Denaturing gradient gel electrophoresis (DGGE) is a powerful and convenient tool for analyzing the sequence diversity of complex natural microbial populations. DGGE was evaluated for the identification of ammonia oxidizers of the ␤ subdivision of the Proteobacteria based on the mobility of PCR-amplified 16S rDNA fragments and for the analysis of mixtures of PCR products from this group generated by selective PCR of DNA extracted from coastal sand dunes. Degenerate PCR primers, CTO189f-GC and CTO654r, incorporating a 5 GC clamp, were designed to amplify a 465-bp 16S rDNA region spanning the V-2 and V-3 variable domains. The primers were tested against a representative selection of clones and cultures encompassing the currently recognized ␤-subdivision ammonia oxidizer 16S rDNA sequence diversity. Analysis of these products by DGGE revealed that while many of the sequences could be separated, some which were known to be different migrated similarly in the denaturant system used. The CTO primer pair was used to amplify 16S rDNA sequences from DNA extracted from soil sampled from Dutch coastal dune locations differing in pH and distance from the beach. The derived DGGE patterns were reproducible across multiple DNA isolations and PCRs. Ammonia oxidizer-like sequences from different phylogenetic groupings isolated from gene libraries made from the same sand dune DNA samples but prepared with different primers gave DGGE bands which comigrated with most of the bands detected from the sand dune samples. Bands from the DGGE gels of environmental samples were excised, reamplified, and directly sequenced, revealing strong similarity or identity of the recovered products to the corresponding regions of library clones. Six of the seven recognized sequence clusters of ␤-subdivision ammonia oxidizers were detected in the dune systems, and differences in community structure between some sample sites were demonstrated. The most seaward dune site contained sequences showing affinity with sequence clusters previously isolated only from marine environments and was the only site where sequences related to the Nitrosomonas genus could be detected. Nitrosospira-like sequences were present in all sites, and there was some evidence of differences between Nitrosospira populations in acid and alkaline dune soils. Such differences in community structure may reflect physiological differences within ␤-subdivision ammonia oxidizers, with consequent effects on nitrification rates in response to key environmental factors.

The community structure of beta-subclass Proteobacteria ammonia-oxidizing bacteria was determined in semi-natural chalk grassland soils at different stages of secondary succession. Both culture-mediated (most probable number; MPN) and direct nucleic acid-based approaches targeting genes encoding 16S rRNA and the AmoA subunit of ammonia monooxygenase were used. Similar shifts were detected in the composition of the ammonia oxidizer communities by both culture-dependent and independent approaches. A predominance of Nitrosospira sequence cluster 3 in early successional fields was replaced by Nitrosospira sequence cluster 4 in late successional fields. The rate of this shift differed between the two areas examined. This shift occurred in a background of relative stability in the dominant bacterial populations in the soil, as determined by domain-level polymerase chain reaction-denaturing gradient gel electrophoresis (PCR-DGGE). Molecular analysis of enrichment cultures obtained using different ammonia concentrations revealed biases towards Nitrosospira sequence cluster 3 or Nitrosospira sequence cluster 4 under high- or low-ammonia conditions respectively. High-ammonia MPNs suggested a decease in ammonia oxidizer numbers with succession, but low-ammonia MPNs and competitive PCR targeting amoA failed to support such a trend. Ammonia turnover rate, not specific changes in plant diversity and species composition, is implicated as the major determinant of ammonia oxidizer community structure in successional chalk grassland soils.

No abstract

Changes in the community structure of chemolitho-autotrophic ammonia-oxidising bacteria of the beta-subgroup Proteobacteria were monitored during nutrient-impoverishment management of slightly acidic, peaty grassland soils, which decreased in pH with succession. Specific PCR, cloning and sequence analysis, denaturing gradient gel electrophoresis (DGGE) and probe hybridisation were used to analyse rDNA sequences directly recovered from successional soils. Four previously characterised ammonia oxidiser sequence clusters were recovered from each soil, three associated with the genus Nitrosospira and one with the genus Nitrosomonas. All samples were dominated by Nitrosospira-like sequences. Nitrosospira cluster 3 was the most commonly recovered ammonia oxidiser group in all fields, but a greater representation of Nitrosospira clusters 2 and 4 was observed in older fields. Most probable number (MPN) counts were conducted using neutral and slightly acid conditions. Neutral pH (7.5) MPNs suggested a decrease in ammonia oxidiser numbers in later successional fields, but this trend was not observed using slightly acid (pH 5.8) conditions. Analysis of terminal MPN dilutions revealed a distribution of sequence clusters similar to direct soil DNA extractions. However, an increased relative recovery of Nitrosospira cluster 2 was observed for acid pH MPNs compared to neutral pH MPNs from the most acidic soil tested, in agreement with current hypotheses on the relative acid tolerance of this group.

Marram grass (Ammophila arenaria L.), a sand-stabilizing plant species in coastal dune areas, is affected by a specific pathosystem thought to include both plant-pathogenic fungi and nematodes. To study the fungal component of this pathosystem, we developed a method for the cultivation-independent detection and characterization of fungi infecting plant roots based on denaturing gradient gel electrophoresis (DGGE) of specifically amplified DNA fragments coding for 18S rRNA (rDNA). A nested PCR strategy was employed to amplify a 569-bp region of the 18S rRNA gene, with the addition of a 36-bp GC clamp, from fungal isolates, from roots of test plants infected in the laboratory, and from field samples of marram grass roots from both healthy and degenerating stands from coastal dunes in The Netherlands. PCR products from fungal isolates were subjected to DGGE to examine the variation seen both between different fungal taxa and within a single species. DGGE of the 18S rDNA fragments could resolve species differences from fungi used in this study yet was unable to discriminate between strains of a single species. The 18S rRNA genes from 20 isolates of fungal species previously recovered from A. arenaria roots were cloned and partially sequenced to aid in the interpretation of DGGE data. DGGE patterns recovered from laboratory plants showed that this technique could reliably identify known plant-infecting fungi. Amplification products from field A. arenaria roots also were analyzed by DGGE, and the major bands were excised, reamplified, sequenced, and subjected to phylogenetic analysis. Some recovered 18S rDNA sequences allowed for phylogenetic placement to the genus level, whereas other sequences were not closely related to known fungal 18S rDNA sequences. The molecular data presented here reveal fungal diversity not detected in previous culture-based surveys.