Recombination-Induced Tag Exchange (RITE) Cassette Series to Monitor Protein Dynamics in Saccharomyces cerevisiae

Proteins are not static entities. They are highly mobile, and their steady-state levels are achieved by a balance between ongoing synthesis and degradation. The dynamic properties of a protein can have important consequences for its function. For example, when a protein is degraded and replaced by a newly synthesized one, posttranslational modifications are lost and need to be reincorporated in the new molecules. Protein stability and mobility are also relevant for the duplication of macromolecular structures or organelles, which involves coordination of protein inheritance with the synthesis and assembly of newly synthesized proteins. To measure protein dynamics, we recently developed a genetic pulse-chase assay called recombination-induced tag exchange (RITE). RITE has been successfully used in Saccharomyces cerevisiae to measure turnover and inheritance of histone proteins, to study changes in posttranslational modifications on aging proteins, and to visualize the spatiotemporal inheritance of protein complexes and organelles in dividing cells. Here we describe a series of successful RITE cassettes that are designed for biochemical analyses, genomics studies, as well as single cell fluorescence applications. Importantly, the genetic nature and the stability of the tag switch offer the unique possibility to combine RITE with high-throughput screening for protein dynamics mutants and mechanisms. The RITE cassettes are widely applicable, modular by design, and can therefore be easily adapted for use in other cell types or organisms.

For example, protein dynamics can influence the mobility and inheritance of proteins, the exchange of subunits of macromolecular complexes, access to otherwise-occupied interaction sites of proteins or, when proteins are degraded and replaced by new ones, in resetting posttranslational modifications (Hager et al. 2009;Radman-Livaja et al. 2011;Hotz et al. 2012;Menendez-Benito et al. 2013). During the past few years, several techniques have been developed to measure or visualize protein dynamics. Some of these techniques, such as FRAP (fluorescence recovery after photobleaching), TimeStamp, or derivatives thereof, make use of fluorescent fusion proteins to follow the movement or stability and synthesis of proteins in single cells (Lin and Tsien 2010;Butko et al. 2012). Other methods involve differential labeling of old and new proteins by using SILAC (Stable Isotope Labeling with Amino Acids in Cell Culture), radioactive labels, or labeling of specific proteins by using SNAP tags or FlAsH-ReAsH technology (Jansen et al. 2007;Adams and Tsien 2008;Sweet et al. 2010;Zee et al. 2010;Ray-Gallet et al. 2011). Most of these methods allow detection of old and new proteins, but only few methods provide the opportunity to specifically purify old and/or newly synthesized proteins by biochemical methods. This aspect is particularly relevant for the study of chromatin protein dynamics, where affinity purification allows mapping of protein occupancy and dynamics on specific regions of the genome.
Several methods recently have been developed to measure chromatin protein dynamics (recently reviewed in Deal and Henikoff 2010a). One is the use of inducible overexpression of a tagged version of the protein of interest in the presence of an endogenously expressed untagged (or differentially tagged) copy (Korber et al. 2004;Dion et al. 2007;Jamai et al. 2007;Kim et al. 2007;Rufiange et al. 2007). Another method (CATCH-IT; covalent attachment of tags to capture histones and identify turnover) involves the labeling of newly synthesized proteins by amino acid analogs that can be coupled to biotin and thereby used for selective purification (Deal and Henikoff 2010a,b). In Physarum, the dynamics of histone proteins can be monitored at the singlecell level by microinjection of small amounts of labeled histone proteins (Thiriet and Hayes 2005;Ejlassi-Lassallette et al. 2011). We recently developed a versatile and flexible method called recombinationinduced tag exchange (RITE), in which epitope tags on an endogenous protein of interest can be swapped in a conditional manner by an inducible Cre recombinase (Verzijlbergen et al. 2010;De Vos et al. 2011;Radman-Livaja et al. 2011;Hotz et al. 2012).
