SEQanswers

Go Back   SEQanswers > Applications Forums > Sample Prep / Library Generation



Similar Threads
Thread Thread Starter Forum Replies Last Post
Illumina Directional RNA-Seq flobpf Illumina/Solexa 58 11-19-2015 03:18 PM
Strand specific RNA-seq data from dUTP protocol ZhengXia Bioinformatics 8 10-16-2014 04:36 AM
RNA-Seq: FDM: A Graph-based Statistical Method to Detect Differential Transcription u Newsbot! Literature Watch 0 08-10-2011 02:00 AM
RNA-Seq: SAW: A Method to Identify Splicing Events from RNA-Seq Data Based on Splicin Newsbot! Literature Watch 0 08-14-2010 02:00 AM
Illumina directional RNA-seq protocol Herve Illumina/Solexa 10 06-13-2010 07:18 AM

Reply
 
Thread Tools
Old 05-16-2013, 07:46 AM   #21
ZWB
Member
 
Location: Livermore, CA

Join Date: Apr 2011
Posts: 27
Default

This is a recently published technique that we developed to make directional RNA-Seq libraries. It's similar to one of the techniques in the Levin et al paper, but we made some significant improvements.

http://www.ncbi.nlm.nih.gov/pubmed/23558773
ZWB is offline   Reply With Quote
Old 05-29-2013, 03:13 AM   #22
MLog
Member
 
Location: Russia

Join Date: Jan 2010
Posts: 36
Question

Hi all,
I started to use Illumina TruSeq Stranded mRNA kit: http://support.illumina.com/sequenci..._prep_kit.ilmn
and I noticed a couple of thing that puzzled me a lot.
First, the protocol for first strand synthesis suggested in the manual is 25C - 10 minutes, 42 C - 15 minutes and 70C 15 minutes. Only 15 minutes, is this really sufficient? Second, there is no purification after first strand synthesis, the second strand mix is added immediately after RT. I always thought that for strand-specific protocol it is critically important to completely eliminate dTTP before second strand synthesis. And last, there is no stage of UDG treatment at all! After adapter ligation and purification the library goes directly to PCR enrichment.
Does anyone have an idea how this could work?
MLog is offline   Reply With Quote
Old 05-29-2013, 05:01 AM   #23
jazz
Member
 
Location: us

Join Date: Nov 2008
Posts: 28
Default

This is the problem with companies in general. They sell kits with 'proprietary' components and won't disclose any other information.
15' does sound insufficient but maybe they are using a modified version of RT enzyme. However, given that it requires 42C temp, my bet is on Superscript II, which is an older generation. It's up to you if you want to modify the program, since regular RT reactions are 50 min-1hr.
You don't need to purify the Ist reaction products before 2nd strand synthesis. Just modifying the buffer conditions is enough. Check the Parkhomchuck 2009 NAR paper where they described their version of directional RNA-seq. No UDG step is required in Illumina protocol since they are ligating adapters first. This is the major difference between their protocol and UDG based strand detection. If I were you, I would follow the Illumina protocol as is, since it is standardized, and hey, you already have spent a ton of money getting the kist, might as well take advantage of that. Otherwise, have a look at the Parkhomchuck paper, which is what I follow for strand-specific library preps.
jazz is offline   Reply With Quote
Old 05-29-2013, 09:14 AM   #24
kcchan
Senior Member
 
Location: USA

Join Date: Jul 2012
Posts: 184
Default

The TruSeq RNA kits don't come with the RT enzyme. You have to supply your own SuperScript II.
kcchan is offline   Reply With Quote
Old 05-30-2013, 12:06 AM   #25
MLog
Member
 
Location: Russia

Join Date: Jan 2010
Posts: 36
Default

Hi jazz,
indeed, the enzyme used in this kit is user-supplied SuperScript II, as kcchan already mentioned. So 15 minutes looks like a typo.
As for the method, it seems that Illumina uses the method similar to Parkhomchuk's one: "Strand specificity is achieved by replacing dTTP with dUTP in the Second Strand Marking Mix (SMM)": http://support.illumina.com/sequenci...questions.ilmn so the absence of purification after first strand synthesis is still surprising.
MLog is offline   Reply With Quote
Old 05-30-2013, 12:22 AM   #26
ETHANol
Senior Member
 
Location: Western Australia

Join Date: Feb 2010
Posts: 308
Default

Quote:
Originally Posted by jazz View Post
No UDG step is required in Illumina protocol since they are ligating adapters first. This is the major difference between their protocol and UDG based strand detection.
Not sure what that means so....

