Re: Deleting 100kb from the genome

[Elph]

Hi Badger, 

In my experience this is pretty inefficient for such big fragments. One trick though is to transfect your cells and dilute in non limiting conditions, meaning that instead of growing single cell clones you will grow heterogenous pools from 20-30 initial cells (you can play around with this dilution factor depending how frequent is your modification). You grow a bit these pools, screen them to find out which ones contain cells harboring your deletion and subclone the pool but this time with limiting dilution so you grow single clones. A second round of screening will allow you to identify which of these clones harbor your deletion.

This is a strategy I applied inspired by this paper: http://www.ncbi.nlm.nih.gov/pubmed/22183967

Alternatively you can throw oligos with LoxP sites and induce the deletion with CRE. I personnally had little success with this method in cells.

The last but most efficient option is to knock in LoxP-PuroR and NeoR-LoxP at both of your targets using donor constructs with homology arms and co-select with Puro and Neo. Then you treat with Cre. This will induce the deletion. But like the previous option you will be left with a LoxP site.

The last thing is if you are simply doing a coding gene knock out you do not need to take out all 20 exons; targeting a frameshift mutation in the first coding exon is enough to abrogate protein production. This will also avoid knocking out regulatory elements for other genes etc that might reside in your gene's intron.

Hope that helps,

E

Le vendredi 28 mars 2014 17:15:13 UTC-7, badger491 a écrit :

Hi all,
I am using CRISPR/Cas9 to delete 20 exons from the mouse genome. I designed 5' target and 3'target upstream and downstream of the coding exons. Based on your experience, will this work? And is there a better way to knockout my gene.

thanks in advance for your response

[Andrea D'Osualdo]

I was able to remove 70kb using 2 CRISPRs in HeLa, THP1 and K562 cells.

Best

Andrea

Sent from my iPhone

On Mar 28, 2014, at 5:15 PM, badger491 <prashant.r...@gmail.com> wrote:

Hi all,
I am using CRISPR/Cas9 to delete 20 exons from the mouse genome. I designed 5' target and 3'target upstream and downstream of the coding exons. Based on your experience, will this work? And is there a better way to knockout my gene.

thanks in advance for your response

--
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.

[dosu74 .]

Hi Badger,

sorry for the delay answering.

I designed the two CRISPRs (in the pX330) targeting the region near the ATG and the STOP codon with the original idea to put an N-term or C-term tag because for my GOI I do not have a nice antibody. Although the T7 was very nice for both, I did not have luck with the repair using ssODN and I decided to use the 2 CRISPR at the same time in co-transfection to remove the entire locus (without any template repair plasmid/oligos).

I genotyped at the beginning the pool of transfected cells. I observed a decreased intensity of both PCRs at the 5'UTR and 3'UTR compared to the wilde-type cells. More important, I saw a PCR product product (of about 470bp) spanning the 70Kb region using the forward on the 5'UTR and reverse on the 3'UTR only in the CRISPRs transfected pool and not in the parental cells (see pdf). Encouraged, I screened many clones (about 100 total) obtained by single cells sorting, looking for clones that have the hybrid PCR product but not the PCR at the 5'UTR neither at the 3'UTR (some clones still have one of the two…).

Finally, I sequenced the hybrid PCR to confirm that corresponded to the 5'UTR "ligated" to the 3'UTR) and I re-confirmed the result by qPCR.

The efficiency was better in K562 (20% - 8/40 clones total) and THP-1 (17% - 2/12 clones total) compared to HeLa (3% - 3/96 clones total).

Best,

Andrea

2014-03-31 22:00 GMT-07:00 badger491 <prashant.r...@gmail.com>:

Thanks for your response. I am taking out isoforms of a protein so I cannot make indels as there is high sequence homology between my protein and isoforms.
But with this deletion, I am not getting positive clones. I agree it is inefficient as both CRISPR target have to cleaved around the same time, otherwise DNA repair machinery would start repairing before other target get cleaved.

Any thoughts on how to make it efficient? I am transfecting pX330 vector with sgRNA.

Thanks

--

[Cosmas Giallourakis]

Dear  Andrea

Can u tell me how you are transfecting THP-1s and which guide/Cas9 plasmids u are using?

Many thx!

Cosmas

[аня]

Dear Anderea, 

Can you please share a bit what kind of transfection strategy are you using for a THP1 cells? 

Cheers 
Anna

[Andrea D'Osualdo]

Hi Anna,

I used the Neon system, although the efficiency was quite poor (about 25%). But then I sorted for GFP positive transfected single cells.

I guess also Amaxa is working in a similar way.

Cheers

Andrea

Sent from my iPhone

[mario.sche...@gmail.com]

Hi Andrea,

I read with great interest that you were successfull to transfect THP-1 with your plasmid. I am working already 6 month on this problem using pX461 and pX462. My transfection efficancy with Nucleofector is ower than 5% with a very weak GFP expression. Is it possible to get in touch per e-mail to discuss my problems with you? I am sure you have the answer to that.

with kind regards
Mario

[Bird Giant]

Hi, Andrea,
  Which plasmid you used? and you used the nikase? I want remove 10kb from lymphoblast cells or fibroblast cells,
and I am a new guy in this field. and could you share your protocol to me? gian...@163.com
  thanks very much.
Bird

[dosu74 .]

Hi Bird,

I used the pX330 (wt Cas9).

I transfected HeLa with Lipo2000 (6-well plate, 1ug each CRISPR plasmid + 40ng GFP).

For K562 and THP-1 I used the Neon system (protocol life tech for K562).

Let me know if you want more details.

Good luck!

Andrea

[Ayşegül Kaymak]

Hello all, 

I have been following this CRISPR forum and I have the same issue. I tried to knock out a gene in mESC cells (E14). I used px330 plasmid and designed two guide RNAs to delete a region around 672 bp. I planned to delete the exon 2 and create a shift at the coding region. After I checked my clones, I have realized that one of the functional domain is still transcribed. Therefore I am planning to do another transfection with 6 plasmids to delete 3 functional domains at the same time. Do you think this is doable?

Andrea, I read your post about 70 kb deletion. I also once planned to do such an experiment. But I was scared that this kind of two guide RNA would only make a point mutation.Because the deletion region is too far away from each other and CRISPR plasmids are working independently. My gene has 17 exons and do you think this could be one of the knock strategy for my experiment? Also if it is ok for you, I would like to discuss with you in a more detailed way via e-mail. 

Thanks in advance for your responses.

 Aysegul

29 Mart 2014 Cumartesi 07:10:53 UTC+1 tarihinde dosu74 yazdı:

[Joe Miano]

On Friday, March 28, 2014 8:15:13 PM UTC-4, badger491 wrote:

Hi all,
I am using CRISPR/Cas9 to delete 20 exons from the mouse genome. I designed 5' target and 3'target upstream and downstream of the coding exons. Based on your experience, will this work? And is there a better way to knockout my gene.

thanks in advance for your response

 

 

To All who wish to make large deletions,

I would be very careful with such plans given how pervasive ncRNAs are, especially snoRNAs.  Recall the original presumed KO of Egfl7 in Nature was actually a phenotype attributed to the removal of a critical miR gene (miR126).  I would be very conservative about deleting genomic DNA.

Joe

[dosu74]

Hi Aysegul,

actually you will have a lot of ind/del only, but in few clones you should get allelic large deletions and hopefully some of them should be in homozygosity. I would recommend to check first the presence of large deletions in a pool of transfected cells by genomic PCR and then proceed to screen single cell clones. Unfortunately, I have only one case experience so far (although in very different cell lines). You are welcome to contact me by email of course!

Cheers

Andrea

[Jonathan Geisinger]

Joe raises a very good point:  if you're generating large deletions, especially novel ones that have not been associated with a disease phenotype , the innocent bystander elements that will be taken along for the ride are very important to consider.  

Regarding Aysegul's question, the greater the number of gRNAs you transfect, the greater the number of off-target effects you have to worry about.  Based on what I've observed with similar cells to yours, you will get the deletion, but probably only in ~1% of alleles.  If this is a essential gene, well, that complicates matters.  Additionally, I suspect that you actually won't get a lot of small indels.  The repair machinery in cells of this type seems pretty efficient.

Additionally, you may want to consider combining your approach with maybe a more traditional knock-out approach.  

[coolsc...@yahoo.com]

Hi Andrea, I am currently screening single cell clones for a large >100kb deletion. I have some clones that have the hybrid PCR product that suggests that the deletion has occured and also with the 5" and 3" cut sites gone. But I am seeing the internal PCR products (amplifying regions that were in the 100kb deletion segment). Was wondering if you had done similar PCRs to look at the deleted region and if this suggests that the segment has translocated to another region in the genome.

[dosu74]

I amplified two different amplicons of the transcript corresponding to the deleted gene and they were not there anymore.

Translocation could be a possibility, as well as aspecific PCR amplification (especially for shorter amplicons). The easy way to check I guess it is PCR purification followed by sanger sequencing: at that point you should be able to tell if it's your gene translocated or just another part of the genome. Hope it makes sense.

Best,

Andrea

[coolsc...@yahoo.com]

thanks for the suggestions I'll clone the product and sequence.

Also wanted to ask you about any advice you may have for single cell cloning of THps. I'm having some difficulty getting clones to survive.

[dosu74]

If you have a fairly clean and unique band, you do not even need to clone your PCR product, nor to purify it from agarose gel. You can simply directly purify the PCR product using columns or ExoSAP reagent (http://www.affymetrix.com/catalog/131310/USB/ExoSAP-IT-For-PCR-Product-Cleanup#1_1).

Regarding the THP-1, I had also a lot of problems to grow them as single cells. Conditioned media could help a bit in my experience but also the sorting conditions are probably important (I will try to get the one I used from our core facility). I would try to test sorting first not-treated cells in order to find good settings for you and then use the CRISPR edited cells.

Best,

Andrea

[coolsc...@yahoo.com]

Thanks so much!

[apes...@gmail.com]

Hi  Andrea,

 

I've been trying to use the Lenticrispr v2 to modify THps and have had no success. Dont see any modification using either nucleofection of the plasmid or transduction with concentrated vector.

 

I was wondering if you have used this lenticrispr v2 system at all for any of the thp work. I'm wondering if the promoter in this plasmid (the modified EF1a promoter - "EFNS") may be an issue ?

 

Any advice you may have would be very appreciated.

[dosu74 .]

Hi,

I never used any lenti system to make ko cell line since I don't want to make CRISPR-Cas9 stable cells. I always used the pX330. When you say you don't see any genome editing, are you checking the pool of infected cells with T7/Surveyor assay or looking for single cell with no expression of your target? In my experience, if you have a negative surveyor, it will be almost impossible to find a ko cell clone. If you actually have a positive surveyor and you sort single cells, you should be able to find it.

Another problem could be if you select with puro after transduction and then you don't sort single cell. In this case, it could be that you still have an heterogeneous population with stable-cells that have in-frame genome editing and still express your gene. And even if you sort for single cell, since the CRISPR and Cas9 are supposed to be stable, I wonder if they could still work and insert in-frame indel in your gene. That is why I would use the lentisystem only for screening purpose but not to create ko cell lines (unless I am missing something…)

Regarding the promoter I don't know the modified EFNS, but the EF1 is my favorite to do lentiviral protein over expression in THP-1.

Not sure if it makes sense..

Cheers

Andrea

[apes...@gmail.com]

Thanks Andrea, I just wanted to confirm that the Cbh promoter works in Thps based on your success with this in these cells. I'm not sure how the EF1a was modified but that does not seem to be working in this cell line as well as another T cell line and I'd never heard of the Cbh promoter.

 

I was using integrase defective lentiviral vectors so also not trying to make stable cells, just transient expression of cas9 and gRNA.

[dosu74 .]

Hi,

I thought you were using the lenticrispr v2 (addgene#: 52961) that should not be "integrase defective" lenti.

Can you tell me a little bit more about your integrase-deficient lenti? did you clone it by yourself or is deposited in addgene? Is it similar to lentiCRISPR (meaning that has both gRNA-trcrRNA and hSpCas9) ? Have you successfully used it in human/mouse cells other than THP-1 ?

Thanks a lot!

Cheers

Andrea

[Krishna K]

Hi Andrea,

I've tried to transfect THp cells using the neon system and have had very poor viability. I tried using their optimization protocol and none of the conditions seem to work, cells die within a day. I used 5 ug DNA for 10^7 cells using their 100ul tips, and have tried both crispr and non crispr plasmids to see if this was a crispr specific issue.

Any insight and advice on what conditions to use for transfecting them would be greatly appreciated.

Krishna

[dosu74 .]

Hi Krishna,

I also had (and still have…) a viability issue with THP-1 cells, even when I electroporate just cells without DNA.

I do not think is a problem of DNA amount. I believe THP-1 cells are very "tricky" cells... you should try to use low passage and very healthy cells (split them periodically, avoid cell "clumps", etc…)

But, as Sungtae noticed in another post, the few cells that survive should be enough in order to get single clones.

I saw Ying suggested also a protocol for the Neon system.

Sorry for not being of help!

Good luck!

Andrea

[Daniel Ibrahim]

Hi everyone,

I couldn't quite figure out how recent the posts were (some seemed over a year old, others much younger). We have very successfully deleted (and inverted and duplicated) over 1Mb fragments in ES-cells. Dels usually are a little more frequent than Dups or Invs. 70kb, 200kb etc are all no problem. you can find more here ()

two things you kan keep in mind are

1. if you add two guides, all sort of things happen, so optimally, you derive a cell line in which you should then really check for your desired type of mutation and if there isn't something else on additionally.
2. we had a good experience with putting guides into DHS-sites specific for ES-cells. so if there's DHS-data for your cell-type of interest it might be worthwhile trying to put them there if it is OK with your experimental plan.

Best

d

[Daniel Ibrahim]

I forgot to add the linkt to the paper: so here it is http://www.cell.com/cell-reports/pdfExtended/S2211-1247%2815%2900029-7

[Diego León]

Hi Bird

Any success using crispr/cas9 in lymphoblast cells?

Regards,
Diego

[Megha Amar]

Hi Daniel,

I am interested in doing single gene duplication using CRISPR, the only concern for me right now is whether I should design paired gRNA to target the entire ORF or just coding region??

Looking forward to your response,

Best,

Megha