The RITE system has been developed in budding yeast and is composed of two parts; a tandem-tag cassette that can be integrated behind the gene of interest for conditional C-terminal tagging, and a stably integrated and constitutively expressed hormone-dependent Cre recombinase that allows control of epitope switching. RITE cassettes encode for one epitope tag (Tag 1 or old tag) between two LoxP recombination sites and a second, orphan, epitope tag (Tag 2 or new tag) downstream of the second LoxP site ( Figure 1A). Upon activation of Cre recombinase activity by the simple addition of the hormone estradiol, a tag switch occurs: the first tag is removed from the genome by recombination between the two LoxP sites and replaced by the second tag. To prevent background recombination, the RITE cassettes contain a selectable marker between the LoxP sites. Note that the LoxP recombination sequence is part of the protein coding sequence in the RITE cassettes, resulting in an in-frame tag following the LoxP sites and allowing for switching by recombination. RITE has several advantages over some of the other methods to measure protein dynamics. The proteins of interest are expressed form their endogenous promoter, avoiding potential problems with overexpression. In addition, a RITE switch does not require specific media changes and is permanent, which allows monitoring of protein dynamics under many different physiological conditions (De Vos et al. 2011;Radman-Livaja et al. 2011;Hotz et al. 2012;Ouellet and Barral 2012;Menendez-Benito et al. 2013). Importantly, old and newly synthesized proteins can be monitored simultaneously. The RITE system is flexible, widely applicable, and compatible with switching between different short epitope tags as well as fluorescent tags. Furthermore, RITE allows selective tagging and following one protein of interest in the context of all other unlabeled protein. Finally, RITE allows dynamics measurements in the context of a genetic screen, allowing identification of proteins controlling protein turnover Hotz et al. 2012). Of note, proteins that are subject to very high turnover may not be suitable for analysis by RITE because Creinduced recombination of the LoxP sites takes several hours to complete (see Results and Discussion and Verzijlbergen et al. 2010).
Here we present a comprehensive toolbox for the RITE assay. The availability of a RITE cassette series containing diverse biochemical and fluorescent tags allows for selection of tag pairs that are optimal for the protein of interest or for the experimental setup. Furthermore, new RITE cassettes with additional invariant tags are presented that bypass the need for protein-specific antibodies and allow for simultaneous detection of old and new protein ( Figure 1B). The RITE cassettes are modular by design. Therefore, they can be easily adapted to modify tags or to adjust cassettes for use in other cell types or organisms for which inducible Cre recombinases are available.

Strains and growth conditions
Yeast and bacteria were cultured under standard conditions . Escherichia coli strain DH5a was used for plasmid preparations. All yeast strains constructed and used Figure 1 Outline of RITE. (A) After integration of a RITE cassette behind the gene of interest (GENE), recombination between LoxP sites is induced by Cre-Recombinase, causing a permanent switch from old Tag 1 to new Tag 2 on the protein of interest. S, spacer; L, LoxP recombination sites; T ADH1 , ADH1 terminator; HphMX, Hygromycin resistance cassette. (B) RITE cassette including an invariant tag (i) upstream of the first LoxP site. The invariant tag is present pre-and postrecombination and can be used for simultaneous detection of the old and new protein of interest.

Construction of RITE cassettes
All plasmids constructed in this study were derived from the previously described pFvL100 (Verzijlbergen et al. 2010) and are listed in Figure 2A and Supporting Information, File S1. pFvL106 contains a cassette switching from an HA (Hemagglutinin) to a T7 epitope tag (HA/T7), but the HphMX selection marker (the Hygromycin B phospho transferase gene under control of the AgTEF1 promoter and terminator) is replaced by a URA3 selection marker. pFvL118 and pFvL119 (HA/T7) are nearly identical except for small sequence differences around the HphMX gene (Verzijlbergen et al. 2010). For pFvL160, oligos were created containing either 2xT7 with BsrGI overhangs or 2xHA with SpeI overhangs. Single-stranded oligos ( Table 2) of matching tags were incubated in 40 mM Tris, pH 8, and 100 mM NaCl in a thermal cycler set at 94°for 3 min with a gradual cooling down to 15°. The double-stranded fragments were purified over a microspin G-50 column (Amersham Biosciences). Oligos were sequentially cloned into pFvL119 digested with the appropriate restriction enzymes. pTW073 was constructed by amplifying a fragment of pFvL118 with primers containing a V5 tag extension and BsrGI restriction sites. Subsequently, this fragment and pFvL118 were digested with BsrGI and ligated together, creating an HA/V5 cassette.
