An anti-cancer binary system activated by bacteriophage HK022 Integrase

Cancer gene therapy is a great promising tool for cancer therapeutics due to the specific targeting based on the cancerous gene expression background. Binary systems based on site-specific recombination are one of the most effective potential approaches for cancer gene therapy. In these systems, a cancer specific promoter expresses a site-specific recombinase/integrase that in turn controls the expression of a toxin gene. In the current study, we have developed a new HK022 bacteriophage Integrase (Int) based binary system activating a Diphtheria toxin (DTA) gene expression specifically in cancer cells. We have demonstrated the efficiency, and the high specificity of the system in vitro in cell cultures and in vivo in a lung cancer mouse model. Strikingly, different apoptotic and anti-apoptotic factors demonstrated a remarkable efficacy killing capability of the Int-based binary system compared to the conventional hTERT-DTA mono system in the LLC-Kat lung cancer mice model; we observed that the active hTERT promoter down regulation by the transcription factors Mad-1 is the cornerstone of this phenomenon. The new Int-based binary system offers advantages over already known counterparts and may therefore be developed into a safer and efficient cancer treatment technology.


INTRODUCTION
Cancer encompasses a large group of diseases characterized by the unregulated proliferation/apoptosis and metastasis, which represent one of the major healthcare problems.
Thus, there is a strong unmet medical requirement for the development of novel therapies that provide improved clinical efficiency and longer survival time period in patients suffering with the different cancer types (1). For cancer therapies to be increasingly successful, however, the major obstacle that must be overcome is the safety efficacy of the cancer treatment technologies. Cancer therapy is approached by four main directions: i. Radiation therapy (XRT); ii. Chemotherapy; iii. Immunotherapy; and iv. Gene therapy (2). Despite that XRT is in common use it is associated with significant side effects on normal tissues and organs that limit the dosages and locations used (3). Similarly, use of the conventional cancer drugs in chemotherapy also has seriously side effects on the healthy cells, organs and whole organisms (4). The main reason of these harmful consequences is lack of tumor specificity. Moreover, the primary intrinsic and/or acquired multidrug resistance is one of the main obstacles to successful cancer treatment (5;6). Successful targeting of immune checkpoints to unleash anti-tumor T cell responses results in durable long-lasting response. However, such cancer immunotherapy helps only in a fraction of patients (7). Probably the combination of therapies is the most promising and effective approach in the cancer treatment (8).
Gene therapy is definitely one of the most important developing fields of the cancer treatment in since the early nineties, [one of the first studies was done by Zvi Ram (9)], and the technologies of many researches are currently in the advanced stages of clinical trials. The main advantage for gene therapy approach is the potential use of targeted delivery that mainly classified into two different strategies: passive targeting and active targeting; enhanced permeation and retention (EPR) is the basis of passive cancer targeting and which has been widely applied in numerous drug delivery systems for cancer targeting (10;11). Whereas, in active targeting, the basic concept is to utilize molecular targeting agent to specifically target the biomarkers or receptors on the cancer cells (12;13). Additional advantage of the gene therapy approach is the tumorspecific expression strategies to avoid harmfulness to normal tissue. However, the development of these tools still represents a significant challenge (2).
Toxin therapy approach as a part of cancer gene therapy is based on toxic gene expression targeted exclusively in the cancer cells resulting in killing through apoptosis without affecting healthy cells [as showed by Zvi ram in 1993, (9)]. One of the most used toxins for cancer gene therapy is the Diphtheria toxin (DT) of Corynebacterium diphtheria (14)(15)(16). DTA belongs to the A-B (Type III) group of exotoxins that catalyzes the ADP-ribosylation of elongation factor-2 (EF-2), hence arrests protein synthesis and induces cell death (17). DTA has a very efficient killing activity rate where a single DTA molecule is sufficient to kill the cell (18;19). Former studies demonstrated selective expression of DTA in tumor cells using a cancer specific promoter (5,(17)(18)(19)(20)(21). The main drawback of this approach is the incomplete specificity of the promoter activity in cancer cells, whereas residual expression resulted in cytotoxic effect also in healthy cells (20;21).
The site-specific Integrase (Int) of coliphage HK022 catalyzes phage integration and excision into and out of its E. coli host chromosome. The mechanism of these site-specific recombination reactions is identical to that of the well documented Int of coliphage Lambda (λ) (29). HK022's host recombination site attB is 21 bp long and the phage attP recombination site is over 200 bp long. It is composed of a 21 bp core that is very similar to attB and flanked by two long arms (P and P') that carry binding sites for Int and for some accessory DNA-bending proteins including the bacterial integration host factor (IHF). Integrative attB x attP recombination results in the integration of the phage into the host genome results in a prophage that is flanked by the recombinant attL and attR sites. The integration could be reversed in a process termed excision, which is also catalyzed by Int (30). We have demonstrated that Int of HK022 is active in human cells (31).
Site specific recombinases are also exploited in binary systems for the toxin specific expression in cancer cells that are based on two DNA plasmids. One is a recombinase substrate carries toxin gene silenced by a transcription terminator located between two Recombination sites, the second plasmid carries the recombinase gene expressed by a cancer specific promoter. This promoter provides the expression of the recombinase to the cancer cells where it excises the terminator and thus activates the toxin expression (9)(10)(11). We have shown previously that Int is much less active in eukaryotic cells compare to the Flp and Cre Site specific recombinases that are currently/ widely used for genome manipulations of higher organisms (32), and unpublished). This fact designates that the activation of a silent substrate of the binary system in case of Int demands considerably higher quantity of this enzyme that is reached only in cancer cells at the high level of activity of the cancer specific promoter. It is not the case in healthy cells where possible leakage of the cancer promoter may lead to basic level of Int expression that is obviously insufficient for such silent substrate activation. In case of much more active recombinase even insignificant level of such basic expression in healthy cells may lead to substrate activation and the toxin expression in the healthy cells. The described Int activity finetuning in the binary system insures the high efficacy safety level in the toxin based cancer therapy technology. We recently developed an Int based binary expression system and demonstrated its increased specificity in human cancer cells and in a lung cancer mouse model using luciferase as reporter (33). Here we demonstrate the application of the Int based binary system using a DTA toxin expression to specifically kill tumor cells for cancer therapy.

