PCR of GeCKO gDNA for deep sequencing

[jde...@gmail.com]

Hi Neville and other GeCKO members,

Thanks for helps in this group. I have problem to get my PCR done with genomic DNA(gDNA) of GeCKO library-infected cells. I optimized my PCR with gDNA of baseline cells and then I decided just to do PCR2 to get sgRNA recovery (340bp).   I did PCR2 (primers: F2_01~09 (mixture of different staggers) and R2 (with different barcode) with gDNA or plasmid as template) or PCR1 (F1 and R1 with gDNA or plasmid as template). So I have 5 samples for PCR test, 1) gDNA of baseline cells, 2) gDNA of DMSO-treated cells, 3) gDNA of drug-treated cells, 4) GeCKO plasmid (positive control) and 5) negative control (no template). I tested different conditions,in which 1) gave a nice 340 bp band with same size of 4), but couldn't get bands for 2) and 3). I used NEB phusion enzyme, added gDNA according to 10 ug/100 ul reaction and PCR for 24 cycles. Cells are H358 cells with cas9 infected with GeCKO A and B and selected with puro (1-2 ug/ml) for 10 days (30-40% survival), at this point called baseline cells, and then did drug screening for 12 days, with splitting every 2 days and DMSO treated as control. 

I wonder if sgRNA were removed during later 12 days drug screening since I didn't include puro again. But this usually should not be, because once puro selection was done and stable cell should keep it. I also tried PCR1 and then PCR2, but gave me a non-specific bands in negative control. Maybe I should decease cycles of PCR1, but in PCR1 sample 2) and 3) in the above didn't show an amplification over 5). Do you have any suggestions?

Thanks a lot,

David  

[Neville Sanjana]

Hi David,

Sorry to hear about your troubles with sgRNA amplification from gDNA. For your GeCKO library, are you using lentiCRISPRv2 (one-vector system) or lentiGuide-Puro (two-vector system)? 

Best wishes,

Neville

--
You received this message because you are subscribed to the Google Groups "Genome Engineering using CRISPR/Cas Systems" group.
To unsubscribe from this group and stop receiving emails from it, send an email to crispr+un...@googlegroups.com.
To post to this group, send email to cri...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[jde...@gmail.com]

Hi Neville,

Thanks for helps. I used two-plasmid system. So I first made clonal cas9 (checked by flag blot) cells with blasticidin. 

David

[Neville Sanjana]

For lentiGuide-Puro (2-vector system) readout from gDNA, we first PCR with the PCR1 primers from the Science (2014) paper:

F1 AATGGACTATCATATGCTTACCGTAACTTGAAAGTATTTCG

R1 CTTTAGTTTGTATGTCTGTTGCTATTATGTCTACTATTCTTTCC 

This should yield a 312 bp amplicon. 

After PCR#2 using our primers (see GeCKO website for a Excel spreadsheet with 12 forward and 12 reverse barcoded primers), the intended amplicon should be ~370 bp. You can alternatively do just one PCR using the PCR#2 primers (for lentiGuide-Puro-based libraries) but, since these longer primers are more expensive (e.g. IDT Ultramers), we tend to do the PCR in two steps. If cost is not an issue, feel free to combine the PCR steps (while making sure to use enough gDNA to keep full screen representation). This can result in a cleaner PCR with less spurious bands. Others have also reporter that gel extraction of the PCR#1 product can reduce spurious bands after PCR#2, but I have not personally tried this.

I'd also recommend varying the amount of input gDNA. For example, we found that Phusion Flash did not work well with 10 ug of gDNA per 100 ul reaction but Herc2 and Taq were OK. Perhaps you can use the gDNA from cells 1-3 days after infection with the GeCKO library to optimize your 2-step (or 1-step) PCR readout and make sure you are seeing the correct band sizes.

