How many of you run into problems of false- positive PCR or lack of sensitivity in qPCR when trying to use 16S primers because of the background genomic DNA in your PCR enzyme mixes?

Well, then check out this paper by a team on Switzerland on how to remove residual contaminating genomic DNA from the enzyme mix, reported in the November 2009 issue of Journal of Microbiological Methods. The article is titled: Development and clinical validation of a modified broad-range 16S rDNA PCR for diagnostic purposes in clinical microbiology, and is authored by Abdessalam Cherkaoui, Stephane Emonet, Dimitri Ceroni, Bruno Candolfi, Jonathan Hibbs, Patrice Francois, and Jacques Schrenzel, from the University of Geneva Hospitals in Switzerland.

Introduction:

The report describes a new method for removal of the contaminating DNA found in PCR enzymes that makes using a broad range microbial 16S assay difficult to use with any sensitivity.  Typically, following broad range PCR, the amplicons are sequenced and then compared with known sequences in public databases to identify the microbes.  However, false-positives caused by contamination, which may be introduced from several sources, makes the use of this standard technique impossible for clinical work. The most common source for contamination with microbial genomic DNA is the DNA polymerase enzyme.  Although other methods for removal of gDNA from DNA Polymerases have been tried (ex. Sau3AI, restriction enzymes, DNase I, ultrafiltration of the PCR mix and UV irradiation), none of these methods are entirely effective. Reported in this paper, the authors show the effectivness of using Exonuclease III to remove the presence of gDNA contamination from the PCR mix for use with clinical samples.

Methods:

Samples:

The authors used clinical samples for this study. Patient samples were collected between 2007 and 2009 and samples were accepted for a broad-range analysis if they met the following criteria:

- High clinical suspicion of infection despite previous negative culture
- Sample was a biopsy, tissue sample or fluid aspirate from a normally sterile site
- Sample was cleared by a clinical microbiologist after discussion with the physician

Samples included specimens such as lung and spine biopsies, pleural fluid, heart valve vegetations, and cerebrospinal fluids.

 PCR Enzyme Treatment:

The authors used Perkin Elmer Amplitaq LD and Exonuclease III from Boehringer Mannheim. To treat the PCR enzymes, they added 10 units of Exonuclease III to 37 µl of Taq for 30 minutes at 37^oC and then inactivated the enzyme at 70^oC for 20 minutes. The dNTPs, primers, and template were added next for a final reaction volume of 50 ul.  PCR products were analyzed on an Agilent DNA 7500 LabChip and compared to the same runs without the Exonuclease III.

PCR Reaction:

PCR was carried out with primers to the 16S rRNA gene (BAK11w and BAK2) and the reaction performed for 40 cycles. Five microliters of the first round PCR was used in a second PCR with hemi-nested primers (BAK11w and BAK553r) for another 40 cycles. Controls were 500 pg of E.coli DNA for the positive and a mixture of reaction components without DNA for the negative control. All reactions were performed in duplicate.

Sequencing:

PCR products were sequenced with the BAK11w primer on the ABI 3130. Products were diluted 1:10 and then 1 µl was used for cycle sequencing. Sequences were compared using the BLAST program and needed to be 98-100% homologous to assign to a species.

Results:

Before testing the technique with clinical samples, the authors evaluated different enzymes and determined that Amplitaq LD gave the lowest genomic DNA background after two rounds of PCR (data was not shown).  Since contamination is visible only after the second round of PCR, the mastermix used for the first round of PCR was Exonuclease III treated but not the second.

The first experiment was a titration of Exonuclease III from 0 to 20 units of enzyme.  Ten units of enzyme gave a clean result after two rounds of PCR.  They next determined the sensitivity of their assay using E.coli and S. aureus genomic DNA and determined it to be 300 cfu per PCR reaction or 30,000 CFU per ml of material.

Patient Samples:

144 patient samples were evaluated in this study which were negative based on direct examination and/or culture. Of these, 17%  (24/144) were positive with broad range PCR (only 3 of the 24 were positive by direct examination and 1/24 were positive by culture). 83% were negative in PCR and no sample that was negative in PCR was positive using other methods.

Some of the organisms that could only be identified using PCR included Bartonella henselae, Coxiella burnetti, Mycoplasma salivarium, and Bartonella Quintana, all of which are notoriously difficult to culture. The most common organism detected by PCR in clinical samples was Streptococcus pneumonia which may have gone undetected by culture due to its tendency to undergo autolysis or because of the antibiotics administered to the patient. The authors explain that Streptococci are the leading agents causing unexplained deaths and identifiable only with molecular assays.

Summary:

The authors demonstrate a way to increase the accuracy and speed of molecular testing of clinical microbiology samples using conventional PCR and sequencing typing techniques. By eliminating the genomic DNA contamination in the PCR mix using an exonuclease enzyme compatible with PCR buffers, they eliminated false-positive results and the assay only adds on 50 minutes. Their method increases the confidence in reporting of positive cases, although for negative cases, they agree that a lower level of detection would be better.

Final Note:

The authors have performed the groundwork for a method that can now be optimized for use in other methods that do provide lower levels of detection sensitivity, such as qPCR or fluorescent assays. If the assay can be sensitive and quantitative, it would make a powerful tool for performing total bacterial counts on any sample- environmental and clinical.

~Suzanne

This entry was posted on Sunday, December 6th, 2009 at 10:42 am and is filed under New papers in microbiology, News. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.