Uncovering the fine print of the CreERT2-LoxP system while generating a conditional knockout mouse model of Ssrp1 gene

FAcilitates Chromatin Transcription (FACT) is a complex of SSRP1 and SPT16 that is involved in chromatin remodeling during transcription, replication, and DNA repair. FACT has been mostly studied in cell-free or single cell model systems because general FACT knockout (KO) is embryonically lethal (E3.5). FACT levels are limited to the early stages of development and stem cell niches of adult tissues. FACT is upregulated in poorly differentiated aggressive tumors. Importantly, FACT inhibition (RNAi) is lethal for tumors but not normal cells, making FACT a lucrative target for anticancer therapy. To develop a better understanding of FACT function in the context of the mammalian organism under normal physiological conditions and in disease, we aimed to generate a conditional FACT KO mouse model. Because SPT16 stability is dependent on the SSRP1-SPT16 association and the presence of SSRP1 mRNA, we targeted the Ssrp1 gene using a CreERT2- LoxP approach to generate the FACT KO model. Here, we highlight the limitations of the CreERT2-LoxP (Rosa26) system that we encountered during the generation of this model. In vitro studies showed an inefficient excision rate of ectopically expressed CreERT2 (retroviral CreERT2) in fibroblasts with homozygous floxed Ssrp1. In vitro and in vivo studies showed that the excision efficiency could only be increased with germline expression of two alleles of Rosa26CreERT2. The expression of one germline Rosa26CreERT2 allele led to the incomplete excision of Ssrp1. The limited efficiency of the CreERT2-LoxP system may be sufficient for studies involving the deletion of genes that interfere with cell growth or viability due to the positive selection of the phenotype. However, it may not be sufficient for studies that involve the deletion of genes supporting growth, or those crucial for development. Although CreERT2-LoxP is broadly used, it has limitations that have not been widely discussed. This paper aims to encourage such discussions.

Introduction deletion of genes that negatively affect cell growth or viability because of the positive selection of the cells in which recombination has taken place. However, many studies require a higher rate of excision to fully understand the function of a gene (e.g., studies involving the deletion of genes that may have a positive effect on cell growth) because cells with recombination may be easily overgrown by cells without recombination. In our efforts to develop an inducible conditional Ssrp1 KO mouse model using the CreER T2 -LoxP system, we encountered multiple problems due to its low efficiency and leakiness. Here, we present these limitations and ways to resolve them.

Transfection and transduction
Lentiviruses and retroviruses were generated in the 293T cell line (Clontech) transfected with the appropriate expression vector along with packaging plasmids using PolyJet™ In Vitro DNA Transfection Reagent (catalog no. SL100688 SignaGen 1 Laboratories) according to the manufacturer's instructions. Virus-containing media from 293T cells were collected 48 h and 72 h later, syringe-filtered through a 0.45 μm filter, and transferred to target cells in the presence of polybrene (10 μg/ml).

Animal housing and maintenance
Animals were maintained at the Roswell Park Comprehensive Cancer Center. All experiments were conducted with approval from the Institute Animal Care and Use Committee at Roswell Park. The facility is certified by the American Association for Accreditation of Laboratory Animal Care in accordance with the current regulations and standards of the U.S. Department of Agriculture and the U.S. Department of Health and Human Services.

