Team:SYSU-China/Project/Result/element test

From 2013.igem.org

Revision as of 23:26, 27 September 2013 by SporeG (Talk | contribs)

ipsc

UPDATE 09/22/2013

Suicide Gene: The Cruel Guard and Its Ancillary Facility

1.Comparison between Suicide Genes: who is the most tough killer?

We have selected several Suicide Gene candidates which functions in different pathways and patterns. In order to choose one that is most capable of mediating death in cancer cell, we first carried out a parallel experiment to compare their performance. The candidates were cloned into plasmid pcDNA3.0 in the MCS after CMV promoter just as shown in picture A, and then transfected into HEK293(PEI) and HepG2(Lipo2000). A GFP processed in the same way was used as a negative control to both indicate the transfection efficiency and normalize the death effect bring about by transfection process.Pictures were taken 48 hr after the transfection process.

To achieve a better resolution of the lethal effect, we tried two techniques with the 48hr cells. The first one is DAPI staining done in RIP1 and RIP3 groups. DAPI can specificly binds to DNA but is rejected by living cells to some extent. With DAPI staining, more cells were darkly dyed in the group of RIP1 and RIP3 compared to GFP control. Broken nuclei membrane can be noticed in cells transfected with RIP1 because DAPI blue spreaded all over cytopkasm The second one is flow cytometrycounting(FCM counting). Cells were stained by PI and then analysed by FCM counting. Dead cells binded more PI per cell in a similar reason to DAPI staining. These two techniques were complementary to each other in a certain degree, for DAPI superior in morphological observation while FCM counting is good at counting.

Figure 1. Performance of different Suicide Genes candidates in causing apoptosis.(A).plasmids construction. (B,C) Wells transfected with RIP1\RIP3 both showed significant phenotype of apoptosis compared with GFP control, 48 hr after transfection. After DAPI staining, dead cell showed deep blue color and the phenotype was better visualized. (D)Apoptosis phenotype of Apoptin(VP3). The protein was expressed in fusion with a GFP. Compared with 0 hr , apoptosis rose 48 hr after transfection, and GFP vision indicated the transfection efficiency.(E)DAPI staining of RIP1, zoomed in vision.We can clearly observe boundary of cells which indicate a broken nucleus membrane and DNA spread all over the cells(F).PI staining and FCM data.

The figure shows that Both RIP1 and RIP3 had a outstanding performance in mediating apoptosis. Although Apoptin also killed cell in the results, its performance wasn’t stable since we cannot reproduce the results with another construction( data not shown.) However, according to many paper, this is a powerful protein with a specificity in killing cancer cell, which persuade us to keep working with it in the latter experiments. Moreover, we also test Bax/Bax S184A with the same method. However, no significant phenotype could be observed(data not shown) and thus we abandoned them in our Suicide Gene list.

The PI staining and FCM data showed that the apoptosis rate of RIP1 and RIP3 were about 51.7% and 47.6%. However, there could be some problems about the results such as a low expression level of GFP control, which made the data not so convincing. And we had planned a repeat for the experiment.

2.Suicide Gene working with its ancillary facility

Although we have successfully proved that each device of our pathway worked well, we wanted to further make use of the convenience of transient transfection due to time-consuming quality of Lenti-Virus infection. So we designed combinational experiments, centered on our protagonist, the Suicide Gene, to further confirmed the devices’ performance in working coordinately. Although we didn’t have enough time to finished every one of them before Regional Jamboree, we have put them into our schedule of October if they’re still needed at that time(in situation where we still don’t have enough convincing data from Lenti-Virus work.)

A.Controlling the switch: a combination of Suicide Gene and Tet-off system.

Tough as the Suicide Gene is, what we expect from it was to protect instead of killing the innocent. And the 2 kinds of “innocent” cells in our designing were iPSC and Hepatocyte. For iPSC, we have assigned it with a controlling panel: the Tet-Off system. Thus we designed an experiment to test the ability of Tet-off system to drive or turning off the expression of Suicide Gene. The designing and expected results were both shown below and Tet-Off construction driving GFP expressiong was taken as the positive control.

Figure2.Combinational test: Tet-Off system and Suicide Gene

B.Rescue scheme: a combination of Suicide Gene and MicroRNA-Target system.

For Heptocyte the protection comes from functioning of MicroRNA-Target system since the miR122 level was very high in Heptocyte(see Design). Due to technical bottleneck in transfection of Heptocyte, we decided to first simulated the environment and thus designed a mini-orthogonal experiment crossing variables of Target and miR122. The designing and expected results were shown below.