From the Illumina website.
/truseq_stranded_mrna_lt_sample_prep_kit/questions.ilmn

How is strandedness maintained after DNA amplification?
Strandedness is maintained via the directionality of the adapters. The p7 adapter will be on the 3' end of the cDNA strand. As a consequence, the cDNA strand is sequenced. Second strand synthesis is performed using dUTP in place of dUTP and a high fidelity polymerase that cannot process through dUTP template is used for PCR enrichment. As a result, only the first strand product is amplified and the strand information is retained based on the p5 and p7 adapter orientation.
__________________
--------------
Ethan
ETHANol is offline   Reply With Quote
Old 06-04-2013, 05:17 AM   #27
pmiguel
Senior Member
 
Location: Purdue University, West Lafayette, Indiana

Join Date: Aug 2008
Posts: 2,317
Default

We have seen a lower yield with these stranded mRNA kits. More than would be expected from just the loss of the 2nd strand during amplification.

We nearly always fragment for only 4 minutes (instead of 8 minutes) to allow isolation of longer cDNA and produce larger insert libraries. This is because a large percentage of our samples are de novo transcriptome.

Anyway, I had not previously noticed the decrease in the length of the reverse transcription step. We may have been relying on this longer extension to produce our longer cDNAs.

Does anyone think that bumping the RT step back to 50 minutes would cause a loss of strandedness of the resulting libraries? I mean, the SuperScript II will gladly do 2nd strand as well as 1st under normal conditions. The presence of Actinomycin D would inhibit that DNA-templated DNA polymerase activity, but the fact that the step was shortened to 15 minutes worries me. Like maybe a longer incubation allows the SuperScript to circumvent the ActD inhibition?

Thoughts anyone?

Or better, has anyone gone back to 50 minute incubations with SuperScriptII and then observed whether some of the strandedness of the resulting library is lost?

--
Phillip
pmiguel is offline   Reply With Quote
Old 06-04-2013, 06:48 AM   #28
Genohub
Registered Vendor
 
Location: genohub.com

Join Date: Mar 2013
Posts: 210
Default

Increasing the time to 50 minutes should not cause a loss of strandedness. Are you doing any additional cleanups with the Directional protocol. If UDG is being added to the polymerase mix, this would result in a lower yield. It would be interesting to perform a typical RNA-Seq library, a stranded library, with the polymerase from the typical RNA-Seq library, then compare yields.
Genohub is offline   Reply With Quote
Old 06-04-2013, 08:39 AM   #29
pmiguel
Senior Member
 
Location: Purdue University, West Lafayette, Indiana

Join Date: Aug 2008
Posts: 2,317
Default

Quote:
Originally Posted by Genohub View Post
Increasing the time to 50 minutes should not cause a loss of strandedness. Are you doing any additional cleanups with the Directional protocol. If UDG is being added to the polymerase mix, this would result in a lower yield. It would be interesting to perform a typical RNA-Seq library, a stranded library, with the polymerase from the typical RNA-Seq library, then compare yields.
We don't do additional cleanup, but since dUTP is not added until 2nd strand synthesis, no additional cleanup is actually needed.

The real problem to maintaining strandedness is just making sure the reverse transcriptase does no 2nd strand synthesis. As long as that does not happen, the libraries should be stranded.

BTW, we use this method now and it appears to work fine -- with the 15' RT incubation. No one seems to think that doubling that incubation time will result in RT doing DNA-templated polymerization.

Spec on the kit is 15 cycles of amplification. We used to typically use 8 cycles. But with the stranded kits we are going to 12-15 cycles. 15 cycles tends to take us about 10 nM, though.

But we scale back the Ampure cuts to mostly remove molecules with inserts smaller than 200 bp. So if the RT is not getting out that far in 15 minutes, that could be eating into our yields.

