Team:Freiburg/Notebook/lab epigenetics

Targeting

Effector

Activation

Epigenetics

June

July

August

September

Repression

Effector Control

Modeling

uniBAss

Standardization

Material and Methods

June

05.06.13

G9a mus musculus DNA arrived from Stuttgart

Lyophilized DNA on Whatman paper was incubated with 50µl sterile H₂O (10min at RT) and eluted from the paper by centrifugation: 51.7ng/µl (NanoDrop).
20µl were sent for sequencing.

Transformation of G9a

4µl of G9a-DNA were transformed into 25µl Top10 competent E. coli. 250µl LB medium was added and cells were grown on 33°C for 1h30min before plating on an ampicillin agar plate.

06.-07.06.13

Clone selection

Because no single clones could be picked, a dilution plating of the transformation was performed and two single clones were selected the next day and streaked out for mini preps.

08.06.13

Mini preps of G9a clones

Clones were prepped with the Roche High Pure Plasmid Isolation Kit. Both clones were sent for sequencing with oIG8007 and oIG8008.

09.06.13

Sequencing results

Both clones contain G9a from mus musculus.
New primers to amplify G9a(mm) were designed and ordered.

insert primer list here?

14.06.13

New primers arrived (oIG8009-oIG8012)

PCR to amplify G9a-SET domaine

aIG8000 (linker-G9a-NLS): oIG8009 (fw) and oIG8010 (rev) with G9a-Mini prep 1 (8.6.) as template.
aIG8003 (NLS-G9a-linker): oIG8011 (fw) and oIG8012 (rev).

PCR aIG8000 and aIG8003

Bands run too low, can be due to gelred. Check size after gelextraction by loading less DNA.

Gel extraction of PCR products aIG8000 and aIG8003d

PCR to amplify Cas9 (on pX334a)

We need the mutated non-cleaving dCas9. Therefore we use pIG2004 from Max as template to produce aIG8001 (oIG2000/oIG8002) and aIG8002 (oIG2006/oIG2009).
5µl of Q5 buffer were used instead of 10µl. Elongation: 4min20sec.

PCR aIG8001, Gelex aIG8003(size correct), marker, aIG8002

Gel extraction of PCR products aIG8001 and aIG8002

15.06.13

PCR for aIG 8004, aIG8008, aIG 8007, aIG 8010 and aIG 8011

• aIG8004: primers: oIG2008 and 8005; template: pIG2004
• aIG8007: primers: oIG8009 and 8014; template: G9a(mm)
• aIG8008: primers: oIG8013 and 8010; template: G9a(mm)
• aIG8010: primers: oIG8014 and 8011; template: G9a(mm)
• aIG8011: primers: oIG8013 and 8012; template: G9a(mm)

PCR reaction was performed as described below

PCR aIG8001, Gelex aIG8003(size correct), marker, aIG8002

19.06.13

Gibson Cloning for pIG8002 and pIG8004

Even though aIG8001 did not yield good results by gelEx, we performed Gibson cloning, as PCR was not repeatable. Gibson mix was prepared as described below.

• pIG8002
• aIG8006: 0.35µl
• aIG8001: 1.85µl
• aIG8002: 2.80µl
• pIG8004
• aIG8001: 1.75µl
• aIG8002: 2.64µl
• aIG8009: 0.617µl

20.06.13

Clones were obtained for pIG8002, but not for pIG8004. clones from pIG8002 were streaked out on plates for miniprep.

21.06.13

test digest of pIG8002

Clones were prepped using the Roche kit and eluted in 50µl ddH₂O. Concentrations were roughly 230ng/µl. A test digest was performed using EcoRV and NotI. If the correct plasmid is obtained, three bands will be visible (at 5850bp, 3400bp and 1700bp).

Results of test digest

15 out of 17 colonies were positive. Clone 3 and 5 were sent for sequencing with oIG0009, oIG8007 and oIG8008.

22.06.13

Gibson for pIG8004

Sequencing results of pIG8002

Sequencing confirmed clone 5 as pIG8002. The catalytically active dCAS9:G9a-SD is complete. Clone 3 showed a deletion in between promoter and ATG. We do not expect this mutation to influence the protein, but, nevertheless, clone 5 is completely correct.

Repetition of Gibson for pIG8004 (dCAS9:dG9a)

As the first Gibson did not yield any clones we repeated it. DNA mix was prepared as described below.

