Our environmental microbiology group has been dealing with what appears to be negative-PCR contamination issues, however we are scratching our heads to figure out what is going on.
1) We have used one DNA extraction kit to get genomics from filters containing air samples, and added a non-sampled filter (filter blank) and a no-filter blank (these two are referred to as extraction negs). We performed this on the lab bench. We proceeded to conventional PCR targeting 16s, using primer pair 1, triplicates per sample -> We found negative controls having band of expected size after pooling the triplicates using DNA clean up kit. PCR-blank (which only contains PCR-grade water as template and performed in hood) is clean (i.e. Neither PCR hood nor any of the PCR reagents are contaminated).
2) We switched to a new box of the same genomic DNA extraction kit, and performed as discussed in step one, but in a fume hood -> Same false-positives showed after pooling triplicates on DNA clean up kit
3) We tested the DNA CLEAN UP kit to check whether THAT kit is contaminated. The test comes out clean -> no contamination in the kit.
4) We switched to a different genomic DNA extraction kit (one less commonly used in the field), performed extraction on the open bench, PCR with primer pair 1. -> Negative is clean
5) Using samples from step 1, we performed PCR targeting 16S using primer pair TWO. This primer pair binds to regions FLANKING that of primer pair 1 (i.e. amplicon size larger than primer pair 1) -> Extraction negs and PCR neg is clean.
This combination of results is really conflicting for us. We can rule out it's the PCR hood as we never had any problems with the PCR-negs. Somehow using primer pair one gives us contamination, but primer pair two targeting the same gene at a different location gives us clean results. One possible reason is amplicon contamination, but again the PCR hood is clean, and so is the PCR clean up kit. Seems like working on the open bench or not is irrelevant, as using a different extraction kit gives clean results.
Essentially, all these negatives are much less intense than that of the positive controls and the samples we are interested in. However, as they are negative, we should see NO band at all. Nonetheless, if we are going to sequence these samples, it is unlikely that we will get any results for the negatives anyways.
Sorry for the lengthy post, but would be great if someone can provide some insight, and if there is not enough "contamination" in the negatives for sequencing, does it really matter? It's similar to the concept of qPCR having a high Ct value to be considered as "noise."
Any opinion is greatly appreciated.
Thank you very much
Edited by Pseudopneumo, 25 July 2013 - 12:05 AM.