A Nucleolus-Predominant piggyBac Transposase, NP-mPB, Mediates Elevated Transposition Efficiency in Mammalian Cells

PiggyBac is a prevalent transposon system used to deliver transgenes and functionally explore the mammalian untouched genomic territory. The important features of piggyBac transposon are the relatively low insertion site preference and the ability of seamless removal from genome, which allow its potential uses in functional genomics and regenerative medicine. Efforts to increase its transposition efficiency in mammals were made through engineering the corresponding transposase (PBase) codon usage to enhance its expression level and through screening for mutant PBase variants with increased enzyme activity. To improve the safety for its potential use in regenerative medicine applications, site-specific transposition was achieved by using engineered zinc finger- and Gal4-fused PBases. An excision-prone PBase variant has also been successfully developed. Here we describe the construction of a nucleolus-predominant PBase, NP-mPB, by adding a nucleolus-predominant (NP) signal peptide from HIV-1 TAT protein to a mammalian codon-optimized PBase (mPB). Although there is a predominant fraction of the NP-mPB-tGFP fusion proteins concentrated in the nucleoli, an insertion site preference toward nucleolar organizer regions is not detected. Instead a 3–4 fold increase in piggyBac transposition efficiency is reproducibly observed in mouse and human cells.


Introduction
First cloned from the cabbage looper moth Trichoplusia ni, piggyBac is a class II DNA transposon that mobilizes DNA segments in a ''cut-and paste'' manner [1]. The piggyBac transposase (PBase) system has been widely applied as a genomic manipulation tool to various mammalian cell lines and model organisms, such as plants, cattle, pig, mouse, rat, rabbit, chicken, worms, fly, mosquito, planarian, yeast, protists, and several nonmodel insects . Major features of the piggyBac system include a high transposition efficiency in different species, large cargo size, seamless removal, and relatively low insertion site preference (other than the conserved TTAA integration sequence) [3,19,[24][25][26].
The piggyBac system is also a candidate tool for regenerative medicine applications [42][43][44]. For induced pluripotent stem cell research, piggyBac can carry reprograming factors that enter and exit the genome without changing any nucleotides [45][46][47][48]. The piggyBac system has been applied to in vitro gene correction research designs in stem cells, to aid in the complete removal of a piggyBac inverted terminal repeat (ITR)-flanked drug selectable marker sequence from an exon without changing an encoded amino acid after genomic manipulations [49].
The transpositional function of mammalian codon-optimized PBase (mPB) can be maintained after mPB is fused with other proteins [34,50]. For example, Cadinanos and Bradley fused PBase with a mutant estrogen receptor variant. Through this fusion, PBase was able to access the nucleus and mediate transposition, but only upon treatment with a steroid compound (tamoxifen) [50]. In another study, the AAV Rep-PBase fusion protein exhibited enriched capability for transposon insertion at Rep recognition sequences in the human genome [51]. Wilson et al. fused a site-specific synthetic zinc-finger DNA-binding domain (ZNF) to the N-terminus of mPB. The chimeric ZNF-mPB transposase exhibited a higher rate of site-directed genomic integration than the native mPB [52,53]. Owens et al. fused the Gal4 DNA-binding domain (DBD) to mPB, and the chimeric Gal4-mPB facilitated transposon integration near artificially introduced upstream activating sequences [54].Transcription activator-like effector (TALE) is a new DNA-binding protein derived from the Xanthomonas sp [55,56]. TALEs contain 34-amino acid tandem-repeat modules, which can be rearranged to target new DNA sequences [57,58]. A TALE-linked PBase was demonstrated to direct transposition to a target region [59]. Protein function can be executed more efficiently by targeting the protein to a specific subcellular compartment. For example, a nuclear localization signal is used to direct recombinases and transcription factors to the nucleus, thereby increasing their access to DNA and other cofactors. Protein transduction domains (PTDs) are used to transfer cargo proteins efficiently across the biological membrane [11,60,61].
Here, we report a fused mPB with an 11-amino acid peptide from the HIV-1 transactivator of transcription (TAT) protein, which contains a short stretch of signal peptide (GRKKR) that is essential to direct TAT predominantly to the nucleolar compartment [62,63]. This fusion transposase is designated as ''nucleoluspredominant (NP)-mPB''. The nucleolus is a three-layer structure composed of nucleolar factors, such as fibrillarin, and chromosomal regions consisting of tandem-repeat sequences of ribosomal genes (rRNAs), called nucleolar organizer regions (NORs). In human, NORs include the 5S, 8S, 18S, and 28S rRNA gene clusters on the short arms of chromosomes 13, 14, 15, 21, and 22. Because there are hundreds of rRNA repeats in the genome, a transposon that jumps into an NOR is unlikely to disrupt a unique, essential gene and cause a phenotypic consequence. Therefore, we chose to target our PBase to the proximity of NORs by adding an NP signal peptide to its N-terminus. Although the subcellular distribution of this modified NP-mPB was concentrated (but not limited to) nucleoli, transposon insertion site preference towards NORs was not observed. Interestingly, both mouse and human cell lines demonstrated a reproducible 3-to 4-fold increase in transposition efficiency compared to mPB.

