Accurate Estimation of Fungal Diversity and Abundance through Improved Lineage-Specific Primers Optimized for Illumina Amplicon Sequencing

• D. Lee Taylor ^[1] ; William A. Walters ^[2] ; Niall J. Lennon ^[4] ; James Bochicchio ^[4] ; Andrew Krohn ^[3] ; J. Gregory Caporaso ^[3] ; Taina Pennanen ^[5]
  1. [1] University of New Mexico

     University of New Mexico

     Estados Unidos

     
  2. [2] Cornell University

     Cornell University

     City of Ithaca, Estados Unidos

     
  3. [3] Northern Arizona University

     Northern Arizona University

     Estados Unidos

     
  4. [4] cBroad Institute of MIT & Harvard, Cambridge, Massachusetts, USA
  5. [5] eNatural Resources Institute Finland (Luke), Vantaa, Finland

  Mostrar afiliaciones +
• Localización: Applied and Environmental Microbiology, ISSN 0099-2240, Vol. 82, Nº 24, 2016, págs. 7217-7226
• Idioma: inglés
• Enlaces
  • Texto completo
• Resumen

  • While high-throughput sequencing methods are revolutionizing fungal ecology, recovering accurate estimates of species richness and abundance has proven elusive. We sought to design internal transcribed spacer (ITS) primers and an Illumina protocol that would maximize coverage of the kingdom Fungi while minimizing nontarget eukaryotes. We inspected alignments of the 5.8S and large subunit (LSU) ribosomal genes and evaluated potential primers using PrimerProspector. We tested the resulting primers using tiered-abundance mock communities and five previously characterized soil samples. We recovered operational taxonomic units (OTUs) belonging to all 8 members in both mock communities, despite DNA abundances spanning 3 orders of magnitude. The expected and observed read counts were strongly correlated (r = 0.94 to 0.97). However, several taxa were consistently over- or underrepresented, likely due to variation in rRNA gene copy numbers. The Illumina data resulted in clustering of soil samples identical to that obtained with Sanger sequence clone library data using different primers. Furthermore, the two methods produced distance matrices with a Mantel correlation of 0.92. Nonfungal sequences comprised less than 0.5% of the soil data set, with most attributable to vascular plants. Our results suggest that high-throughput methods can produce fairly accurate estimates of fungal abundances in complex communities. Further improvements might be achieved through corrections for rRNA copy number and utilization of standardized mock communities.

    IMPORTANCE Fungi play numerous important roles in the environment. Improvements in sequencing methods are providing revolutionary insights into fungal biodiversity, yet accurate estimates of the number of fungal species (i.e., richness) and their relative abundances in an environmental sample (e.g., soil, roots, water, etc.) remain difficult to obtain. We present improved methods for high-throughput Illumina sequencing of the species-diagnostic fungal ribosomal marker gene that improve the accuracy of richness and abundance estimates. The improvements include new PCR primers and library preparation, validation using a known mock community, and bioinformatic parameter tuning.