Hope that helps,

Neville

[Multi Nucleated]

If I just want to check for good representation (no experimental samples), do I need all 12 barcodes or can I just use 1?  Or if I want to check both initial representation and biological samples at the same time, can I use 1 barcode for initial pooled library plus additional unique N barcodes for N biological samples in the same HiSeq run?

(Happy new year!)

[Neville Sanjana]

Hi MN,

 

If I just want to check for good representation (no experimental samples), do I need all 12 barcodes or can I just use 1?  

With a single library, you can just use a single barcode but the problem is with sequence diversity when you run MiSeq/HiSeq/NextSeq (monotemplate). So, if you sequence just 1 sample for the entire run, we recommend you mix 4 or more forward primers (i.e. 4 different staggers) to avoid monotemplate.

 

Or if I want to check both initial representation and biological samples at the same time, can I use 1 barcode for initial pooled library plus additional unique N barcodes for N biological samples in the same HiSeq run?

Yes, this will work fine and avoid monotemplate as long as N is greater than 3 or 4.

 

(Happy new year!)

Same to you. Best wishes for a successful screen!

Best,

Neville

[Ben Boward]

Hi all,

I have a question that is an extension of this conversation, that I was hoping someone could clear up. PCR 1 is divided into multiple reactions (10ug each). Should the PCR 1 reactions be combined before using 5uL of it for PCR 2, or should I be doing the same number of PCR 2 reactions as PCR1 reactions, using 5uL of each PCR 1 for each PCR 2 reaction?

Thank you for being so helpful!

Ben

[Neville Sanjana]

Hi Ben,

After PCR1, we combine the reactions and then use the mixture as template for PCR2 (where barcodes for individual samples are added). We have followed previous pooled screening recommendations where one PCR2 reaction is done for every 10K constructs in the library. 

For example, if you need 10 PCR1 reactions to keep full representation and you are using a library with 80K constructs, we would pool the 10 PCR1 reactions and then do 8 PCR2 reactions using the pooled PCR product. There may be better approaches to this but this has worked for us.

Hope that helps,

Neville

[jde...@gmail.com]

Hi Neville,

I ordered Herc2 and look like the result is better. Maybe it's right NEB phusion doesn't have good results with high content of genomic DNA. 

However, I am a little confused about tube numbers for PCR2, following Ben's post. As you said, you pooled PCR1 and performed PCR2 for reactions/10k library, such as 11 PCR2 reactions/110k GeCKO A and B, is this correct? Since there're still a lot of tubes for PCR2 it would not save many F2 and R2 primers for PCR1+2 vs PCR2 only. So what's the point with PCR1+2, maybe for better amplification and specificity? I'm not arguing this, just want to understand this better.

Thanks a lot for helps.

David

[Neville Sanjana]

Hi David,

Either way can work.... if it makes sense for your particular case (representation & library size), it might be more economical to do 1-step PCR. (It certainly is simpler.) For higher representation screens (e.g. 1000x coverage), it might be better to do 2-step. Also, if you are using libraries cloned into different vectors, then 2-step can allow you to re-use the same PCR2 primers for readout from different vectors. 

There is definitely a lot of room in the PCR/primer design for innovation and improvement!

Hope that helps,

Neville

[jde...@gmail.com]

I get it, Thank you. Updating my PCR, with Herc2 polymerase I perform 1-step PCR and they all works well, just a few of samples need titrate genomic DNA amount.  Thank you very much for your helps!

David

[Shirley]

Neville,

What if I am using lentiCRISPRv2 (one-vector system)? It has been a bit frustrating for me to identify the correct band after 2-step PCR. I am sending library to check presentation first and I followed your Science paper and using the same primers. Shall I do additional steps to add adaptor priming sites for lentiCRISPRv2?

Thanks.

