Elimination of host and dead pathogenic organism DNA for improved diagnostic outcome

Abstract

Molecular-based diagnostic techniques used to detect bacterial infections in clinical samples are often unable to discriminate between living and dead bacteria and extracellular DNA. This presents a major problem, leading to reduced specificity and sometimes in a failure to properly monitor the effectiveness of antimicrobial therapy. An additional problem faced by molecular diagnostic assays as well as next generations sequencing (NGS)- based diagnostic approaches is the presence of high quantities of contaminating host DNA, which often obscures the presence of and changes in microbial populations due to the majority of sequencing reads aligning to host DNA. Here we describe the effect of two buffers used to achieve live/dead differentiation and the removal of contaminating host cell DNA. We evaluated these buffers in two different scenarios; one focusing on a previously validated TB diagnostic and one following a NGS-based approach. For the diagnostic method, 101 fresh sputum samples were treated with the live/dead differentiation buffer before lysis and molecular diagnosis by quantitative polymerase chain reaction (q-PCR). For future NGS applications, samples will first be treated with a differential lysis buffer to selectively lyse any host cells present. This will be followed by treatment with a DNA removal buffer and consequent q-PCR and sequencing. The q-PCR results were evaluated against Sputum smear microscopy (S), Culture (C) and GeneXpert MTB/RIF Ultra. Of the 101 samples analysed, 13 were culture contaminated and 7 were Non-Tuberculosis mycobacteria (NTM). The basic test identified 12 and the live/dead differentiation test 10 samples respectively as positive from the culture negative samples. A total of 11 samples indicated the presence of DNA from dead bacilli, where the ΔCt value was more than the threshold value of 2.5. The molecular test, both with and without live/dead differentiation, was able to identify 2 positive MTBC samples from the 13 contaminated cultures. Both the basic and the live/dead differentiation test, were able to identify 3/4 first time GeneXpert Ultra samples and had a 100% specificity for NTM. The inclusion of a live/dead differentiation buffer improves the specificity of the molecular test but may have a negative effect on the sensitivity especially on samples with a low bacilli starting concentration. A few false positives were indicated but could be due to the fact that some of the culture results were false negative as confirmed by follow-up tests. Further optimization of the live/dead differentiation process can enhance the specificity. This includes an increase in the concentration of the buffer or an initial liquefaction step of the sputum before proceeding with the live/dead differentiation step. The inclusion of a live/dead differentiation buffer indicated that DNA from dead bacilli was present in some of the samples