirmed by western blot with antibodies specific to HsTRBP, HsAgo2 and HsDicer. For tubulin, HsTRBP and HsAgo2 detection, anti-tubulin (cloneYL1/2 AbD, Serotec: MCA77G), anti-HsTRBP (clone46D1: bsm-50266M) and anti-HsAgo2 (clone11A9, Merck-Millipore: MABE253) antibodies were labeled with peroxidase using a Peroxidase Labeling Kit-NH2 (Dojindo: LK11) and used at 1/2,500, 1/10,000 and 1/1,000 dilution, respectively, in phosphate-buffered saline with 0.005% Tween 20 and 5% skim milk (PBS-T-milk). HsDicer was detected with a 1/300 dilution of anti-HsDicer antibody (Santa Cruz Biotechnology: sc-30226) in PBS-T-milk and 1/10,000-diluted anti-rabbit IgG-HRP secondary antibody (GE Healthcare: NA934) in PBS-T. The primary antibodies were incubated overnight at 4 ℃ and washed twice with PBS-T at room temperature for 30 min. The secondary antibody was incubated at room temperature for 2 h and washed twice with PBS-T at room temperature for 30 min. Western blot signals were detected as chemiluminescence with Immobilon-P substrate (Merck-Millipore: WBKLS0500) using the Fusion SL chemiluminescence imaging system (Vilber-Lourmat).

Detection of RNA

To prepare total RNA samples from diploid strains (YTT603–YTT618), cells expressing human silencing factors were cultured to log phase in YP galactose medium at 30 ℃ and then harvested. RNA extraction was performed as previously described (Iida and Kobayashi, 2019). For northern analysis of siRNA, 5 μg of total RNA samples were denatured in 50% formamide at 65 ℃ for 10 min and separated by 12% urea-PAGE with 0.5 × Tris-borate-EDTA. The separated RNA samples were transferred onto a Hybond N+ nylon membrane (GE Healthcare) on a Transblot SD cell (Bio-Rad) at 4 ℃ and 400 mA constant current for 2 h. The membrane was optimally UV-crosslinked (120,000 μJ/cm²) by a UV crosslinker (Stratagene: StrataLinker 1800) and dried after soaking in 5 × SSC. Prehybridization was performed with 10 ml of Rapid-Hyb buffer (GE Healthcare: RPN1635) at 42 ℃ for 2 h. Five picomoles of oligo probes (siGFP01–13) were radiolabeled with [γ-³²P]ATP (6,000 Ci/mmol; PerkinElmer: NEG502Z) by T4 polynucleotide kinase (Takara) in 30-μl reactions at 37 ℃ for 30 min. The purified probes from a MicroSpin G-25 column were mixed with prehybridization buffer to start overnight hybridization at 25 ℃. The membrane was washed twice with 5 × SSC/0.1% SDS buffer at room temperature for 10 min and once with 2 × SSC/0.1% SDS buffer at 42 ℃ for 10 min. After stringent washing, the membrane was exposed to an imaging plate, and siRNA signals were detected on an FLA7000 phosphorimager (GE Healthcare).

GFP and hairpin RNAs were detected by conventional northern blot. In the northern blot, 5 μg of total RNA samples were separated on 1.2% agarose gels with formaldehyde denaturing buffer (1 × MOPS and 6% formaldehyde). Separated RNA samples were stained with ethidium bromide and their images were captured on a Fusion SL imaging system. The RNA samples were capillary-transferred onto a Hybond N+ membrane in 10 × SSC at room temperature overnight. After optimal UV-crosslinking, the membrane was prehybridized with 10 ml of Rapid-Hyb buffer at 65 ℃ for 2 h. To detect GFP and hairpin RNAs, the gfp2 fragment was randomly labeled and incubated with the membrane at 65 ℃ for 2 h. The membrane was washed once with 2 × SSC/0.1% SDS buffer at room temperature for 30 min and washed twice with 0.1 × SSC/0.1% SDS buffer at 65 ℃ for 30 min. The blot was exposed to an imaging plate and analyzed on an FLA7000.