Measurement of Transposition Efficiency
Procedures for electroporation and drug selection for the mouse [65] and human [16] ES cells were performed according to previous studies. Briefly, to determine the difference in transposition efficiency between mPB and NP-mPB transposase, 1610 7 mouse ES cells were electroporated with 10 mg of donor plasmids (pXL-T3-Neo-UGm-cHS4X; UGm) and 2 mg of helper plasmids (pTriEX-mPB or pTriEx-NP-mPB). Electric pulses were provided by a BTX Electro Square Porator EM830 (Harvard Apparatus, Inc., Holliston, MA, USA; 230 V, 0.77 ms). After electroporation, each cuvette of mouse ES cells was seeded onto a 10-cm plate with a feeder layer. One day after electroporation, sequential drug selection was conducted with 200 mg/ml G418 for 2 weeks. Mouse ES cells were fixed by methanol and stained by 0.5% toluidine blue for 15 min. Colonies were washed twice with PBS and dried at room temperature for 30 min. Transposition efficiency was evaluated by the number of surviving colonies per plate.
To determine the PBase transposition efficiency in human ES cells, H9 cells were cultured with 10 mM of rho-associated kinase (ROCK) inhibitor (Y-27632; Calbiochem, San Diego, CA) 2 to 4 h before electroporation, to prevent apoptosis during electroporation. For each electroporation, 10 mg of donor plasmid (UGm) and 10 mg of helper plasmid (pTriEX-mPB or pTriEX-NP-mPB) were mixed with 10 7 H9 cells. After electroporation, each cuvette of human ES cells was seeded onto a 10-cm plate with a feeder layer. The cells were cultured with medium containing 10 mM ROCK inhibitor. On day 2 after electroporation, cells were cultured with medium without ROCK inhibitor. On days 3 to 7, cells were cultured with a selection medium containing 50 mg/ml G418. On days 8 to 9, the G418 dosage was increased to 100 mg/ ml. On day 10, surviving colonies were stained by 0.5% crystal violet for 15 min. The colonies were washed with PBS three times and dried at room temperature for 30 min.
To determine the PBase transposition efficiency in a human cancer cell line, 6610 6 Hela cells were transfected in a 10-cm plate with 10 mg of donor plasmids (pXL-T3-Neo-UGm-cHS4X; UGm) and 2 mg of helper plasmids (pTriEX-mPB or pTriEx-NP-mPB) using a transfection reagent (Maestrogen Inc., Las Vegas, NV, USA). After transfection, drug selection was conducted with 200 mg/ml G418 for 10 days. Hela cells were fixed by methanol, stained by 0.5% crystal violet, and the transposition efficiency was evaluated by the number of surviving colonies per plate.
A similar transposition efficiency assay was performed in HEK293T cells. Because HEK293T cells are intrinsically neomycin-resistant, the UGm transposon was replaced by the pGG134 transposon carrying the hygromycin resistance gene cassette [35]. To compare the transposition efficiency mediated by different amounts of PBase, 5610 6 HEK293T cells were transfected with 10 mg of donor plasmids (pGG134) and 0.25 to 2 mg of helper plasmids (pTriEX-mPB or pTriEx-NP-mPB). One day after transfection, one-fifth of the cells were split into another 10cm plate. After selection by 200 mg/ml hygromycin for 7 days, HEK293T cells were fixed by methanol, stained with 0.5% crystal violet, and the transposition efficiency was evaluated by the number of colonies per plate. In the comparison of transposition efficiency, Student's t test was applied to determine the significance of the difference.

