Ninhydrin test

The elution products were put onto a thin layer chromatography (TLC) plate.

The 15% ninhydrin solution (solved in methanol) was then put onto the TLC plate.

After 10 minutes incubation, the TLC plate was pictured.

2.5.3 Dot-blotting

The elution products were applied onto the nitrocellulose (NC) paper (Pall, USA) which was rinsed with 1X PBS buffer on a dot-blot machine (Bio-East, Taiwan). Samples were gently for 10 minutes. The NC paper was blocked by 5% of skin milk at 4°C overnight or 37°C for 3 hours with shaking. The paper was then washed with 1X PBS containing 0.05% Tween 20 three times at room temperature for 5 minutes. The anti-HA antibody (Roche, Basel, Switzerland) were diluted 200X in 5% blocking buffer (5% skin milk in 1X PBS buffer) and applied onto the NC paper gently at room temperature for 1 hour with shaking. The mixture was then discarded. The NC paper was washed with 0.05% PBST three times at room

temperature for 5 minutes. The 2^nd antibody conjugated with HRP (DakoCytomation, Denmark) was diluted 100X in 5% blocking buffer and applied on the NC paper in dark for 1 hour with shaking. After 1 hour incubation, the NC paper was washed as above and the substrate was applied onto the NC paper for 10 minutes in dark. The NC paper was covered in the lead blocker (Okamoto, Japan) with the film for several minutes depended on the intensity of the signals. Then, the film was developed in the developer for 1 minute. The film was washed in water and then stained in the fixer for 1 minute.

2.6 PEI-peptide complex transfection

The enhancement of transgenic expression in vitro by RGD-4C-HA was performed by PEI-peptide complex transfection. The PEI and RGD-4C-HA were incubated at room temperature for 5 minutes at different molar ratios. After 5 minutes incubation, the PEI-peptide complex was used as the native PEI. The transfection was done as described above.

2.7 Statistical analysis

Results were expressed as mean ± SE. Statistical significance of differences between mean values was estimated using the Student’s t-test (Microsoft Excel). p

< 0.05 was considered significant.

Chapter 3 Results

3.1 Establishment of the transcription factor-based mini-promoter (TSP) system

3.1.1 Screening of the activities of several transcription factors in different cells

Previous literatures have indicated that the activities of transcription factors are different in different cell types. In order to character the activities of the transcription factors in tumor cells, the plasmids pAP-1-hrGFP, pCRE-hrGFP, pCRII-hrGFP, pNF-κB-hrGFP, pNFAT-hrGFP, MZF-1-hrGFP, and a control plasmid (pARE-hrGFP) were respectively co-transfected with a reporter plasmid (pAsRed2-N1) into different cells. ARE was a binding site of a prokaryotic transcription factor ampR and it was used as a negative control group. In addition, the reporter plasmid (pAsRed2-N1) with a transgene encoding the red fluorescent protein driven by CMV promoter was used to normalize the transfectant efficiency between each sample. B16-F10 cells (mouse melanoma cells with high metastatic potential), Balb/3T3 cells (mouse immortalized fibroblast cell with high proliferating activity) and HeLa cells (human cervical carcinoma cells) were assayed their certain activities of transcription factors as described above. The

expression index of each transcription factor was calculated according to the following formula:

TFI of TFBS-hrGFP / TFI of AsRed2 in TFBS Expression Index =

TFI of ARE-hrGFP/TFI of AsRed2 in ARE

TFI = Total fluorescence intensity

The results showed that the expression indexes of HIF-1 and NF-κB were higher than other transcription factors in all cells. NF-κB activities were 6-fold, 12-fold, and 4-fold higher than ampR (ARE) in B16-F10, Balb/3T3, and HeLa cells, respectively. The expression indexes of HIF-1 were 15-fold, 43-fold, and 9-fold higher than ampR (ARE) in B16-F10, Balb/3T3, and HeLa cells, respectively.

Except of HIF-1 and NF-κB, the expression indexes of CREB were higher than ampR in all cell types, but the data were not significant after statistical calculation (Figure 2).

