RNA-seq from single cells?

[daniel007]

Hi Simone,

With reference to your trials of 2ul tagmentation, did you have any luck reducing the size of your libraries? Even with 250pg input, I find that my final libraries are average 600bp on the BA, which makes our sequencing core apprehensive about loading them onto the HiSeq. I guess the limiting factor is loaded enzyme - did you ever try "re-loading" the commercial product with adapters?

Thanks in advance.

[adamreid]

We are trying to sequence RNA from individual small cells.

We are sorting by FACS and using SmartSeq2 and Nextera XT to make libraries.

We have had several problems.

We get rRNA contamination from both E. coli (perhaps due to contaminated reverse transcriptase) and from our target organism. This is strange because SmartSeq2 should only amplify polyA mRNAs.

Secondly when we sequence the libraries we get adaptor contamination, sometimes from the strand-switching oligo and also from the Nextera transposase.

Has anyone noticed similar problems?

Cheers,

Adam

[Simone78]

Originally Posted by daniel007

Hi Simone,

With reference to your trials of 2ul tagmentation, did you have any luck reducing the size of your libraries? Even with 250pg input, I find that my final libraries are average 600bp on the BA, which makes our sequencing core apprehensive about loading them onto the HiSeq. I guess the limiting factor is loaded enzyme - did you ever try "re-loading" the commercial product with adapters?

Thanks in advance.

Hi,
we routinely perform tagmentation in 2 ul volume and my input is generally 100 pg (and used 12 cycles PCR). Even then, sometimes I get libraries that are a bit too long, so I am thinking to reduce it to 50 pg (and 14 cycles PCR post-tagm).
I never tried to re-load the commercial Tn5. I am not sure it´s possible or, at least, I never thought about it!
Best,
Simone

[Simone78]

Originally Posted by kobeho24

Hi Simone,
Between these two, which one's better? It seems that SMARTscribe is similar to SSII and Maxima H- is similar to SSIV. Besides, as SMARTscribe is 100U/ul while Maxima H- is 200U/ul, do we still make it 100U in the 10ul RT reaction as it is in the Smart-seq2 protocol? Am I right?

Best!
Gary

sorry for the very late reply. If you are still interested, I can say that Maxima H- works well and now I would use that. I recently tried again with SMARTscribe but I didn´t get great results and I don´t really know the reason since it used to work quite well.
Best,
Simone

[immpdaf]

Hi everyone,
I would be very grateful if anyone could give me some suggestions in our single-cell RNA seq data analysis part.

we have 2 groups of single cells (one normal single cells and one disease single cells), we performed single-cell RNA sequencing. Our library is made using SMART-SEQ2 protocol and it is single-end. We have around 4 million reads / single cell.

Now, using Differential gene Expression analysis, we are going to find significant genes which are upregulated or downregulated in disease cells group with regards to normal group.
So, which normalization technique could you recommend? Our bioinformatician uses TMM to normalize raw counts and he applies R package Monocle to perform DE.
He believes that if we use RPKM, we will get many false positive genes, since we are not comparing genes in one sample, but we are comparing different samples. Do you think it is right?

Many thanks in advance.

[skwek1]

[QUOTE=jwfoley:

Also, you mention using iso nucleotides as your 5' blocker. Have you done the test using biotin instead? That seems more common in the literature, and is a lot cheaper (plus it's difficult to get LNAs and iso bases from the same company). In my experience it works just fine to eliminate the "hedgehog".[/QUOTE]

Hi jwfoley,
Did you use the same TSO primer sequence published by Simone with LNA but biotinylate the nucleotides in the front? Can you please send me the sequence and indicate which nucleotides are biotinylated? Did you order this from Exiqon?
Has anyone compared iso-TSO vs biotinylated-TSO vs iso-TSO-LNA vs biotinylated-TSO-LNA?
Thanks!
Serena

[skwek1]

Hi Simone,
I will like to sequence single immune cells and there are concatamers in the H2O control. Can you please give me the sequence for ISO-TSO that you use to make your library? Does it still have LNA or just all rG? Also where do you order the iso-primers from?
Thanks,
Serena

[seqal]

Hi Simone (or anyone else),

Can you explain to me the benefit of having the same ISPCR sequence on the polyT and TSO? Does it help with the amplification to have one primer instead of two? Would a single strand with the ISPCR sequence at the 5' end and the reverse complement of ISPCR at the 3' end not loop around and bind to each other, blocking both ends?