To quantify GFP reporter RNA, reverse transcription–quantitative PCR (RT-qPCR) of total RNA samples was performed with comparative Ct (∆∆Ct) analysis of GFP reporter and ACT1 genes. The RT-qPCR was performed with a KAPA SYBR FAST One-Step qRT-PCR Kit (Kapa Biosystems: KK4650) on an Eco Real-Time PCR System (Illumina). For GFP-reporter-specific RT-qPCR, 20 ng of total RNA was analyzed with primers pPCRsplitGFPFw and pPCRsplitGFPRv. The reaction cycle consisted of incubations at 42 ℃ for 10 min and 95 ℃ for 3 min followed by 40 cycles of 95 ℃ for 5 s, 52 ℃ for 20 s, and 72 ℃ for 30 s. For ACT1 control RT-qPCR, total RNA was analyzed with primers OLI1126 and 1127, with a reaction cycle of 42 ℃ for 10 min, 95 ℃ for 3 min, and 40 cycles of 95 ℃ for 5 s, 50 ℃ for 20 s, and 72 ℃ for 30 s. Each reaction was normalized by ROX low dye and quantified by ∆∆Ct analysis.

RESULTS

Construction of a chromosome vector

To clone non-host external genes, we constructed an rDNA repeat with ~600-bp unique barcode sequences (BCs) that are used as the target for integration. One hundred forty-three BCs were constructed by random oligo DNA, strand synthesis and ligation (see Materials & Methods, Fig. 1B), and cloned into plasmids (pBCs). All BC sequences constructed in this study are available in the DNA data bank of Japan (DDBJ) database with accession numbers LC597665–LC597807. We started with a yeast strain with low rDNA copy number (~15 copies; Iida and Kobayashi, 2019) and added rDNA units with BCs to the end of the rDNA repeat by homologous recombination. First, a plasmid with the 3′ end of 35S rDNA, BC1, LYS2 and a non-rDNA region was constructed (p1^stBC, Fig. 1C). p1^stBC was digested with AvrII and PstI, transformed into the low rDNA copy strain, and LYS⁺ cells were selected (Fig. 1D). After checking the integration by PCR, p2^ndBC, having an rDNA unit with BC2 and the URA3 selection marker, was digested with AvrII-PstI and integrated by homologous recombination into the BC1–non-rDNA region. URA⁺ lys⁻ cells were selected. This second integration makes it possible to use the LYS2 marker for the third integration with BC3. In the third integration, p3^rdBC, having an rDNA unit with BC3 and the LYS2 selection marker, was digested with AvrII-PstI, and integrated by homologous recombination into the BC2–non-rDNA region. ura⁻ LYS⁺ cells were then selected. This third integration allows us to use the URA3 marker for the next integration. By repeating this process, we can continue to add rDNA units with BCs until the wild type rDNA copy number of budding yeast (~150 copies) is reached. Here, we stopped at three copies (BC1–3) and tested whether this rVEC system worked. In Fig. 1E, the presence of three BCs in the rDNA was confirmed by PCR and the construct was named the “mini rDNA chromosome vector (mini rVEC)”.

Integration of human silencing genes into the mini rVEC

To clone external genes into the mini rVEC, we designed an integration system as follows (Fig. 2). Into the middle of the BC cloned in a pBC plasmid, a galactose-inducible promoter (GAL1), a transcription terminator and a selection marker (URA3) were integrated, yielding pBC-GAL. On both sides of URA3, R-recombinase recognition sequences were attached (Araki et al., 1985; Raghuraman et al., 1997). These are used for removal of the URA3 marker and allow reuse of the marker (Fig. 2A). pBC-GAL was amplified by PCR with primers that have 50-bp identity to both ends of the external gene at the 5′ end and at the BC end. Together with the PCR-amplified external gene, three fragments were transformed into budding yeast. These fragments recombine in the cell via the homologous sequences at the ends and become integrated into the target BC site in the mini rVEC (Fig. 2B).

Fig. 2.

Insertion of an external gene (EG) into a mini rVEC. (A) The GAL1 promoter/terminator and URA3 marker were integrated into the center of BC in pUC vectors pBC GAL-n (where n = 1–143). The URA3 marker has R-recombinase recognition sequences on both sides that are used to remove the marker sequence after selection. pBC GAL-n is used as a PCR template to amplify two fragments, BC (left half)-URA3-GAL1 terminator-EG tail and EGI tail-GAL1 promoter-BC (right half). (B) Integration of three PCR products into a mini rVEC. Two DNA fragments from pBC GAL (A) and an EG were transformed into a cell with a mini rVEC. The three fragments recombine in the cell and integrate into the target BC (upper panel). After integration, URA3 was removed by R-recombinase to allow reuse of the marker gene (lower panel).