Cells, growth conditions, mice, plasmids and oligomers
The bacterial host used was E. coli K12 strain TAP114 (lacZ) deltaM15 (34). It was grown and plated on Luria-Bertani (LB) rich medium with the appropriate antibiotics. Plasmid transformations were performed by electroporation (35). Plasmids and oligomers are listed in Tables 1 and 2 All mice procedures were performed in compliance with Tel Aviv University guidelines and protocols approved by the Institutional Animal Care and Use Committee. C57BL/6 strain was used for the mice experiments.
All the experiments were at least repeated tri times.
The plasmid pKS1161 carrying CMV-attR-Stop-attL-DTA cassette (Stop-transcription terminator) was constructed by a ligation of the AgeI-NotI DTA PCR fragment obtained using pIBI30-176 (kindly provided by Dr. Maxwell F.) as template and primers oEY557+oEY558 (Table 2, List of oligomers) into the same sites of pMK189. The plasmid pNG1805 carrying DTA gene under the control of hTERT promoter was constructed by ligation of the AgeI-blunted-NotI DTA fragment from pKS1161 with HindIII-blunted-NotI fragment of pNA1263. The plasmid pAE1808 carrying the DTA gene under the control of CMV promoter was constructed by a ligation of the AgeI-NotI DTA fragment from pKS1161 into the same sites of pEGFPN1. The plasmid pAE1850 used as a site-specific recombination substrate for DTA toxin assays under the control of hTERT promoter (hTERT-attR-Stop-attL-DTA) was constructed by a ligation of the NheI-blunted-AgeI attR-Stop-attL fragment from pMK189, AgeI-NotI DTA fragment from pAE1808 with HindIII-blunted-NotI fragment of pNA1263. Plasmid constructs were verified by DNA sequencing. (Table 1, List of plasmids).

DTA toxin activity assay in cell cultures
For the estimation of Int dependent DTA toxin activity the appropriate cells culture (described above) were cotransfected by the GFP reporter plasmid (pEFGPN-1) (1 g), site-specific recombination substrate plasmid pAE1850 (300 ng) and the Int expressing plasmid pNA1263 (1 g). As a positive control of 100% GFP expression, the cells were transfected by plasmid pEFGPN-1 (1 g). As a positive control of the DTA activity the cells were cotransfected by the plasmid pEFGPN-1 (1 g) and the plasmids carrying DTA gene under the control of CMV (pAE1808) as non-tissue specific expression or hTERT promoter (pAE1805) as cancer specific expression (300 ng). As a negative control of Int dependent DTA toxin activity the cells were cotransfected by the reporter plasmid pEFGPN-1 (1 g) and site-specific recombination substrate plasmid pAE1850 (300 ng). All cotransfection were performed in the total DNA quantity 2.3 g.  Forward and side-scatter profiles were obtained from the same samples.