Shirley

On Thursday, December 31, 2015 at 7:16:40 AM UTC-5, Neville Sanjana wrote:

[Neville Sanjana]

Hi Shirley,

Yes, the PCR can sometimes be a bit challenging, but the dominant band should be the right one. Perhaps you can double-check your gels before sending for deep sequencing? Using our primers (v2_Adaptor F/R), your PCR#1 product (on lentiCRISPRv2) should be 283 bp. And, using our PCR#2 (barcoded Illumina) primers, your PCR#2 product on lentiCRISPRv2 should be 360-370 bp.

All primers sequences can be found here (see item 3): http://genome-engineering.org/gecko/?page_id=15 

best wishes,

Neville

[мария фомичева]

Hi Neville and other GeCKO users,

Thank you so much for this forum! It really helped to get to the point where I am now:)

I am using one vector mouse GeCKO v2.0 system. Currently I am running PCRs. PCR#1 worked beautifully, but PCR#2 looks a little worrying: I got the expected band (370 bp), but there is some smear around it (see the attached picture). Do you think it is OK and I can cut out my bands for gel purification and sequencing? Or I need to change my PCR conditions to get rid of the smear? 

Thank you so much,

Maria

P.S.: PCR conditions are

98C for 3 min

24 cycles of

98C for 15 s

I tried annealing T 65-70C for 30s - I see the smear for all of them

72C for 30s

Then, 72C for 2 minutes.

Reagents	1 reaction
DNA	5 µL
Illumina F (10 uM) (I used one primer set for the pcr test)	5 µL
Illumina R (10 uM)	5 µL
Herculase II	1 µL
5x Herc. Buffer	20 µL
dNTP (10mM)	2µL
Water	Up to 100µL

[jizha...@gmail.com]

Hi, Neville

I have a similar problem. I can't get the PCR1 work for a specific condition (fortunately, it is a untreated control). I have tried to prepare the gDNA several times, it still does not work for this condition specifically. I used Herculase II, tried various DMSO, Tm, primer concentration, just no any band. Really no clue why?

Any suggestion for trouble shooting, or why there is a big difference between different samples for PCR1?

Thanks

Ji

[Neville Sanjana]

Hi Maria,

Your gel looks quite similar to what mine look like. I would do the equimolar combination of PCR#2s (usually I use gel quantification) and then gel extract the final mixture product. If you do a gel extraction of the correct size band, it should sequence well.

Best wishes,

Neville

--

[Neville Sanjana]

Hi Ji,

For PCR1, I typically only amplify for 15-20 cycles and usually the band is not visible. Normally, I proceed directly to PCR2 and, only if I don’t see a band then, would I be worried. Can you let us know what happens after PCR2?

Hope that helps,

Neville

--

[Julian Spagnuolo]

Hi Neville,

I have a question on primer design for the 1st round PCR to amplify the gRNA regions from extracted gDNA - instead of using staggered primer sequences to increase library diversity (and avoid problems with cluster calling) would it also be possible to just include a degenerate sequence in between the illumina adapter and the primer-binding-sequence?

e.g. 5'-[illumina seq read1/2]-[N6]-[primer-binding-seq]-3'

The purpose is to simplify the 1st round PCR (only 1 primer pair required) and increase through-put for screens that require a large number of replicates or conditions and therefore a large number of sequencing libraries. My understanding from the illumina literature and discussion on seqanswers is that low library complexity is only a problem in the first 6 or so base-calls which are required for cluster identification on the illumina platforms.

Thanks in advance,

Julian

[Omar aa]

I see the  same thing after the first PCR but when I run the second PCR for the same smaples I got almost same intensity 

On Thursday, October 27, 2016 at 6:14:48 AM UTC-6, Bengül Çetin wrote:

Hi Neville,

Firstly, I would like to thank you for this interactive group with beneficial feedbacks.

I need a suggestion about nested PCR. I can see the bands after PCR2 with the right size, but band intensities vary highly among samples for the same library probably affecting gRNAs representation in sequencing  as well. Is there a way to get around this situation? Also, I was not able to find a detailed PCR protocol of yours from Science paper.

