PCR condition to prepare amplicon library for NGS of GeCKO v2 library

[Hiro]

Hi Neville,

Thanks for helping scientists to use epoch-making whole-genome CRISPR/Cas GeCKO library.
I have one question regarding the amplicon library preparation for Illumina sequencer.
I successfully finished GeCKO v2 library amplification, viral package, transduce them into cells, put selective pressure on them in vitro and extracted their gDNA.
So now is the time to prepare amplicon library and do NGS.
Thanks to your support to upload necessary primer sequences in your website, I already succeeded 2 step PCRs using v2Adaptor primer for 1st and F01-F12 primers for 2nd to amplify GeCKO v2 library plasmid.
I also sequenced them using MiSeq and confirmed that complexity of obtained Gecko v2 library was maintained.
At that time, I used 10ng of plasmid and did PCR at 16 cycles each of 1st and 2nd PCRs. Finally, I gel purified and obtained about 0.8ug of amplicon. All these process seems to work well.
However, I never tried these PCRs using 10ug of gDNA containing Gecko library as you did in the paper.
So I may need to optimize PCR condition again.
But, if I need to evaluate PCR condition each time by actually sequencing amplicon library using illumina Hiseq2500, it costs too much and also waste lots of my samples (10ug of gDNA).
So my question is how I can optimize the PCR condition in order to generate amplicon library from gDNA samples to the level suitable for NGS.
I know that I can obtain more than enough amount of amplicon if I set very high number of PCR cycles.
But I'm specifically afraid to make amplicon too much saturated by unnecessary higher PCR cycles.
So I appreciate if you tell me the exact number of cycles of 2 PCRs and their annealing temperatures that you use, because the condition described in the Science paper is GeCKO v1 library so you may change PCR condition for v2 library.
Do you also have some idea how much amplicon from 10ug of gDNA is ideal for NGS? Then, I may be able to just quantify the amplicons from different number of PCR cycles in order to optimize the condition.

Sorry for long e-mail,
Takahiro

[Jian Huang]

Hi Hiro,

Did yo use the PCR#1 and PCR#2 primers in the supplemental of Science paper?

PCR#1: F1 AATGGACTATCATATGCTTACCGTAACTTGAAAGTATTTCG

R1 CTTTAGTTTGTATGTCTGTTGCTATTATGTCTACTATTCTTTCC

[Neville Sanjana]

Hi Takahiro,

Using Herculase2, a good number as a starting point for optimization is 20 cycles for PCR#1 and 20 cycles for PCR#2. We typically use an annealing temperature of 60C. You can also try this with another polymerase but the numbers might be slightly different for cycles and annealing temperature.

For NGS, we do want to avoid overamplification, so it's best to do the least number of cycles possible. If you find you get very strong bands with the protocol I've given, then you might want to try lowering the number of cycles in both PCRs. Typically for Illumina sequencing you actually need a very small amount of DNA to sequence but we typically start with more so that we can quantify, pool samples, etc.

best,

- 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.
For more options, visit https://groups.google.com/d/optout.

[Jian Huang]

Neville,

Thanks for all the information.

I am still a bit confused about the PCR#1 and PCR#2. Based on the map of the vector, PCR#1 amplify region 2479 -2815 and PCR#2 amplify 2469-2857. I thought PCR# 2 should be within the region of  PCR# 1. Why #2 is a little bit outside of #1? 

The new reverse primer for PCR#1 discussed in another topic is the combination of primers target 4878-4897 and 2469-2493.I don't really understand why it is better. How does it work?

Thanks very much!

[Neville Sanjana]

Hey Jian... PCR#2 primers definitely anneal inside of PCR#1 amplicon. For LCv2, the PCR#1 primer adds a PCR#2 annealing site on the amplicon and this might be the source of confusion. Take a careful look at the PCR#1 primers for LCv2. :)

- Neville

[Hiro]

Hi Neville,

Thanks for your information.
This helps a lot!

