Precaution for using nucleic acid-based methods to detect Aeromonas.

Aeromonas is ubiquitous in aquatic ecosystems but can cause various human infections, including gastrointestinal tract syndromes and soft tissue infections (3). Its role as a diarrheal agent remains controversial, but several studies show a substantial association (1, 2, 4, 6). For this reason, we included Aeromonas in our recent multiplex PCR-Luminex assay for several bacterial enteropathogens (5). This letter is intended to urge caution when using such molecular methods to detect Aeromonas spp. In our work (5), we demonstrated the analytical performance of the PCR-Luminex assay using fecal samples that were spiked with relevant bacterial pathogens and extracted with the QuickGene DNA tissue kit (Fujifilm, Tokyo, Japan). However, after press, when we subsequently used the QIAamp DNA stool minikit (Qiagen, Valencia, CA), 71% (25/35) of previously negative stool samples yielded low-level amplification of Aeromonas (e.g., quantitative PCR [qPCR] threshold cycle [CT] range of 30.7 to 37.4). This was confirmed with two distinct singleplex real-time PCR assays, one that used a TaqMan probe to detect the same region of the aerolysin gene that was targeted by the PCR-Luminex assay and another that used SYBR green to detect a 16S rRNA region. Primers and probes are published in reference 5. We dissected the Qiagen extraction components into (i) elution buffer (AE) alone, (ii) wash buffers (AW1 and AW2) and elution buffer, (iii) lysis buffer (AL) and ethanol, followed by wash buffers and elution buffer, (iv) stool lysis buffer (ASL), followed by lysis buffer, ethanol, wash buffers, and elution buffer, or (v) InhibitEX tablets and stool lysis buffer, followed by lysis buffer, ethanol, wash buffers, and elution buffer. These preparations were prepared in duplicate, DNA was extracted following the manufacturer’s instructions, and real-time PCR was performed using both Aeromonas assays. Extracts from the first four procedures were negative (16/16), while those from the fifth procedure were positive (4/4). We then tested an additional 8 tablets of InhibitEX (from 4 lots of kits), and all eight of these extracts were positive for Aeromonas (using both assays). These aerolysin amplicons were sequenced and were 100% identical to the amplicon deposited in GenBank under accession number EF189591, a sequence of Aeromonas hydrophila strain ZN1 that was distinct from our spiked A. hydrophila strain (28SA) with only 92% identity, arguing against contamination. We revisited the clinical samples that we evaluated in our study (5). This revealed that 3/10 clinical specimens were positive for Aeromonas at low levels (e.g., CT 30.7) and contained the questionable ZN1 sequence. The corrected sensitivity/specificity remained similar at 89%/95%. Our interpretation is that traces of DNA in certain molecular kit components may exist, since Aeromonas is common to water. We do not know the extent of Aeromonas detection using other extraction procedures, but we would urge one to pay special attention to this step when trying to detect Aeromonas using molecular methods.