Thanks!

[kobeho24]

Originally Posted by seqal

Hi Simone (or anyone else),

Can you explain to me the benefit of having the same ISPCR sequence on the polyT and TSO? Does it help with the amplification to have one primer instead of two? Would a single strand with the ISPCR sequence at the 5' end and the reverse complement of ISPCR at the 3' end not loop around and bind to each other, blocking both ends?

Thanks!

Hi seqal,
For you first question, I read some literature (Chenchik et al. 1998)about template switching, which is the main mechanism of SMART-seq2. It suggests that reverse transcriptase can search for potential acceptor template sites when cDNA synthesis is interrupted. To efficiently catalyze template switching, the acceptor template should have a homologous region with the 3' end of cDNA.
I think it is due to some intrinsic properties of RTase in the course of retroviral replication. That's why we make part of both sides' sequence identical.
I might be wrong, please correct my misunderstanding if so

Best!
Gary

[jwfoley]

Originally Posted by skwek1

Hi jwfoley,
Did you use the same TSO primer sequence published by Simone with LNA but biotinylate the nucleotides in the front? Can you please send me the sequence and indicate which nucleotides are biotinylated? Did you order this from Exiqon?

No, my design was somewhat different. I remain skeptical of LNAs as discussed previously in this thread. As also discussed previously, Exiqon is the only vendor of LNAs in North America.

But if you biotinylate the oligo, just stick it on the 5' end. IDT's code for this is /5Biosg/.

[SunPenguin]

Hey guys,

In doing the smart-seq 2 protocol for single cell RNA seq, my colleagues and I find that sometimes we see very uneven preamplification. There are often these "twin peaks" at around 1200bp and 1700bp. I don't see this problem if we repeat the same protocol, but with regular TSO (without LNA). I know in the beginning of this thread, someone mentioned something similar, but it seems like people don't know what it is. Can anyone comment further on the phenomenon?

For bulk samples, LNA works well for us, and is our default. For single cells, it seems rather different.

PS these are immune cells, if that's relevant.

[skwek1]

Originally Posted by jwfoley

No, my design was somewhat different. I remain skeptical of LNAs as discussed previously in this thread. As also discussed previously, Exiqon is the only vendor of LNAs in North America.

But if you biotinylate the oligo, just stick it on the 5' end. IDT's code for this is /5Biosg/.

Thank you for your response jwfoley! Can you lease share the sequences of your TSO, oligodT and IS primers? Should I biotinylate all?

Please email me so I can ask you more questions. I have problems amplifying single cell using the Smartseq2 protocol, though 10 cells and 100 cells worked fine. I am using biotinylated TSO(LNA), biotinylated oligodt and biotinylated IS primers. I'm getting quite desperate and would appreciate your help!

Thanks,
Serena
serena.kwek@ucsf.edu

[SunPenguin]

Hey Simone,

I recently saw a paper, where they put at 3-carbon spacer on the 3' end of the TSO for template switching RT. Supposedly it's to reduce "primer founded amplification." I have never seen this before. Do you have any opinion on that?

[skwek1]

Originally Posted by SunPenguin

Hey guys,

In doing the smart-seq 2 protocol for single cell RNA seq, my colleagues and I find that sometimes we see very uneven preamplification. There are often these "twin peaks" at around 1200bp and 1700bp. I don't see this problem if we repeat the same protocol, but with regular TSO (without LNA). I know in the beginning of this thread, someone mentioned something similar, but it seems like people don't know what it is. Can anyone comment further on the phenomenon?

For bulk samples, LNA works well for us, and is our default. For single cells, it seems rather different.

PS these are immune cells, if that's relevant.