Polymerase Chain Reaction (PCR) Assays for Integration Site Analysis
Mouse ES cells were subjected to inverse PCR, as described previously [16]. Briefly, Spel was used to digest 1.5 mg of genomic DNA and then was inactivated. UGm primers (UGm-I and UGm-N) and Bac primers (BacE and BacF) were used to perform nested PCR reactions on self-ligated Spel-restricted genomic DNA ( Figure 1C). Amplified DNA fragments were fractionated and purified for direct sequencing with the primer BacF.  genome. These sites fulfilled the following three criteria: they (i) contained the ITR sequence from the nested primer to the end of the 59 ITR, (ii) showed $96% identity to the genomic sequence over the high-quality sequence region, and (iii) had the highest alignment score. In this study, the gene region was specified as the genomic location containing the coding region and the extended region. The coding region was defined as the coordinates from the first base of exon 1 to the last base of the final exon of the longest isoform, to avoid ambiguous counts. The extended region included the sequences extending 5-kb upstream and downstream of a gene. The non-gene region was the genome interval between gene regions. All of the bioinformatics data of exons, introns, and non-gene regions were obtained from the UCSC Genome Browser website.

Protein Stability Assay
To test the protein stability of PBase variants, HEK293T cells were transfected with PBase-expressing plasmids (pTriEX-mPB-2A-eGFP and pTriEX-NP-mPB-2A-eGFP) using the Maestrofectin transfection reagent (Maestrogen Inc.). Transfected cells were harvested 2 days later and stained with allophycocyanin (APC)conjugated anti-His tag antibody (1:500, GeneTex, Inc.). The specimens were strained through BD Falcon cell strainers and kept on ice before acquisition. Flow cytometry analysis was performed with a FACS AriaII (BD Bioscience Inc., San Jose, CA, USA) facilitated by FlowJo software (Tree Star, Inc., Ashland, OR, USA). In all samples, .10,000 cellular events from the defined cell cluster were analyzed per tube.

Cytotoxicity Assay
To analyze the potential cytotoxicity caused by mPB and NP-mPB, HEK293T cells were transfected with different amounts of PBase-expressing vectors, and cell mortality was measured by a (3-

Immunocytochemistry and Fluorescent Confocal Microscopy
Hela cell explants were fixed with 4% paraformaldehyde in PBS for 20 min at room temperature, washed twice with PBS, permeabilized for 20 min using 0.3% Triton X-100/PBS, and blocked for 30 min in blocking buffer (10% normal goat serum and 2% bovine serum albumin/PBS). Hela cell explants were incubated with anti-cyclin T1 (1:200; Santa Cruz Technology, Santa Cruz, CA, USA) and anti-fibrillarin (1:200; Abcam plc) antibodies, diluted in blocking buffer at 4uC overnight, and incubated for 1 h at room temperature with Alexa Fluor 568-and Alexa Fluor 647-conjugated secondary antibodies (Invitrogen). Fluorescence microscopic images were captured and processed using a Leica TCS SP5 confocal spectral microscope imaging system (Leica Microsystems GmbH, Heidelberger, Germany).

Construction of the NP-mPB-expressing Vector
To direct the mPB toward the nucleolar compartment, we added an NP signal peptide sequence (GRKKR) from the HIV-1 TAT protein to the N-terminus of mPB. The resulting NP-mPB fusion was constructed with a pTriEX-HTNC-derived plasmid [60]. The pTriEx-NP-mPB construct was composed of a CAG promoterdriven mPB variant with a preceding (His) 6 tag and an NP signal peptide ( Figure 1A, bottom). In the control pTriEx-mPB plasmid, the coding sequences for the promoter and mPB were kept the same, except that the NP peptide-encoding sequence was removed ( Figure 1A, middle).
To elucidate whether the NP peptide directs the transposase to nucleoli as predicted, constructs encoding mPB and NP-mPB fused in-frame with tGFP were made, and the subcellular localization of the transposase was traced under confocal microscopy ( Figure 1B, top two constructs). We also made constructs that linked either mPB or NP-mPB to an eGFP moiety using a self-cleaving T2A peptide, and compared whether the addition of the NP peptide altered the transposase protein level through the translation or degradation of the peptide ( Figure 1B, bottom two constructs). The transposition efficiency mediated by each transposase variant was tested with a previously described, approximately 12-kB piggyBac transposon, pXL-T3-Neo-UGm-cHS4X (UGm), which carried both dual fluorescence and neomycin-resistance expression cassettes. By using this transposon, PBase-mediated transposition events could be selected with G418 and verified by the dual fluorescence appearance of the surviving colonies ( Figure 1C, Figure S1) [16].

