|Home | About | Journals | Submit | Contact Us | Français|
A promoter that augments gene expression in response to stimulation of ionizing radiation would be a desired tool for radiogenetic therapy, a combination of radiotherapy and gene therapy. Although various promoters occurring naturally or artificially have been used for researches, one showing higher reactivity to ionizing radiation is desirable. In the present study, we attempted to improve a radiation-responsive promoter of the p21 through a technique called DNA shuffling. A library of DNA fragments was constructed by re-ligation of randomly digested promoter fragments and improved promoters were chosen out of the library. We repeated this process twice to obtain a promoter showing 2.6 fold better reactivity to ionizing radiation compared with its parent, p21 promoter after 10 Gy γ-ray irradiation. Nucleotide sequence analyses revealed that the obtained promoter was densely packed with some of the cis-acting elements including binding sites for p53, NF-κB, NRF-2, AP-1 and NF-Y more than p21 promoter. In addition, it was shown that its induction by ionizing radiation was dependent upon p53 status of a cell line, suggesting that the promoter retained properties of the p21 promoter. This technique is simple and efficient to improve a promoter responsive to other stimulus of interest besides IR.
Radiotherapy is one of the main strategies used to treat solid tumors. In recent years, techniques such as conformal radiotherapy and intensity-modulated radiation therapy have increased the accuracy of dose delivery to the tumor, keeping the surrounding normal tissue less exposed. However, regarding certain tumor types, curative radiotherapy still remains challenging (1). Hence, various strategies to improve local tumor control are currently under investigation. Among them, a radiogenetic therapy, the combination of radiotherapy with gene therapy, seems to be a promising strategy because it exploits the fact that ionizing radiation (IR) can spatially and temporally control the expression of antitumor transgene in a radiation field by use of radiation-responsive promoters as well as cell killing effect.
The cellular response to IR induces the transcriptional activations of a number of genes involved in immediate early responses to mitogenic stimulation or cellular stress. These transcriptional activations are achieved through transcription factor binding to the cis-acting elements responsive to IR. CArG elements and nuclear factor-kappa B (NF-κB) binding sites have been identified as radiation-responsive elements within the 5′-flanking regions of Egr1 and c-IAP2, respectively (2, 3). However, although various genes and proteins induced by IR have been identified, relatively few radiation-responsive elements within the gene regulatory region of those genes have been fully characterized with the exception of a few elements (4).
In the past, artificial promoters responsive to IR for use in suicide gene therapy vectors have been developed employing promoters of radiation-responsive genes. Construction of the plasmid with p21 promoter driving the expression of the inducible nitric oxide synthase (5), which leads to the generation of the vasoactive, cytotoxic and radiosensitising product, nitric oxide (NO), resulted in more effective tumor growth delay than a constitutive CMV promoter in vivo (6). Moreover, Scott et al. have investigated the parameters needed to enhance promoter activation by IR with CArG elements and demonstrated that increasing the number of it, up to certain level, increased the induction activity (7). Furthermore, Ueda et al. have constructed the plasmids with four tandem repeats of the NF-κB binding sites to drive the expression of the apoptotic sucide gene BAX and presented the evidence that the human tumor cells transfected with the plasmids significantly reduced the number of cells with IR treatments (3). Initially, p53-dependent radiation-responsive genes (e.g., p21 and Gadd-45) seemed to be not ideal to utilize as a promoter for cancer gene therapy, because p53 mutations are observed in many tumors and their promoters may not function efficiently in this context (1). On the contrary, it has been reported that the expression of p21 was independent of p53 status in various tumor types and its transcriptional activation has been observed in tumor cells with IR treatments (8, 9), suggesting that p21 promoter can be utilized for cancer gene therapy with IR treatments.
In the development of a radiogenetic therapy, driving high-level transgene expression in a tumour-specific manner remains a key requirement to decrease damage to normal tissue and increase the expression of trans-suicide gene to tumor, therefore the construction of an artificial promoter responsive to IR with the big induction would be desired. We successfully constructed radiation-responsive promoters using the combination techniques of random cis-acting element elongation and error prone PCR (10). However, relatively a few kinds of cis-acting elements utilized in this study as radiation-response elements might influence some limitation for the induction activity by IR. On the other hand, DNA shuffling enables to construct the radiation-responsive promoter by random rearrangement of a great number of cis-acting elements encoded within gene regulatory regions of radiation-responsive genes without identification of radio-responsive cis-elements. In addition, this technique could take in the sequence that contributes to 3D conformation of the promoter as well as radiation-responsive cis-acting elements when compared to random cis-acting element elongation method.