3.1.2 Construction of the transcription factor-based synthetic promoter (TSP)

The activities of NF-κB and HIF-1 were found that they were relatively higher in these tumor or rapid-proliferating cells. In addition to CREB, it was considered as the important role in cell transformation such as myeloid leukemia and

contributing to tumor metastasis and invasion. Based on above, the binding sites of NF-κB, CREB, and HIF-1 were assembled to create a novel synthetic mini-promoter. The CMV promoter of pAAV-MCS-hrGFP was replaced by the transcription factor-based synthetic promoter (TSP) and the new plasmid (pD5-hrGFP) was verified by restriction enzyme digestion (Figure 3).

3.1.3 Transcription factor-based mini-promoter activity in different cells

The pD5-hrGFP contained a novel mini-promoter TSP with 9 transcription factor binding sites of 3 NF-κB response elements, 3 CREB elements and 3 HIF-1 elements. In order to verify the efficiency of this promoter, the plasmids pD5-hrGFP, pCRE-hrGFP, pCRII-hrGFP, pNF-κB-hrGFP, and a control plasmid (pARE-hrGFP) were co-transfected with a reporter plasmid (pAsRed2-N1) into different cells. ARE was a binding site of a prokaryotic transcription factor ampR and it was used as a negative control group. In addition, the reporter plasmid (pAsRed2-N1) with a transgene encoding the red fluorescent protein driven by CMV promoter was used to normalize the transfectant efficiency between each sample. The expression index of each transcription factors was calculated according to the following formula:

TFI of TFBS-hrGFP / TFI of AsRed2 in TFBS Expression Index =

TFI of ARE-hrGFP/TFI of AsRed2 in ARE

TFI = Total fluorescence intensity

The results showed that the expression indexes of TSP were 6-fold, 36-fold, and 4-fold higher than ampR (ARE) in B16-F10, Balb/3T3, and HeLa cells respectively.

The results indicated that TSP was truly active in these tumor or rapid-proliferating cells. In addition, the expression indexes of HIF-1 were 15-fold, 54-fold, and 15-fold higher than ampR (ARE) in B16-F10, Balb/3T3, and HeLa cells respectively. It was noticed that the HIF-1 remained the highest activity in all three cells which meant the promoter might be also utilized in cancer gene therapy (Figure 4).

3.1.4 Inhibition effect of TSP in HeLa cells

We had proven that TSP was active in several types of cell including tumor cells in previous study. The activity of TSP was apparently related to the activities of the transcription factors: NF-κB, CREB, and HIF-1. Generally, the activity of some transcription factor might be reduced under several physiological conditions.

For example, NF-κB inhibitors were occasionally utilized in the treatment of cancer.

In addition, tumor progression was largely depended on angiogenesis. When the

angiogenesis in tumor was completed, the activity of HIF-1 would reduce because of the sufficient oxygen supply. Since NF-κB and HIF-1 responsive element were partial components of TSP, the activity of TSP might be lost in these circumstances.

In order to verify whether the activity of TSP was affected under such circumstance, several compounds were used to inhibit the activities of certain transcription factors.

The experiment was performed as described above. The samples were treated with the corresponding inhibitors after 24 hours transfection. After drug treatment for 16 hours, the gene expression was determined by FACScan flow cytometer (Becton Dickinson, Moutain View, CA). The results indicated that the activity of NF-κB was inhibited by 25μM hydroquinone and the activity of the promoter was reduced to 83.7%. Similarly, the activity of HIF-1 was inhibited by D609 (50μg/ml) and the activity of the promoter was reduced to 69.07%. However, the activities of TSP were not significantly lowered than control group at the presence of inhibitors (Figure 5).

