DNA-Based Characterization and Identification of Arbuscular Mycorrhizal Fungi Species

Abstract

Arbuscular mycorrhizal fungi (AMF) are obligate symbionts of most land plants. They have great ecological and economic importance as they can improve plant nutrition, plant water supply, soil structure, and plant resistance to pathogens. We describe two approaches for the DNA-based characterization and identification of AMF, which both can be used for single fungal spores, soil, or roots samples and resolve closely related AMF species: (a) Sanger sequencing of a 1.5 kb extended rDNA-barcode from clone libraries, e.g., to characterize AMF isolates, and (b) high throughput 454 GS-FLX+ pyrosequencing of a 0.8 kb rDNA fragment, e.g., for in-field monitoring.

Appendix: Examples of Command Lines for 454 Sequencing Data Analysis

Bioinformatics pipeline for analyzing 454 sequence reads

• 1. De-multiplex.

  For example, when sequencing a full 454-plate split into four gaskets (physically separated compartments), we use the following command to de-multiplex the first gasket:

  split_libraries.py -m Mapping1.txt -f 1.TCA.454Reads.fna -q 1.TCA.454Reads.qual -l 500 -o split_Library_Run1_Output/ -n 1000000

  and this command to de-multiplex the second gasket:

  split_libraries.py -m Mapping2.txt -f 2.TCA.454Reads.fna -q 2.TCA.454Reads.qual -l 500 -o split_Library_Run2_Output/ -n 2000000

  Consider that the parameters (such as sequence length) can be modified according to your needs; in the previous example we set the minimum length of sequences to be implemented in the clustering to 500 bp.

• 2. Combine your de-multiplexed sequences in a single file:

  cat split_Library_Run1_Output/seqs.fna split_Library_Run2_Output/seqs.fna > Combined_seqs.fna

• 3. Cluster your sequences.

  First you have to prepare a text file containing the parameters of the clustering. We use the following parameters and save the text file as parameters.txt:

  pick_otus:otu_picking_method uclust

  pick_otus:similarity 0.98

  pick_otus:enable_rev_strand_match True

  We afterwards perform the clustering by using the following command:

  pick_de_novo_otus.py -i combined_seqs.fna -p parameters.txt -o uclust_picked_otus/

• 4. Remove singletons.

  After clustering, you obtain a biom table with your “OTUs” (representative sequences of 98 % similarity clusters).

  First remove the singletons (sequences represented only once) from the table:

  filter_otus_from_otu_table.py -i otu_table.biom -o otu_table_no_singletons.biom -n2

  Afterwards remove singletons from the fasta file:

  filter_fasta.py -f combined_seqs.fasta -o biom_filtered_seqs.fasta -b otu_table_no_singletons.biom

• 5. Remove non-AMF sequences.

  The previously created file “biom_filtered_seqs.fasta” contains your combined sequences without singletons. However, it still contains non-AMF sequences which in most cases have to be removed before further analysis.

  To remove these sequences we use Blast2GO (https://www.blast2go.com/b2ghome) which takes individual sequences and finds similar sequences in NCBI.

  The output of Blast2GO is an Excel-format file with the hits of your query sequences. We normally order the hits alphabetically and simply manually delete the non-AMF rows from the Excel table. After deleting the non-AMF rows, copy the remaining names of sequences, which will be kept for further analysis, and paste them into a text file. Name the text file as seqs_to_keep.txt (or according to your naming system).

  To remove the non-AMF sequences from the FASTA file write in QIIME:

  filter_fasta.py -f seqs_no_singletons.fasta -o seqs_no_cont.fasta -s seqs_to_keep.txt

  To remove the non-AMF sequences from the OTU table write:

  filter_otus_from_otu_table.py -i otu_table_no_singletons.biom -o otu_table_nosingletons_nocontaminants.biom -e seqs_to_keep.txt --negate_ids_to_exclude

• 6. Convert the file otu_table_nosingletons_nocontaminants.biom into a tabulator delimited table:

  convert_biom.py -i otu_table_nosingletons_nocontaminants.biom -o otu_table_nosingletons_nocontaminants.txt -b

  This table contains information about the samples, AMF “OTUs” (98 % similarity clusters), and read amounts.