I have begun preparing libraries for the sequencing. The Sultan (2012) protocol is nice and clear, but there are two buffers that it incorporates which are quite expensive. These buffers are parts of a second strand synthesis kit and it costs upwards of 1000$ for the kit - but the cost of the kit is due to the enzyme (which I don't use for Sultan's protocol) not the buffer. Instead of buying the entire kit, I ended up making the buffer myself (all the components cost less than 200$ total). Other than this issue with the buffer, I am quite pleased with the protocol in general.
the recipes I used are here:
5x second strand synthesis buffer invitrogen cat #11917-010
formula: 100 mM Tris-HCl (pH 6.9), 450 mM KCl, 23 mM MgCl2, 0.75 mM beta-NAD+, 50 mM (NH4)2SO4
for 10ml:
-1ml Tris HCl pH6.9
-0.3355g KCl
-0.23ml 23mM MgCl2
-0.0050g beta NAD+
-0.0661g (NH4)2SO4
bring to 10ml with DEPC H20
10x Reverse transcription buffer invitrogen cat #18080051
formula: 200 mM Tris-HCl, pH 8.4, 500 mM KCl
for 10ml:
-2ml 200mM Tris HCl pH 8.4
-0.3728g KCl
bring up to 10ml with DEPC H2O
As for the TruSeq protocol, the final step is a PCR enrichment of the adapter-ligated fragments. The protocol calls for 15 cycles to enrich the product, but that will likely place the reaction in the plateau phase (really bad because it results in "PCR duplicates" and reduces the overall complexity of the library). To optimize the number of cycles that I'll end up using I am running a qPCR. This will give me an understanding of how efficiently the library amplifies and therefore how many cycles to run so that my library is in the exponential phase and not plateau.
I started with 0.5ul of ligated product in the qPCR and the concentration was well below what the Nanodrop can read (not surprising as I haven't enriched it yet, but normally a bad sign). The qPCR results show that my library plateaued in 20 cycles, so I will be shooting for 14 cycles which should be mid-exponential phase. The next trick is that when I do the actual enrichment PCR I will have a much higher amount of starting DNA (~10ul instead of 0.5ul) and so plateau will be achieved much sooner. As I will be starting with 20x more DNA, I will need 4-5 fewer cycles (2^4 < 20 < 2^5). Assuming that the amplification will procede with the same efficiency (and that the efficiency is ~2, though often slightly lower, 1.8ish), this means that in my enrichment PCR I should use 10 cycles.
I will actually run two qPCRs to test the difference between 10 and 13 cycles. This should help me hone in on the exact number of cycles to use. I can afford to do this with this library as it is a "test." I won't be doing different cycles on the future libraries that I plan to sequence.
I will need to do this same qPCR experiment on each of my library samples and determine an "average" number of cycles to use. I could enrich each sample independently, but that is a bad idea since this is transcriptome work and the relative expression levels will depend on the number of cycles. I will probably do qPCR on the first 8 to get a good average, and then use that as the standard number of cycles for the entire experiment.
--------------------------------------------------------------------------------------------------------------------
Sultan, M., S. Dökel, V. Amstislavskiy, D. Wuttig, H. Sültmann, H. Lehrach and M.-L. Yaspo. 2012. A simple strand-specific RNA-Seq library preparation protocol combining the Illumina TruSeq RNA and the dUTP methods. Biochemical and Biophysical Research Communications 422:643–646. Elsevier Inc.
