I would like to manage a mostly open-access e-notebook and share my findings! My inspiration for that idea comes from Alistair Boettiger’s notebook.
Walk today: IMR90 Cells.
18th run: FISH wasn’t great, but found a reasonable cell. Had bubbles. Can first see them at the pipe coming out of valve 0, and heading to the 1st bridge. This caused the vfocus to be inaccurate. Tomorrow, when washing the pipes, should connect an empty coverslip, and run the fluidics with the experiment parameters to check where are the bubbles coming from. Also check if falcon tubes are empty.
Good ‘walking’ run (from the 17th). Setting the max allowed correction at 50 um fixed the z-drift issue.
Reduced secondaries concentration from ~1.4 uM, to ~0.7 uM.
Modified parameters: When first introducing flow to the flow cel use 8 RPM for 6 minutes to fill the flow cell nicely.
Flow rate: 12 RPM. Washing time: 5 minutes. Other flow times: 4 minutes. Image pause (before imaging): 2 minutes.
We found that when increasing the piezo respond time, the columns disappear. carl reasoned that the piezo is jittering.
The source of the Z-columns we are seeing seem to originate from scanning in Z. When I had a bead sample, I saw that when I was scanning in Z i saw columns, while without scanning, the columns disappeared.
For last couple of ‘walking’ experiments I placed the imaging buffer in a 4c cooler. That seems to be a bad idea… At both times I obtained few blinks. The walking worked much better when I either put the buffer in ice (which melts overnight), or in room temp.
I’ve tried using 200 nm crimson beads for fiducial, but those are too bright. Can try 100 nm beads.
First fully-automated step! There is a slight focal shift, and seems like a slight drift as well.
Because of bubbles, had to re-seal some of the needles, and fix bridge #1. It should be a good idea, to run the wetting and cleaning of the pipes while the flow cell is connected to always check for bubbles. So far I fixed 111-711, 887, and 888.
Re-checked my HOPs2_12_49 sublibrary. Should work well with 0.4 uM primer concentration, and 2 ng/ul template concentration.
Ordered Cynthia’s chip, Wren’s chip is on the way.
Should order the oligos for Cynthia’s chip.
- Made some adjustments, including reorienting the pump, and attempting to get two different pipes in a bridge pipe as close as possible. It helped very much! Now, at 20 rpm, for the Bioptechs flow cell, the rate is ~ 1 ml per 3 minutes. That means, it takes 3 minutes for 1 ml to accumulate in the waste tube, after going through the Bioptechs flow cell.
- Walking works, at least by operating the pump manually for now. Did a 1 step proof-of-principle. Imaged C08 and C09. 3 minutes per wash. 30 minutes secondaries hybe.
40 mm coverslips FISH:
- Works well. 35 ul hybe. 2.5 uM per ~ 2 Mb probe concentration.
- Use ~ 7 ml in 10 cm dish, for 2 40 mm coverslips. In waterbath, place each coverslip in glass-bottom dish or coplin jar lid with parafilm on top.
- Hybed with 22X30 mm rectangular #1.5 coverslips to seal. Give ~ 7 minutes at 42c for rubber cement to dry. For denaturation I placed the hybe sandwich on a western gel glass plate.
We sent the 12k chip to be synthesized at CustomArray. Tomorrow morning will order oligos from IDT.
- Make an estimation of the tubing volume.
- Gilson tubing ID = 0.38 mm = 0.015″.
- McMaster tubing ID = 0.02″.
- Dead volume is about 300 μl. ibidi channel plus reservoirs is about 200 μl. All together 500 μl.
- Measure the time it takes to flush 1 ml.
- at 20 rpm, the rate is about 1/7 ml/min, which is ~ 140 μl/min.
- Pipette ~ 1 ml 2X SSC into positions 888, 887, and 111, and flush for wetting the pipes.
- Make ~ 3 ml imaging buffer, and replace tube at position 887.
- Make ~ 5 ml washing buffer, and replace tube at position 888.
- Make ~ 1 ml hybridization buffer, and replace tube at position 111.
- have ~ 50 μl of imaging buffer without BME in channel 2.
- Connect channel 2 to the fluidics.
- Choose a cell for imaging.
- Start wash-image-hybridize cycle.