Generation of mutant mice
The Ssrp1 targeting vector was obtained from the European Mutant Mouse Consortium (EUCOMM) (PG00202_Z_1_D01) to create a conditional knockout of the Ssrp1 gene. This vector was electroporated into the G4 ES cell line (129:B6 from Andras Nagy's lab) by the Gene Targeting and Transgenic Shared Resource at Roswell Park Comprehensive Cancer Center. Clones of the cells with an integrated insert were selected using G418. Clones positive for homologous recombination were identified via Southern blotting and then expanded for injection into C57BL/6J albino host embryos. Embryos were transferred into CD-1 pseudopregnant foster mice. Chimeras were generated from three ES cell clones, and germline transmission was achieved.
ES cell electroporation. ES cell electroporation was performed by the Gene Targeting and Transgenic Shared Resource at Roswell Park Comprehensive Cancer Center. Briefly, SSRP1 mutant clones were grown in complete ES cell medium on a feeder layer in a 100 mm dish. On the day of electroporation, the ES cells were trypsinized to prepare a single cell suspension. A circular plasmid (11 μg) containing a puromycin resistance cassette was electroporated into the cells. Puromycin selection (0.6 μg/mL) was carried out for 3 to 4 days beginning the day after electroporation. Colonies were picked and transferred to individual wells of a 24-well plate one week after electroporation. ES cell clones were tripsinized, and protein and DNA were isolated to assess the excision of FRT sites.
Screening of ES clones for homologous recombination using Southern blotting. DNA was isolated from the ES cells using the Wizard 1 Genomic DNA Purification Kit (Promega, catalog no. A1120). Southern blotting was performed as previously described [27]. Genomic DNA was digested with either BsrG1 or Ace1/Nde1. DNA was separated on a 0.8% agarose gel, and transferred to a nylon membrane for Southern blotting. The 5' hybridization probe (included region size: 8821-9200) was generated by PCR using the following primers: forward primer CGCTTGATGCACTTTTGCTA; reverse primer CAGGACGTTGATGGAAGACA (product size: 347). The 3' hybridization probe (included region size: 28590-30625) was generated by PCR using the following primers: forward primer GCATTGGTTTTCACCCTTGT; reverse primer CCCCACGTGGATAAATTCTG and a product size: 417 Genotyping. Genotyping was done using PCR. Tail tips (1 to 2 mm) were dissolved in 200 μL DirectPCR Lysis Reagent (catalog no. 102-T Viagen) containing 0.2 mg/mL Proteinase K (catalog no. AM2546 Ambion) and incubated overnight at 55˚C. The crude lysates were incubated at 85˚C for 45 min to inactivate Proteinase K and then centrifuged at 10,000 rpm for 1 min to collect the crude DNA sample in the supernatant. The Ssrp1 fl/fl , Ssrp1 fl/+ and Ssrp1 +/+ genotypes were determined by PCR using 50 to 100 ng DNA. PCR was carried out using 2x Taq mix (catalog no. FBTAQM, Froggabio) under the following conditions: 95˚C 1 min (1 cycle); 35 cycles of 95˚C for 30 sec, 61.2˚C for 30 sec, and 72˚C for 40 sec; 72˚C for 7 min (1 cycle). The following primers were used to identify wild-type (Ssrp1 + ) and mutant allele (Ssrp1 NeoLacZfl ): forward primer (P1): TCCTGGAACTCACTCTGTAGACC, reverse primer (P2): TTAAGTGCTAGAGACCTGATGGC and a second reverse primer (P3A): CTTCACTG AGTCTCTGGCATCTC. The PCR products were run on a 2% agarose gel (1x TAE) at 133 V and the bands were observed as following: P1:P2 (~350 bp) for the wild-type allele, P1: P3A (~592) for mutant allele.
To determine the excision of the synthetic cassette via Flp recombination, PCR was performed as follows: 95˚C 1 min (1 cycle); 35 cycles of 95˚C for 30 sec, 59.6˚C for 30 sec, and 72˚C for 40 sec; 72˚C for 7 min. (1 cycle) The primers were: forward primer P1: TCCTGGAAC TCACTCTGTAGACC and reverse primer P5: AGTTATCTCGACGAAGTTC. The excised band for the synthetic cassette was~161 bp. The mutant allele without the synthetic cassette (Ssrp1 fl ) was identified P1 and P2 primers (same as above) using the PCR protocol described above. PCR products for wild-type Ssrp1 + were observed at~350bp and for Ssrp1 fl were observed at 533bp. The CreER T2 genotype was determined using the Jackson Laboratory genotyping protocol (https://www2.jax.org/protocolsdb/f?p=116:5:0::NO:5:P5_MASTER_PROTOCOL_ID, P5_JRS_CODE:20407,008463). To determine the excision of the critical exons (exons 4 to 9) (Ssrp1 Δ ) via Cre recombination, forward primer P1 was used in conjunction with reverse primer Pcre2 (CTGGTCTGCATTTCACCCTT). The P1:Pcre2 product was approximately 500 bp.