The goal of this study was to construct the radiation-responsive promoter through a DNA shuffling and to examine the induction property of the constructed promoter to IR. This technique started with a construction of a promoter library by digesting a promoter fragment with a known property and reconstructing DNA fragments by random ligation of the digested fragments. Cis-acting elements in the promoter would be shuffled and recombined to change their numbers and positional relationships in resultant fragments while some of them are kept intact. Better promoters with the induction activity to IR could be generated in the library as a result. p21 has been reported as one of the genes responsive to IR (11). Therefore, we hypothesized that the improvement of 5′-flanking region that regulates the expression of p21 would result in the construction of promoter with bigger induction activity to IR and employed it as an initial DNA fragment of DNA shuffling. We have successfully constructed radiation-responsive promoters through a DNA shuffling after two cycles of the process compared to their parent, p21 promoter. In addition, the artificial promoter with highest induction activity was appeared to drive the gene expression through p53 responsive element (p53RE). This DNA shuffling is considered to be simple and efficient technique to improve a property of promoter responsive to other stimulus of interest besides IR.
MCF7 (human breast carcinoma), DU145, PC-3 (human prostate carcinoma), HT-29 (human colon carcinoma), and HeLa (human cervical carcinoma), were grown in RPMI1640 medium supplemented with 10%(v/v) heat inactivated fetal calf serum and maintained at 37°C in humidified air with 5% CO2.
Cells were irradiated using an Eldora 8 60Co teletherapy unit (Theratronics International Ltd., Ontario, Canada) at dose rates between 200 and 250 cGy/min. Decay corrections were done monthly, and full electron equilibrium was ensured for all irradiations.
A promoter probe plasmid, pGL4.11-FseI was designed to possess a unique FseI site into the multiple cloning site, derived from pGL4.11 (Promega Corp., Madison, WI), a promoter-less plasmid. pGL4.11 possesses its FseI site downstream of the luciferase gene. Therefore, the FseI site was destroyed by T4 DNA polymerase (TAKARA BIO INC., Ohtsu, Japan) treatment and self-ligation after FseI digestion. Using this plasmid without FseI site as a template, PCR was performed with a pair of primers 5′-ATGGCCGGCCAAGCTTGGCAATCCGGTAC-3′ and 5′-ATGGCCGGCCGAGGCCAGATCTTGATATCCTC-3′ containing an FseI recognition sequence. After digestion of PCR products with FseI, self-ligation was carried out to construct pGL4.11-FseI.
To amplify a 5′-flanking region from the transcription start site to 4579 bp of p21, PCR was carried out in 50 μl reaction volume containing 2.5 units LA Taq (TAKARA BIO INC., Ohtsu, Japan), 5 μg extracted genomic DNA from MCF7 cells as its template, and a pair of primers 5′-GCTCTCTTCCGTGGAGGTGGATCCCTGTAG-3′ and 5′-CTGAGTGCCTCGGTGCCTCGGCGAATCCGC-3′. Its cycling condition was an initial denaturation at 94°C for 1 min followed by 30 cycles of 98°C for 10 sec, 69°C for 5 min, and followed by a final extension cycle for 7 min at 72°C. After checking the PCR amplification with electrophoresis, nested PCR was performed with a pair of primers 5′-ATGGTACCAGATGCTCAGGCTGCTGAGGAGGGCGCG-3′ and 5′-ATGCTAGCTCCGGCTCCACAAGGAACTGACTTCGGC-3′ with the same cycling condition described above. Purified PCR fragments were digested with KpnI and NheI and then inserted into the plasmid pGL4.11-FseI to construct a plasmid pp21-Up. As for the cloning of a 5′-flanking region from the translation start codon to the transcription start site of p21 with approximately 5.5 kbp in length, PCR was also carried out with a pair of primers 5′-ATGCTAGCTGGGCGCGGATTCGCCGAGGCACCGAGG-3′ and 5′-ATGCTAGCGGCGCCTGCAGCAGAGATACAAGGAAGGCC-3′. The PCR products extracted from agarose gel after the electrophoresis were cloned into a plasmid pCR-XL-TOPO (Invitrogen Co., CA, USA) following the manufacture’s instructions. The DNA fragments carved out with NheI were cloned into a plasmid the pp21-Up to construct a plasmid pp21. Prior to use, the orientations, integrity and sequences of these constructed plasmids were confirmed by nucleotide sequencing analysis.