To test the transformation system, we cloned three human genes that encode the RNA interference (RNAi) system (RISC complex: Dicer, Ago2, TRBP) that silences gene expression post-transcriptionally by cleaving mRNA (Gregory et al., 2005). In the RNAi pathway, which has been found in most eukaryotic cells (Hannon, 2002), double-stranded RNA (dsRNA) is processed into small dsRNA (siRNA) that guides the RISC complex to the target mRNA and cleaves it. Interestingly, this RNAi pathway does not exist in budding yeast (Harrison et al., 2009). However, when the three genes are transformed into the yeast, they are able to function (Suk et al., 2011). Initially, using pBC1-GAL, which contains the BC1 sequence and the GAL1 promoter, a terminator and URA3 in the middle of the sequence, we amplified two DNA fragments using primers with 50-base identity to the tail and head of the TRBP gene at the 5′ end, and to the BC1 end specific sequences (Fig. 2A). The PCR fragments were transformed into budding yeast containing the mini rVEC with three BCs (BC1–BC3) in the rDNA (Fig. 2B, top), and URA⁺ colonies were selected. After checking the insertion into the BC1 site by PCR, the URA3 marker was removed by R-recombinase to allow reuse of the marker (Fig. 2B, bottom). We repeated the same procedure with the Ago2 and Dicer genes using pBC2-GAL and pBC3-GAL as the templates, respectively, and thus established a strain containing the mini rVEC and three external human genes (Fig. 3A).

Fig. 3.

Expression of human RNAi genes in yeast with a mini rVEC. (A) Structure of three human genes in the mini rVEC. (B) Western blot analysis to confirm expression of the human genes. Yeast strains with different combinations of the three human RNAi genes were tested for expression. Tubulin was used as the internal control.

Human silencing genes function in the yeast system

We tested the functionality of the three human RNAi genes integrated into the mini rVEC of yeast. Initially, gene expression was checked by western blotting. As shown in Fig. 3B, expression of each gene was detected in all combinations. In addition, for the duration of the experiment, the three cloned genes were stably maintained and no deletional recombination occurred.

We next tested the in saccharo function of the RISC complex, which is known to cleave dsRNA to small fragments in vivo and in vitro (MacRae et al., 2008) (Fig. 4A). For this analysis, a modified GFP gene whose transcript forms a hairpin structure was transformed (Fig. 4B). The hairpin-structured RNA is known to be a good substrate for small interfering RNA (siRNA) production (MacRae et al., 2008). The results are shown in Fig. 4B. In the cell with all three proteins expressed, siRNA (gfp-siRNA) was most clearly observed. The second most effective combination was Dicer and TRBP, and third was Dicer and Ago2. In cells expressing only Dicer, a faint product band could be observed. These results are similar to those from in vitro analysis (MacRae et al., 2008).

Fig. 4.

Small RNA production in yeast. (A) Schematic explanation of small RNA production by human RNAi proteins. gfp-hairpin RNA is the substrate for the small RNA production. (B) Yeast strains with different combinations of human RNAi genes were tested for the production of small RNA. Upper panel: rRNA and tRNA were detected by ethidium bromide staining as a loading control. Lower panel: gfp-siRNA was detected by northern analysis.

siRNA-silenced gene expression in saccharo

In human cells, siRNA recognizes and cleaves target mRNA to repress expression. We tested whether such a reaction occurs in our in saccharo experimental system. For this purpose, another GFP gene was transformed as a reporter (Fig. 5A). Both gfp-hairpin RNA for siRNA and GFP-reporter for monitoring mRNA cleavage were expressed in cells with the mini rVEC, and the non-cleaved RNA (GFP-mRNA) was detected by northern analysis (Fig. 5B, middle) and RT-qPCR (Fig. 5B, bottom). As shown in Fig. 5B (middle), in cells expressing all three proteins (D, A, T), the band from the GFP-reporter (asterisk) is slightly reduced. This reduction was confirmed by RT-qPCR (Fig. 5B, bottom). In this assay, in cells expressing all three proteins (D, A, T), the amount of non-cleaved GFP-mRNA was most prominently reduced (Fig. 5B, bottom, asterisk). Taken together, these results indicate that the chromosome vector can provide active enzymes that function in the in saccharo system.