Generation of primary, immortalized, and transformed cells
For the generation of primary fibroblasts from Ssrp1 fl/fl , Ssrp1 +/+ , Ssrp1 fl/+ , Ssrp1 fl/fl CreER T2+/-, Ssrp1 fl/fl CreER T2+/+ , and Ssrp1 fl/+ CreER T2+/+ mice, mice were anesthetized with isoflurane, and an approximately 2 cm piece of the tail tip was collected. The samples were washed in sterile 1xPBS containing 1% AA. The tail tips were minced into small pieces using a pair of scissors. F-12K medium containing 1% AA and 2mg/ml Collagenase A (10 ml) was added to the samples, which were incubated for 2 h at 37˚C with shaking. The samples were centrifuged at 1200 rpm for 10 min at room temperature. The pellet was resuspended in 7 mL F-12K medium containing 10% FBS and 1% AA and placed in a 100 mm plate (catalog no. 40-229690, Krackeler). The plates were kept in a sterile humidified tissue culture incubator at 37˚C and 5% CO 2 for 4 to 5 days to allow cells to adhere to the plate.
To generate immortalized fibroblasts, primary fibroblasts were infected with genetic suppressor element 56 (GSE56) virus and subsequently selected with 500 μg/ml neomycin (G418 Sulfate). Post-selection, the cells were maintained in DMEM medium supplemented with 10% FBS and 1% penicillin-streptomycin.
For the generation of transformed fibroblasts, the immortalized fibroblasts were infected with oncogenic H-RAS V-12 virus and selected with bleomycin (30 ug/ml) for approximately one week.

Excision of Ssrp1 (in vitro and in vivo)
For the excision of Ssrp1 in vitro, primary, immortalized, or transformed cells (1x10 6 ) were plated in 150 mm plates. The cells were treated with 2 μM 4-OHT for 96 h (primary cells) or 120 h (immortalized and transformed cells). The medium was replaced every 48 h with fresh 2 μM 4-OHT containing medium. The excision of Ssrp1 was confirmed at both the DNA and protein level.

Immunoblotting
Protein extracts were prepared by lysing cells in Pierce 1 luciferase cell lysis buffer (2x) (Thermo Scientific, catalog no. 16189) for in vitro studies. Protein extracts from mouse spleen were prepared by lysing with 1x RIPA buffer (150 mM sodium chloride, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris pH 8.0). Protein concentrations were determined using Quick Start ™ Bradford 1x dye reagent (Bio-Rad, catalog no. 500-0205). Equal amounts of protein were run on precast 4 to 20% gradient gels (Bio-Rad, catalog no. 3450028), and transferred to PVDF membranes (Bio-Rad, catalog no.1620177). The membranes were blocked with 5% nonfat milk-PBS buffer and incubated with primary and secondary antibodies according to the manufacturer's instructions. Western Lightning Plus-ECL (PerkinElmer catalog no. NEL104001) substrate was used to detect the proteins.

Immunofluorescence staining
Cells were washed with PBS and fixed with 4% paraformaldehyde for 15 min at room temperature. The fixed cells were blocked with PBS containing 3% BSA and 0.2% Triton X-100 for 1 h. The cells were incubated with primary antibody (1:200, anti-SSRP1 or anti-CRE recombinase) in blocking solution for 1 h. The samples were washed three times in wash solution (6x dilution of blocking solution in PBS) for 10 min each and then incubated with secondary antimouse Alexa Fluor 1 antibody (1:500) in washing solution for 45 min. The cells were washed three times in washing solution for 10 min each and then counterstained with Hoechst 33342 (1 μg/mL) for 10 min. Images were acquired using a Zeiss Axio Observer A1 inverted microscope with N-Achroplan 100x/1.25 oil lens, Zeiss MRC5 camera, and AxioVision Rel.4.8 software. Image analysis and quantification were performed using ImageJ software. Each image was divided into four quadrants, and ten cells in each quadrant were analyzed to determine the corrected total cell fluorescence (CTCF) using the following equation: CTCF = integrated density-(area of the selected cell x the mean fluorescence of background readings).