• aIG8001: 1,75ul
• aIG8002: 2,64ul
• aIG8009: 0,617ul

DNA mix was added to the aliquoted Gibson reaction mix and was incubated for 60 minutes at 50°C, 3 minutes on ice and 3 minutes on RT. 5µl of the mixture was transformed into competent E. coli and plated on Amp(+) plates.

23.06.13

11 Clones were obtained and picked. They were streaked out on plates for further analysis.

24.06.13

Test digest of pIG8004

Clones were prepped using the Roche kit. Test digest was performed using EcoRV and NotI. See table below for composition of test digest.

Results of digest

11 out of 12 clones were positive, as all 3 bands were visible as expected. Numer 3 and 5 were sent for sequencing with oIG0009, oIG8007 and oIG8008.

July

02.07.13

Sequencing results of pIG8004 and starting of target site ligation

pIG8004 was confirmed as positive. The next step is to ligate target sites designed by Manuel into both constructs (from here on called VEGF1-4, respectively pIG8002_1-4 and pIG8004_1-4), for functional testing of our construct. To do so, the construct will be opened with BbsI and afterwards the annealed oligos will be ligated into the construct with T4 ligase. Digest was performed as follows.

Results of digest

loading scheme: marker - pIG8002(digested) - pIG8002(control) - pIG8004(digested) - pIG8004(control). No big shift is visible when loading only undigested control.

03.07.13

Ligation of the annealed oligos into pIG8002 and pIG8004. 50ng of opened vector is ligated with 150-200ng of insert. A strong excess of small oligo is suspected to work more efficiently than less.

• pIG8002+oKM511/512:pIG8002_1
• pIG8002+oKM513/514:pIG8002_2
• pIG8002+oKM515/516:pIG8002_3
• pIG8002+oKM517/518:pIG8002_4
• pIG8002+oKM511/512:pIG8004_1
• pIG8002+oKM513/514:pIG8004_2
• pIG8002+oKM515/516:pIG8004_3
• pIG8002+oKM517/518:pIG8004_4

ligation was performed as follows:

5µl of each transformation assay was transformed into competent E. coli (batch 8.5.), following the standard protocol.

04.07.13

Test digest of ligated crRNAs

Clones were obtained for all constructs. Minipreps were performed and yielded roughly 230ng/µl. Test digest was performed with BbsI and XhoI as described several times above. Results are inconclusive, so pIG8002_1 and pIG8004_1 were sent for sequencing with oIG0015.

Test digest - no difference is visible

05.07.13

pIG8002_1 was sequenced as correct, pIG8004_1 was negative, BbsI seemed to have cut too much of the plasmid, one Direct Repeat is missing. Strange. rest was sent for sequencing with oIG0015.

09.07.13

Sequencing revealed positive clones for pIG8002_1 and 3, as well as for pIG8004_4. Negative colonies were discarded. More preps were done to screen for further clones.

• 8002_2.2
• 8002_4.2
• 8004_1.2
• 8004_2.2
• 8004_3.2

10.07.13

8004_3.2 is positive, rest negative. For the missing constructs one of each clone was sent.

11.07.13

positive clones were obtained for 8002_2.3, 8002_4.3, and 8004_1.3. New clones were sent for sequencing.

12.07.13

positive clone for 8004_2.4 was found. All constructs are done. Midi preps were inoculated in 150ml LB with 150µl Amp. As well as control plasmids without target sites. Flasks were directely inoculated by giving a little bit of colony with a sterile tip into the flask.

15.07.13

Midi preps were performed using the JetStar Midi Kit. Pellet was resolved in 100µl of ddH₂O. All preps yielded acceptable results (500-1900 ng/µl).

HEK293T cells were seeded into 24 well plates. Our construct should be able to efficiently repress VEGF expression by methylating H3K9. As HEK293 cells have an open locus we should see a reduction in VEGF levels. 65,000 cells were seeded per well.
Each construct will be transfected in biological triplicates, let grow for 48 hours and the supernatant harvested. ELISA measurments will be performed using the peprotech VEGF ELISA Development kit.

16.07.13

Transfection of the cells was performed as described in the table below. PIF-GFP was transfected as a transfection control to have a fast impression of transfection efficiency. Next transfection should be calculated for 2-3 wells more to have a buffer for pipetting mistakes. Recalibrate pipette?

Transfection scheme 16.07.

17.07.13

Transfection seems to have worked, as we see green fluorescence in the nuclei of the cells.