NP-mPB Mediates Elevated Transposition Efficiency in Mouse and Human Cells
To evaluate whether the PBase-mediated transposition efficiency was affected by the additional NP peptide, either pTriEx-mPB or pTriEx-NP-mPB was co-electroporated with the pXL-T3-Neo-UGm-cHS4X plasmid into either mouse or human ES cells. The number of colonies surviving G418 selection was used to measure transposition efficiency mediated by different PBase variants. The expression of dual fluorescence under fluorescent microscopy was used to reconfirm the transposon integration events ( Figure  S1).
In mouse ES cells, the transposition efficiency mediated by NP-mPB was about 3.5-fold that by mPB (Figure 2A, 2D). In human ES cells, the transposition efficiency mediated by NP-mPB was 3.2-fold that by mPB ( Figure 2B, 2E). In Hela cells, the transposition efficiency mediated by NP-mPB was about 3.2-fold that by mPB ( Figure 2C, 2F). Clearly, there was a significant, .3-fold increase in transposition efficiency after the NP peptide was added to mPB. In HEK293T cells, the transposition efficiency was studied by transfection with gradient concentrations of PBase. The transposition efficiency of NP-mPB was 2.5 to 6 times that of mPB ( Figure 3).
For mPB and NP-mPB, no obvious overexpression inhibition was found within the range of PBase-expressing vector amounts in our study because the transposition efficiency was higher when more PBase was transfected. The differences of transposition efficiency between the NP-mPB and mPB groups were smaller when larger amounts of plasmid DNA were transfected.

NP-mPB does not Mediate Transposon Insertional Preference Towards NORs
To examine whether addition of the NP peptide drove the transposon insertional preference towards NORs, we sequenced the NP-mPB-mediated insertion sites and located them on the mouse genome sequence. Single mouse ES cell-derived colonies were picked for inverse PCR to identify the transposon insertion sites. Twelve clones were expanded, PCR-amplified, and sequenced ( Table 1).
The mouse genome BLAT search program revealed that the 12 insertion sites were distributed through 10 of the 20 chromosomes in the mouse genome (i.e., chromosomes 3-5, 9, 10, 12-14, 16, and 17). The distribution was relatively random, except for three sites in chromosome 4. Five of the 12 insertion sites were in the gene region; the other 7 sites were in the non-gene region. None of the insertion sites were in the NOR-defined rRNA gene clusters.

The NP-mPB Protein Level does not Explain the Increased Transposition in Mammalian Cells
To decipher the cause of the increased transposition by NP-mPB in mammalian cells, the PBase protein expression profile was surveyed. One hypothesis to explain the elevated transposition efficiency is that there was an increased PBase protein accumulation in the NP-mPB-transfected cells that resulted from either elevated expression or decreased degradation of PBase. To test this hypothesis, HEK293T cells were transfected with pTriEx-mPB or pTriEx-NP-mPB by a liposome-based protocol. The transfection efficiency was .90%, as revealed by fluorescent microscopy of cotransfected UGm plasmid expression. We confirmed the specificity of the anti-His tag antibody to identify mPB and NP-mPB and their molecular weights on SDS-PAGE by an immunoprecipitation experiment followed by a LC-MS/MS assay ( Figure S2). However, the total transposase protein of NP-mPB was less than that of mPB ( Figure 4A). Therefore, the total amount of transposase expression did not account for the increased transposition efficiency of NP-mPB.
Another hypothesis is that NP-mPB was more concentrated in the subcellular compartment(s) where transposition occurred. Because adding NP signal peptide theoretically leads to nuclear accumulation, Western blot was performed after nucleo-cytoplasmic separation to determine the relative amount of protein in the nucleus. For both NP-mPB and mPB, most of the transposase proteins were in the nucleus. Compared to mPB, the absolute amount of NP-mPB protein in the nucleus was comparatively less. Therefore, the elevated transposition efficiency mediated by NP-mPB was not due to an increased nuclear accumulation of PBase protein.

