Pollen DNA extraction and amplicon library preparation
Pollen DNA was extracted from sample pollen balls (2015) and anther pollen (2015 and 2016) using a CTAB-Chloroform DNA preparation protocol (as in Agrawal et al. 2013), and stored at -20°C until amplification and quantification. We first quantified the DNA concentration in each sample using a Qubit fluorometer, and diluted each DNA sample to ~2 ng/uL. We also created standard dilutions from 1:10 to 1:10000 in triplicate from a single sample of known origin. We assayed 152 unknown origin pollen ball DNA samples split between two sets. Each set included pollen DNA from 76 unknown samples, as well as the same 8 DNA samples of known origin (two from each species), and 12 samples from the standard dilution in triplicate.
Each set of 96 was then transferred to four identical 96-well plates, where we ran a PCR amplification. We conducted 10uL reaction volumes, including: 6.4 uL of ddH20, 1 uL of MgCl2, 1 uL of dNTPs, 0.2 uL of each forward and reverse primers (above), 0.1 uL (0.25 units) of JumpStart Taq (Sigma-Aldrich), and 1 uL of template (at 2 ng/uL). For each set of four plates, we then used four identical thermal cyclers to run the following protocol simultaneously: 94oC for 3 minutes, and then for plates 1, 2, 3 and 4 we had 20, 25, 30, or 35 cycles of 94oC for 30 seconds, 55oC for 30 seconds, and 72oC for 1 minute, respectively. We then used a final extension time of 5 minutes.
This resulted in eight 96-well plates—four for each set—representing the different cycle treatments. We then cleaned up each reaction with 1.8X volume SeraPure beads: 10uL of sample with 18uL of beads, and performed two 70% etOH washes. We eluted in 20 uL of resuspension buffer (Illumina).
We then ran an individual indexing reaction for each sample within each set (i.e. 384 randomly chosen, unique indexes for each set). The 20 uL indexing reaction included 4 uL of ddH20, 10 uL of HiFi Master Mix (KAPA Biosystems), 1 uL of each the forward and reverse i5 or i7 indexes, and 4 uL of template DNA from the amplification step. This was run with the following thermal cycling conditions: 95oC for 3 minutes, 98oC for 30 seconds, followed by 8 cycles of 98oC for 30 seconds, 63oC for 30 seconds, and 72oC for 30 seconds. We had a final extension time of 3 minutes.
Within each sample set, we pooled 5uL of each indexed sample from across the four-cycle treatments, resulting in one plate for each of the two sample sets. As before, we used a 1.8X SeraPure bead cleanup for the 20uL pooled samples, and completed two 70% etOH washes. We eluted samples into 20uL of resuspension buffer. An equal volume of each sample was then pooled—within each set—into the final library. We sequenced each of the two final libraries separately across two lanes of an Illumina MiSeq, with 2x150 paired end sequencing chemistry.