Let me ask you how to determine "very strong bands".
Is it determined by the final PCR product (PCR#2)? or do you even consider it just after the first PCR (PCR#1)?
To gel purify the amplicon of PCR#2 for NGS, I may load 100ul PCR#2 reaction into a well (or split into 2 wells?).
So do you mean that it is overamplification if you see very strong bands from this well?
Or do you load a portion of PCR samples?
I'm sorry to ask you a basic question but I think that the band intensity may vary depend on how much sample I load in a well.
It is also very helpful if you remember the quantity of PCR amplicon from each reaction (or pool you did in Science).

To Jian,

Neville already answered your question but it was also discussed in another thread in this forum.
https://groups.google.com/forum/#!searchin/crispr/sequence$20v2$20gecko/crispr/U9mHAtXD3dU/iwFwLRmsfLAJ
You can also refer the PCR primers that Neville uploaded his GeCKO website.

Thanks,
Hiro

2014年11月5日水曜日 21時16分04秒 UTC-6 Neville Sanjana:

[Neville Sanjana]

Thanks for your information.
This helps a lot!

Happy to help.

 

Let me ask you how to determine "very strong bands".
Is it determined by the final PCR product (PCR#2)? or do you even consider it just after the first PCR (PCR#1)?

PCR#2 on gel.... the ideal situation is a band that is just visible (so you can quantify the PCR product for pooling with other conditions). But it is OK to have a somewhat dark band.... just don't run it many cycles past that point, which will create some amplification bias.

 

To gel purify the amplicon of PCR#2 for NGS, I may load 100ul PCR#2 reaction into a well (or split into 2 wells?).

I combine all my PCR#2 products unpurified first to pool samples and then gel purify the pooled product.

 

So do you mean that it is overamplification if you see very strong bands from this well?
Or do you load a portion of PCR samples?

I look at every individual PCR#2 to check for overamplification, which is necessary if you're using gel quantification for pooling.

 

I'm sorry to ask you a basic question but I think that the band intensity may vary depend on how much sample I load in a well.
It is also very helpful if you remember the quantity of PCR amplicon from each reaction (or pool you did in Science).

Not sure what you mean here.... this will vary for each experiment depending on how you pool samples. In our case, we pooled experimental conditions to have equal representation of each condition (sample barcode) in the sequencing mix. 

- Neville

[Jian Huang]

Neville and Hiro,
Thanks very much to both for answering my questions.
I definitely read the other thread and to be honest, it was very confusing because someone said they got 2123 bp and Hiro said that he got 340bp, which makes more sense to me.

Anyway just to be clear, the primers for PCR#1 is:

Forward PCR#1 primer for lentiCRISPRv2: aatggactatcatatgcttaccgtaacttgaaagtatttcg

Reverse PCR#1 primer for lentiCRISPRv2: TCTACTATTCTTTCCCCTGCACTGTtgtgggcgatgtgcgctctg

And the primers for PCR#2 is uploaded at GeCKO website.

Am I correct?

Thanks very much again!

[Hiro]

Hi Neville,

Thanks again for your quick and detailed answers.
I may misunderstand about pooling of amplicons from different experimental conditions (different barcodes).
I thought that I should gel purify PCR products of each condition separately, check their concentration individually by nano drop and finally pool them at the same concentration for NGS.
But you seem to pool them together before gel purification.
So in this case, you run the small portion of each PCR product in the gel and quantify them individually based on the band intensity.
Then, you pool these products based on the band intensity, run all pooled products in the gel and purify them for NGS.
Do I understand your way correctly?
I guess that in this way you may be able to reduce the number of samples in the gel purification process, right?

Thanks,
Hiro

 

2014年11月6日木曜日 10時16分34秒 UTC-6 Neville Sanjana:

[Hiro]

Hi Jian,

I use the primers that you wrote.
Final amplicon size is around 350bp (vary dependent on stagger sequence of primers).

