Bacterial contamination of in vitro plant cultures: confounding effects on somaclonal variation and detection of contamination in plant tissues

Bacterial contamination represents a serious problem for plant tissue culture research and applications. Bacterial interference with normal plant physiology and morphology can generate misleading conclusions if the presence of bacteria is ignored. Bacterial contaminants in in vitro plant culture are typically detected by direct observation; thus, it is assumed that cultures without visible symptoms are bacteria free. Here, we demonstrate that contaminating Bacillus DNA in plant DNA solutions from asymptomatic plants can interfere with the analysis of somaclonal variation in chrysanthemum. We studied somaclonal variation in chrysanthemum using short semi-specific PCR primers based on conserved motifs in NBS–LRR disease resistance genes and in mobile elements. Instead of true somaclonal variation we found three polymorphic bands derived from contaminant bacterial DNA in plant extracts. Although the detection of asymptomatic bacteria in in vitro plant cultures is a major issue, we found that it has not been adequately addressed to date, particularly for studies on somaclonal variation. We reviewed the most commonly cited contaminant bacteria in in vitro plant culture and designed specific 16S rRNA gene-based PCR primers for the main genera causing contamination (Bacillus, Pseudomonas, Staphylococcus, Lactobacillus, Erwinia/Enterobacter and Xanthomonas). Using a panel of pure bacterial DNAs, artificial mixes of bacterial/plant DNAs, and in vitro plant cultures with and without visible contamination we demonstrated that our primers are in most instances both reliable and sensitive, and appropriate for the identification and tracking of the most frequent bacterial contaminants in plant in vitro cultures. Implications of bacterial identification to molecular analysis of somaclonal variation and plant culture decontamination are discussed.