A 3-step polymerase chain reaction (PCR) approach was used to construct pTW081 and pTW087. Fragments containing either HA flanked by MluI and SphI restriction sites, or T7 flanked by ClaI and NsiI restriction sites were amplified from pFvL118. These fragments and pFvL118 were digested with the respective restriction enzymes and ligated to construct a cassette with swapped tags (pTW081; T7/HA; Verzijlbergen et al. 2010). For pTW087 a fragment containing HA was amplified using primers with a tail of 6xHis, creating an HA-6xHis fragment flanked by ClaI and NsiI restriction sites. The fragment and pFvL118 were digested with ClaI and NsiI and ligated to construct an HA-6xHis/T7 cassette (pTW087; De Vos et al. 2011).
n  An HA-6xHis fragment flanked by BglII and BsrGI restriction sites was amplified by PCR from pTW087. The fragment and pTW081 were digested with these enzymes and ligated, constructing pTW094 (T7/HA-6xHis). For pMT001 (HA-6xHis/V5), plasmids pTW087 and pTW073 were digested with NsiI and MluI. The fragment of pTW087 containing spacer-LoxP-HA-6xHis-T ADH1 -HphMX was ligated into the fragment of pTW073 containing LoxP-V5. For pMT004-006 a fragment of double stranded DNA containing V5, 2xT7, or 2xFLAG with KpnI and SpeI overhangs were made by oligo annealing (Table 2). Oligos were ligated into pFvL119 digested with KpnI and SpeI. The tags of the resulting plasmids were combined with the HA-6xHis tag of pTW094 using AscI and SphI restriction sites to generate pMT004 (V5/HA-6xHIS), pMT005 (2xT7/HA-6xHIS), and pMT006 (2xFLAG/HA-6xHIS). pKV005 was constructed by a 3-step PCR on pKV001 (a pFvL119 derivative containing additional restriction sites), introducing an invariant MYC tag between the spacer and LoxP site. The resulting cassette is MYCi-HA/T7. For pKV006, a V5-fragment with Eco47III restriction sites was amplified by PCR. The fragment and pKV001 were digested by Eco47III and ligated, constructing a cassette with an invariant V5 tag between the spacer and LoxP site (V5i-HA/T7). For pKV014, monomeric red fluorescent protein (mRFP) was amplified by PCR with primers containing BsrGI and HindIII restriction sites at the ends. This fragment and pFvL160 were digested with BsrGI and HindIII and ligated, resulting in a cassette with a fluorescent tag post-recombination (3xHA/3xT7-mRFP). For pKV015, yEGFP and mRFP were amplified with primers containing restriction sites for SpeI, and for BsrGI and HindIII, respectively. These fragments and pKV006 were digested with the appropriate enzymes and ligated, resulting in a cassette that combines epitope tags with fluorescent tags, pKV015 (V5i-HA-yEGFP/T7-mRFP, (Verzijlbergen et al. 2010)). For pVM013, pTW081 and pVM012 were digested with BsrGI and HindIII. The small fragment of pVM012 containing yEGFP was ligated into the large fragment of pTW081 (T7/HA), resulting in a T7/HA-yEGFP cassette (Menendez-Benito et al. 2013).

