Sequencing a CO1 metabarcoding library on Illumina MiSeq using spike in of custom primers leads to only PhiX reads

[kbkubow]

Hi Everyone,

I have been struggling with an issue I am hoping someone could give me some insight on. I have been constructing and sequencing 16S and ITS metabarcoding libraries using the Schloss 16s protocol on my Illumina MiSeq. I started with 16s using the original Schloss 16s primers. I spike in custom sequencing primers so that I sequence both my 16s library and the Illumina PhiX library I use to increase diversity. Then I substituted the primer specific portion for ITS and remade the custom sequencing primers accordingly. This worked fine, I was able to get both the ITS sequences as well as PhiX. However, when I try doing the same thing with CO1 primer sequences, I only get PhiX reads out of the run. I followed the same procedure as when I modified the primers and custom sequencing primers for ITS, but I don't get any CO1 reads at all. I am unclear why this is occurring. I tried modifying the CO1 custom sequencing primers to be closer in length and melting temperature to the original 16s custom sequencing primers, but it still doesn't work. Does anyone have any insight into why this might be occurring?

On a similar note, does anyone have any primers/protocols they are already using for CO1 metabarcoding? If you do and would be willing to share you primer sequences with me, that would be great (including the custom sequencing primers if you are using those). I am currently using a 1-step PCR approach, but am considering trying a 2-step PCR approach if that might help. Two-step PCR approaches do appear to be more common in the literature, as least for CO1. However, if that would still require the custom sequencing primers, then I feel like I would be right back where I am right now.

Thank you so much for any advice you might have!

In case it helps, I am including examples of my PCR primers and the custom sequencing primers below:

i7: CAAGCAGAAGACGGCATACGAGAT[AACTCTCG]AGTCAGTCAGCCTTCAGGRTGRCCRAARAATCA
i5: AATGATACGGCGACCACCGAGATCTACAC[ATCGTACG]TATGGTAATTGTGGWACWGGWTGAACWGTYTAYCCYCC
Read1: TATGGTAATTGTGGWACWGGWTGAACWGTYTAYCCYCC
Read2: AGTCAGTCAGCCTANACYTCNGGRTGNCCRAARAAYCA
Index: TGRTTYTTYGGNCAYCCNGARGTNTAGGCTGACTGACT

[ADoetsch]

Hi kbkubow,
I don't have the solution but I have some experience with customization of the Schloss protocol for different targets. We never tried CO1 though.
Checking the melting temperature is important, but you considered this already.
How much primer do you add? After consulting Illumina's tech support, we always add 3.4 µL of primer (100 pmol/µL).

From your sequences I noticed the high number of ambiguous bases. Your read2 primer has 3x N and 5x Y/R, which means that your primer mix has potentially 2048 (4^3 x 2^5 = 2^11) different molecules! In other words, you reduce the probability of an exact match between primer and template by three orders of magnitude. Another effect ist a very broad range of potential melting temperatures (the NEB Tm calculator gives 64 - 76°C for the Hot Start Taq enzyme).
I am not sure if this is your actual problem but we never used this level of ambiguity in a sequencing primer.

Are you sure that all these 2048 are likely? If you can find that only a small fraction of these are present in actual sequences, it might be better to prepare a custom mix of exact primers instead, choosing the most abundant sequences.
Check out the Illumina tech note for ITS sequencing, which uses this approach: https://emea.support.illumina.com/co...0064940-01.pdf

It's not a definite answer but I hope it helps...

[kbkubow]

Hi ADoetsch,

Thank you for your reply! I am realizing I accidentally posted the wrong custom sequencing primer sequences. Apologies for the mistake!
Here are the actual sequences:
Read 1: TTGTGGDACWGGNTGRACHGTBTAYCCTCC
Read 2: AGTCAGTCAGCCTTCAGGRTGRCCRAARAATC
Index: GATTYTTYGGYCAYCCTGAAGGCTGACTGACT

But still, I think you are correct, that there are a large number of ambiguous bases. And this could be contributing to the issue. Its a good idea to see if I need all possible combinations or not, and if I instead can prepare of custom mix of exact primers. I will look into this!

Thank you!!!
Karen

[Tom Barker]

Hi kbkubow,

I don't have anything to add regarding your current one step PCR approach but should you decide to switch to a two step approach I can confirm that you wouldn't need custom sequencing primers however you would have to sequence through the CO1 specific part of your primers which will use up some of you read lengths. Illumina has a protocol for this which is fairly straight forward, you would just have to swap out the 16S part of the primer design for your CO1 primer sequences. I've attached their protocol in case you decide to go down this route and it's of any use.

[tweakyaustin]

Hey tom, have you sorted yet? I have done C01 using 2 step PCR. I Tried the one step and it didn't work for some reason.
On another note ate you having any Q-score issues with your 600 cycle kits?