single nuclei RNAseq from frozen tissues

Posted By: jingjingao, on Sep 26, 2017 at 3:24 PM

We want to develop a method of 10X single cell RNAseq for frozen tissues. To do that, we have to isolate intact nuclei from frozen tissues. Any of you have experience in extracting whole nuclei from frozen tissues?

Any suggestions will be much appreciated!



56 Replies

Re: single nuclei RNAseq from frozen tissues

Posted By: jens-10x, on Oct 2, 2017 at 9:43 AM

Hi Jingjin,


Thanks for you interest in our 10x Single Cell Application.

We do have a Demonstrated Protocol on Isolating Single Nuclei; however, the samples we used were from frozen cell lines and fresh mouse brain tissue, only. We do not have any experience in extracting single nuclei from frozen tissue.


From our experience with fresh tissue, we have seen that lysing the tissue directly (as opposed to dissociating it into single cells first), worked better in our hands. This may be a reasonable approach for you as well in your efforts to optimize the protocol for your tissue type. 


Frozen tissue will be more susceptible to cell (and nuclei) death and accumulation of debris/aggregate. With that in mind, I would recommend to follow the best practices outlined in Protocol step 3 of our Demonstrated Protocol. It includes a density gradient and myelin removal (in case you work with neuronal tissue) step that will definitely help to further clean up your nuclei suspension. 


Hope that helps. Keep us posted on your progress!





Re: single nuclei RNAseq from frozen tissues

Posted By: Luciano2018, on Mar 7, 2018 at 10:49 PM

Hi all,


I have been working on "a" single nuclei RNAseq protocol from frozen material. The consistent issue I have been having however, is the resuspension of the nuclei in the PBS + 1% BSA buffer (see also note below). While this buffer has been good at avoiding clumpling of whole live cells, it does a not so good job at keeping the nuclei in isolation and therefore tend to clump. reading the literauture on nuclei rpeservation, there are no nuclei stabilizer ions in it (e.g. Ca++/Mg++ ). I tried 2% BSA (as suggested in the video) and the results were marginally better nut not near good enough. Mg++ or Ca++ (at least 3 mM) help maintain nuclei integrity and therefore makes them less sticky and in consequence keep them in isolation. So, here is my quetsion, I would like to know whether adding one of these cations would interfere with downsteams steps (Jens, any thoughts?). It'll make things way easier at resuspending nuclei pellets. Nuclei look not clumpled at all and happy in the lysis buffer but as soon as I spin them down (300-500 rcf @ 4 deg) and attempt resuspension in PBS + 1% BSA buffer, clumps would form and that was it. It does not happen 100% of the time but it happens a lot. This nuclei prep has lots of moving parts and would be great if we can brainstom and have something ready for other users.


Note: I assume here the PBS used by 10x recomended protocol for nuclei prep is PBS is Ca++ abd Mg++ free. The one recommended from SIgma is not sold in Australia for instance as its GE-IVD so I could go for the overpriced UltraPure BSA from Thermo.   


Any help or advice would be very nice.




Re: single nuclei RNAseq from frozen tissues

Posted By: jens-10x, on Mar 8, 2018 at 12:01 AM

Hi Luciano,


Thanks for checking back in with us. Can I ask, what tissue type are you working with? Are those nuclei from brain tissue or any other tissue type?


If those new buffer conditions work well for your nuclei suspensions, I suggest to stick with them and see if you can maintain a single nuclei suspension before you're ready to add them to the RT master mix. I'd also recommend to keep the 2% BSA in your resuspension and wash buffer, even though the disaggregation only improved marginally. Also, we do not recommend to use > 3mM Magnesium in the final nuclei or cell suspension buffer as this may inhibit the RT reaction. That risk will further increase if the nuclei volume that is added to the RT mix is on the upper end (20-30 ul). If you do have to use higher Magnesium concentrations (e.g. 3.5 mM), then I suggest to keep the nuclei concentration relatively high (> 1000 nuclei/ul) so that you can use < 10ul nuclei volume and still target a high number of nuclei for your analysis.


Let me know if you have any more questions.




Re: single nuclei RNAseq from frozen tissues

Posted By: Luciano2018, on Mar 8, 2018 at 12:16 AM

Hi Jens, Thanks for your quick response! No brain, no. I wish...ahahaa. These are from multiple sources as are random tumours types we are truing to analyse. Unfortunately accessing to fresh material for us is nearly impossible so frozen is the way to go for now. The disegregation and lysis steps works like a charm, lots of nice intact nuclei. But as soon as I change the buffer (from lysis to resuspension and wash) the nuclei clump. So, I am a bit stuck in this part. Not sure what else to do to keep them separated. I have read in the "dark net" (hahaha) that some people add DNAse I, aiming to trim whatever DNA stand is sticking ourt of the nuclei and make it sticky and reduce any circulating DNA that leaked form the nuclei. After that, they wash twice as usual to remove the DNAse I. However, I am kind of may or it may not work at all and if it does and some DNAse I stays there then that can't be any good. I like the idea of maintaining the nuclei concentration while having Mg in the buffer so the nuclei volume is <10 uL.


So, you don't recommend Mg++ > 3 mM, but do you recommend 3 mM? or some at least? what's the safest amount other than none?