Assessment of bacterial community structure in soil by polymerase chain reaction and denaturing gradient gel electrophoresis.

Bacterial community structure was studied in a Flevo silt loam (FSL) soil microplot, as well as in 15 other soils, by using DNA extraction followed by molecular fingerprinting. Total community DNA was extracted and purified by a direct method, which yielded amplifiable DNA of high molecular weight for all soils. A variable region of the 16S rRNA gene was then amplified by PCR with bacterial primers, resulting in a mixture of amplicons separable via denaturing gradient gel electrophoresis (DGGE). The DGGE profiles of FSL soil were indicative of dominant soil bacterial types, as evidenced by assessing the amplification of Enterobacter cloacae and Arthrobacter sp. targets in a soil DNA background. These targets produced barely detectable bands when present in soil DNA at roughly 5 x 10(6) genome equivalents per g dry soil, and strong bands at 27-fold higher levels. The PCR-DGGE analysis of the FSL soil was highly reproducible. Furthermore, different single versus composite topsoil samples yielded similar DGGE profiles with respect to major bands. In addition, samples taken along vertical soil cores (0-45 cm depth) revealed relative stability of the DGGE profiles. The profiles produced with DNA obtained from different aggregate size fractions of this soil were also similar with respect to the main bands. Moreover, FSL topsoil samples taken over a 1-year period (fallow soil) yielded stable profiles. These data suggested that the soil bacterial communities thus determined were dominated by a limited number of stable and ubiquitous types. The 16 soils, representing varying types and geographical locations, were assessed for differences in their bacterial DGGE profiles. There were striking differences between the profiles obtained for these soils. Evidence was found for the hypothesis that similar soil types tend to contain similar structures of the dominating bacterial types as revealed by the DGGE profiles.