The Protein Stability of NP-mPB is Decreased Compared to that of mPB
Although the cis-transcriptional regulatory elements were identical on pTriEX-mPB and pTriEX-NP-mPB, the pTriEX-NP-mPB-transfected cells showed less PBase protein on Western blot analysis. To examine whether this finding was due to a difference Figure 7. Subcellular localization of mPB and NP-mPB. Hela cells were transfected with mPB-tGFP or NP-mPB-tGFP expression vectors and immunostained with an anti-cyclin T1 antibody (red) and an anti-fibrillarin antibody (blue). (A) Fluorescence microscopy revealed that mPB was localized in the nucleus, but largely outside the nucleoli, as indicated by the fibrillarin. The cyclin T1 signal was dotted and spared the nucleus. (B) In contrast, NP-mPB was localized in the nucleus, with a strong staining in the nucleoli. There was a dotted pattern of NP-mPB in the nucleus beside the nucleoli. Some of the dots were overlapped with the cyclin T1 signal. doi:10.1371/journal.pone.0089396.g007 in the translation and/or degradation of PBase protein, we created plasmids expressing the mPB-2A-eGFP and NP-mPB-2A-eGFP fusion proteins. Then, the PBase protein stability was determined by flow cytometry of Hela cells transfected with either pTriEx-mPB-2A-eGFP or pTriEx-NP-mPB-2A-eGFP. The presence of transposase (mPB and NP-mPB) was evaluated by using an APC-conjugated anti-His tag antibody. Because the eGFP moiety was first translated in the PBase fusion proteins, but then separated from PBase by self-cleavage of the 2A peptide, the eGFP intensity provided an estimate of the mPB and NP-mPB production independent of subsequent degradation events.
The two transfected populations only showed a slight difference in eGFP intensity distribution on flow cytometry, indicating that similar amounts of mPB and NP-mPB were translated ( Figure 5A, 5C). However, compared to cells from the pTriEx-mPB transfection experiment, close to 4-fold fewer cells from the pTriEX-NP-mPB transfection experiment displayed an APC fluorescent intensity that exceeded the threshold ( Figure 5A, 5B). Furthermore, when we only considered the APC fluorescent intensity of cells from the eGFP+ population from both experiments, the mean APC fluorescent intensities were 671 arbitrary units (a.u.) for pTriEX-mPB and 302 a.u. for pTriEX-NP-mPB transfectants. These findings might imply an increased protein degradation of NP-mPB ( Figure 5D).

NP-mPB Overexpression Causes Little Cytotoxicity
To rule out the possibility that NP-mPB causes cytotoxicity, HEK293T cells were transfected with pTriEx-mPB, pTriEx-NP-mPB, pTriEx-mPB-tGFP, or pTriEx-NP-mPB-tGFP separately ( Figure 6). Cells transfected with a tGFP expression vector were used as a negative control. In transfection experiments using different amounts of PBase-expressing vectors, the survival rates of the control, mPB, and NP-mPB groups were not significantly different. Thus, we found no evidence to suggest that NP-mPB overexpression caused an increase in cytotoxicity.

The NP Signal Directs NP-mPB to a Predominantly Nucleolar Distribution
For transposase to mediate transposition events, it should localize to somewhere near its substrate (genomic DNA in the nucleus). Therefore, it was reasonable to assume that a more efficient delivery of PBase to its target site could increase the transposition rate. NP-mPB theoretically has a preference to locate in the nucleolus. To define the distribution of mPB and NP-mPB in the subcellular compartments, Hela cells were transfected with either pTriEx-mPB-tGFP or pTriEx-NP-mPB-tGFP, and then immunostained with anti-cyclin T1 and anti-fibrillarin antibodies. Fluorescence microscopy revealed that both mPB and NP-mPB proteins located mainly in the nucleus. However, whereas NP-mPB was obviously concentrated in the nucleoli and other dots overlapping with cyclin T1 signals (Figure 7), mPB was evenly distributed in the nucleus, relatively sparing the nucleoli.