3.2 Design of RGD-4C-HA and the functional regions

Many peptides had been identified by phage display to react strongly to certain receptors or molecules. The RGD-4C (CDCRGDCFC) peptide was discovered to specifically bind to integrin α_vβ₃ expressed on the surface of B16-F10 cells. The binding activity of RGD-4C peptide was utilized to improve the DNA delivery

efficiency of PEI to certain cells. For the purpose, this peptide must contain several additional functional domains. The peptide RGD-4C-HA (CDCRGDCFCGGGYPYDVPDYAGGGDDDEC which was purchased from MDBio, Taiwan, ROC) was designed as a multi-functional peptide. The RGD-4C (CDCRGDCFC, underlined) sequence is the targeting region to direct the molecule binding to integrin α_vβ₃ on B16-F10 cell surface. The HA tag (YPYDVPDYA, bolded) was designed to act as a spacer to separate from the absoption domain and as an epitope for antibody detection. The absorption region (DDDE, dotted) was designed for absorption with PEI. The four continuous amino acids (DDDE) sequence contained negatively charged residues would absorb to the positively charged PEI by electrostatic forces. The final amino acid cysteine with the sulfyl group can be used to couple with the primary amine group of PEI. The two GGG sequences were spacers to separate the functional domains (Figure 6).

3.3 The binding affinity of RGD-4C-HA

The multi-functional peptide RGD-4C-HA was determined whether the other functional regions interfered with the activity of RGD-4C to abolish the targeting activity. The RGD-4C-HAs with different concentrations were mixed with target cells to determine the binding activity. The results revealed that the total fluorescence intensities increased significantly for B16-F10 cells compared to

negative control under all concentrations. However, the total fluorescence intensities were no differences compared to negative control in Balb/3T3 or HeLa cells (Figure 7a). Generally, the total fluorescence intensity was obtained by the events time the fluorescence mean. The events represented the binding percentage of RGD-4C-HA to the population and the fluorescence mean represented the binding strength on a single cell. When we focused on the two parameters respectively, it was found that the binding percentages of RGD-4C-HA increased significantly in Balb/3T3 cells under 2μM and 200nM. The binding percentages increased from 45.30% (NC) to 47.61% and 49.28% under 2μM and 200nM RGD-4C-HA in Balb/3T3 cells. It was noticed that the binding percentages increased from 44.53% (NC) to 53.84%, 59.13%, and 52.53% under 2μM, 200nM, and 20nM RGD-4C-HA in B16-F10 cells at the same time. The binding percentages were no differences compared to negative control (Figure 7b). In other way, the total fluorescence mean didn’t increase in Balb/3T3 and HeLa cells as in B16-F10 cells (Figure 7c). These results indicated that RGD-4C-HA bound to B16-F10 and Balb/3T3 cells significantly since the binding percentages increased, but it was shown that the binding percentages were apparently larger in B16-F10 than in Balb/3T3 cells. However, the differences of the total fluorescence mean showed that there were more ligands on a single cell in B16-F10 rather than in Balb/3T3 cells.

3.4 The absorption of RGD-4C-HA to PEI

RGD-4C-HA peptide could bind to the surface of B16-F10 cells that represented the targeting region still remained the activity as in its native RGD-4C form. Moreover, the absorption region was determined whether it had the activity to absorb to PEI. The PEI and RGD-4C-HA were incubated at room temperature for 30 minutes at a molar ratio 6:1. After incubation, the PEI-peptide mixture was separated by a gel filtration column Sephacryl S-200 (GE Healthcare, Chalfont St., UK) and the elution products were collected to test whether the PEI could form complex with RGD-4C-HA. The ninhydrin test was used to identify the existence of PEI (Figure 8a) and the dot immunoblotting was used to determine the existence of RGD-4C-HA by anti-HA antibody (Figure 8b). Ninhydrin reacted to primary or secondary amine group to give a colored product (usually yellow to brown).

Although the RGD-4C-HA also has the amine group, the ninhydrin test in this experiment would not be false positive because the concentration of the amine groups in RGD-4C-HA was far below the sensitivity range of ninhydrin. The elution product was double positive in ninhydrin test and dot immunoblotting that revealed the negatively charged absorption region might act with the positively charged PEI to form complex by electrostatic forces.