Methylene blue staining
Primary (1x10 6 ) or immortalized or transformed (5x10 5 ) fibroblasts were seeded in duplicate in a 6-well plate and allowed to grow to confluency (~3 to 4 days). Cells were fixed and stained with 0.5% methylene blue staining solution (1 ml per well) in 50% methanol (, catalog no. M4159 MilliporeSigma) for 10 min at room temperature with shaking. The stained cells were washed with water, and the plates were air dried.
Expression of the excised gene was determined using F1 and R2 (F1:R2). qPCR was performed using Applied Biosystems Power SYBR 1 Green PCR Master Mix (, catalog no. 4367659 Ther-moFisher) according to the manufacturer's instructions using the default parameters of the 7900HT sequence detection system (ABI PRISM; Applied Biosystems).

Statistical analysis
The unpaired t-test (Mann-Whitney test) was used for comparisons of quantitative data from the control and treated groups. Analyses were conducted using GraphPad Prism 7.03. All pvalues are two-sided.

I. Generation of a mouse model with conditional deletion of Ssrp1 (Ssrp1 fl/fl )
Previously, we showed that FACT stability is dependent on the dimerization of its two subunits (i.e., SPT16 and SSRP1) [8]. Furthermore, SPT16 is only stable when SSRP1 mRNA is present in the complex [8]. Therefore, a targeting vector that allowed for conditional KO of Ssrp1 was used to generate a conditional FACT KO mouse model. The EUCOMM targeting vector consisted of the complete Ssrp1 gene with two LoxP sites flanking exons 4 to 9 of Ssrp1 and a synthetic cassette. The synthetic cassette consisted of a promoter-less LacZ gene and neomycinresistance gene (Neo r ) under the control of the β-actin promoter. A third LoxP site was located upstream of the β-actin promoter. This synthetic cassette, which was flanked by two FRT sites, was within the intronic region between exons 3 and 4 of Ssrp1 (Fig 1A). Similar to the Cre-LoxP system, the Flp-FRT site-specific recombinase system uses FRT sites that are recognized and recombined by Flp recombinase [28]. The targeting vector was electroporated into G4 (B6.129s) mouse ES cells. Following selection with G418, Southern blot-based screening for the 3' and 5' recombination arms was used to identify clones containing the homologous recombination of the mutant (Ssrp1 Neo LacZ fl ) Ssrp1 allele in place of the endogenous wild-type allele (Ssrp1 + ) (Fig 1C). Chimeras were generated and subsequently bred with wild-type mice (Ssrp1 +/+ ) to obtain an F1 generation of heterozygous (Ssrp1 Neo LacZ fl/+ ) mice. Their genotypes were determined by PCR (Fig 1D).
Ssrp1 is not expressed ubiquitously [3] in adult mice. LacZ did not have its own promoter and could be expressed under the control of the endogenous Ssrp1 promoter. β-galactosidase staining was performed to compare the expression of mutant and wild-type alleles. Positive staining was observed in tissues from heterozygous mice that expressed SSRP1, including testes (high FACT expression), colon (moderate FACT expression), and pancreas (weak FACT expression). Staining was absent in tissues that do not express SSRP1 (e.g., lung). Thus, the Ssrp1 Neo Lacz fl allele had a similar expression pattern to that of the wild-type Ssrp1 + allele. Tissues from wild-type mice were negative for LacZ staining (Fig 1E).
For future studies, we wanted to use LacZ as a reporter of mutant Ssrp1 expression. Thus, we generated mice expressing the LacZ gene but lacking the expression of the Breeding of Ssrp1 Neo LacZ fl /+ to Ssrp1 Neo LacZ fl /+ resulted in 59% Ssrp1 Neo LacZ fl /+ and 40.9% Ssrp1 +/+ pups based on 22 pups from three litters. However, homozygous pups (Ssrp1 Neo LacZ fl/ Neo LacZ fl ) were not obtained. Because the expression of at least one Ssrp1 allele is necessary to maintain embryonic viability at early stages of embryonic development [2], we hypothesized that the lack of generation of Ssrp1 Neo LacZ fl/ Neo LacZ fl mice might be due to the lack of SSRP1 expression from the mutant allele.
To test this hypothesis, we compared the SSRP1 and SPT16 protein levels in the spleens (a FACT-positive organ in adult mice) of Ssrp1 +/+ and Ssrp1 Neo LacZ fl /+ mice from two different pairs of littermates (15 weeks and 28 weeks of age). The results showed that the expression of both subunits was approximately two times higher in the Ssrp1 +/+ mice compared to the Ssrp1 Neo LacZ fl /+ mice. These data suggested the absence of expression from the Ssrp1 Neo LacZ fl allele, and could explain the absence of homozygous Ssrp1 Neo LacZ fl pups (Fig 2A). We hypothesized that the strong β-actin promoter used for Neo expression might be interfering with Ssrp1 expression. However, we later learned that the targeting vector provided by the EUCOMM was predicted to generate a null allele through splicing to the lacZ trapping element contained in the synthetic cassette. Unfortunately, this characteristic of the synthetic cassette was not explicitly explained by EUCOMM and has only been discussed in one publication [29].
To generate homozygous floxed Ssrp1 mice, we had to remove the synthetic cassette using Flp-mediated FRT recombination. Ssrp1 Neo LacZ fl /+ mice were crossed with mice expressing Flp recombinase (Flpe). From this cross, only~28% of the progeny had the synthetic cassette excised based on 14 pups from two litters (Fig 2B), which suggested that the rate of excision at the FRT sites by Flpe was low. To overcome this problem, we obtained mice expressing FlpO recombinase. According to the literature, FlpO has a higher recombination efficiency compared to Flpe [30,31]. Use of the FlpO recombinase resulted in~43% of the progeny undergoing recombination based on 14 pups from two separate litters (Fig 2C).
To determine if the excision efficiency could be improved in vitro, ES cells were electroporated with either Flpe or FlpO plasmid and tested for the excision of the synthetic cassette DNA from F1 progeny of wild-type (Ssrp1 +/+ ) crossed with chimera (Ssrp1 Neo LacZ fl /+ ). Two bands are present in heterozygous pups (Ssrp1 Neo LacZ fl /+ ) (red). Numbers indicate individual pup IDs in one litter. E) β-Galactosidase assay to determine LacZ activity (Blue-Green stain for positive activity) on SSRP1 positive (colon, pancreas and testis) and negative (lung) tissues isolated from Ssrp1 +/+ and Ssrp1 Neo LacZ fl /+ mice. https://doi.org/10.1371/journal.pone.0199785.g001 The fine print of CreER T2 -LoxP system PLOS ONE | https://doi.org/10.1371/journal.pone.0199785 June 28, 2018 using PCR (S1 Fig). Similar to the in vivo data, after selection of the ES cells for the expression of Flpe or FlpO, only 22% and 50% of the clones, respectively, showed excision of the synthetic cassette. Thus, mosaic excision of the synthetic cassette from the mutant allele was observed both in vitro and in vivo.

