Effects of Anaerobic and Aerobic Digestion at Mesophilic and Thermophilic Temperatures on the Quantity of Carbapenem Resistance Genes in Wastewater Sludge

Antibiotic resistance genes (ARGs) encode resistance against antibiotic drugs in bacteria. Carbapenem antibiotics are one of the last lines of defence against bacterial infections and carbapenem resistance genes (CRGs) accumulate in sludge during wastewater treatment. Anaerobic and aerobic digestions are commonly used for sludge treatment. Due to the public health relevance of the land application of digested sludge, this research aimed to investigate the effects of the digestion on the abundance and removal of CRGs. To reach the aim, sludge samples were obtained from mesophilic and thermophilic digesters, sludge DNA was isolated and purified, and target genes were quantified. The target genes included CRGs (blaGES, blaOXA-48 and blaIMP-27), 16S rRNA and Int1. Sharp declines of up to 96.9% in DNA extraction yields were observed during biological digestion. As well, the purity of the DNA isolates declined up to 79.5% in 10 days. These observations indicate that the change in DNA extraction yield was at least partly due to the change in DNA extraction efficiency. Therefore, the absolute quantity of the target genes of each sludge sample was defined as copy number of the gene per unit mass of DNA. The absolute quantity declined during the process, and it was analyzed by kinetic models. The absolute value of the first-order kinetic rate constants followed the order blaOXA-48 > blaGES > Int 1 > 16S rRNA during batch anaerobic digestion. Because gene quantification is time-consuming, costly and sensitive to contamination, multiple regression models were proposed to estimate the abundance of target genes from sludge parameters that are routinely measured for monitoring and design of biological sludge digesters. The multiple regression models were constructed in linear (LM) and non-linear (NLM) fashions. Adjusted R2 ranged 0.352-0.988 and 0.916-0.996 for the LM and NLM models, respectively. Overall, the NLM models predicted the abundance more accurately than the LM models. This research proposed the first NLM models to simulate the abundance of target genes during aerobic digestion of sludge. The NLM models may be useful tools for researchers and practitioners of biological sludge digestion to estimate the removal rate of target genes under different conditions.