DNA delivery in mice
DNA delivery in C57BL/6 mice was performed using the linear polyethylenimine based delivery reagent in vivo-jet PEI (Polyplus transfection, France) as previously described (33).

Quantitative Real-Time PCR
Total RNA was isolated from murine lungs with EZ-RNA Isolation kit (Biological Industries, Beit Haemek, Israel). cDNA was synthesized using Verso cDNA kit (Applied Biosystems, Foster City, CA). mRNA expression was assessed by StepOne quantitative Real-Time PCR system (Applied Biosystems, Foster City, CA). qRT-PCR for mouse Katushka, Cas-3, Bax and Bcl-2 mRNA were carried out. Details of the primers are given in Table 2.
In order to normalize the expression of Cas-3, Bax and Bcl-2 genes according to Katushka cells distribution in murine lungs, Absolute Quantification (AQ) method was applied. This method is based on standard curve of known quantity, with comparison of unknown quantities to the standard curve and extrapolation of the expression value. First, total number of mRNA molecules of Katushka, Cas-3, Bax and Bcl-2 were calculated according to its expression using the formula: µg DNA x pmol/660 pg x 10 6 pg/1 µg x 1/N = pmol DNA. Then the number of Cas-3, Bax and Bcl-2 molecules was divided by the number of Katushka mRNA molecules in order to normalize their gene expression to one Katushka molecule.

DNA manipulations
Plasmid DNA from E. coli was prepared using a DNA Spin Plasmid DNA purification Kit (Intron Biotechnology, Korea) or a NucleoBond TM Xtra Maxi Plus EF kit (Macherey-Nagel, Germany).
General genetic engineering experiments were performed as described by Sambrook and Russell (35).

Statistical analysis
Data were presented as the mean ± SD. P-values of less than 0.05 were considered statistically significant by Student's two-tailed t test assuming equal variance.

Integrase based binary DTA expression system for specific cancer cell killing
We previously demonstrated the validity of the Int activated cancer specific binary cell expression system with the luc reporter (33). Here we report the application and the level of specificity of the Int activated binary system using the DTA toxin to kill cancer cell (henceforth the binary system, Fig.1A). This binary system consists of two plasmids: a pNA1263 that expresses the int gene under the control of the cancer specific hTERT promoter that is active practically in all types of tumors and immortal cells, but is silent in somatic tissues (38;39), and a second plasmid pAE1850 that carries the silent open reading frame of the dta gene separated from the hTERT promoter by a transcription terminator (40)(Stop) flanked by tandem attR and attL HK022 recombination sites (Fig.1A). Int-catalyzed attR x attL recombination excises the terminator sequence, thereby allowing expression of the toxin from the promoter. In this binary system the expression of both genes (int and dta) are regulated by the hTERT promoter. As a positive control of the DTA toxicity in cell cultures we used plasmid pAE1808 that expresses dta gene from the constitutive and strong human cytomegalovirus promoter (CMV). To exanimate the efficiency of the binary system versus the conventional mono DTA-based cancer gene therapy approach; we used the plasmid pNG1805 that carried dta gene under the control of hTERT promoter.
The efficiency and specificity of the Int activated binary system in tissue cultures First, we have examined the toxic activity of the less active dta mutant (G128D) based on its ability to abolish the translation of the green fluorescent protein (GFP) as reporter in cell cultures (37). The cell lines: HEK293, LLC-Kat and BJ were cotransfected with three plasmids: the silent dta substrate plasmid that carried the hTERT promoter-attR-Stop-attL-dta cassette (henceforth the sub), the hTERT-Int expressing plasmid and the EGFP-expressing plasmid (pEGFP-N1).
The same cells were transfected with two control plasmids: one to verify the DTA activity using CMV-dta and the second carries the hTERT-dta expressing plasmid, each cotransfected with the EGFP reporter plasmid. To assess the essential silencing of the sub cells were  (Fig. 1B, grey columns). These cells were used in our previous report (33) to develop lung-metastatic tumors in mice (33). The data with this cell line also have demonstrated no toxicity in cells transfected with the silent sub alone (92% surviving cells that shows non-significant difference from the positive 100% control) (Fig. 1B, gray column in Sub). The cells showed strong toxicity (44% surviving cells) with the CMV-dta plasmid and significant toxicity (55% and 73%) with the mono hTERT-dta and the binary systems (Fig. 1B, gray columns). It should be noted that DTA activity in vitro in LLC-Kat cells was lower than in HEK293 cells (Fig. 1B, grey and black columns in CMV-dta).
The third treated cells line (BJ) consisted of normal human foreskin fibroblast cells aimed to examine the cancer specific efficacy of the hTERT promoter (Fig. 1B, white columns). The data in this cell line also have showed no toxicity in cells transfected with the silent sub alone (100% surviving cells) (Fig. 1B, white column in Sub). The cells showed strong toxicity (45% surviving cells) with the CMV-dta plasmid and significant toxicity (82%) with the mono hTERT-dta (Fig.   1B, white columns). On the other hand, the binary system driven by hTERT promoter did not demonstrate any toxicity compared to the cells transfected by the pEGFP alone (Fig. 1B, white column in Sub+ hTERT-int). These results clearly showed that Int based DTA expression system is highly efficient to kill lung cancer cells in vitro while presented highly safety functionality in normal cells as the DTA expression was found to be essentially silence probably due to the transcription Stop sequence.

