Pyrosequencing of 28S rRNA gene amplicons
28S rRNA gene was applied to investigate fungal taxonomy in this study. After combining with a single sample-specific 8-mer tag sequence separated by a 2-bp spacer sequence, the primers LR0R (5’-ACCCGCTGAACTTAAGC-3’) and LR3 (5’-CCGTGTTTCAAGACGGG-3’) were applied to amplify the D1-D2 region of 28S rRNA gene, which generated a gene amplicon of about 650 bp. The 100 µL PCR system was composed of 50 mM KCl, 10 mM Tris-HCl at pH of 8.3, 0.1% Triton X-100, 2.5 mM MgCl2, 200 µM dNTPs (Invitrogen Life Technologies, Carlsbad, CA, USA), 5 units of Taqpolymerase, 0.1 µg BSA (New England Biolabs Inc., Ipswich, MA, USA), 0.4 µM each primer and 12.5 ng DNA template. PCR thermocycling conditions were as follows: 3 min at 94 °C for initialization, followed by 35 cycles of 1 min at 94 °C for denaturation, 45 secs at 53 °C for annealing, 1 min at 72 °C for extension, and finally 15 min at 72 °C before maintained at 4 °C. For each sample, five to ten PCR reactions were performed. PCR products were combined and purified by agarose gel electrophoresis, and then collected with QIAquick Gel Extraction Kit (Qiagen, Venlo, Netherlands). PCR products were quantified by PicoGreen using a FLUOstar Optima (BMG Labtech, Jena, Germany). Equimolar uniquely-tagged amplicons from each sample were then combined into one library. Tags separated raw sequences via the function “Split by tag” of the IEG Galaxy pipeline (http://zhoulab5.rccc.ou.edu:8080). They were further cleaned via LUCY based on the following thresholds: maximum average error at 0.01, maximum error at the ends at 0.01, minimum score at 21, window size at 50, minimum length at 200, and allowable ambiguous base number (‘N’) at 10. Only sequences with no ‘N’ were accepted and were clustered into operational taxonomic units (OTUs) using UCLUST at the 97% identity. OTU singletons were removed and OTUs annotation was performed by the RDP Classifier with the latest Fungal LSU training set 11 (http://rdp.cme.msu.edu/classifier/classifier.jsp) at the 50% confidence interval. We used raw OTU abundance in the following data analyses.