Thanks,
Hiro

2014年11月6日木曜日 10時54分24秒 UTC-6 Jian Huang:

[Neville Sanjana]

Hi Hiro... Yes, that's correct. Just one gel extraction after pooling based on band intensity.

- Neville

--

[Hiro]

Hi Neville,

Sorry for bothering you again but could you look at the gel image of PCR amplicon attached?
I followed the PCR protocol you told me (20 cycles, Annealing Tm 60C, Herculase2 polymerase) and tried to amplify 5ug of gDNA in one tube (100uL reaction). Your protocol works pretty well. This time, I didn't see any non-specific band.
Gel image shows 5uL and 15uL of PCR amplicon. Do you think that this is overamplification or appropriate?

Thanks,
Hiro



2014年11月6日木曜日 13時59分38秒 UTC-6 Neville Sanjana:

[Neville Sanjana]

Looks good to me! (I'm assuming that this is a 1kb+ ladder and that this band is at 350bp.)

Good luck with sequencing,

- Neville

[Hiro]

Ni Neville,

Thank you so much for your help.

Hiro


2014年11月7日金曜日 0時23分50秒 UTC-6 Neville Sanjana:

[Hiro]

Hi Neville,

New questions come up to me about NGS condition using Illumina sequencer and the PCR step of amplicon library preparation.
How much concentration of library do you use for HiSeq2500 sequence?also how much PhIX spike do you use?
Last time when I sequenced Gecko library plasmid using MiSeq, I used 2nM of amplicon library and it seemed to work.
But I used 50% PhIX and wasted lots of read. So I want to know the appropriate condition.

About the preparation of amplicon library by 2 step PCR (#1 and #2), I believe that one experimental sample needs 13 PCR reactions (10ug gDNA each) in order to maintain library complexity at x300 as you did in Science paper. But do you also perform 2nd PCR (#2) of these 13 reactions, individually (in individual 13 tubes)? or after 1st PCR, do you combine 13 reactions, take 5 ul of PCR products and perform 2nd PCR in 1 tube?
If you perform the 2-step PCR amplification of 13 tubes individually, total post PCR amplicon volume may be 1.3ml.
In this case, do you just mix them, take a portion of them and pool them into other experimental samples and do gel purification?

I'm bothering you many times... but you suggestion and answer are invaluable to me.

Thanks
Hiro

2014年11月9日日曜日 11時17分04秒 UTC-6 Hiro:

[Neville Sanjana]

Hi Hiro,

How much concentration of library do you use for HiSeq2500 sequence?also how much PhIX spike do you use?

For HiSeq, our sequencing core determines the concentration. I deliver it at 10nM and then they dilute. I know they also spike-in PhiX but not sure how much. 

 

Last time when I sequenced Gecko library plasmid using MiSeq, I used 2nM of amplicon library and it seemed to work.
But I used 50% PhIX and wasted lots of read. So I want to know the appropriate condition.

For MiSeq, I use a final library concentration of 14pM with 10% PhiX spike-in.

 

About the preparation of amplicon library by 2 step PCR (#1 and #2), I believe that one experimental sample needs 13 PCR reactions (10ug gDNA each) in order to maintain library complexity at x300 as you did in Science paper. But do you also perform 2nd PCR (#2) of these 13 reactions, individually (in individual 13 tubes)? or after 1st PCR, do you combine 13 reactions, take 5 ul of PCR products and perform 2nd PCR in 1 tube?

Combine the PCR#1 and then do PCR#2 from the combined products. We normally do one PCR#2 reaction for every 10,000 constructs in the library but I have not rigorously tested if this is necessary. This is from a shRNA pooled screening protocol and we followed it.

 

If you perform the 2-step PCR amplification of 13 tubes individually, total post PCR amplicon volume may be 1.3ml.
In this case, do you just mix them, take a portion of them and pool them into other experimental samples and do gel purification?

That's right. We gel normalize the unpurified products, mix, and then gel purify the mix.

 

I'm bothering you many times... but you suggestion and answer are invaluable to me.

No problem.... glad I can help!

- Neville 

[Hiro]

Hi Neville,

Thanks a lot!
Hopefully, I will be able to obtain my screening data real soon.

Thanks,
Hiro

2014年11月10日月曜日 22時24分49秒 UTC-6 Neville Sanjana:

[Chao Wu]

Hi, Neville and Hiro,

I'm doing the gDNA deep sequencing PCR at the moment, and got lots of trouble, the most important is, the gDNA PCR with the 1st set of primers:

 F1 AATGGACTATCATATGCTTACCGTAACTTGAAAGTATTTCG

R1 CTTTAGTTTGTATGTCTGTTGCTATTATGTCTACTATTCTTTCC

I used Tm 60Co, Herculase2 DNA polymerase,40uL PCR reaction with different amount of gDNA as marked, and loaded everything, with 2ug DNA, I can see a band around 340bp, which is not very sharp, I used 1.2% gel, I'm not sure whether this band is good enough so that I can proceed the 2nd PCR with less cycle of the 1st PCR, please take a look at the image and give some suggestions, thank you very much!

And about the 2nd PCR, I have another question, do you purify the 1st PCR with PCR clean kit or something else then do the 2nd PCR, or you have another way to do this?

Thanks a lot! I've been stuck here for quite a while and can't move on to next step, any suggestions are appreciated!

Chao

[Klaus]

Hi Hiro,

can you share your PCR protocol? I am not sure how did you mix the F01-F12 primers, and how much you use for the PCR reaction.

thanks

在 2014年11月4日星期二 UTC-5下午4:53:18，Hiro写道：

[NJ CRISPR]

Hi Neville,

  I just got my sequencing results of GeCKO v1 library transducted into cells.

  I got 47000 sgRNA aligned to reference, which finally aligned to 17500 genes. It seems that the representation of sgRNA is not ideal enough.

  As you have mentioned that you normally do one PCR#2 reaction for every 10,000 constructs in the library. In my work, I just did 4 PCR#2 reaction and nearly 4/6 library was detected. So does this mean that more PCR#2 reaction is needed? That is to say, for 64751 sgRNA, 6~7 PCR#2 reaction will be OK?

  Hope for your help.

Sharp

在 2014年11月11日星期二 UTC+8下午12:24:49，Neville Sanjana写道：

[Neville Sanjana]

It's difficult to know where the loss of representation is happening. Did you sequence the plasmid used to make the virus and verify the full library representation? 

If the plasmid had close to full representation, then the loss of representation could be biological (e.g. dropout of essential genes) or it might occur during PCR#1 or PCR#2.

Hope that helps,

Neville

[Neville Sanjana]

No problem.... we did 7 PCR#2 reactions for each biological sample (i.e. following the same rule you mentioned in your earlier post: one PCR#2 rxn per 10k constructs in the library). This rule was recommended to us by others who had done readout from large pooled shRNA libraries and is not something we optimized ourselves, so it is possible that less reactions could also work but this is what we did.

best,

Neville

On Wed, Mar 25, 2015 at 4:43 AM, NJ CRISPR <xiap...@gmail.com> wrote:

Thank you so much for your help, Neville.

We validated the representation of plasmid and got nearly 64500 sgRNA. 

The representation in your Science paper was very good. I think the biological might not be the main reason, although we used the different cell type (mine is HepG2)

So I think the main reason might be the PCR#1 or PCR#2. 

Would you mind offer me the actual number of PCR2# reactions when you conducted the GeCKO v1 validation of transducted cells? 

Sharp

在 2015年3月25日星期三 UTC+8下午1:18:02，Neville Sanjana写道：

[NJ CRISPR]

That's it! Thanks!

在 2015年3月25日星期三 UTC+8下午9:11:31，Neville Sanjana写道：

[shashank tripathi]

Hi Hiro,

Could you please elaborate the way you validated the lentiCRISPRV2 Plasmid Library prep by deep sequencing.

I assume for first PCR you used

v2Adaptor_F	AATGGACTATCATATGCTTACCGTAACTTGAAAGTATTTCG
v2Adaptor_R	TCTACTATTCTTTCCCCTGCACTGTtgtgggcgatgtgcgctctg

What did you do for second PCR. Did you perform 12 individual PCRs using F1-R1 to F12-R12  primer sets, or you pooled the primers and performed single reactions, and pooled PCR products later on ?

Thanks
 Shashank

[Sung]

Hello

I'm curious about the primer sequence information about first round PCR reverse primer.

The sequence is

v2Adaptor_R

TCTACTATTCTTTCCCCTGCACTGTtgtgggcgatgtgcgctctg

TCTACTATTCTTTCCCCTGCACTGT is partial sequence which can bind to upstream of U6 promoter.

tgtgggcgatgtgcgctctg is partial sequence which can bind to outside chRNA.

Is it correct? Where does vsAdaptor_R binds in LentiCRISPR version2?

If you can help me, I really appreciate for that.

2014년 11월 5일 수요일 오전 6시 53분 18초 UTC+9, Hiro 님의 말:

[Deathcrafter 2000]

Hi.

 

How much PCR product is in a single 100ul PCR2 reaction?  I'm writing SOP for our group and may need to concentrate and reduce the volume of PCR2 before running on a gel due to the number of PCR reactions I will have.

 

Thanks

[Ben Boward]

Hi Hiro,

I am working on this step of the experiment right now and I have a lot of the same questions that you had. Could you share any protocols that you have or have made for this process? I am confused about how the PCR works in terms of how much reagent to use (polymerase, dNTPs, ect.) and how to use the F01-F12 primers for PCR #2. Do I need all 12 of those primers and should I be mixing them or doing each separately?

Thank you for helping other scientists!

Ben

[Shirley]

Hiro,

Congratulations on your success in your experiment. It was very impressive you got perfect bands for PCR runs. Since we are using the same CRISPRv2, I think you might answer my questions. I want to send my library for sequencing, but have been running some problems. I saw ~285bp product from PCR#1, but I did not see the product in PCR#2 which I used 20 cycles. What was your experimental conditions for PCR run 2. BTW, I used the F2 and R2 from the science paper except that my barcodes were on reverse primers. Thanks.

Shirley

On Tuesday, November 4, 2014 at 4:53:18 PM UTC-5, Hiro wrote:

Hi Neville,

[Oriana Genolet]

Dear all,

I am doing the first round of PCR from the gDNA isolated after the GeCKO screen for guide amplification (Shalem et al., 2013). When I do the PCR (with 10µg gDNA and 100µl total PCR volume) and after loading 1/5 of the PCR reaction per lane in an agarose gel, I see degradation of the gDNA (see attached image). I was wondering if anybody has observed the same and could tell me whether this is normal. Since normally I would load 100 ng of gDNA for amplification (which is normally not visible) I have never assesed the state of the DNA after a PCR and wouldnt know. I tested several reactions to see where the problem could lie: gDNA only with H2O in the Cycler, Reaction without enzyme, less DNA per PCR reaction, and another enzyme (see attached image). I already see some degradation in the Cycler of the DNA just with H2O, but I do amplify the 300bp amplicon in one of the reactions.

Anyway (sorry for the long mail :-)), if anybody could give me some feedback on this it would really be appreciated!

All the best,

Oriana

[Xiarong Shi]

I guess the long DNA is sheared to some degree by high temperature after rounds of denaturing and annealing in the cycler. you probably don't have to worry about this as long as you get your PCR product.