PCR-mediated gene tagging
To target RITE cassettes to the gene of interest, the cassette was amplified using integration primers that contain 40 bp of sequence of the gene of interest for homologous recombination ( Figure 2B). The forward primer (F1) sequence of the cassette is 59-GGT GGA TCT GGT GGA TCT-39. For pFvL106 to pKV006, the reverse primer used to amplify the cassette is 59-AGGGAACAAAAGCTTGCATG-39 (R1), which anneals 54 bp downstream of the cassette. For pKV014 to pKV016 the reverse primer sequence of the cassette is 59-TCAGGCGCCGGTGGAGTGGCG-39 (R3). This primer is mRFP specific and anneals at the end of mRFP to introduce a stop codon (underlined sequence), which is missing in the mRFP sequence. The reverse primer used for pVM013 is 59-TGATTACGCCAAGCTCG-39 (R2), which anneals further downstream than R1 (90 bp downstream of the cassette) and which can be used when the sequence of R1 is absent. For one strain in this study (SPC42G), a reverse primer was used that anneals even further downstream of the cassette. When designing primers to target RITE to the gene of interest, it is important that the reading frame of the ORF and the fused cassette is maintained. The PCR products were transformed into strain NKI2176, NKI2036 or Y7092 using standard transformation protocols (Gietz and Schiestl 2007). After transformation, cells were plated onto YEPD plates and incubated overnight at 30°; the following day plates were replica-plated onto selection plates (YEPD containing 200 mg/mL Hygromycin). Integration of the cassettes was checked by colony PCR.

Cre recombinase vectors
Two plasmids were constructed to integrate Cre recombinase in the yeast genome. pTW040 was constructed by cloning a P TDH3 -Cre-EBD78 fragment digested with PvuII into pRS303 digested with SmaI. For the P TDH3 -Cre-EBD78 fragment, P TDH3 was cloned with ApaI and BspEI upstream of Cre-EBD78, where the P GAL had been replaced by a multiple cloning site. pSS146 was constructed by cloning the P TDH3 -Cre-EBD78 fragment of pTW040 into pRS306 with EcoRI and NotI. Both plasmids contain unique restriction sites that can be used to integrate Cre-Recombinase at the HIS3 locus, the URA3 locus or the CYC1 terminator ( Figure 2C).

Detection of recombination by Southern blot
For Southern blotting 5 · 10 8 cells were spun and frozen at 280°. A histone H3 (HHT2)-specific probe was made by PCR amplification using primers HHT2_HindIII_for (GAATCTTTCTGTGACGCTTGG) and HHT2_HindIII_rev (GGGGAAGAACAGTTGGAAGG), resulting in a 650-bp amplicon covering the region 576,144-576,794 of chromosome XIV. When used on genomic DNA that was digested with HindIII, the three bands recognized are specific for before the switch (3000 bp, pre), after the switch (931 bp, post), or as an internal control (1538 bp, control). Radioactive Southern blotting was performed using 50 mCi of 32 P-dCTP; incubation occurred overnight at 65°.

Chromatin immunoprecipitation (ChIP)
For ChIP, cells were grown to mid-log phase in YEPD with 200 mg/ mL Hygromycin for preswitch strains or YEPD for postswitch strains. Samples of 123 · 10 9 cells were taken, fixed for 10 min in 1% formaldehyde, and washed with cold TBS. Pellets were stored in 12-mL flat bottom-tubes or 2-mL screw-cap tubes at 280°until further processing. Cells in 12-mL tubes were disrupted in 300 mL of breaking buffer (100 mM Tris, pH 7.9; 20% glycerol; protease inhibitor cocktail EDTA-free; Roche) with 400 mL of glass beads in a multivortex for 20 min at 4°. Cells in screw-cap tubes were disrupted in 200 mL of breaking buffer with 500 mL of glass beads in a bead beater in a 4°c old block for 2 min. Lysis was at least 70%, as determined by microscopy. Lysates were transferred to 1.5-or 2-mL tubes and 1 mL of FA buffer (50 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic n  acid-KOH (HEPES-KOH), pH 7.5; 140 mM NaCl; 1 mM EDTA; 1% Triton X-100; 0.1% Na-deoxycholate; protease inhibitor cocktail EDTA-free) was added. The mixture was centrifuged for 1 min at 20,817 x g at 4°, and the pellet was washed once more with FA buffer. The pellet was resuspended in 450 mL of FA, divided over two 1.5-mL tubes, and sonicated for 627 min in a Bioruptor (Diagenode) with 30 sec on-off cycles on high power. Lysates were cleared by centrifugation for 5 min at 4°at 20,817 x g. Supernatant containing chromatin was transferred to 1.5-mL tube and 1 mL of FA was added to samples of the 12-mL tubes. Screw-cap tube chromatin samples were treated with micrococcal nuclease to generate mononucleosomes. For these samples, 800 mL of final buffer (15 mM Tris, pH 7.4; 50 mM NaCl; 1.5 mM CaCl 2 ; 5 mM b-mercaptoethanol; 5 mM MgCl 2 ) was added. Samples were incubated with 30 units of micrococcal nuclease (Fermentas) at 37°for 20 min. The reaction was stopped by adding ethylene glycol tetraacetic (EGTA) acid and EDTA to a final concen-tration of 10 mM and placing tubes on ice. The chromatin solution was centrifuged for 15 min at 20,817 x g at 4°; the supernatant was transferred to a new 1.5-mL tube and stored at 220°.
Magnetic dynabeads coupled with Protein G (Dynal) were incubated in PBS containing 5 mg/mL bovine serum albumin (BSA) with antibody for at least 4 hr at 4°. Then, 200 mL of soluble chromatin was added to prepared dynabeads and incubated rotating overnight at 4°and 1 mL of FA buffer was added and samples were incubated rotating for 5 min at room temperature. The samples were washed twice with each of the buffers FA, FA-HS (50 mM HEPES-KOH, pH 7.5; 500 mM NaCl; 1 mM EDTA; 1% Triton X-100; 0.1% Na-deoxycholate), RIPA (10 mM Tris, pH 8; 250 mM LiCl; 0.5% NP-40; 0.5% Na-deoxycholate; 1 mM EDTA). Finally, the samples were washed once with TE (10 mM Tris, pH 8; 1 mM EDTA). Then 100 mL of elution buffer (50 mM Tris, pH 8; 10 mM EDTA; 1% SDS) was added to the samples and incubated for 10 min at 65°. Subsequently  the samples were centrifuged 1 min at 20,817 x g and 80 mL of supernatant was collected. Then, 70 mL of TE was added to samples and cross links were reversed in 0.625 mg/mL ProtK and 3 mg/mL RNaseA incubated for 1 hr at 50°and subsequently overnight at 65°. For input samples, 40 mL of chromatin solution was combined with 60 ml elution buffer and 70 mL of TE and treated in the same manner as IP samples to reverse cross links. DNA was purified by using the High Pure PCR Product Purification Kit (Roche). Alternatively, DNA was extracted by using Chelex-100 resin (Bio-Rad) (Walsh et al. 1991;Nelson et al. 2009).
Quantitative PCR Quantitative real time-PCR (qPCR) was performed with SYBRgreen master mix (Applied Biosystems or Roche) according to the manufacturer's manual. IP samples were diluted 10 times, and input samples were diluted 100 times before analyzing by qPCR on a 7500 Fast Real-Time PCR system (Applied Biosystems) or LightCycler 480 II (Roche). qPCR primers are shown in Table 3.

Microscopy
Microscopy samples were fixed with 4% formaldehyde, stained with 1 mg/mL Hoechst 33342 (Invitrogen) and mounted with Vectashield solution (Vector Laboratories) onto concanavalin A2coated cover slides. Samples were analyzed on a Leica SP5 confocal system using a 405-nm laser to excite Hoechst, 488 nm for yEGFP and 561 nm for mRFP.

RESULTS AND DISCUSSION
RITE cassette and Cre vector construction RITE cassettes containing different combinations of small epitope tags or fluorescent tags were generated by combining tag modules of previously described RITE cassettes or with new epitope tags as described in the section Materials and Methods. The RITE cassettes are shown in Figure 2A. Many of the RITE cassettes contain unique restriction sites between the different elements, which facilitates modification of the constructs for other applications. Some of the RITE cassettes contain an invariant tag for simultaneous detection of old and new protein using commercially available antibodies. The cassettes can be integrated by homologous recombination behind any gene of interest ( Figure 2B).
Two vectors are available for integration of the hormonedependent Cre recombinase ( Figure 2C). Unique restriction sites in the Cre vectors allow integration of the constructs by single crossover at the CYC1 terminator region, or at the HIS3 or URA3 regions. These options and the efficiency of integration depend on the auxotrophic alleles present in the target strain. For example, targeting of the HIS3-Cre cassette (pTW040) to the commonly used his3Δ200 allele (see http://wiki.yeastgenome.org/index.php/Strains) is inefficient because of the relatively short region of homology on one end. The advantage of targeting to this locus is that the integrated Cre is less prone to be lost by pop-out recombination. When other strategies are used it is recommended to maintain selection for the integrated Cre vector to select against pop-out events. Cre recombinase is expressed constitutively under control of the TDH3 promoter and CYC1 terminator and fused to the human Estrogen Binding Domain (EBD), which makes the nuclear activity of Cre dependent on the hormone b-estradiol (Logie and Stewart 1995). Estradiol releases the EBD from cytosolic heat shock proteins and allows entry of the EBD-Cre chimeric protein into the nucleus for recombination (Logie and Stewart 1995). This allows timed control of the RITE-tag switching. Here we use a deriv-ative of Cre-EBD (Cre-EBD78) that is tightly dependent on the addition of b-estradiol in budding yeast (Lindstrom and Gottschling 2009;Dymond et al. 2011). Background recombination before induction of the tag switch and recombination efficiency after induction can be determined by Southern blot analysis or by plating cells on nonselective media and then replica-plating the colonies to media containing Hygromycin (see Figure 3 and Verzijlbergen et al. 2010). In a typical experiment, the average background recombination is 10% or less, whereas the Cre-induced recombination efficiency is 95% or more (see Figure 3 and Verzijlbergen et al. 2010). The completion of a recombination-induced tag switch in a population of cells generally takes a few hours (Figure 3). This time window should be taken into consideration when RITE is applied to study the dynamics of proteins with a very high turnover rate. Of note, the rate of induced vs. background recombination can vary between strains, tagged genes, cell cycle stages, and experimental conditions. We occasionally encountered strains that showed high background recombination or low levels of induced recombination (data not shown), the reasons of which are unknown. We also found low levels of induced recombination in synthetic media containing monosodium glutamate instead of ammonium sulfate as the nitrogen source (data not shown).

Immunoblot detection of RITE tags
To demonstrate the functionality of the short biochemical epitope tags in the new RITE cassettes, the respective RITE cassettes were PCR amplified and targeted to the histone H3 gene HHT2 ( Figure 2B). To avoid interference of nontagged histone H3, the other gene encoding Figure 5 Validation of RITE tags in ChIP assays. RITE-ChIP of H3 and H2A.Z followed by qPCR analysis of three loci: A = PTC6, B = RSA4, C = SPA2. The regions around the transcription start sites of PTC6 and RSA4 have high H2A.Z occupancy; the SPA2 region is located at the 59 end of the SPA2 coding sequence and has low H2A.Z occupancy (primer sequences are listed in Table 3). ChIP signals were normalized over the corresponding inputs. The relative ChIP efficiency (H2A.Z/H3 on each of the three loci examined) varied between the four epitope tags. Average of three biological replicates 6 SEM. Results were obtained from: HA-ChIP on strains NKI2178 (H3-HA-6xHIS/T7) and NKI8030 (H2A.Z-HA-6xHIS/T7), V5-ChIP on strains NKI8050 (H3-V5/HA-6xHIS) and NKI8053 (H2A.Z-V5/HA-6xHIS), FLAG-ChIP on strains NKI8052 (H3-2xFLAG/HA-6xHIS) and NKI8055 (H2A.Z-2xFLAG/HA-6xHIS), and T7-ChIP on strains NKI8051 (H3-2xT7/HA-6xHIS) and NKI8054 (H2A.Z-2xT7/HA-6xHIS. H3 (HHT1) was deleted. We analyzed strains prior to the tag switch as well as strains that had undergone a permanent tag switch. Immunoblot analysis of whole-cell extracts showed that both tags (preswitch shows old Tag 1, postswitch shows new Tag 2) can be detected with the use of tag-specific antibodies (Figure 4, A and B). The old Tag 1 was detected before the switch, whereas the new Tag 2 was detected after the switch. Due to background recombination, low levels of the new Tag 2 were detected before the switch in some of the samples. RITE was also applied to 3-phosphoglycerate kinase (Pgk1), a housekeeping protein. In this case, the RITE cassette contained combinations of short epitope tags and fluorescent tags (pKV015). The old and new tags could be detected on immunoblots before and after the switch, respectively ( Figure 4C). In addition, the RITE cassette also harbors an invariant V5 tag, which should be present pre-switch as well as postswitch. Pgk1 containing the invariant V5 tag could be detected in both samples ( Figure 4C).

Detection of RITE tags by ChIP
One of the advantages of RITE compared with other methods for measuring protein dynamics is the possibility of applying it to affinity purification. We determined how the various tags in the RITE cassettes perform in ChIP experiments. For this purpose, we tagged histone H3 (HHT2, in the absence of HHT1) and histone variant H2A.Z (HTZ1) with different RITE cassettes and performed ChIP-qPCR experiments on preswitch strains (see Materials and Methods). Three loci were examined (primer sequences are listed in Table 3), representing high (PTC6, RSA4) and low (SPA2) levels of steady-state H2A.Z and similar levels of H3 (Kobor et al. 2004;Albert et al. 2007). Indeed, using the RITE tags, we found that H2A.Z levels relative to H3 levels differed according to the previously shown steady-state levels ( Figure 5). Note that although the trends were similar, the relative ChIP efficiency (H2A.Z/H3) was not the same for the different RITE tags ( Figure 5). Apparently, the short epitope tags do not work equally well for each protein. Fortunately, the RITE assay is very flexible. The expanded RITE cassette series that we describe here affords the selection of epitope tag pairs that work efficiently for the protein of interest. Furthermore, because the RITE cassettes are modular by design, other tag sequences can be readily incorporated.
Combining RITE with fluorescence microscopy The addition of fluorescent markers to the RITE technology adds the possibility of spatiotemporal monitoring of proteins of interest (Verzijlbergen et al. 2010;Hotz et al. 2012;Menendez-Benito et al. 2013). We previously described the HA-yEGFP/T7-mRFP and 3xHA-yEGFP/3xT7-mRFP RITE cassettes (see Figure 2A) (Verzijlbergen et al. 2010;Menendez-Benito et al. 2013). An mCherry/GFP RITE-like cassette has been described by the Barral lab (Hotz et al. 2012). Two additional RITE cassettes can be used to switch from a small epitope tag to a larger fluorescent tag (Figure 2A). They can be used for conditional knock-ins when larger tags cause growth defects. As the protein of interest is only fluorescent after the switch, these RITE cassettes are particularly suited for monitoring synthesis and the spatiotemporal behavior of the new protein by microscopy. Figure 6 shows Spc42, a component of the yeast spindle pole body, tagged with these RITE cassettes ( Figure 6, A and D). The spindlepole body could be observed after the tag switch as a bright dot next to the nucleus ( Figure 6, B and E), whereas before the tag switch only background signal was observed. Expression of the proteins was confirmed by immunoblot analysis (Figure 6, C and F). Of note, such color switch tags can be combined with microscopy-or flow cytometry-based genetic screens to identify proteins that control protein or organelle dynamics (Hotz et al. 2012).