--
Phillip
pmiguel is offline   Reply With Quote
Old 06-04-2013, 04:27 PM   #30
ETHANol
Senior Member
 
Location: Western Australia

Join Date: Feb 2010
Posts: 308
Default

The original dUTP protocol required massive amounts of starting material because the ActD. If you removed ActD from the protocol you still got pretty good strandedness but required a lot less starting material. So my guess was that the Illumina stranded kit would give a lot lower yield. I'll find out today as I'll be using it for the first time. They do say that you can start out with 4 ug of RNA vs. 1 ug from the unstranded kit so that is some testament to lower efficiency. But I was thinking the yield was going to be a lot lower then that 4-fold difference in requested starting material suggested. From you experience that seems like the case.

I'm not sure why Illumina is recommending the same 15 cycles of PCR for both kits. In my experience with the unstranded kit that was way too many. I guess from your experience for that stranded kit that is more in the correct range.

Phillip sorry to hijack your thread a little, but how are you quantitating your libraries for pooling?
__________________
--------------
Ethan
ETHANol is offline   Reply With Quote
Old 06-04-2013, 08:58 PM   #31
Jon_Keats
Senior Member
 
Location: Phoenix, AZ

Join Date: Mar 2010
Posts: 279
Default

We are testing the Agilent stranded kit in the next couple of weeks. I'll try to get back to this thread to mention the outcome. Good to see ETHANol again.
Jon_Keats is offline   Reply With Quote
Old 06-04-2013, 09:13 PM   #32
MLog
Member
 
Location: Russia

Join Date: Jan 2010
Posts: 36
Default

Quote:
Originally Posted by pmiguel View Post
Or better, has anyone gone back to 50 minute incubations with SuperScriptII and then observed whether some of the strandedness of the resulting library is lost?

--
Phillip
I used the same program as for standard TruSeq RNA kit (50 minutes at 42C). I will sequence these libraries soon and will write here about results.
MLog is offline   Reply With Quote
Old 06-05-2013, 08:39 AM   #33
pmiguel
Senior Member
 
Location: Purdue University, West Lafayette, Indiana

Join Date: Aug 2008
Posts: 2,317
Default

Quote:
Originally Posted by ETHANol View Post
The original dUTP protocol required massive amounts of starting material because the ActD. If you removed ActD from the protocol you still got pretty good strandedness but required a lot less starting material. So my guess was that the Illumina stranded kit would give a lot lower yield. I'll find out today as I'll be using it for the first time. They do say that you can start out with 4 ug of RNA vs. 1 ug from the unstranded kit so that is some testament to lower efficiency. But I was thinking the yield was going to be a lot lower then that 4-fold difference in requested starting material suggested. From you experience that seems like the case.

I'm not sure why Illumina is recommending the same 15 cycles of PCR for both kits. In my experience with the unstranded kit that was way too many. I guess from your experience for that stranded kit that is more in the correct range.

Phillip sorry to hijack your thread a little, but how are you quantitating your libraries for pooling?
You want to watch for adding too much RNA -- the probes/beads can become overloaded and not remove enough rRNA. Contrariwise, if not enough RNA is there, or you don't aliquot out enough SA beads, you get probe contamination of your libraries. Bummer! Get used to eyeing those magnetic bead pellets to make sure they are big enough.

About pooling -- erg, sore subject. Well, we do a bioanalyzer chip first to make sure we have something and determine its size after amplification. Then we do qPCR on each of the components of the pool. Then we make the pool. Then we run another high sensitivity chip and do a final pool qPCR.

I think everything goes much smoother if your libraries are size selected on a gel and thus have a tight distribution. Clustering itself has this intense size bias, with the shortest amplicons seeming like they always cluster and the longest ones seeming like they never do. Correcting for that is unnecessary with tight bands, nearly impossible with wide ones. Well, speculation on my part entirely, though.

--
Phillip

Then we cluster at 15 pM. And... well we usually hit reasonable cluster densities on the flow cell. But frequently the spread on the number of reads for various components of the pool is 3x or higher.
pmiguel is offline   Reply With Quote
Old 06-06-2013, 03:46 AM   #34
Genohub
Registered Vendor
 
Location: genohub.com

Join Date: Mar 2013
Posts: 210
Default

Phillip, In my experience the ActD prevents DNA synthesis during first strand synthesis, but does not stop RNA synthesis. Haven't seen a significant difference between 15-50 minutes.