Int-based binary system for DTA expression selectively kills mice cancer cells
At the next stage, we wanted to examine the cancer specific efficacy in killing as well as in safety of the Int-based binary system in whole organism in vivo. Mice were treated with the binary system and the therapeutic effect was assessed after 24 hours. Comparative quantitative analysis of DTA expression level by IHC in LLC-Kat mice lungs has revealed 33 folds of DTA immune detection of the developed Int-based system over the hTERTdta mono system (Fig. 3, e). We detected a substantial increased DTA expression in healthy mice that were treated with hTERT-dta plasmid compare with same mice that treated with the Int-based binary DTA expression system demonstrating the highly efficacy safety of the binary system (Fig. 2, e).
Since DTA expression induces apoptosis (18;19) in the following experiments the levels of apoptotic effect was analyzed in the lung specimens of healthy and LLC-Kat lung cancer mice using IHC, Western blot and quantitative Real-Time PCR (qRT PCR) methods. and Sub+hTERT-int, respectively). Fig. 4A shows that the Katushka patterns were detected in both cancer lung samples, confirming the cancer development. ERK-Ph proliferation marker abundance was decreased in the lungs of the mice treated with the binary system compare to untreated mice (Fig. 4A). On the other hand, both apoptotic markers Cas-3 and JNK-Ph showed essential increased abundances in the lungs of the mice treated with the binary system (Fig.   4A). The differential levels of the apoptotic and proliferation markers in LLC-Kat cancer mice treated with the Int-based system compared to the untreated mice presented the efficiency of the applied cancer treatment.
Since DTA expression induces apoptosis as described before (18;19), using qRT PCR we wanted to analyze the mini-transcriptome of the following genes: proCas-3, Bax that is functionally characterized as an apoptosis-promoting factor and the Bcl-2 that is characterized as an apoptosis-suppressing factor (52;53) in LLC-Kat cancer lungs treated by the binary system (Fig.4B, a, b, c, respectively). This analysis demonstrated that the level of Bax, Bcl-2 and proCas-3 gene transcription in cancer treated lungs were much higher than in the cancer untreated lungs (Fig.4B). Moreover, the qRT-PCR analysis also demonstrated an increased ratio of Bax/Bcl-2 in the LLC-Kat cancer lungs treated by the binary system compared to the untreated lungs. Bax/Bcl-2 ratio is described as a cellular `rheostat' of apoptosis sensitivity in the sense that the intracellular ratio of bax/bcl-2 protein can profoundly influence the ability of a cell to respond to an apoptotic signal (54;55). According to this concept, a cell with a high bax/bcl-2 ratio will be more sensitive to a given apoptotic stimuli when compared to a similar cell type with a comparatively low bax/bcl-2 ratio (56). (Fig. 4B, d).