II. Excision of Ssrp1 with ectopically expressed CreER T2 in vitro
While generating mice for in vivo experiments, we tested the excision of the Ssrp1 fl allele in vitro using fibroblasts isolated from the tails of Ssrp1 +/+ , Ssrp1 fl/+ , and Ssrp1 fl/fl mice. We previously observed very low FACT levels in primary non-tumor cells. However, FACT levels were increased upon oncogene-induced transformation of the primary cells [32]. Therefore, to establish in vitro FACT-positive cells, we immortalized primary tail fibroblasts using the dominant negative p53 genetic suppressor element 56 (GSE56 [32]) and then transformed the immortalized cells using H-Ras V12 oncogene. As expected, immortalized and transformed cells expressed higher levels of the FACT subunits (Fig 3A). Immortalized and transformed cells were infected with retroviral CreER T2 recombinase (MSCV CreER T2with puromycin selection marker [33]) followed by puromycin selection. After selection, cells were treated with 4-OHT (2 μM), an active metabolite of tamoxifen. SSRP1 protein levels were slightly reduced at some time points after the start of treatment with 4-OHT (48 to 120 h), but no clear time-dependence was observed (Fig 3B, 3C and 3D). PCR analysis confirmed excision of the critical exons of Ssrp1 (Ssrp1 Δ ) following treatment with 4-OHT. However, the presence of an unrecombined Ssrp1 fl allele was also detected ( Fig 3E). These data suggested that the excision was happening only in a fraction of cells in the population or only one Ssrp1 fl allele was excised. These results may be due, in part, to the heterogeneity among cells in the population with respect to Cre-recombinase expression levels.
To test this hypothesis, single cell clones of transformed Ssrp1 fl/fl MSCV CreER T2 fibroblasts were generated by limiting dilution. Cloned cells were subjected to 4-OHT treatment, and SSRP1 immunofluorescence was performed to identify the clones with complete excision of Ssrp1. Variable excision was observed within each clone, and no clones demonstrated uniform excision in all cells in response to 4-OHT (Fig 4 and S2 Fig). Thus, excision did not occur in all cells and, moreover, excision of both alleles occurred in almost none of the cells.
This low recombination efficiency could be due to the inaccessibility of the LoxP sites to the Cre recombinase. To test this hypothesis, we assessed the excision efficiency of CreER T2 in cells with a more open chromatin state achieved using treatment with trichostatin A (TSA), an inhibitor of histone deacetylase (HDAC). Immortalized Ssrp1 fl/fl MSCV CreER T2 fibroblasts were treated with a range of 4-OHT concentrations in the absence or presence of TSA (100 ng/ ml) for 48 h. SSRP1 protein levels were reduced following 4-OHT treatment in the presence of TSA. However, a reduction in SSRP1 protein levels was also observed with TSA alone ( Fig  5A). Genomic DNA PCR showed no improvement in the excision efficiency, suggesting that the opening of chromatin was not sufficient to improve the recombination efficiency of ectopically expressed CreER T2 (Fig 5B).
Another possibility for the low excision efficiency was low Cre-ER T2 expression. To test this hypothesis, we evaluated the cytoplasmic and nuclear distribution of CreER T2 protein with and without 4-OHT treatment. Immunofluorescence staining for Cre and SSRP1 in transformed These results suggested that CreER T2 was unstable in the absence of 4-OHT. Although all the cells in the population were resistant to puromycin, some had suppressed CreER T2 expression.

III. Excision of Ssrp1 in cells from mice with germline expression of CreER T2
The results presented thus far suggested that ectopic expression of CreER T2 did not lead to efficient homozygous excision of Ssrp1 in Ssrp1 fl/fl fibroblasts. Hence, we determined whether germline-expressed CreER T2 was able to excise Ssrp1 effectively. Ssrp1 fl/fl mice were crossed with homozygous Rosa26 CreER T2 (CreER T2+/+ ) mice. Since the literature suggests that even one allele of CreER T2 is sufficient for the excision of a gene of interest [34][35][36], heterozygous mice for both alleles (Ssrp1 fl/+ CreER T2+/-) were bred to get mice homozygous for floxed-Ssrp1 with at least one allele of CreER T2 (Ssrp1 fl/fl CreER T2+/-). Fibroblasts were isolated from tail tissue of Ssrp1 fl/fl CreER T2+/mice and then immortalized and transformed. To test the efficiency of germline CreER T2 , immortalized Ssrp1 fl/fl CreER T2+/cells were treated with 4-OHT for 120 h. Ssrp1 excision in these cells was assessed using PCR ( Fig 5C). The results showed the presence of a recombined Ssrp1 allele in both 4-OHT-treated and untreated samples, which suggested that germline CreER T2 is leaky. Moreover, the Ssrp1 fl allele was still observed in 4-OHTtreated cells, which suggested that like retroviral CreER T2 , germline CreER T2+/may not be active in all cells. Alternatively, there was only a low chance of excising both alleles together or calculate the Corrected Total Cell Fluorescence (CTCF). E) PCR of the genomic DNA to determine the excision of Ssrp1 after 4-OHT treatment of immortalized and transformed Ssrp1 fl/fl fibroblasts with or without transduction with MSCV CreER T2 .
To determine if the excision efficiency of germline CreER T2+/could be improved when only one allele was excised, we treated immortalized Ssrp1 fl/fl CreER T2+/fibroblasts with 4-OHT (round 1) for 120 h. Genomic DNA was isolated from both treated and untreated cells to assess the excision of the Ssrp1 allele. At the same time, the treated cells were re-plated and subjected to a second round of 4-OHT treatment for 120 h. Consistent with the previous results, a single round of 4-OHT treatment resulted in incomplete excision as indicated by the presence of both Ssrp1 fl and Ssrp1 Δ bands on Part A in S4 Fig. Moreover, the second round of treatment did not lead to excision of both alleles. These results suggest that the inefficiency of a single copy of germline-expressed CreER T2 may not be because of its inability to excise two alleles but may be due to inaccessibility of the LoxP sites to the Cre recombinase. To test this hypothesis, we treated cells with 4-OHT in the presence or absence of TSA in an attempt to make the chromatin more accessible., The efficiency of the single germline copy of CreER T2 following treatment with 4-OHT and TSA increased compared to the ectopically expressed CreER T2 as seen by the decrease in the Ssrp1 fl band in genomic DNA analysis (Fig 5E). In addition, the absence of SSRP1 protein upon TSA treatment showed that the opening of the chromatin enhanced the excision rate of endogenously expressed CreER T2 . However, SSRP1 was absent following TSA treatment even in the absence of 4-OHT. These data suggest that either CreER T2+/is leaky or the inhibition of HDAC causes a reduction of FACT subunit protein levels or both (Fig 5D). Since the CRE-mediated recombination efficiency was low when only one allele of CreER T2 was present, we tested whether the presence of two alleles of CreER T2 would be more efficient. We generated mice homozygous for both floxed Ssrp1 and CreER T2 (Ssrp1 fl/fl CreER T2+/+ ), and then generated primary, immortalized, and transformed cells as described above. 4-OHT treatment significantly reduced SSRP1 levels in all cell types (Fig 6A). PCR analysis of genomic DNA also confirmed excision of the critical exons of Ssrp1 and absence of the mutant allele of SSRP1 upon treatment with 4-OHT ( Fig 6B). Thus, expression of CreER T2 from two alleles resulted in a much more efficient recombination, which in turn was more efficient than recombination from ectopically expressed CreER T2 (MSCV CreER T2 ) (Fig 6C).

