The current threat of antimicrobial resistance can not be adequately determined strictly through a surveillance of bacterial pathogens. The majority of bacteria in the ecosystem of humans and animals are nonclinical and exist in a commensal relationship with their host. These commensal organisms may serve as reservoirs of antibiotic resistance genes if they are able to transfer resistance genes to and from other bacteria, including pathogens. How do we estimate the importance of this enormous commensal bacterial population? Cultivation techniques are inherently biased, and many of the organisms are unculturable. When probing an entire sample for resistance genes, these genes can often be found in diverse locations with varying levels of measurable selection pressure. Analytical models based on these data may not accurately assess the relationship between selection pressures and the observed resistance. How should a study be designed to accurately investigate the reasons why antimicrobial resistance genes persist? Without this knowledge, it is impossible to design effective interventions to reduce the risk of antibiotic resistance gene transfer between animals and humans. We believe that resistance gene load in a sample will provide a greatly superior means of assessing the strength of various selection pressures being exerted on the bacterial population within the sample. These quantitative estimates will be reflective of the entire bacterial community in the sample and will enable us to assess the short and long-term selection pressures being applied to the bacteria in that community.
Consequently, the Specific Aims of this project are:<OL> <LI> To optimize and validate real-time PCR in the quantification of antimicrobial resistance genes in fecal samples.<LI>To apply these quantitative PCR methods to a sample set that is already well-characterized. <LI>To assess the accuracy and precision of analytical techniques for quantifying the association between antimicrobial use and antimicrobial resistance.
NON-TECHNICAL SUMMARY: The increasing rate of development of bacterial resistance to antimicrobials has been well-documented, and this has major consequences for human and animal health. The results of this study will improve our ability to relate antibiotic use to antibiotic resistance in an accurate manner. This will enable the development of rational approaches to antibiotic use policy.
APPROACH: Real-time PCR methods will be validated to quantify the amount of specific resistance genes in the total bacterial population of a sample. These methods will then be applied to a sample set collected during a 2.5-year longitudinal study of antimicrobial resistance in dairy cattle. The results of culture, conventional PCR and quantitative PCR will be compared. Finally, we will use computer simulation to evaluate various analytical methods for associating antibiotic use with antibiotic resistance. In these models, data from culture, conventional PCR and quantitative PCR will be compared for their relative accuracy in predicting the true underlying association. The development of these quantitative data along with a thorough evaluation of different analytical models for predicting antibiotic resistance will enable more accurate, focused and effective intervention strategies for reducing the human and animal health risks associated with antibiotic resistance.