Test the FISH in the Nikon downstairs. All channels work. They are organized from highest (2) to lowest (5) SNR .
Test walking with 1 hybridization.
4:00 pm – Run T7 with HOPs dye-conjugated primers, and NH Beta2.
The system is connected as follows:
15 ml tubes are connected to the valves. Valves are chained, where valve # 1 is connected to the pump. The pump is connected to the flow cell input, and the flow cell output is connected to the other end of the pump tubing. This edge of the pump tubing is waste.
The fluidics system is compatible with both the ibidi u-slides, and Bioptechs flow cell (FCS2).
For the ibidi u-slides, I connected the tubing and saw the flow-through is good, with no bubbles.
u-slide and 40 mm coverslip ready to image.
Once I started counting the blinks I’m getting at different compartments, and realized I’m not getting enough blinks (N-STORM, Bruker’s Vutara, Xiaowei Zhuang’s scope), I performed qPCR and realized that the library is not reliable (I believe the synthesis wasn’t good, 90k chip and 160 nt probe length seems too much). However, my other library (149 nt, 12k chip) seems to give good qPCR results (with the same set of primers).
Based on the qPCR, I came up with the following guidelines for my C42 linear PCR:
10X Kappa Buffer – 10 μl (1X)
10 μM F primer (OX10-07) – 4 μl (0.4 μM)
10 μM R primer (OX10-07) – 4 μl (0.4 μM)
10 mM dNTPs – 4 μl (0.4mM)
Kapa taq – 1 μl
2ng/μl Raw library – 1 μl
Water – 76 μl
Yield: 14 ng/μl in 30 μl
14*30=420 ng; Template was 2 ng; 210X amplification; 25 rounds.
Efficiency = (1.238/2)*100 ~ 62%
The 100X reaction gave the same efficiency, in average, per well.
Add fiducial. Check FISH.
Linear for HOPs1. Use 2 ul. If not working well do emulsion PCR.
If working, run 50X T7 PCR for HOPs1 and nonHOPs.
Clean 5 slides.
FISH for STORM.
Channel1: C8. 1:35,000 – Orange Fiducial.
Channel2: C7. 1:35,000 – Orange Fiducial. No good.
Channel3: C13C14. 1:35,000 – Orange Fiducial. Cy3 channel no good.
Channel4: C8 and C7.1:50,000 – Orange Fiducial. Noisy. Cy3 better than channel 3, but still noisy.
Channel5: HOPs and C1.1:50,000 – Orange Fiducial. HOPs didn’t work.
Channel6: HOPs and C4. 1:35,000 – Orange Fiducial. HOPs2 bind both. HOPs1 sometimes one. Very noisy.
Pack for STORM.
10-12: STORM with Talley. Chromatic aberration and 3D-STORM, and 2-reporter imaging.
FISH – C42 HOPs and test PAINT. Also test tracing.
Cynthia finish FISH and image.
Split cells for Cynthia, and help her fix.
Finish FISH and image.
Plan for this week:
1. 10am: Olympus meeting.
2. FISH for STORM.
Channel1: C01 (C1) + HOPs (Uni). Try photo-bleach HOPs after imaging.
Channel2&6: C01 (C1C2C3).
Channel3: C01 (C1C2C3) +C42 (C45).
Channel4: C01 (C1C2C3) +C42 (C43).
Channel5: C01 (C1C2C3) +C42 (C44).
3. PCR35 gel: all samples, 200 ng per sample.
Saw a smear (at least for some samples), so ran T7PCR35 while titrating template concentration: 160 ng, 80 ng, 40 ng, 20 ng, 10 ng.
4. T7 HOPs.
10-12: Jenny’s DC.
Chip42: Used universal bridges in the 647 channel, which showed bright signal.
C45, which should have over 3,000 probes, with over 3 probes/kb density, targeted with P5, didn’t paint. C44, with ~350 probes, ~1.5 probes/kb, imaged with P1, painted in regions, which don’t co-localize , with the 647 channel, and gave dimmed signal.
Next: repeat the FISH with region-specific bridges, to see if co=localizing (at least C44).
Try different conditions for hybing and washing.
MatTeks with nail polish or rubber cement, although preventing the glass from dislodging, seemed not to work, probably as the glue dissolved, the matteks became leaky, and the probe was lost.
LabTek chamber slides – Seems to withhold the temperature. Probe might have evaporated when incubated. Next time: Cover with 2 parafilm tapes, and than add the lid. Place in 42c incubator on Sonny’s bench.
Ordered BB414, BB515 and BB575, for human repetitive regions.
Group meeting – 10 am.
Meeting Diego at 1 pm.
Last time 150 pmole total for hybridizing entire chip (17 Mbp), worked at least for the first compartment. Had some issues with the Mattek dishes, apparently, the bottom glass slide may dislodge at temperatures higher than 45c. Seems like it happens more often with dishes that were stored in the cold room for more than 2 weeks following fixation and permeabilization (which I tend not to use anyway). I pre-sealed the glass with rubber cement or nail polish, and so far the slide didn’t get dislodge. The rubber cement can be removed easily. Also, ordered ibidi flow cells, which should be stable at least until 80c. And, trying some LabTeks on slides.
FISH with new chip:
Trying 64,100 and 150 pmol in total. Also trying 78c denaturation with 100 and 150 pmol in total.
FISH – testing hybridizing of an entire library (C01~17Mbp). 40 μl reaction with a total of 64/100/150 pmol of Oligopaints.
Orange beads require multi-wavelength activation to get the “bleed-through”, which is time-consuming.
Tetraspeck beads (d=100 nm) – 1:500 dilution in PBS+ (PBS + Mg and Ca) is a bit too dense (but still ok). Try 1:1,000.
Fiducial protocol: Wash with 0.2X SSC, aspirate, add ~ 50 ul TS 1:500 beads in PBS+ for 30 seconds. Wash with 0.2X SSC.
Few protocol improvements that allow 75% yield:
- Start T7 with ~1 ug DNA/reaction. Last time I had ~1.3 ug/reaction.
- Use Maxima RT H minus for the RT, and incubate for 2 hours at 50c.
Fiducial – 1:20,000 dilution for the 0.2 orange microfluors was a bit too high. Next time try 1:50,000.
To add fiducials:
- Aspirate cell’s buffer.
- Add fiducials (try 1:50,000) for 30 seconds.
- Wash with 0.2X SSC or 2X SSCT.
DNA-PAINT with Hiroshi and Brian – P13 didn’t show rapid blinking. P7,10 did show, however Hiroshi wasn’t able to co-localize our diffraction limited spots with the PAINT ones. Next time try an imager sequence which we already know work well.
STORM with Alistair – Learned Alistair’s pipeline for imaging and analyzing, and was able to install 3D-DAOSTORM on our windows computer.
Brian is not filling well, so postponed DNA-PAINT imaging for tomorrow. We have to samples: 1) TAD44 2) i/TAD44. Strategy: use 2nd sample, image TAD44, rehybe (use 37c solution) with toeholds, bridge and secondaries for iTAD44 and image again.
Next week – amplify HOPS with Jenny, image in our scope.
- HOPS1 entire library and TAD44 and TAD17.
- HOPS2 entire library and TAD44 and TAD17.
- HOPS1 and HOPS2, TAD44 and TAD17.
Fix overnight cells in MatTek dishes.
FISH for 3D-STORM with Alistair. Samples:
FISH with Brian and Hiroshi.
12/10/15 – 1/4/2016
Gave a talk at the Israel Physics Society 61st meeting in Bar Ilan University, Ramat-Gan
Made 4 35 mm MatTek dishes with 7 mm imaging well (glass bottom #.0) to be fixed today, to compare auto fluorescence/noise with MatTek dishes where cells were grown over for a few days. Used a 1:5 dilution and seeded 100 ul to the imaging well. Didn’t use Poly-L-Lysine to coat the dishes. At the same time seeded 8 plates with 1:40 dilution.
Optimizing FISH in MatTek protocol:
- Try lay MatTek dishes in coplin jars and fill jars with water up to the MatTek dish.
- Don’t dry the imaging well completely while washing.
- 2nd hybridization (bridges and secondaries) worked without using a coverslip to press the hyb mix.
Aim: Optimizing FISH MatTek protocol (for future STORM)
Experiment details: 35 mm MatTek dish with d=10 mm for the imaging well. Overnight hybridization at 42c, 100 ul hybe mix.
FISH in MatTek dish:
Got the alpha-sats to work, although need to adopt a more gentle way to lay a coverslip over the dish and avoid a bubble. Suggestion: use a smaller coverslip and/or tweezers that can grab the coverslip in a horizontal alignment so I can gently place it over the imaging well.
T7 method to amplify probes:
- Start with 4 DNA PCR reactions and collapse them into 22 ul, which is enough for 2X T7 reaction plus 1 ul for specing.
- Following overnight T7 reaction, split into two tubes and do RT on both.
- Digest RNA.
- Do probe cleanup with vacuum and collapse into one tube; elute in 200 ul UPW.
- Speedvac and resuspend to 40 pmol/ul.
Lower denaturation temp didn’t work.
1 pmol/25 ul bridges worked better than 10 pmol/25 ul bridges. More cells showed spots and the SNR was better. 42L (less stringent) with 10 pmol bridges didn’t work at all.
Now try in MatTek dish with 1 pmol bridges: Instead of the 40 pmol Oligopaints concentration i use for slides, I used 100 pmol in 100 ul hyb buffer volume.
Need to amplify 44TAD – 32-mers and 42L.
A. FISH with 32-mers, 42N and 42L (@ 40 pmol Oligopaints concentration):
- 32-mers, 10 pmol Oligobridges concentration, 37c hyb.
- 32-mers, 1 pmol Oligobridges concentration, 37c hyb.
- 42L, 10 pmol, 42c hyb.
- 42L, 10 pmol, 47c hyb.
- 42L, 1 pmol, 42c hyb.
- 42L, 1 pmol, 47c hyb.
- 42N, 10 pmol, 42c hyb.
- 42N, 10 pmol, 47c hyb.
- 42N, 1 pmol, 42c hyb.
- 42N, 1 pmol, 47c hyb.
- 42N, 10 pmol, denaturation: 72c for 20 minutes, 47c hyb.
- 42N, 10 pmol, denaturation: 72c for 20 minutes, 42c hyb.
- 42N, 10 pmol, denaturation: 78c for 3 minutes, 42c hyb.
A total of 2 hybs for the 32-mers and 4 hybs for the 42-mers.
B. Freeze PGP1F cells and seed into MatTek 35mm dishes. Now in -80c.
Meeting with Brian, Hiroshi and Sonny
- Use 42/47c for hybridization of 42-mers.
- Lower bridge concentration.
- Hybridize bridge and secondaries at the same time.
- Email Brian and Hiroshi the highest density Oligopaints for TAD44.
- DNA-PAINT: Alpha-sats at 42/47c and TAD44. LabTak 2 (#1.5 glass, 8 cells, use the inner ones). 150-200 ul per hyb. Don’t leave it for more than 2 days.
- Brian denaturate @ 78c.
Prepared freezing media to freeze 21 vials of PGP1F from 7 100 mm dishes.
Seeded 8 35 mm dishes with IMR90 cells for super-resolution imaging.
Writing a script for spot analysis (MATLAB).
Made more 42 masked to N (42N) Oligopaints for TAD44 and stored them at 40 pmole/ul in a total of 27.7 ul.
Compared 42L (less stringent, first 4 images) and 42N (more stringent, last 3 images). 42N gave a more clear result with higher efficiency.
Received PGP1F from Coriell in flasks and put them in 37c after viewing on the scope.
Oligobridges seems to work, although optimization is required and since the complete library I have now is not continues (different regions in the chromosome), I should amplify one spot to get clear spots. Therefore, chose to amplify TAD44 and increase probe concentration.
Had some problems with amplifying probes, DNA binding buffer or wash buffer were contaminated. Qiagen and Zymo kit gave similar results when eluted in 30 ul UPW.
Probes yield was low, 10-13 pmole/ul in 10 ul total. Most likely T7 3 hours reaction is not enough. Will do overnight.
Oligopaints library testing:
a. Oligopaints oligos testing: 32-meres/42 less stringent/42 more stringent (masked to N)
b. Visualization methods: direct labeling/spot-specific secondaries/spot-specific bridges and universal secondaries.
Added: Fixation and permeabilization protocol (thanks Sonny).
Testing TADs chip:
Aim 1: Validate the chip (32 and 42N mers).
Aim 2: Test the bridge-oligo method.