3.5 The enhancement of transgenic expression by PEI-peptide complex

The binding affinity to B16-F10 and the absorption ability to PEI were proved in the previous study. In this section, the PEI-peptide complex was used as a modified transfection reagent to determine whether it could increase the level of transgenic expression or not. The plasmid pAAV-MCS-hrGFP with a transgene encoding the green fluorescent protein driven by CMV promoter was used as reporter gene. The results showed that the levels of expression were 2.8-fold, 2.3-fold, and 4.8-fold higher than PEI alone in B16-F10, Balb/3T3, and HeLa cells respectively at 10μM RGD-4C-HA (Figure 9). When the concentrations of RGD-4C-HA were decreased, the levels of expression were also reduced for B16-F10 cells. However, the levels of expression were reduced to 63% and 71%

than PEI alone in HeLa cells at 1μM and 100nM RGD-4C-HA respectively. The results indicated that RGD-4C-HA combined with PEI increased the expression of transgene in B16-F10 cells and the effects on variant cells were different.

3.6 The enhancement of transgenic expression by PEI-peptide complex (HIF-1)

The PEI-peptide complex was proved to increase the transgene expression in

B16-F10 cells under all concentrations of RGD-4C-HA. However, the transgene expression also increased in Balb/3T3 and HeLa cells under 10μM RGD-4C-HA.

The previous study indicated that the increase was only partial specific for B16-F10 cells under some conditions. In previous study, the HIF-1 had been shown to have higher activity in tumor or rapid-proliferating cells (Figure 2 and figure 4). It might be a potential mini-promoter for cancer gene therapy. Therefore, the pCRII-hrGFP which contained 7 copies of HIF-1 responding site was used to determine whether it had the activity of specific expression for B16-F10 cells. The results indicated that the levels of expression were 2.7-fold, 4.6-fold, and 4.4-fold higher than PEI alone for B16-F10, Balb/3T3, and HeLa cells respectively under 10μM RGD-4C-HA (Figure 10). When the concentration of RGD-4C-HA reduced, the levels of expression were 1.3-fold and 1.5-fold than PEI alone in B16-F10 under 100nM and 10nM RGD-4C-HA respectively. However, the levels of expression were no differences than PEI alone in Balb/3T3 and HeLa cells under other RGD-4C-HA concentrations. The results indicated that RGD-4C-HA combined with PEI increased the transgene expression under HIF-1 mini-promoter in B16-F10. However, the transgene expression still increased for all cell types without specificity under 10μM RGD-4C-HA.

3.7 Enhancement of transgenic expression by PEI-peptide complex combined with TSP in B16-F10 cells

The previous study had proved that the PEI-peptide complex could enhance the tansgene expression especially in B16-F10 cells in a partial specific fashion. The combination of HIF-1 mini-promoter with PEI-peptide complex failed to achieve specific expression for target cells. It was shown that TSP had higher activity in tumor or rapid-proliferating cells in previous study. In this section, the pD5-hrGFP which contained TSP was combined with the PEI-peptide complex to determine whether it could achieve specific therapy in B16-F10. The transfection experiments were performed as described above. The results indicated that the levels of expression were 6.7-fold, 2.4-fold, and 1.7-fold higher than PEI alone for B16-F10 cells at 10μM, 100nM, and 10nM RGD-4C-HA respectively (Figure 11).

Surprisingly, the levels of expression were almost no differences than PEI alone for Balb/3T3 and HeLa cells under the same condition. The level of expression even reduced to 41% than PEI alone in HeLa cells at 1μM. The results indicated that TSP combined with RGD-4C-HA could achieve specific enhancement of transgenic expression for B16-F10.

Chapter 4 Discussion

The measurements of several transcription factors associated tumorigenecity were rapid and convenient in laboratory. Since the activities of transcription factors varied in different types of tumor cells, the response elements of TSP can be varied according to the transcription factor profiles in target cell. Moreover, the activities of various transcription factors in target cells can be obtained by high-through-put screening systems or reference searching. The information for transcription factors should be helpful in the designation of different TSP.

The novel mini-promoter TSP contained three copies of NF-κB, CREB, and HIF-1 response elements and was active in tumor and rapid-proliferating cells such as B16-F10, Balb/3T3, and HeLa cells. In this study, the activity of NF-　B mini-promoter was similar to TSP whereas the HIF-1 mini-promoter was higher than TSP (Figure 4). However, the copy numbers were different among these mini-promoters. The NF-κB mini-promoter contains four copies of NF-κB response elements, CREB mini-promoter contains 5 copies of CREB response elements, and the HIF-1 mini-promoter contains 7 copies of HIF-1 response elements. The copy number of each response element on the mini-promoter may affect the activity of expression. In addition, the spacers which separate different response elements also have influences on the activity of whole promoter.