DISCUSSION
In this paper, we have developed an anti-cancer binary system based on a site-specific recombination reaction catalyzed by Int that activates the expression of the dta toxin gene specifically in cancerous cells without affecting neighboring normal cell.
Comparative analyses of the Int-based binary and hTERT-dta mono systems in healthy C57BL/6 mice has revealed an unexpected substantial DTA expression in mice treated with hTERT-dta plasmid as a result of non-specific hTERT promoter leakage that led to the activation of apoptosis indicated by the apoptotic markers Cas-3 and P-53 and TUNEL, demonstrating the poor safety of the mono system [ Fig. 2, panels (e)]. In contrast, the Int-based binary system did not cause the similar non-specific effect. These results demonstrate the efficacy and safety of the binary system towards normal vs. cancerous cells [ Fig. 2, panels (e)].
As described above the comparative quantitative analysis of DTA expression level in LLC-Kat mice lungs Int-based binary system revealed elevation of 33 folds more DTA over the mono system (Fig. 3, e). Actually Int-catalyzed attR x attL recombination removes the terminator sequence from the silent substrate that produce the hTERT-attB-dta product plasmid which differs from hTERT-dta plasmid only the presence of the short (50 bp) attB sequence between the promoter and dta gene. We proposed that the difference in DTA expression in these two plasmids may be explained by hTERT promoter differential regulation in the binary and mono systems in LLC-Kat cancer cells.
hTERT promoter regulates the expression of human telomerase reverse transcriptase (hTERT) gene which product is a part of enzyme telomerase together with an RNA subunit and telomerase-associated proteins (57). Whereas the RNA subunit of telomerase is expressed in most cells, hTERT expression is repressed in the normal cells, while in more than 85% of all tumor cells hTERT transcription is upregulated. Thus, hTERT expression resulting in telomerase activity is critical for tumorigenesis (58). Furthermore, inhibition of telomerase activity leads to senescence or apoptosis in tumor cells (59)(60)(61), indicating that telomerase activity is required for the long-term viability of tumor cells. One of the most significant regulatory elements in hTERT To test the proposed suggestion we have performed the analysis of the Mad1 expression level in both healthy and lung cancer mice by IHC using Ab against Mad1 (Fig. 5 A, B). These data unambiguously demonstrates our suggestion that active hTERT intracellular down regulation is caused by activation and elevation of the suppressor regulator Mad1, different transcription factors that regulates hTERT promoter might be involved.
One of the Int distinctiveness that can significantly increase specificity and efficiency of the binary system in human cells is Int activity dependence on the host encoded integration host factor (IHF) and the phage-encoded excisionase (Xis) in E. coli (30).
We have shown that in contrast to the requirement of Int for IHF and Xis in its natural milieu, the wild type Int was active in human cells without the need to supply any accessory proteins (67).
However the addition of these proteins improved the Int activity in human cells (68). Similar results have been also received for Int Lambda (Int-λ) in human cells. The expression of single chain IHF strengthened the wild type Int-λ activity to the levels of the IHF-independent mutants (69;70). In vitro experiments with Int-λ have also showed that human chromatin-associated proteins HMG1 and HMG2 can substitute to some extent for the requirement of the accessory proteins (71).
The results stated above allow assuming that definite histone, histone-like or chromatinassociated proteins can render the strengthening effect on Int activity in human cells.
Intriguingly the expression of some human histones and histone-like proteins (H2A, H2B, H3, H4, CENP-A) are upregulated in numerous cancers (72)(73)(74). All these data allow suggesting that Int activity may be strengthened in the human cancer cells compare to healthy cells by the putative histone or histone-like Int accessory proteins upregulating. This phenomenon of Int cancer specific activation via such accessory proteins can significantly increase specificity and efficiency of the Int-based binary system that provides its additional advantage compare to conventional mono systems for human cancer therapy. For further cancer gene therapy studies we suggest to consider the regulation elements as a crucial and significant factor for the activity and efficiency of the developed approaches.
Our binary anti-cancer system belongs to the toxin gene delivery approach but has the obvious advantages compared to conventional toxin delivery approaches: it shows a higher safety efficacy in healthy cells and a higher killing efficacy in cancer cells. Moreover, unlike other systems using DNA vectors for mammalian gene manipulations ours is viral-free. This binary system may have a strong impact on human cancer curing. The preclinical trials including the examining of the binary system in survival experiments are in progress.