IV. Excision of Ssrp1 in mice with germline CreER T2
To compare the efficiency of Ssrp1 fl excision by one or two copies of germline CreER T2 in vivo, we treated Ssrp1 fl/fl CreER T2+/+ and Ssrp1 fl/fl CreER T2+/mice with 1 mg/day tamoxifen and two Ssrp1 fl/fl CreER T2+/mice with vehicle control intraperitoneally for five days. DNA was collected The fine print of CreER T2 -LoxP system from the mice seven days after the final treatment to determine the excision efficiency. PCR analysis showed complete excision of the Ssrp1 fl gene in the Ssrp1 fl/fl CreER T2+/+ mice (Fig 7A). One of the Ssrp1 fl/fl CreER T2+/mice was treated twice with tamoxifen to mimic the in vitro study. However, two rounds of treatment did not result in complete excision but there was a slight decrease in the intensity of the Ssrp1 fl band after the second round of treatment ( qPCR of genomic DNA from these mice showed that there was a 2.5-fold higher rate of Ssrp1 excision in Ssrp1 fl/fl CreER T2+/+ mice compared to Ssrp1 fl/fl CreER T2+/mice ( Fig 7B). Because of the small sample size, the differences did not reach statistical significance. However, these observations potentially indicate that, like in vitro, two alleles of Rosa26 CreER T2 are required to achieve complete excision of Ssrp1 in mice.