Discussion
The function of mPB has been demonstrated to remain intact when mPB is fused with different proteins, such as with the Gal4 or zinc-finger DBD, to create a targeted transposition system [53,54]. In this study, PBase was fused with the PTD of the HIV-1 TAT protein (N'-GRKKRRQRRRPPA-C'). The PTD traverses the cell membrane to promote cellular uptake of the fusion protein [66,67], and may enhance membrane penetration directly or by endocytosis [68,69]. Additionally, the fusion cargo influences the permeabilization mechanism [70], and the delivery protocol impacts the transmembrane efficiency and final subcellular localization of the fusion protein. Studies have demonstrated that liposome-based transient transfection of TAT-GFP in mammalian cells permits nuclear localization of the fusion protein [71,72]. It has also been shown that the PTD of TAT protein contains nucleolar localization signal [62,63,[73][74][75][76].
Our initial goal was to create a modified mPB that would mediate transposition in the NORs, where gene disruption would typically not be harmful. Eleven amino acids of the PTD peptide from TAT sequences were added to the N-terminal region of mPB as an NP signal peptide, to target PBase to the proximity of NORs in a transient transfection system. However, analysis of the insertion site did not reveal a transposition preference for NORs. The transposition efficiency mediated by NP-mPB was increased by 3-to 4-fold, with about 40% of integrations occurring in the gene region.
Although modification of transposase by ligation with a signal peptide or protein can theoretically change the transposition preference, it is not always altered as expected. One example includes the newly developed excision competent/integration defective PBases. When these PBases were fused with a ROSA26binding ZNF to restore the integration ability, transposon insertion into ROSA26 target sites was not obviously increased [77]. Likewise, in this study, NP-mPB was not successful at targeting a specific gene region (i.e., NORs). However, the significant increase of transposition events mediated by NP-mPB was an advantageous finding.
To investigate the mechanism of increased transposition efficiency, we assessed the PBase protein expression profile. Compared to mPB, the NP-mPB was associated with a higher integration rate for a lower amount of protein present. The Western blot analysis of the nucleo-cytoplasmically separated samples also indicated that NP-mPB was less enriched in the nucleus compared to mPB when equal amounts of the plasmids were transfected. To rule out the possibility that the reduced NP-mPB production was due to toxicity with plasmid transfection, a cell survival assay was performed. The data did not demonstrate a significant cytotoxicity difference between mPB and NP-mPB at various concentrations. However, the protein stability assay suggested that NP-mPB is less stable than mPB and, therefore, may have a higher degradation rate, which could explain its lower amount of protein. Nevertheless, from the transposition efficiency tests using different amounts of PBase-expressing vectors in HEK293T cells, overexpression inhibition is unlikely to explain the observed transposition efficiency differences between the PBase species in our experiment.
It is worth probing the mechanism for the higher transposition efficiency but lower protein presentation of NP-mPB. One possible explanation is the targeted delivery ability of the fusion protein. The TAT PTD permits the fused protein to localize to the nucleus, specifically in the subnuclear compartments [76]. Two factors that are closely associated with TAT in the nucleus are fibrillarin and cyclin T1. TAT was reported to associate with the fibrillarin-U3 snoRNA complex in Drosophila nurse cell nucleoli [74]. Although fibrillarin colocalized with TAT in the nucleoli in another study, no direct protein-protein interaction was found by fluorescence resonant energy transfer (FRET). Cyclin T1 is a transcriptional elongation factor that interacts with TAT and the transactivating response element to mediate activation of some lentiviral promoters [78][79][80][81]. Visualization of FRET between TAT and cyclin T1 proved the direct proteinprotein interaction [76].
Because of the many advantages of the piggyBac system for mammalian functional genomics and regenerative medicine research, substantial effort has been put forward to improve the expression level and enzyme efficiency of PBase. Transcriptional abundance can be ensured through the use of cis-regulatory elements, such as strong promoters mediating high-level transcription and effective polyadenylation signals to increase mRNA stability. Studies have sought to optimize codon usage for translation in mammals [50]. Using error-prone PCR and high-throughput screening, researchers generated a hyperactive mPB, named hyPBase [82], which featured a 9-fold increase in transposon integration and a 17-fold increase in excision compared to the original mPB. Fusion of an NP signal peptide to mPB could contribute to an increase in transposition efficiency.
In this study, we demonstrated that NP-mPB was restricted to the nucleus and associated with fibrillarin and cyclin T1. Unexpectedly, the transposon insertion sites were relatively random, except for a tendency to be in the gene region, which was similar to a previous report for mPB [44]. There was no obvious preference for the (fibrillarin-associated) NORs. Therefore, the purpose of constructing NP-mPB to target the transposon to the NORs and to prevent essential gene disruption was not successful. However, the resulting NP-mPB exhibited improved transposition efficiency and relatively random transposon integration.
The PBase was revised in several ways to improve its transposition efficiency. The developed NP-mPB exhibits enhanced efficiency without increased toxicity, which is desirable for transgenesis, insertional mutagenesis, and gene therapy. In addition, the NP-mPB does not alter the random insertional preference of the transposon and, therefore, maintains the advantage for establishing a mutagenesis library. In the future, linkage of the NP signal peptide to other modified transposases could possibly be used to increase the efficiency further.