The activity of TSP was obviously related to the activities of the selected transcription factors: NF-κB, CREB, and HIF-1. In the inhibition experiment, it was shown the activities of TSP were more resistant to inhibitors and were not significantly lowered than control group at the presence of inhibitors (Figure 5).

TSP consists of three kinds of response elements and it may result in when the activity of one transcription factor is reduced under certain physiological conditions, the others are still active and maintain the promoter activity. The resistance to such inhibitors may be beneficial for the therapy combined with other therapy such as chemotherapy or certain micro-environment in tumor such as highly angiogenic condition.

TSP mini-promoter consists of only 110 bp which can be modified to improve the activity regulation of expression such as insertion of other transcription factor response element or enhancer. Moreover, the small size of TSP can also be utilized for gene therapy by viral delivery systems. The viral vectors usually use CMV promoter that is always larger than 1kb, thus it may have limitations to result in the fail of package of viral particle containing the large size of therapeutic gene.

RGD-4C-HA was designed as a multi-functional peptide to bind to integrin α_vβ₃ and to absorb to PEI. The RGD-4C (CDCRGDCFC) sequence of RGD-4C-HA could bind to integrin α_vβ₃ expressing cells (B16-F10) as in its native form (Figure 7). Besides, the absorption domain of RGD-4C-HA could bind to PEI

(Figure 8). These results indicate that the functional domains could act without interfering to each other.

In the binding assay of RGD-4C-HA to different cells, RGD-4C-HA strongly bound to B16-F10 rather than others and the total fluorescence intensity was 1.5-fold higher than negative control group under 200nM RGD-4C-HA. Focus on the binding percentage of RGD-4C-HA for Balb/3T3 cells, it was increased significantly but the increase was far lower than for B16-F10 (Figure 7b). However, the total fluorescence mean was no significant differences between negative control group and Balb/3T3 group (Figure 7c). These results may mean that Balb/3T3 cells express few integrin α_vβ₃ on the surface but the level should be below the high sensitivity range.

It was noticed that the total fluorescence intensity was highest under 200nM RGD-4C-HA rather than 2μM. This phenomenon may result from the formation of the inter- or intra-molecular disulfide bond between RGD-4C-HA. There are five cysteines in the RGD-4C-HA sequence and the rate of spontaneous formation of disulfide bond may probably increase at high concentration. The cysteines can form certain disulfide bond to fold proper or improper structures for ligation to integrin α_vβ₃. It was reported that the affinity of RGD-4C to integrin α_vβ₃ is seriouly affected by its correctly secondary structures. Thus RGD-4C-HA may have improper secondary structures to impede its affinity at high concentration.

The PEI-peptide complex can be produced in an easy mixture without complicated chemical coupling reactions. The PEI-peptide complex can enhance the transgene expression in target cells and reduce it in other cells under certain conditions (Figure 9). It was showed that the levels of expression were higher than PEI alone in B16-F10 cells since the RGD-4C-HA can preferentially bind to B16-F10. At the same time, the levels of expression reduced to 63% and 71% than PEI alone in HeLa cells at 1μM and 100nM RGD-4C-HA respectively. The RGD-4C-HA absorbed to PEI and acted as a targeting molecule which may impede

The PEI-peptide complex can be produced in an easy mixture without complicated chemical coupling reactions. The PEI-peptide complex can enhance the transgene expression in target cells and reduce it in other cells under certain conditions (Figure 9). It was showed that the levels of expression were higher than PEI alone in B16-F10 cells since the RGD-4C-HA can preferentially bind to B16-F10. At the same time, the levels of expression reduced to 63% and 71% than PEI alone in HeLa cells at 1μM and 100nM RGD-4C-HA respectively. The RGD-4C-HA absorbed to PEI and acted as a targeting molecule which may impede