Discussion
CreER T2 -mediated gene inactivation is a powerful tool to study a gene of interest with temporal and spatial control. However, to fully understand the function of a gene, it is important that the Cre recombination be efficient and complete. Here, we report the limitations of the CreER T2 system encountered while generating the Ssrp1 KO mouse model. The results of this study highlight the inefficiencies of this system and provide insights on how to increase the recombination efficiency. We understand that each case of genetic engineering potentially has specific limitations, which may result from specific gene or genomic features. Thus, the experience that we report here may only be fully applicable to Ssrp1. Nevertheless, we believe it is The fine print of CreER T2 -LoxP system pertinent to disclose such observations to develop a better understanding of the most widely used system for genetic manipulation.
The excision efficiency of CreER T2 is very low when expressed ectopically through retroviral transduction. Differential excision efficiency of each floxed allele was seen after treatment with 4-OHT, which may have been due to: 1) mosaic expression of CreER T2 (i.e., not all cells express CreER T2 ) even though they were puromycin resistant and incomplete excision of the gene occurs [37]. It is possible that CreER T2 was active in the absence of any ligand or was activated by estrogen or other steroids present in the FBS, which might lead to excision of Ssrp1 in untreated cells (i.e., negative selection for the cells with active CreER T2 in the absence of tamoxifen). 2) CreER T2 might not be in an active state in the cells. In particular, clonal variation in the excision of Ssrp1 indicated the loss of CreER T2 activity in certain cells generated from a single cell clone. Analysis of the cytoplasmic and nuclear levels of CreER T2 suggested that this protein was not stable in the absence of its ligand as only a few cells stained positive for CreER T2 in the cytoplasm in the absence of 4-OHT even though they were puromycin resistant. However, the number of cells with positive nuclear stain increased in the presence of 4-OHT. These results are not unprecedented as many steroid receptors (e.g., androgen receptor) are unstable in the absence of their ligand [38]. Thus, due to the lack of CreER T2 activity, partial excision of Ssrp1 might occur. 3) Ectopically expressed CreER T2 might not be able to access the floxed gene due to a closed chromatin state. Upon treatment with an HDAC inhibitor, the chromatin might become more accessible, leading to complete excision by CreER T2 . However, our results suggest that this explanation is less likely because complete excision did not occur even in the open chromatin state caused by TSA (Fig 5A and 5B).
Expression of a single germline copy of Rosa 26 CreER T2 (CreER T2+/-) also resulted in incomplete excision of Ssrp1. We found that CreER T2+/was leaky (in vitro) as both Ssrp1 Δ and Ssrp1 fl were present in the absence of 4-OHT treatment (Fig 5C). Although partial excision of Ssrp1 using ectopically expressed CreER T2 was not due to inaccessibility of Cre to the chromatin, this was not true for the single germline copy of Cre. Treatment of Ssrp1 fl/fl CreER T2+/immortalized cells with TSA and 4-OHT increased the degree of excision of Ssrp1 (Fig 5D and  5E). We also saw the presence of Ssrp1 Δ in cells treated with TSA alone (i.e., no 4-OHT treatment), confirming that CreER T2+/was leaky. FACT expression is critical for the survival of transformed cells and, to some extent, immortalized cells [4]. Therefore, cells expressing CreER T2 might have undergone complete excision of Ssrp1 in the absence of 4-OHT but did not survive. Only cells that had incomplete excision of Ssrp1 survived. Thus, we could see both Ssrp1 Δ and Ssrp1 fl bands in our assay and the presence of SSRP1 in these cells. Therefore, expression of a single copy of Rosa 26 CreER T2 allele was inefficient in yielding homozygous excision of Ssrp1. However, we found that two copies of Rosa26 CreER T2 expressed in the germline led to complete excision of Ssrp1 both in vitro and in in vivo. Hence, to achieve homozygous excision of Ssrp1, two endogenously expressed alleles of Rosa26 CreER T2 are required. Furthermore, determination of excision frequency is necessary for correct assessment of results.
The CreER T2 -LoxP system was developed in the late 1990s to early 2000s. It has led to many novel and interesting discoveries, and this system prevails even today in the field of gene targeting. With the development of CRISPR technology, the use of Cre-LoxP has been reduced. However, the use of CreER T2 -LoxP remains constant as a frequently used recombinase system to generate a KO mouse model with temporal and spatial control. Our study describes important factors to consider when employing this system to generate a conditional KO mouse model, particularly of a gene that may play a necessary role in tumor promotion. Our results may not be applicable to studies involving inducible KO of a tumor suppressor gene or the knock-in of a gene.