VEGF ELISA kit arrived. Antibodies were aliquoted following the instructions and stored at -20°C. Buffers were prepared as follows:

• PBS: dilute 100ml 10xPBS in 900ml ddH₂O.
• PBS-T: 1xPBS+0,05% Tween-20
• Blockbuffer: 31ml PBS+ 0,31g BSA
• Diluent: 45ml PBS-T+0,045g BSA

Every buffer containing BSA is stored at 4°C. ELISA plate was coated with capture antibody as described in the instructions.

18.07.13

ELISA detection of VEGF content of supernatant

All steps were performed following the protocol of Peprotech, except the following steps: Blocking time was elongated to 2 hours, and washing was performed 3 times with 300µl wash buffer (PBS-T). Samples were used diluted and 1:2 diluted for ELISA detection.

The loading scheme of the ELISA plate is shown below.

ELISA and plate scheme from 16.07.

results of ELISA

Results of the first VEGF ELISA from 16.07., error bars represent standard deviation.

Direct comparison of 8002 and 8004 constructs show less VEGF in 8002 (active G9a) as expected. Controls show very low amounts as well. This is rather strange, but gives a hint that our primary idea is correct. Several possible problems and solutions wre identified.

• VEGF is secreted when cells are exposed to oxygen stress. The distribution of O2 could be varying inside the plate, messing with our results. The use of filters instead of lids could be a solution.
• The use of an internal standard could be a possible solution for reducing cell toxicity of G9a and gives us as well a hint on cell number/transfection efficiency. A SV40:SEAP plasmid will serve as this.
• Triplicates should be mandatory, if possible

22.07.13

HEK293T cells were seeded at 70,000 cells/well in 24-well plates.

23.07.13

Transfection was performed as described in scheme and protocol from 16.07.. Each well was co-transfected with 0.0375µg of SV40:SEAP (commercial pK2 SEAP control plasmid by CloneTech (R)) that will serve as internal standard. The cells were quite dense, so we will have to evaluate transfection efficiency carefully.

Transfection scheme for repeated experiment from 16.07. 5% SV-40 will serve as internal standard.

24.07.13

Cells were too dense, transfection efficiency was low => Cells were discarded. Seeding of cells will be repeated.

27.07.13

HEKs were seeded at 65,000 cells per well in 24-well plates. We were advised by Bea to assure good distribution of cells.

28.07.13

Cells show strange behaviour: They form a clump in the middle of the well, whereas the outer space is too lose. This will reduce the transfection efficiency strongly in the middle of the well, making our experiment quite difficult. Cells were discarded. again, cells were seeded with the help of Adrian.

29.07.13

Same procedure as everyday. We suspect the cells to be misbehaving. We acquired new cells from AG Warscheidt for testing other cells. A contamination could explain the problems in growth.
for testing we transfected the old cells anyway to check efficiency.

New cells were seeded at 65.000 cells per well into 24 well plates. (HEK293 cells). Cells were resuspended and singled with great(!) caution. We checked the transfection efficiency of the old cells which was below 10%. The cells were discarded.

August

01.08.13

Cells are not dense enough for transfection (~30%), so transfection was postponed to the next day.

02.08.13

Cells were transfected following the standard protocol as described in the following tables at 12.a.m..

Loading scheme of the plates used in the second ELISA experiment.

Transfection scheme for ELISA experiment

The plates were covered with O₂ diffusible filters to prevent oxygen stress to occur and messing up our data.

03.08.13

Medium of transfected cells was changed at 10 a.m., to repress the native VEGF level to a minimum. Transfection efficiency was around 50% when assaying the PIF-GFP construct. We are confident, that this will suffice for our experiment.

04.08.13

Supernatant of cells was harvested (400µl) for further analysis. 200µl are used for the VEGF ELISA, whereas the other 200µl were frozen to perform a SEAP measurement the next days as internal standard. ELISA was performed as described above.
Additionally cell lysis was performed using modified RIPA buffer for expression control of our construct by Western blot. Lysates were cooked for 5 minutes in sample buffer and frozen for later blotting.

VEGF locus 2 and 3 were displaying strong (!) repression this time (see graph) in comaprison to the control bar. But controls are varying strongly in VEGF expression. Even more, locus 1 behanves really strange. Maybe the cells have died in this wells? Western Blot will show the problems.

Results of the second ELISA experiment performed. VEGF is measured using ELISA. Controls are varying strongly. Maybe normalization to SEAP will solve the problem.