To amplify a 5′-flanking region from the translation start site to 4579 bp of p21, approximately 10 kbp in length, PCR was performed using the plasmid pp21 as a template with a pair of primers 5′-ATGGTACCAGATGCTCAGGCTGCTGAGGAGGGCGCG-3′ and 5′-ATGCTAGCGGCGCCTGCAGCAGAGATACAAGGAAGGCC-3′. Its cycling condition was an initial denaturation at 94°C for 1 min followed by 30 cycles of 98°C for 10 sec, 69°C for 10 min, and followed by a final extension cycle for 7 min at 72°C. After confirming the PCR amplification on electrophoresis, DSN digestion was performed at 65°C for 10 min in 50 μl reaction volume containing 1 μg purified PCR product, 0.001 Kunitz-units of DSN (Evrogen Joint Stock Company, Moscow, Russia) and 10 × DSN master buffer (provided in the kit) and was subsequently inactivated by the addition of 2 × DSN stop buffer (provided in the kit). Randomly generated 5′- or 3′-protrusions by DSN digestion were converted into blunt ends with DNA Terminator End Repair Kit (Lucigen Corp., Middleton, WI) according to the manufacturer’s instruction and then randomly assembled with T4 DNA ligase. After that, a double strand synthetic DNA fragment containing the FseI recognition suquence was added into this reaction mixture as a linker. Assembled DNA fragments digested with FseI were inserted into the promoter probe plasmid pGL4.11-FseI. These plasmids were transformed into TOP10 competent bacteria (Invitrogen Life Technologies, CA, USA) to construct a promoter library. Three clones showing the highest fold inductions by IR among variants randomly picked up from the promoter library were used as templates to generate a promoter library of the next generation.
Nucleotide sequence analysis were performed using Big Dye Terminator Cycle Sequencing Ready Reaction kit (Applied Biosystems, CA, USA) with a pair of primers, 5′-TACGCTCTCCATCAAAACAAAACG-3′ and 5′-GTAGAATGGCGCTGGGCCCTTCTT-3′. To search potential binding sites for the transcription factor in the promoter constructed through DNA shuffling, TRANSFAC (http://www.gene-regulation.com/index.html) was used.
The day before transfection, 2.0 × 104 cells were seeded in a 24-well culture plate. To determine the fold induction of constructed promoter activity by IR, the prepared cells were transiently transfected with 100 ng of a constructed plasmid and 1 ng of an internal control vector pGL4.70 (Promega, Madison, WI, USA) using Effecten transfection reagent (QIAGEN, Valencia, CA, USA). The transfected cells were incubated for 24 h and then exposed to radiation. Luciferase activities were measured using dual luciferase assay kit (Promega, Madison, WI, USA). Relative luciferase activities were calculated after normalization with reporter gene expressions of the internal control.
All values are expressed as means ± standard deviations. Significant differences between groups were determined by one-way analysis of variance (ANOVA). If the ANOVA was significant, the Bonferroni/Dunn procedure was used as a post hoc test. p < 0.05 was considered to be significant.