Hi Sunpenguin,
Can you please give me the sequence for your TSO (without LNA) for SMartseq2? Where do you order it from? I am doing single cell RNAseq for immune cells and am having more primers dimers than cDNA for single cell even when I have biotinlylated all the primers. For bulk immune cells, 10 and above, it is fine.
Thanks!
Serena

[skwek1]

Originally Posted by SunPenguin

Hey Simone,

I recently saw a paper, where they put at 3-carbon spacer on the 3' end of the TSO for template switching RT. Supposedly it's to reduce "primer founded amplification." I have never seen this before. Do you have any opinion on that?

Hi,
Can you please give the reference?
Thanks!

[Simone78]

Originally Posted by SunPenguin

Hey Simone,

I recently saw a paper, where they put at 3-carbon spacer on the 3' end of the TSO for template switching RT. Supposedly it's to reduce "primer founded amplification." I have never seen this before. Do you have any opinion on that?

Hi,
as said above, the reference would be appreciated.
I never saw this paper you are talking about and therefore I can´t comment on it. I can just tell you that we tried to block the 3´-end of the TSO in different ways but the cDNA yield after RT was always much lower and we eventually gave up.

[Simone78]

Originally Posted by skwek1

Hi Sunpenguin,
Can you please give me the sequence for your TSO (without LNA) for SMartseq2? Where do you order it from? I am doing single cell RNAseq for immune cells and am having more primers dimers than cDNA for single cell even when I have biotinlylated all the primers. For bulk immune cells, 10 and above, it is fine.
Thanks!
Serena

I had the same problem when working with innate lymphoid cells. Immune cells are small and have much less RNA than, for example, cell lines.
My advice is to block all the oligos with biotin (all!), increasing the number of PCR cycles (23 should be fine), make your own magnetic beads (according to the Rohland & Reich, Genome Res 2010 paper) and using a buffer with a lower % of PEG in order to increase the cutoff (if you are interested I can give you the details), using an oligodT with "V" and not "VN" in the end to avoid weird pairing with the 2 rG in the end of the TSO, as suggested in some papers (for example, look for the "CATS" paper in another thread on this forum). I would also play with oligodT conc (reducing it sometimes help a lot) but I wouldn´t touch the TSO conc.
Some (but not all) of these changes are described in the method part of our last paper --> PMID: 26878113

/Simone

[skwek1]

Originally Posted by Simone78

I had the same problem when working with innate lymphoid cells. Immune cells are small and have much less RNA than, for example, cell lines.
My advice is to block all the oligos with biotin (all!), increasing the number of PCR cycles (23 should be fine), make your own magnetic beads (according to the Rohland & Reich, Genome Res 2010 paper) and using a buffer with a lower % of PEG in order to increase the cutoff (if you are interested I can give you the details), using an oligodT with "V" and not "VN" in the end to avoid weird pairing with the 2 rG in the end of the TSO, as suggested in some papers (for example, look for the "CATS" paper in another thread on this forum). I would also play with oligodT conc (reducing it sometimes help a lot) but I wouldn´t touch the TSO conc.
Some (but not all) of these changes are described in the method part of our last paper --> PMID: 26878113

/Simone

Hi Simone,
Thank you! I will try out your suggestions. Should I try TSO with no LNA?
Thanks,
Serena

[Simone78]

Originally Posted by skwek1

Hi Simone,
Thank you! I will try out your suggestions. Should I try TSO with no LNA?
Thanks,
Serena

that´s a good question! LNA-TSO definitely increases sensitivity, although it comes with some problems (strand invasion). Personally, I am still using the LNA-TSO but many people are of a different opinion (also here in this thread).

[anamar]

Hello to everybody. I've come many times to this thread and I thank everybody for sharing their knowledge.

I prepared Nextera XT libraries from a pool of single cells. I amplify with SMARTER v4, and in some cases had molecules > 1.5kb disturbing my bioanalyzer, migrating to the next well and making it hard to quantify cDNA.

Many of my libraries are OK. But some libraries have come up with a peak around 150bp and I am unsure whether this is "bird nesting", over/undertagmentation or any other artefact that may influence my output. I kindly ask for your feedback. I attached some bioanalyzer profiles. Thank you all in advance.