The results, together with the RT-qPCR data of miR122 level in Heptocyte(see qPCR in Hepatocyte)[linker] , would prove that the device’s knockdown efficiency is sufficient to protect Heptocyte. Also with the mathematical model derived from Simulation experiment with GFP, we can predict the rescue efficiency under different concentration of miR122 and target-site number.

Figure3.Rescue experiment:MiR122-Target system and Suicide Gene

C.The final test-run: combination of all three systems before setting up stable cell line

With a idealized transfection efficiency, theoretically the three systems can be assembled and tested by simply transfect the cells with 3 plasmids(fig4.A). Plasmids TG2T/TSG2T and Et(see the short name in fig4.A) can turn on the Tet-off system and drive expression of GOI(Suicide Gene or GFP), while plasmid CEP can simulate the environment of Hepatocyte and execute the rescue scheme. Although practically it could be difficult due to efficiency and duration time limitation, we still decided to try it as a complement and see how much we can do with the transient transfection strategy. So we designed the following orthogonal experiment which, though seemingly complicated, could perfectly give a cross-reference to each variable and confirm all elements of our pathway.

Figure4.Orthogonal experiment of all three devices.

We carried out the plan in HEK293 cell with PEI. However, due to technical problems no obvious results can be observed. We may try to optimize the technical factors to promote efficiency and repeat the experiment if necessary.

MicroRNA-Target Devices: Security Gate for Liver Cell

1.Performance comparison between target-sites: CULT and Complete

Two kinds of target-sites were taken into consideration just as mentioned in module Design. To make a comparison between their knockdown efficiency under different concentration of miR122, another orthogonal experiment was performed and two plasmids(fig.A) were co-transfected by PEI into HEK293 cell line. Pictures were taken 24 hours after transfection.

Figure5. Comparison of performance of different target-sites in knocking down GFP expression. (A) Plasmids A/B/C/D produced mRNA of GFP with different target-site in its 3’ UTR. Plasmids E produced miR122. Plasmids used in this experiment had the problem of producing a background noise of GFP, which was solved later(see below).(B) Knockdown effect of different target site under different miR122 level. The fields of vision were chosen so that they could well represent the average level of the wells.

The result shows that with the same copy number, complete target-site works better than CULT target-site, which fits the results of the 2010 Heiderburg iGEM team. What’s more, two copies target-site works better than single site, which encourage us to further build target-sites of more copy number and employed more quantitating experiments.

Plasmid E used in the experiment produced background GFP, which was a problem neglected by us. After finding out the problem, we did quick-change to introduce a stop-codon into the CDS of GFP, which proves very efficient. The new miR122 expression plasmids were applied to other experiments just as mentioned. However, we found that although this did interfere our results to some degree, we could still gain insight into the problem so the data was kept.

2.Quantitative characterization and target-site modelling

In order to build a precisely controlled device, we decided to further characterized the complete-target-sites of different copy number and chose the most suitable one for pathway construction. To begin with we designed another orthogonal experiment crossing the two variables, target-site-number N and MiR122 concentration M, with a more precise gradients for MiR122. But soon we found that it would be too labour-costing and decided that a better way was to replace it with 2 dosage-gradient experiments and a mathematic model. (fig.6)

Figure6.Two strategies of characterizing MiR122-Target system.

A.Dosage experiment testing performance of 2 copies complete target-site

Firstly we fixed the target-site number at 2 and set up a concentration gradient for miR122 expression plasmid (fig.7,A, plasmid CEP) to gain insight into its dosage effect. Plasmids A and B were co-transfected using PEI into HEK293 cell line. The gradient were set up from 0 to 0.5ug of plasmid per well. The experiment was done with 2 parallel groups of wells( that is, 6 wells as a group and there were 2 repeats) to leave samples for both Western-Blot and RT-qPCR(since both techniques require destruction of sample). To ensure the parallelity so that the two data can be employed to illustrate the problem together, we had repeated the experiment 2 times till we gained same fluorescence intensity with corresponding wells(data not shown).

Figure7.Dosage effect of miR122 on 2 copy number target-site.(A)Plasmid A produced GFP mRNA with a 2 copies Complete target-site in its 3’ UTR, while plasmid B produced miR122 without making fluorescence noise. (B) A gradient of total plasmid amount per well was set up as shown. A significant positive correlation between knockdown efficiency and plasmid B concentration can be observed.(C) RT-qPCR showed that plasmid B successfully express miR122 in cells, the level of which was in a good linear correlation with the transfection concentration through linear-regression analysis(data not shown). (D).A not-so-successful Western-Blot showing the GFP level in each well. Based on scanning densitometry, the level can be counted and plotted as above.

The results showed a simple and elegant linear correlation between knockdown efficiency and miR122 concentration.What's more, we have planned to repeat the experiment and make a statistic data.

B.Performance of complete target-site of different copy number

Secondly we fixed the miR122 concentration at 0.75ug to ensure an optimal performance of each targets and let target-site number be the variable this time. Since we had already made sure of the relationship of miR122 level and plasmid concentration, RT-qPCR was no longer needed and so was the parallel groups.

Apart from the 2 and 4 copy target-site, we have also constructed a mixed-up plasmid with 2 complete target-site followed by 2 CULT target-site(2+2).

Figure8.Comparison of knockdown efficiency of different target-site.

The results show that 4 copy target-site has the most brilliant knock-down efficiency, which is about under such a miR122 level, which conforms to the data given by 2010 Heiderburg using dual-fluorescent system. What’s more, in Part1 we have already proved that 2 copy target-site works better than 1 copy target-site. Finally we decided to first continue further construction of the pathway with the 2 copy target-site, considering that its moderate knockdown efficiency may well coordinate with the other two systems. And if the efficiency turns out improper, we can replace it with other candidates easily with our molecular designing.

C.Modelling

In order to better understand the quality of our miR122-Target system, we have employed interpolation technique with Matlab to model its performance. The modeling can be seen at Modelling module in Results. Also, we have planned a more well-designed experiment to test and characterize all target-sites(in fact, there are 8 of them, produced by permutation and combination of different numbers of complete and CULT). The data derived from that would better support the modeling work.

Figure9.Modeling of MiR122-Target system.

Tet-Off system: Controlling panel for iPSC

We also used eGFP to test the switching and driving performance of Tet-Off System, which would be a convenient controlling panel to switch pathway state between On and Off and thus provided protection for iPSC. We have collected and constructed three TRE elements of different generations.and thus made a comparison between them.

Figure10. Performance of Tet-off system in switching between different state.Two plasmids(fig.) were co-transfected into Bosc cell line using PEI. The wells with only plasmids B/C or D characterized the basal leaky expression of different pTRE elements, which can be viewed as the first Off-state. The well with both plasmids represents the On-state of the system, and the well further added with dox represents the second Off-state. Dox concentration was chosen according to protocols in order to give a best performance of the system and avoid hazard on cell growth.

The results showed that pTRE-Tight has the least leaky expression and a good switch response between On and Off state. However, in our experiments leaky expression was too high of pTRE2 and pTRE3G elements, and they cannot response to tTA and DOX properly(data not shown),which confused us a lot since according to protocol from Invitrogen, pTRE3G should have the best performance. We have repeated the experiments several times in troubleshooting, but gained no satisfactory results. So we decided to progress with pTRE-Tight in the first place and leave the other ones as alternative schemes.

Figure11. KRAB the new tet off system

Lentivirus

Packaging:

We packaged two batches of lentivirus, one containing the upstream tTA plasmid, the other containing the downstream suicide gene and target plasmid.

Figure . The map of plasmids for lentivirus packaging.

The package of lentivirus vectors was done in HEK 293T cell lines in 10cm cell culture plates, along with two packaging plasmid: psPAX2 and pMD2G. Both of them were 3rd generation packaging plasmid. We then collected the CM supernatant after 48h and 72h, and transfered all of them into tubes.

Centrifugation:

Because these lentivirus vectors were supposed to infect not only Hep G2 cell lines but also mouse iPSCs, which was very difficult to infect, the requirement of the MOI (Multiplicity of Infection) was much higher than normal experiments, so before infecting cells, we ultracentrifugated these viral vectors.

Figure . The steps for lentivirus packaging and centrifugation.

Here is the photo of the 48hr culture medium with lentivirus.

Figure. The colors of culture medium with different lentivirus.

After successfully testing for each parts, it's the right time to think about the assays in different period of cells.

Click the cells to see the results.


The fate of a unfortunate iPSCs.

Sun Yat-Sen University, Guangzhou, China

Address: 135# Xingang Rd.(W.), Haizhu Guangzhou, P.R.China