Any help would be appreciated.

Kind regards,
Bengül

[rs...@georgetown.edu]

Hi all, 

I tried to do PCR1 twice and then visualize the bands by 1.5% agarose gel but it did not work. I used the same PCR conditions in the 2 times but I did 18 cycles the first time and then 24 cycles in the second time. However, in both of them I only see 2 very large bands (more than 10kb), please see the picture attached. I really have no clue. 

Here is the protocol we used: 

PCR#1                                                                                                 Cycling Conditions

Primer F (v2Adaptor_F): 2.5uL (10uM stock)                                     95deg: 2min

Primer R (v2Adaptor_R):2.5uL (10uM stock)                                     98deg: 1 sec   

dNTP: 1uL                                                                                            60deg: 5 sec    18 or 24 cycles

Buffer (5x): 20uL                                                                                  72deg: 35sec

gDNA:  Variable volume (10ug)                                                           72deg: 60sec 

Herc2 polymerase: 1uL

Water: Variable volume. Use enough to bring final volume to 100uL.  

F-Primer: AATGGACTATCATATGCTTACCGTAACTTGAAAGTATTTCG

R:Primer:TCTACTATTCTTTCCCCTGCACTGTtgtgggcgatgtgcgctctg

To me, it seems like the primers did not bind at all and thus the band I have is representing the whole plasmid? However, we are using the same primers you all mentioned. 

Any suggestions/explanations? Should I go ahead and do PCR2 directly even if I do not see the right band in PCR1? 

Thanks,

Reham Ajina 

On Wednesday, December 30, 2015 at 4:24:16 PM UTC-5, jde...@gmail.com wrote:

[Neville Sanjana]

Dear Reham,

I recommend just moving from PCR#1 to PCR#2. We usually do not visualize on gel until after PCR#2 (and then look for specific band size). The primers in PCR#1 do bind the vector and will (with more cycles that the ~20cycles that we recommend) produce a band of size ~280bp. The PCR#2 product (which you should read out on gel) should be ~370bp.

Hope that helps,

Neville

--

[Haiyin Liu]

Dear forum users,

I'm having a problem with my PCR#2 amplicons for NGS. Even though my "pilot PCR" (spot testing few random samples) produced beautiful clear bands, my full scale PCR#2 has many smaller unspecific bands (primer-dimers?).

Reaction:

10ul 10X buffer

0.25ul HotStart Taq 

5ul FOR (10uM)

5ul REV (10uM)

2ul dNTP (10mM)

5ul template (PCR1 product)

72.75ul H2O

95’C 1 min
--

(24x)
95’C 15 sec
60’C 30 sec
72’C 30 sec
--

72’C 7 min

I've (unsuccessfully) tried repeating it and reducing primer amount and increasing template amount. Does anyone have any advice? 

Best regards,

Haiyin

[Marta Kovatcheva]

Hi All,

Sorry to revive such an old topic but I was hoping someone could help me out. I am interested in doing a one-step PCR instead of two, but I was curious about the number of cycles necessary. Is the idea to use the same amount of gDNA (ex. sufficent gDNA for 300X coverage, spread into 10ug reactions) but just use PCR2 primers straight away? In this case, since you are not doing the 16-20 cycles of PCR1, how many cycles would you recommend for PCR2? I am worried about sufficient amplification vs. PCR bias.

Thanks for any insight!!!

Best

Marta

[Julia Joung]

Hi Marta,

Please see our recent Nature Protocol paper for screening that you might find helpful. We describe 1-step PCR protocol with 22 cycles using 2.5ug of gDNA per reaction. Typically I use up all of the gDNA harvested from cells at 500x coverage.

Best,

Julia

--

You received this message because you are subscribed to the Google Groups "Genome Engineering using CRISPR/Cas Systems" group.

To unsubscribe from this group and stop receiving emails from it, send an email to crispr+unsubscribe@googlegroups.com.