Detecting protein dynamics RITE enables the following of protein dynamics under many physiological conditions. Here, we show the application of one of Figure 6 Analysis of fluorescent knock-in RITE cassettes. (A) Spc42, a subunit of the spindle pole body, was tagged with a 3xHA/3xT7-mRFP RITE cassette (strain Spc42R). (B) Fluorescent microscopy of strain Spc42R. In the preswitch sample, no mRFP was detected; in the postswitch sample Spc42-mRFP was clearly visible. (C) Immunoblot analysis of the Spc42R samples described in (B). Pgk1 was used as loading control. (D2F) as in (A2C) but using Spc42 tagged with a 1xT7/1xHA-yEGFP RITE cassette (strain Spc42G). Spc42 containing a single HA or T7 tag could not be detected on blots using the tag-specific antibodies. However, an antibody against spacer-LoxP could visualize old and new Spc42. Scale bars, 2 mm. the new RITE cassettes, which switches from V5 to HA-6xHis and was applied to histone H3 (HHT2, in the absence of HHT1) to measure chromatin protein dynamics ( Figure 7A). Cells were arrested by starvation and a preswitch sample was taken before inducing Cre recombinase ( Figure 7B). After 16 hr of recombination, a postswitch sample was taken before the cells were released in fresh media. Two additional samples were taken during the release, at 2 hr, before cell division, and at 4 hr, when most of the cells had divided once. During the release, the amount of old Tag 1 (V5) decreased, whereas new Tag 2 (HA-6xHis) increased, as expected ( Figure 7C). In the pre-and postswitch samples we detected V5-but hardly any HA-tagged protein. The small amount of HA-tagged protein found in the postswitch sample is consistent with previous results showing that in starved cells there is a low level of histone exchange (Verzijlbergen et al. 2010). Using an antibody against the spacer-LoxP peptide sequence, we could detect old Tag 1 and new Tag 2 simultaneously on a single blot due to different mobility on SDS-PAGE gels.
Next, we performed ChIP on these samples and examined occupancy of old and new histone H3 at the PTC6 gene and the SPA2 gene (Table 3). For both genes we detected a decrease in V5tagged protein (old H3) and an increase in HA-tagged protein (new H3) when switched cells were analyzed four hours after release into fresh media (Figure 7, D and E). The level of histone H3 exchange, i.e., the ratio of new H3 over old H3 (new/old), increased over time ( Figure 7F). Because the cells are released into log phase and undergo DNA replication, the exchange measured here was mostly caused by replication-coupled histone deposition, and to a lesser extent by replication-independent histone exchange, the latter of which can vary from gene to gene (Dion et al. 2007;Jamai et al. 2007;Rufiange et al. 2007;Deal et al. 2010;Verzijlbergen et al. 2010).
RITE is a versatile method than can be used to study protein dynamics by different downstream applications on any protein of interest that tolerates C-terminal tags. The GFP tagged yeast library shows the many examples where C-terminal tagging is allowed , and subsequently released into fresh media for 2 hr (no cell division) or 4 hr (one cell division). An antibody against spacer-LoxP detects both old and new H3. (D-E) H3-V5 and H3-HA-6xHis ChIP analysis of samples described in (B) to measure occupancy (IP/input) of old and new H3 at PTC6 (around transcription start site) and SPA2 (coding sequence) (primer sequences are listed in Table  3). (F) Exchange of histone H3 determined as the ChIP ratio of new H3 over old H3 (HA/ V5). Average of three biological replicates 6 SEM. without significant effects on cell viability or other vital functions (Ghaemmaghami et al. 2003;Huh et al. 2003). The short epitope tags can be used for immunoblot and affinity purification protocols such as ChIP. Thereby, RITE can be combined with proteomics methods, genomics methods, or DNA-based highthroughput screens (Verzijlbergen et al. 2010;De Vos et al. 2011;Radman-Livaja et al. 2011;Verzijlbergen et al. 2011). The fluorescent tags in the RITE cassettes can be applied to measure protein dynamics in single cells by live imaging (Verzijlbergen et al. 2010;Hotz et al. 2012;Menendez-Benito et al. 2013) and further expanded toward high-throughput genetic screening. The suite of RITE cassettes enables a flexible design for many applications. Furthermore, the RITE cassettes are modular by design and can therefore be easily adapted for use in other cell types or organisms. The RITE plasmids and sequence information will be available through EUROSCARF (Frankfurt, http:// web.uni-frankfurt.de/fb15/mikro/euroscarf/index.html).