| Abstract: | Six published fungal specific primer sets (NS1/NS2, SSU‐0817/SSU11‐96, SSU‐0817/SSU‐1536, EF4/EF3, EF4/fung5 and FR1/FF390) were examined for their applicability to the analysis of soil fungal communities using bioinformatic tools as well as real PCR systems. Virtual primer matching for EF4/EF3 and EF4/fung5 revealed good matching with zygomycetous, ascomycetous and basidiomycetous 18S rDNA database entries. Whereas primer EF4/EF3 had no cross matches in the rDNA databases for plant and invertebrate, primer EF4/fung5 gave one signal with the corresponding database. Similar results were obtained for the primer set SSU‐0817/SSU‐1536. Two matches with plant rDNAs and 22 or 12 matches with the invertebrate database could be identified for the primer sets SSU‐0817/SSU‐1196 and FR1/FF390, respectively. Primer pair NS1/NS2 showed only a 70% match with fungal 18S rDNA sequences, but a 75% to 90% match with non‐fungal sequences. Alignments of 2000 eukaryotic sequences using “ARB” confirmed that PCR fragments obtained by the primer sets EF4/EF3, EF4/fung5, SSU‐0817/SSU‐1536 and FR1/FF390 were supposed to include hypervariable regions (V4, V7, V9), whereas the others included regions which were more phylogenetically conserved. Practical PCR approaches affirmed fungal specificity as predicted by virtual primer matching for EF4/EF3, EF4/fung5 and FR1/FF390. However FR1/FF390 amplified only 60% of the fungal samples under investigation. All other primer sets amplified fungal as well as non‐fungal samples. |