Noticed that TruSeq Stranded doesn't have a UDG step. Seems like they rely on their polymerase to not read through dUTP bases. Does anyone know what polymerase this is? Has anyone added the UDG step anyway and compared strandedness? I didn't understand this comment earlier in the thread, "No UDG step is required in Illumina protocol since they are ligating adapters first"


Quote:
Originally Posted by pmiguel View Post
We don't do additional cleanup, but since dUTP is not added until 2nd strand synthesis, no additional cleanup is actually needed.

The real problem to maintaining strandedness is just making sure the reverse transcriptase does no 2nd strand synthesis. As long as that does not happen, the libraries should be stranded.

BTW, we use this method now and it appears to work fine -- with the 15' RT incubation. No one seems to think that doubling that incubation time will result in RT doing DNA-templated polymerization.

Spec on the kit is 15 cycles of amplification. We used to typically use 8 cycles. But with the stranded kits we are going to 12-15 cycles. 15 cycles tends to take us about 10 nM, though.

But we scale back the Ampure cuts to mostly remove molecules with inserts smaller than 200 bp. So if the RT is not getting out that far in 15 minutes, that could be eating into our yields.

--
Phillip
Genohub is offline   Reply With Quote
Old 06-06-2013, 04:55 AM   #35
pmiguel
Senior Member
 
Location: Purdue University, West Lafayette, Indiana

Join Date: Aug 2008
Posts: 2,317
Default

Quote:
Originally Posted by MLog View Post
Hi all,
I started to use Illumina TruSeq Stranded mRNA kit: http://support.illumina.com/sequenci..._prep_kit.ilmn
and I noticed a couple of thing that puzzled me a lot.
First, the protocol for first strand synthesis suggested in the manual is 25C - 10 minutes, 42 C - 15 minutes and 70C 15 minutes. Only 15 minutes, is this really sufficient?
Thanks for bringing this to our attention. Yes the old (non-stranded) kit used the same RT at 70C for 50 minutes.

Is 15 minutes long enough. Well, we have made a bunch of libraries this way and they do work. However we noticed a recent decrease in yield with this kit. That is, previously we would get so much library that we cut back from 15 amplification cycles to 8. Now we tend to do 12-15 cycles.

However, it may well because of the issue you mention. The kit will have been optimized for typical RNAseq "counting" experiment. These are often done with single reads. So insert sizes beyond 100 nt are not needed. However we frequently do de novo transcriptomes. For these we think there is an advantage to using paired end reads and therefore want larger inserts.

With the old (non-stranded) kits, this was not an issue. As long as you reduced the fragmentation time, you could create cDNAs that were >2kb, if you wanted. (Well, you also needed to size select some how to get rid of the shorter insert stuff.)

But it could be that 15 minutes just does not get out to the distance we want.
pmiguel is offline   Reply With Quote
Old 06-06-2013, 10:50 AM   #36
pmiguel
Senior Member
 
Location: Purdue University, West Lafayette, Indiana

Join Date: Aug 2008
Posts: 2,317
Default

Quote:
Originally Posted by Genohub View Post
Phillip, In my experience the ActD prevents DNA synthesis during first strand synthesis, but does not stop RNA synthesis. Haven't seen a significant difference between 15-50 minutes.
RNA synthesis? Reverse Transcriptase, as far as I know, can't synthesize RNA. Both of its pol activities are DNA pol. But one is RNA templated (RT) and the other is DNA templated (2nd strand).
Quote:
Originally Posted by Genohub View Post
Noticed that TruSeq Stranded doesn't have a UDG step. Seems like they rely on their polymerase to not read through dUTP bases. Does anyone know what polymerase this is? Has anyone added the UDG step anyway and compared strandedness? I didn't understand this comment earlier in the thread, "No UDG step is required in Illumina protocol since they are ligating adapters first"
I don't know.

--
Phillip
pmiguel is offline   Reply With Quote
Reply

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off




All times are GMT -8. The time now is 11:54 AM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2020, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO