Transcript-specific Smart-3seq
65 views
Skip to first unread message
Aurel Nagy
Jun 7, 2019, 5:09:06 PM6/7/19
to Smart-3SEQ
Hello,
I'm interested in modifying the first RT primer to selectively prime a specific gene rather than all polyA mRNA. What is an appropriate annealing temperature to aim for? The RT reaction goes at 42C if I'm reading the protocol correctly; should I aim to have the melting temperature close to 42 to prevent non-specific priming?
Thank you,
Joe Foley
Jun 10, 2019, 12:21:50 PM6/10/19
to smart...@googlegroups.com
This seems possible in principle, by replacing the oligo(dT) in the
1S primer with a sequence complementary to your target, though we've
tried it before with poor results.
The annealing kinetics of the first-strand RT primer is probably going to be the least of your problems, because of both the temperature and the unusually high ionic strength in the TS-RT reaction - any primer long enough to have a specific sequence is probably also long enough to anneal. However, that specificity is the bigger problem. The normal version of the protocol uses oligo(dT) to select for polyadenylated RNAs, i.e. to avoid priming from ribosomal RNA, which is the vast majority of total RNA. So you'd want to be very sure your custom primer has no similarity to your species' ribosomes. Most likely the vast majority of the RNA in your sample is not your transcript of interest, so there's also the challenge of avoiding off-target priming on other transcripts, especially those that are more common. And if the abundance of the target transcript is low, you might even be competing with potential priming from genomic DNA via strand invasion. So you should probably treat your sample with DNase, and possibly even an initial rRNA depletion if that signal dominates your data, but regardless you'll probably need to use a lot more RNA because you're aiming for a much smaller target.
A way to go about it might be to test your target-specific 1S primer with large amounts of non-precious input RNA known to express your target transcript, and compare that with negative controls: a no-template control, just to check for primer dimers (which might be substantial), and if possible also a similar RNA sample that does not express your target transcript (e.g. a knockdown), to check for off-target priming from other transcripts. You'll definitely have to reoptimize the relationship between input amount and PCR cycles, and you may also have to reoptimize the concentration of the 1S primer. You can simplify the latter optimization by adding SYBR Green or similar to the library PCR and running it to completion in a qPCR machine; you'll need to watch not just the absolute yield but also the relative yield above the negative controls. When we tried this before, we saw high yields even from the no-template control, presumably primer dimers, so even with 500 ng of total RNA we weren't actually getting more library than we already did with no RNA. Of course this will depend on the target sequence. You might also be able to reduce the formation of TS-RT primer dimers by removing the random bases from the 2S primer, since you'll need to spike in a lot of PhiX to get base diversity for Illumina sequencers in the rest of the read anyway.
Aside from those problems, you might have some flexibility in other parts of the design. Instead of the 3' end you could aim your primer at some comfortable distance from the 5' end of the transcript (e.g. 400 nt). Then your amplicons would already be the right size and you could skip the initial fragmentation. But it might still help to denature the RNA in case there are secondary structures that inhibit RT, so you could keep that first incubation and just move all the magnesium ingredients (which cause fragmentation) into the later step as we do in the FFPE LCM protocol. This could help you exclude off-target products by a more stringent size selection, and depending on your goals you might also want to keep the cDNAs long anyway so you can read more of the sequence.
If that doesn't work, another approach is to make the initial cDNA from all polyadenylated RNAs as usual, and then add the target specificity (and one of the adapters) during PCR instead. Then you wouldn't have to worry so much about TS-RT primer dimers. That appears to be the strategy used in the SMARTer TCR kits from Takara Bio.
The annealing kinetics of the first-strand RT primer is probably going to be the least of your problems, because of both the temperature and the unusually high ionic strength in the TS-RT reaction - any primer long enough to have a specific sequence is probably also long enough to anneal. However, that specificity is the bigger problem. The normal version of the protocol uses oligo(dT) to select for polyadenylated RNAs, i.e. to avoid priming from ribosomal RNA, which is the vast majority of total RNA. So you'd want to be very sure your custom primer has no similarity to your species' ribosomes. Most likely the vast majority of the RNA in your sample is not your transcript of interest, so there's also the challenge of avoiding off-target priming on other transcripts, especially those that are more common. And if the abundance of the target transcript is low, you might even be competing with potential priming from genomic DNA via strand invasion. So you should probably treat your sample with DNase, and possibly even an initial rRNA depletion if that signal dominates your data, but regardless you'll probably need to use a lot more RNA because you're aiming for a much smaller target.
A way to go about it might be to test your target-specific 1S primer with large amounts of non-precious input RNA known to express your target transcript, and compare that with negative controls: a no-template control, just to check for primer dimers (which might be substantial), and if possible also a similar RNA sample that does not express your target transcript (e.g. a knockdown), to check for off-target priming from other transcripts. You'll definitely have to reoptimize the relationship between input amount and PCR cycles, and you may also have to reoptimize the concentration of the 1S primer. You can simplify the latter optimization by adding SYBR Green or similar to the library PCR and running it to completion in a qPCR machine; you'll need to watch not just the absolute yield but also the relative yield above the negative controls. When we tried this before, we saw high yields even from the no-template control, presumably primer dimers, so even with 500 ng of total RNA we weren't actually getting more library than we already did with no RNA. Of course this will depend on the target sequence. You might also be able to reduce the formation of TS-RT primer dimers by removing the random bases from the 2S primer, since you'll need to spike in a lot of PhiX to get base diversity for Illumina sequencers in the rest of the read anyway.
Aside from those problems, you might have some flexibility in other parts of the design. Instead of the 3' end you could aim your primer at some comfortable distance from the 5' end of the transcript (e.g. 400 nt). Then your amplicons would already be the right size and you could skip the initial fragmentation. But it might still help to denature the RNA in case there are secondary structures that inhibit RT, so you could keep that first incubation and just move all the magnesium ingredients (which cause fragmentation) into the later step as we do in the FFPE LCM protocol. This could help you exclude off-target products by a more stringent size selection, and depending on your goals you might also want to keep the cDNAs long anyway so you can read more of the sequence.
If that doesn't work, another approach is to make the initial cDNA from all polyadenylated RNAs as usual, and then add the target specificity (and one of the adapters) during PCR instead. Then you wouldn't have to worry so much about TS-RT primer dimers. That appears to be the strategy used in the SMARTer TCR kits from Takara Bio.
--
You received this message because you are subscribed to the Google Groups "Smart-3SEQ" group.
To unsubscribe from this group and stop receiving emails from it, send an email to smart-3seq+...@googlegroups.com.
To post to this group, send email to smart...@googlegroups.com.
Visit this group at https://groups.google.com/group/smart-3seq.
To view this discussion on the web visit https://groups.google.com/d/msgid/smart-3seq/11812d35-88e9-4e9a-b227-f015ceafca4b%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Aurel Nagy
Jun 13, 2019, 11:32:10 AM6/13/19
to Smart-3SEQ
Thanks for the comprehensive reply! If it is of interest to anyone I will post updates on how this process goes.
Reply all
Reply to author
Forward
0 new messages