Fig. 5.

Gene silencing by the human RNAi system in yeast. (A) Schematic explanation of the silencing assay in yeast. Expression of GFP-reporter (GFP-mRNA) is repressed by gfp-siRNA digestion if the system works. (B) Detection of transcripts of GFP-reporter. Transcripts of GFP-reporter (GFP-mRNA) were detected by northern analysis (middle panel) and RT-qPCR (bottom panel). Top panel: 25S and 18S rRNA were detected by ethidium bromide staining as a loading control. Middle panel: gfp-hairpin RNA and GFP-mRNA were detected by northern analysis. Bottom panel: GFP-mRNA was quantitated by RT-qPCR. The reduction of GFP-mRNA in the strain expressing gfp-hairpin RNA and the complete human RNAi system is highlighted with an asterisk. Values are the average of two independent experiments, and the error bars indicate SEMs. The P-value for the difference in GFP-mRNA levels in gfp-hairpin-expressing wild-type and human RNAi system strains was tested using a two-sided Welch’s t-test (P = 0.0065).

DISCUSSION

rDNA is a highly repetitive sequence in eukaryotic cells and can be used as a target for integration of an external gene (Lopes et al., 1989, 1996). This high-copy-number situation is critical to meet the huge demand for ribosome synthesis and as such the maintenance of the rDNA repeat structure is important for viability of the cell. Indeed, each organism maintains its proper rDNA copy number (Long and Dawid, 1980). We have studied the rDNA maintenance mechanisms and have described a unique recombination system (Kobayashi, 2014). This system contributes to gene amplification and homogenization of the rDNA sequences. Several factors related to rDNA stability have been identified. One such factor is Fob1, which associates near the region of transcriptional termination of the 35S rRNA gene and inhibits a replication fork moving in the direction opposite to the transcription (Kobayashi, 2003). Under FOB1-defective conditions, rDNA recombination is repressed and stability is enhanced (Kobayashi et al., 1998). We have now used the fob1 mutant to construct a mini chromosome vector and succeeded in establishing a stable expression system that provides sufficient amounts of active proteins in an “in saccharo” experimental system.

However, in this study, only weak cleavage activity of target GFP-mRNA was detected (Fig. 5B). One explanation for this reduced activity is that some factors that promote the reaction are absent and yeast proteins are unable to compensate for these missing molecules. Using this yeast system, we will be able to search for the missing factors by introduction of a human gene expression library.

In terms of stability of the chromosome vector, thus far, we have been able to establish a strain with 19 BCs and are continuing to increase the number. Theoretically, as the yeast cell maintains more than 100 rDNA copies, we expect such a level of BC insertion to be possible. Previously, we made a strain with 2.4 kb of external sequence in each of the ~150 rDNA units (Miyazaki and Kobayashi, 2011). To establish this, we initially made a strain that has only two rDNA copies in the chromosome. To support the growth, helper plasmids that express rRNA were used. Into the two-copy rDNA strain (fob1), we inserted the 2.4-kb DNA fragment sequence (URA3, lacO repeat) and amplified the rDNA unit by complementation of the FOB1 gene. The copy number of repeats containing the external sequence increased up to ~150 and was stably maintained. This confirms that the rDNA has the capacity to include many external DNA fragments.

A drawback of the current strategy is that it takes time to insert the rDNA units with the BC because we add them one by one. At the current pace of insertion (two weeks per copy), it would take approximately four years to produce a chromosome vector that could accommodate ~100 genes. Therefore, we have started construction of rVEC in different chromosomes. For example, while the original rDNA is located on chromosome XII, we plan to establish an artificial rDNA repeat with BCs on chromosome VI. It is known that such artificial rDNA is stably maintained in yeast (Oakes et al., 2006). Subsequently, the two strains with rDNA in different chromosomes can be crossed to establish a strain with two rDNA repeats on two different chromosomes. Such independent construction of rDNA with BCs on different chromosomes will shorten the time needed to establish a strain with many BCs in the rDNA.

Challenges in using budding yeast for analysis of molecular mechanisms of different organisms are known. For example, Truong and Boeke (2017) exchanged the histone of yeast for human histones in order to establish a yeast with human-type chromatin. However, after depletion of the original yeast histones, most of the cells died and the survivors had many mutations in the transformed human genes. This suggests that human chromatin cannot function in wild type yeast and that to establish a yeast with human-type chromatin, other human genes that modify the chromatin to a functional form in yeast are required. For the cloning of many genes, our rVEC has the potential to be useful. Once human-type chromatin or other human physiological pathways are established in yeast, we can take easier and less expensive strategies for research using powerful yeast genetics and other established experimental methods. Such a “humanoid yeast” would also be available for drug screening and could reduce the number of experimental animals used. Thus, the impact of this in saccharo experimental system should be high not only in biological studies but also in medical science.

ACKNOWLEDGMENTS

We thank Dr. Frederick Roth for human RNAi genes. We also thank Drs. Marc Gartenberg and Hiroyuki Araki for the R-recombinase gene and technical advice for the recombinase experiments. This work was supported by JST CREST (Grant Number JPMJCR19S3 to T. K.) and in part by grants-in-aid for Scientific Research (17H01443 to T. K. and 20K06600 to T. I.) from the Japan Society for the Promotion of Science.

REFERENCES

•  Araki,  H.,  Jearnpipatkul,  A.,  Tatsumi,  H.,  Sakurai,  T.,  Ushio,  K.,  Muta,  T., and  Oshima,  Y. (1985) Molecular and functional organization of yeast plasmid pSR1. J. Mol. Biol. 182, 191–203.
•  Clancy,  T., and  Hovig,  E. (2014) From proteomes to complexomes in the era of systems biology. Proteomics 14, 24–41.
•  Drinnenberg,  I. A.,  Weinberg,  D. E.,  Xie,  K. T.,  Mower,  J. P.,  Wolfe,  K. H.,  Fink,  G. R., and  Bartel,  D. P. (2009) RNAi in budding yeast. Science 326, 544–550.
•  Ganley,  A. R. D.,  Ide,  S.,  Saka,  K., and  Kobayashi,  T. (2009) The effect of replication initiation on gene amplification in the rDNA and its relationship to aging. Mol. Cell 35, 683–693.
•  Ganley,  A. R. D., and  Kobayashi,  T. (2011) Monitoring the rate and dynamics of concerted evolution in the ribosomal DNA repeats of Saccharomyces cerevisiae using experimental evolution. Mol. Biol. Evol. 28, 2883–2891.
•  Gregory,  R. I.,  Chendrimada,  T. P.,  Cooch,  N., and  Shiekhattar,  R. (2005) Human RISC couples microRNA biogenesis and posttranscriptional gene silencing. Cell 123, 631–640.
•  Guacci,  V.,  Hogan,  E., and  Koshland,  D. (1994) Chromosome condensation and sister chromatid pairing in budding yeast. J. Cell Biol. 125, 517–530.
•  Hannon,  G. J. (2002) RNA interference. Nature 418, 244–251.
•  Harrison,  B. R.,  Yazgan,  O., and  Krebs,  J. E. (2009) Life without RNAi: noncoding RNAs and their functions in Saccharomyces cerevisiae. Biochem. Cell Biol. 87, 767–779.
•  Iida,  T., and  Kobayashi,  T. (2019) RNA polymerase I activators count and adjust ribosomal RNA gene copy number. Mol. Cell 73, 645–654.
•  Jensen,  N. B.,  Strucko,  T.,  Kildegaard,  K. R.,  David,  F.,  Maury,  J.,  Mortensen,  U. H.,  Forster,  J.,  Nielsen,  J., and  Borodina,  I. (2014) EasyClone: method for iterative chromosomal integration of multiple genes in Saccharomyces cerevisiae. FEMS Yeast Res. 14, 238–248.
•  Kilmartin,  J. V. (2014) Lessons from yeast: the spindle pole body and the centrosome. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 369, 20130456.
•  Kobayashi,  T. (2003) The replication fork barrier site forms a unique structure with Fob1p and inhibits the replication fork. Mol. Cell. Biol. 23, 9178–9188.
•  Kobayashi,  T. (2006) Strategies to maintain the stability of the ribosomal RNA gene repeats—collaboration of recombination, cohesion, and condensation. Genes Genet. Syst. 81, 155–161.
•  Kobayashi,  T. (2014) Ribosomal RNA gene repeats, their stability and cellular senescence. Proc. Jpn. Acad. Ser. B Phys. Biol. Sci. 90, 119–129.
•  Kobayashi,  T.,  Heck,  D. J.,  Nomura,  M., and  Horiuchi,  T. (1998) Expansion and contraction of ribosomal DNA repeats in Saccharomyces cerevisiae: requirement of replication fork blocking (Fob1) protein and the role of RNA polymerase I. Genes Dev. 12, 3821–3830.
•  Long,  E. O., and  Dawid,  I. B. (1980) Repeated genes in eukaryotes. Annu. Rev. Biochem. 49, 727–764.
•  Lopes,  T. S.,  de Wijs,  I. J.,  Steenhauer,  S. I.,  Verbakel,  J., and  Planta,  R. J. (1996) Factors affecting the mitotic stability of high-copy-number integration into the ribosomal DNA of Saccharomyces cerevisiae. Yeast 12, 467–477.
•  Lopes,  T. S.,  Klootwijk,  J.,  Veenstra,  A. E.,  van der Aar,  P. C.,  van Heerikhuizen,  H.,  Raúe,  H. A., and  Planta,  R. J. (1989) High-copy-number integration into the ribosomal DNA of Saccharomyces cerevisiae: a new vector for high-level expression. Gene 79, 199–206.
•  MacRae,  I. J.,  Ma,  E.,  Zhou,  M.,  Robinson,  C. V., and  Doudna,  J. A. (2008) In vitro reconstitution of the human RISC-loading complex. Proc. Natl. Acad. Sci. USA 105, 512–517.
•  Miyazaki,  T., and  Kobayashi,  T. (2011) Visualization of the dynamic behavior of ribosomal RNA gene repeats in living yeast cells. Genes Cells 16, 491–502.
•  Murray,  A. W., and  Szostak,  J. W. (1983) Construction of artificial chromosomes in yeast. Nature 305, 189–193.
•  Oakes,  M. L.,  Johzuka,  K.,  Vu,  L.,  Eliason,  K., and  Nomura,  M. (2006) Expression of rRNA genes and nucleolus formation at ectopic chromosomal sites in the yeast Saccharomyces cerevisiae. Mol. Cell. Biol. 26, 6223–6238.
•  Proffitt,  J. H.,  Davie,  J. R.,  Swinton,  D., and  Hattman,  S. (1984) 5-Methylcytosine is not detectable in Saccharomyces cerevisiae DNA. Mol. Cell. Biol. 4, 985–988.
•  Raghuraman,  M. K.,  Brewer,  B. J., and  Fangman,  W. L. (1997) Cell cycle-dependent establishment of a late replication program. Science 276, 806–809.
•  Richardson,  S. M.,  Mitchell,  L. A.,  Stracquadanio,  G.,  Yang,  K.,  Dymond,  J. S.,  DiCarlo,  J. E.,  Lee,  D.,  Huang,  C. L. V.,  Chandrasegaran,  S.,  Cai,  Y., et al. (2017) Design of a synthetic yeast genome. Science 355, 1040–1044.
•  Saka,  K.,  Takahashi,  A.,  Sasaki,  M., and  Kobayashi,  T. (2016) More than 10% of yeast genes are related to genome stability and influence cellular senescence via rDNA maintenance. Nucleic Acids Res. 44, 4211–4221.
•  Stajich,  J. E.,  Dietrich,  F. S., and  Roy,  S. W. (2007) Comparative genomic analysis of fungal genomes reveals intron-rich ancestors. Genome Biol. 8, R223.
•  Stano,  P. (2018) Is research on “synthetic cells” moving to the next level? Life (Basel) 9, 3.
•  Suk,  K.,  Choi,  J.,  Suzuki,  Y.,  Ozturk,  S. B.,  Mellor,  J. C.,  Wong,  K. H.,  MacKay,  J. L.,  Gregory,  R. I., and  Roth,  F. P. (2011) Reconstitution of human RNA interference in budding yeast. Nucleic Acids Res. 39, e43.
•  Truong,  D. M., and  Boeke,  J. D. (2017) Resetting the yeast epigenome with human nucleosomes. Cell 171, 1508–1519.e13.
•  Wakatsuki,  T.,  Sasaki,  M., and  Kobayashi,  T. (2019) Defects in the NuA4 acetyltransferase complex increase stability of the ribosomal RNA gene and extend replicative lifespan. Genes Genet. Syst. 94, 197–206.