We used a technique of DNA shuffling. In the process, a promoter library was constructed and, out of the library, promoters with higher fold inductions to IR than their parent wild type p21 promoter were chosen. First, a DNA fragment covering a 5′-flanking region of the p21 from −4579 bp (4579 bp upstream of the transcription start site) to +5396 bp (translation start site) amplified by PCR was digested into fragments of various lengths with DSN and their fragments were randomly assembled by ligation reaction. Resultant DNA fragments were inserted into upstream of the luciferase gene of a promoter probe plasmid, pGL4.11-FseI, to construct the 1st generation promoter library (Fig. 1). We picked up 31 clones from the 1st generation promoter library and transfected each of them into MCF7 cells. The transfected cells were exposed to γ-ray at 10 Gy and harvested at 6 h after IR because the kinetics of induction activity for the parental p21 promoter reached to the maximum fold induction of 1.3 ± 0.2 (n = 22) at 6 h after 10 Gy IR (Fig. 2B). It was found that 6 clones showed higher fold inductions after IR than the parental p21 promoter. We chose three promoters, clone 1-1, 1-2 and 1-3, out of the 1st generation promoter library as representatives, showing the highest fold inductions among the 31 clones. As shown in Fig. 2B, the fold inductions of each three clones were 1.6 ± 0.3 (n = 9), 1.7 ± 0.1 (n = 8) and 1.8 ± 0.3 (n = 3), respectively. Employing the 3 representative clones of the 1st generation promoter library as the initial material DNA fragments, we constructed the 2nd generation promoter library with a similar process as described in Fig. 1. Out of 25 clones picked up from the 2nd generation promoter library, 4 clones revealed greater fold inductions than the 3 representative clones of the 1st generation promoter library. We chose 3 promoters, clone 2-1, 2-2 and 2-3, of the 4 clones as representatives, showing the highest fold inductions among the four. As shown in Fig. 2B, the fold inductions of each three clones were 2.2 ± 0.2 (n = 17, p < 0.01), 2.2 ± 0.3 (n = 15, p < 0.01) and 2.6 ± 0.6 (n = 13, p < 0.01), respectively. We then constructed the 3rd generation promoter library to obtain a promoter with a higher fold induction than the best promoter found in the 2nd generation promoter library. However, 23 clones taken randomly from the 3rd generation promoter library did not show higher fold induction than clone 2-3 (Data not shown). As for basal promoter activities without IR, the activities of three best promoter clones constructed in each generation promoter libraries were low enough when compared with that of their parental p21 promoter (Fig. 2A).
The time course of radiation-responsive promoter clone 2-3 were evaluated with Dual Luciferase assay. When exposed to a dose of 10 Gy, the fold induction of luciferase gene driven by clone 2-3 at 1, 3, 6, 12, 24 h after IR was 1.1 ± 0.1 (n = 3), 1.7 ± 0.4 (n = 3), 2.6 ± 0.6 (n = 13, p < 0.01), 1.9 ± 0.4 (n = 3), and 1.2 ± 0.1 (n = 3), respectively, showing that the induction activities through IR was observed at 3 h and then peaked at 6 h after IR, and gradually declined thereafter when irradiated at 10 Gy (Fig. 3).
IR is known to activate p53 through pathway including DNA-PK or ATM and then induce p21 expression (12). Therefore, we assumed that p53 could be involved in reactivity of clone 2-3 by IR. To confirm this hypothesis, clone 2-3 was introduced into various cell lines with different p53 status and then evaluated the fold induction of transfected cells after stimulation with IR. As shown in Fig. 4, in MCF7 cells with wild type p53 and in DU145 cells with heterozygous mutant of p53, clone 2-3 led fold inductions of luciferase expression to 2.6 ± 0.6 (n = 13, p < 0.01) and 2.4 ± 0.1 (n = 3, p < 0.01), respectively, in response to 10 Gy γ-ray. Meanwhile, the fold inductions of HT-29 cells with homozygous mutant of p53, and PC3 cells of a p53 null mutant, were 1.0 ± 0.0 (n = 3, p < 0.01) and 0.8 ± 0.1 (n = 3, p < 0.01), respectively, confirming that the induction of luciferase expression by IR was not observed in these contexts. In HeLa cells, the transcriptional activity of p53 is strongly repressed by overexpression of E6 protein from oncogenic human papillomavirus type 16. In agreement with this p53 degradation pathway, the fold induction by IR in HeLa was not observed as well as HT-29 and PC-3, showing the fold induction was 1.0 ± 0.2 (n = 3, p < 0.01). Moreover, to confirm the contribution of p53 to induction activity of clone 2-3 by IR, we evaluated the changes of fold inductions by overexpression of p53 with a plasmid expressing p53 under control of the CMV promoter, TP53 (ORIGENE, Rockville, MD, USA). Overexpression of p53 in MCF7 cells resulted in higher luciferase activity (2.6 ± 0.6 (n = 3, p < 0.01)) compared to that without p53 overexpression, suggesting that the induction activity of clone 2-3 is affected by p53.