05.08.13

SEAP measurement was performed as described in the standard protocol. All VEGF levels were divided by the SEAP level to get a normalization to an internal standard. Error was propagated using Gauß's mistake propagation.

by looking at the data now it is obvious that sample 4_1 can be discarded for further analysis. But nevertheless the rest of the samples is looking great. Even the controls are within range of one standard deviation. Strangely PIF:GFP displayed a very strong SEAP expression, lowering the bar in the normalized plot. But loci VEGF2 and VEGF3 display really strong effects.

Results of the second ELISA experiment performed. VEGF is normalized to SEAP. Error bars represent standard deviation.

For further expression analysis we were using a Western blot approach. To do so, 2 8% Acrylamide gels were pourn with 50% Sucrose added. They were stored at 4°C and will be ran tomorrow.

06.08.13

Protein samples were run on an 8% gel at 130 Volt. Afterwards a semidry blot was performed at 70mA to transfer protein onto an VDF-membrane. Primary antibody (anti-HA and anti-ß-act) was diluted 1:2500 in 2% milkpowder. Blocking was performed using 4% milk in PBS-T for 1 hour. Primary incubation took place overnight at 4°C on a vertical shaker.

07.08.13

Membrane was washed 3x for 5 minutes with 1x PBS-T. Secondary antibody (anti-mouse) was diluted 1:5000 and applied to the membrane for 2 hrs at RT. An additional washing step was performed 3 times for 5 minutes in 1x PBS-T.

Detection of Western Signal<

Membranes were treated with 600µl of ECLI and II each. Signal was detected using the LS imageQuant. Results are displayed in the following figure.

Western blot with the samples of second ELISA experiment. Evident expression of Cas9-G9a is visible at all lanes (even for 4_1, but very weak). Nevertheless, loading controls are varying strongly, indicating badly performed cell lysis.

We see expression of dCAS9:G9a in each well, except 4_1, which supports our hypothesis, that here a problem with the cells occurred. The ß-actin levels are varying strongly what leads to the impression, that cell lysis was not performed perfectly . Nevertheless, we could detect our protein in each well were we saw effect. Today is a good day!

08.08.13

Further planning after discussion: A kinetic should be done, but should be planned carefully.

12.08.13

Plans for Kinetics:

• Kinetics will be performed in two ways: Either measuring production of VEGF in 24hours over 5 days, and VEGF levels over 5 days, without medium change.
• To reduce variability in between plates we will transfect in 10cm dishes, and spread afterwards in 24 well plates.
• 6 24-well plates will be used
• Filters and SEAP standard will be used

14.08.13

All constructs were transfected into HEK293 cells in 10cm-dishes. For DNA mix, see the following table. Every construct was treated the same.

Western blot with the samples of second ELISA experiment. Evident expression of Cas9-G9a is visible at all lanes (even for 4_1, but very weak). Nevertheless, loading controls are varying strongly, indicating badly performed cell lysis.

15.08.13

To detach cells from the plates, medium was removed. Afterwards they were washed in 5ml PBS.
PBS was removed and another 10ml of PBS was added. Cells detached using a cell scraper centrifuged and resuspended in 4ml DMEM. 75% of a plate was diluted in 9.5ml medium. 500µl was put to each well of a 24-well plate. Plates were seeded according to following scheme.

Cell plate scheme and experimental setup.

16.08.13.

Standardisation of G9a in RFC 25

Standardisation of G9a will be performed using a Gibson approach. 2 cutting sites have to be mutated in the protein sequence. Primers were designed that contain the mutation in the overhangs. With those primers PCR will be performed and Gibson will ligate the whole thing in the pSB1C3 backbone with RFC25 prefix and suffix.

All oligos arrived for the standardisation. PCR was performed on pIG8002 with:

• oIG8015 + 8017 => aIG8100
• oIG8018 + 8019 => aIG8101
• oIG8020 + 8016 => aIG8102

µl	type
10	Q5-HF Reaction Buffer
1	Template
1	Primer1
1	Primer2
4	dNTPs
1	DMSO
0.5	Q5-HF Polymerase
Add to 50	H2O

• Annealing: 60°C
• Elongation: 30 sec
• 24 cycles

all PCRs yielded desired products. loading scheme: aIG8100 - marker - aIG8101 - marker - aIG8102

All bands were extracted using the Roche kit. Gibson was performed using the three fragments and the opened pSB1C3 backbone. for DNA mix see the following table. Reaction was performed as described in the standard protocol.

5µl of the reaction mix were transformed into competent E. coli following the standard protocol. Bacteria was spread on Cap containing LB plates.

16.08.13

Red light inducible PhyB-G9A and Cas9-PIF system

PCR amplification of the backbone containing PhyB for Gibson

µl	type
10	Q5-HF Reaction Buffer
1	pKM018 (30ng)
1	oIG8021
1	oIG8022
2.5	dNTPs
1.5	DMSO
0.5	Q5-HF Polymerase
Add to 50	H2O

• Annealing: 60°C
• Extension: 2min 30sec

PCR amplification of G9A with specific overhangs to the PhyB-bb

µl	type
10	Q5-HF Reaction Buffer
1	pIG8002
1	oIG8023
1	oIG8024
2.5	dNTPs
1.5	DMSO
0.5	Q5-HF Polymerase
Add to 50	H2O

• Annealing: 60°C
• Extension: 40sec

Results

Left side: 2log laddder; On the right side: G9A; everything else are different samples

Left: 2log ladder, right: PhyB bb

All bands had the expected size (gelred standard) and were band isolated and purificated

22.08.13

Red light inducible PhyB-G9A and Cas9-PIF system

Another gibson approach of G9A PhyB-bb using a 1/4 ratio

Gibson approach:

Test digest of some Gibson colonies using Sal1 and EcoR1

Left: 2 log ladder; then tested Gibson colonies from 1-10. Colony 5-10 seem to be defenitly possitive.

Colony 5 was send for sequencing

Ligation of two VEGF crRNAs into pIG3010

For light experiments a plasmid containing the tracrRNA and crRNA is needed.

Ligation

µl:	Substance:
1	pIG3010
5	crRNA2/3 (2/3 refere to the different VEGF target sides)
2	10X T4-Ligase Buffer
1	T4-Ligase
11	H2O
20	Total

• 0.5h at RT

17.08.13

Colonies were obtained for the standardized G9a-SD. To screen for positive ones, colony-PCR was performed as described in the standard protocol. The following PCR mix was used.

µl	type
2.5	taq buffer
1	oIG6017
1	oIG6018
2.5	dNTPs
0.125	Taq-polymerase
Add to 25	H2O

• Annealing: 60°C
• Elongation: 45 sec
• 24 cycles

One lane shows the desired band at roughly 1kb. Upper row, left lane.

Positive colony was streaked out for miniprep.

24.08.13

Data was evaluated. We do not see repressive effects in any of the kinetics. The experiment failed miserably (see figure below for details). Possible explanations are:

• VEGF is a stress response gene, that could have been activated by transfering the cells.
• There is no effect of G9a - all effects are simply measuring mistakes
• We screwed up transfection

For the sake of time, the experiment will not be repeated, but the original experiment will be repeated to prove to reproducibility of our data.

Results of kinetic

Results between the kinetic were the medium was not changed. We do not see any differences in between the constructs.

Results of the kinetic where the medium was changed every 24 hours. Again, we do not see any differences in between the constructs. possible explanations above.

08.09.13

Supernatant was harvested, ELISA was performed and SEAP was assayed as described above. We can draw the following conclusions from it (see figure below).

• the results are reproducible, even though not to the extend as last time
• the endogenous VEGF contents are strongly varying - Even though of the usage of an internal standard
• We should stop doing experiments, until the devices are finally built.
• the number of passages may influence the chromatin structure of the VEGF locus - maybe use fresh cells next time

VEGF analysis to confirm our results - again, the first locus (-8) does not show any effect. But the other effects could at least be confirmed, concerning their tendencies.

26.09.13

Cells reached optimal confluence. Transfection as performed following the standard protocol and the following schemes.

Transfesction scheme for final VEGF ELISA.

Transfection calculations (1) for final VEGF ELISA.

Transfection calculations (2) for final VEGF ELISA.

28.09.13

Supernatant was harvested and ELISA detection was performed following the standard protocol provided by Peprotech. The loading scheme of the 96 well plate is done as follows.Evalutaion will be performed after SEAP assay was performed.

Loading scheme for ELISA plate.

29.09.13

SEAP assay was performed, diluted 1:10. Standard protocol was followed. SEAP levels seem to be quite small. ELISA and SEAP data was evaluated and the ratio of VEGF/SEAP was calculated. (see following graph)One can see, that we have the repression of the VEGF level, even more than the CRISPRi. This is interesting and important, as the Cas9 protein is higher expressed than the cas9:G9a fusion. These are very nice results, especially, that the off-target controls do not show any differences. Error bars are standard deviation of VEGF and SEAP propagated via Gauß.

VEGF/SEAP ration, reflecting VEGF levels.