Method: Sonny amplified 2 neighboring TADs and their iTADs (4 sequential regions in total) through the universals of HOPS1, HOPS2 and non-HOPS for the 32 and 42N mers. He then seeded IMR90 cells, fixed and permeabilized them. I hybridized them with bridge-oligos (2nd hyb) and their complementary secondaries.
Followed Interphase FISH protocol. Hybridized overnight at 47c, which might prove to be to stringent, especially to the 32 oligos. Washed with 2X SSCT at 60c for 15 min (step 10), placed the slides tilted to gently dry excess wash buffer while setting up the hyb reaction, wiped slides in the areas surrounding the cells (not where the cells are!) and rehybed for 30 minutes with bridge-oligos and their fluorescent secondaries.
Noisy and we hardly saw any HOPS signal.
- Check HOPS density at this region. – Density is low. Oops – HOPs were designed to PGP1, but we used IMR90.
- Hybridize bridges first, wash and then add secondaries.
- Hybridize at 37c.
UCEs: As their name implies, ultraconserved elements (UCEs) are highly conserved regions of organismal genomes shared among evolutionary distant taxa – for instance, birds share many UCEs with humans. For further information regarding UCEs please refer to ultraconserved website or contact either Ruth McCole or Jelena Erceg.
The figure below was imaged with Olympus IX83 inverted microscope in Ting Wu lab. The probe design that spans ~300 kb and contains ~2,500 Oligopaints oligos was carried by Roxana Tarnita, Ruth McCole and Brian Beliveau (thanks!). Son C. Nguyen instructed me with probe production (thanks Sonny). The UCEs themselves span ~0.5% of the entire imaged region. The entire probe is on Chromosome 10 within a single TAD (thanks Jelena for the bioinformatics!).
1. Analyze spots distribution (size, distance, etc.)
2. Image with STORM.
3. Down-size probe region, add colors (UCEs and close environment vs. a more distanced flanking region), add control, probe more UCEs simultaneously.
Oligopaints occupancy suggestions:
1. Use zipper probes. A color for each strand (so 2 colors) to allocate genomic regions where one strand is blocked.
2. Using the zipper approach, rehybe without denaturation (to avoid probes cross-talk).
3. Bioinformatics: Use a sliding window approach for targets that were disqualified using the binning approach.
Spoke to Caroline about the cell number (and she looked at the microscope), she says the issue is the IMR90 morphology, where they shrink upon trypsin treatment. Therefore, the small diameter which I gated out is actually reasonable.
Expanded 1:6 the Hu iPS cells I received from Renata (Church lab, thanks!) on may 18 (1st passage).
Seeded the IMR90 for FISH. 1 6-well of IMR90 is not enough for FISH. I had to resuspend in 0.5 ml. Cells amount was <3e5.
Imaging analysis: I fixed the python path but there are a few bugs because Alistair uses windows.
Image Analysis – Need to define Python path (version 2.7) for Alistair’s and Jeff’s STORM package to work properly.
Chr6_1500 probe – Prepared using the lambda-exo method. very low yield (2 pmole/ul). Have to try again and measure concentration after every step.
UCE137 probe – Prepared using Sonny’s RT method. A slight variation for not digesting the dye was made. Apparently nanodrop device is only valid for dyes with Absorption maxima over 450 nm. Therefore it is not sufficient for ALEXA405. Probe was validated under gel.
04/01/15 – 04/12/15 – Vacation.
Meetings at BIU (Israel).
Met with Estelle – Discussed guidelines for fellowship applications.
Met with Yuval, Ronen and Anat regarding papers:
a. Anat should go over some of the data she produced and see if she could re-analyze it or perform more measurements.
b. Ronen will finalize the analysis and when Anat sends more data will finish writing the paper.
c. Guy and Anat will write the experimental section once Ronen will send them the part that he wrote.
a. Ronen will perform analysis or finalize the equations and model and will send to Guy to analyze.
b. Guy will write the paper, will send to Ronen to go over it and then to Yuval.
Next imaging session: Start recording only after molecules have gone to the dark state. Otherwise the bin list is enormous.
1:1000 gold nano particles concentration (for fiducials) was too low.
Dax converter for nikon is working.
Downloaded Xiaowei Zhuang’s lab matlab package. For proper use with the mlist function discard initial frames showing initial switching to dark state (saturation…)