To further confirm whether p53RE recognized by p53 were encoded within clone 2-3, the nucleotide sequence analysis was performed and then potential binding sites for transcription factor was searched for with TRANSFAC. Clone 2-3 was found to be 417 bp in length and composed of four p53REs, four NF-kB binding sites, three NF-E2 related factor 2 (NRF-2) binding sites, two activating protein-1 (AP-1) binding sites and CCAAT boxes (Fig. 5, defining cut-offs as matrix similarity = 0.7, core similarity = 1.0).
In the present study, radiation-responsive promoters were constructed through the DNA shuffling technique that 5′-flanking region of p21, approximately 10 kbp in length, was digested and assembled randomly. An increase of induction activity by IR was observed every time the DNA shuffling was carried out repeatedly. These results show that promoter clones showing a higher fold induction than their parental clones could be found in a promoter library generated from digested fragments of the parental p21 promoter, indicating DNA shuffling is a simple and efficient technique to improve not only a function of structural gene, but also a property of a promoter. However, 23 clones taken randomly from the 3rd generation promoter library did not show higher fold induction than clone 2-3. It might be difficult to simply obtain a promoter with a stronger induction with combinations of cis-acting elements contained in the 5′-flanking region of p21. The other reason for this result might be due to the lack of the screening system to pick out the candidates with the strong radiation-responsive property from the constructed library including incomparable numbers of clones. Wright et al. have broken through this hurdle by using flow cytometry as the screening technique for DNA shuffling and succeeded in the construction of the stronger promoter than cytomegalovirus (CMV) promoter, which provides high expression level in multiple cell types (13). We could break through this hurdle employing this screening technique or through DNA shuffling with other DNA fragments derived from different promoters with radiation-responsive properties to construct a new library.
The nucleotide sequence analysis of clone 2-3, which showed the highest induction activity with IR treatment, revealed that it composed of four the p53REs, four NF-kB binding sites, three NRF-2 binding sites, two AP-1 binding sites and CCAAT boxes as potential binding sites for transcription factor. Besides p53, transcription factor nuclear factor-Y (NF-Y) which recognizes the CCAAT box, AP-1 and NF-kB were reported to be activated by IR (14, 15). In addition, Nrf2 activation is often linked to oxidative stress (16). These cis-acting elements were originally derived from a 5′-flanking region of p21 with approximately 10 kbp in length. For example, the number of potential p53REs, NF-kB binding sites, CCAAT boxes, AP-1 binding sites and Nrf-2 binding sites were 2, 45, 12, 15 and 21, respectively, in a 9950 bp DNA fragment of presumably the entire promoter region of p21, detected using TRANSFAC (defining cut-offs as matrix similarity = 0.7, core similarity = 1.0). On the other hand, clone 2-3 was found to contain 4, 4, 2, 2 and 3 elements, respectively, showing that all cis-acting element probably involved in the induction were accumulated in the 417 bp DNA fragment. These transcription factors are considered to be independently involved in the gene expression of clone 2-3, not through the functional p53. In addition, alignments or positional relationships may also be accounting for improvement of reactivity of clone 2-3 to IR as Stefan et al. reported that the spacing between cis-acting elements is influential for promoter reactivity to stimulation (17). Taken together, by optimization of the number and alignments of cis-acting elements responsive to IR through a DNA shuffling, clone 2-3 could possess higher radiation responsive reactivity than the parental p21 promoter.
In the present study, we have successfully constructed radiation-responsive promoters by improving the p21 promoter through DNA shuffling. Moreover, the constructed promoter clone 2-3 with highest induction activity was appeared to drive the gene expression mainly through p53RE. For clinical use, especially for radiogenetic therapy, p53-dependent promoter might be available for the protection of normal tissue from IR. For instance, the vector, which induces a free radical scavenging enzyme gene under the control of the p53-dependent promoter by IR, might enhance the therapeutic effects with few side effects to normal tissue. Finally, this DNA shuffling is considered to be simple and efficient technique to improve a property of promoter responsive to other stimulus of interest besides IR.
This research was supported by the Intramural Research Program of the National Institutes of Health, National Cancer Institute, and also in part by the Grant-in-Aid for Scientific Research on Priority Areas (C) 21591618 from the Ministry of Education, Culture, Sports, Sciences and Technology of Japan.