Developing More Rapid, Accurate and Cost-Effectiveness BRD Diagnostics

Background

Bovine respiratory disease (BRD) is a major cause of morbidity, mortality and economic losses in feedlot cattle. While many factors play a role in the development of BRD, bacterial infections are an important contributor, and antimicrobial resistance (AMR) in these bacteria hampers control efforts.

Complicating things further, not all bacteria are equally capable of causing disease. For example, there are different serotypes (strains) of Mannheimia haemolytica (one of the main BRD bacteria). Serotypes 1 and 6 frequently cause BRD, whereas serotype 2 is typically found in healthy cattle. Some bacteria may be carrying antibiotic resistance genes and others don’t (but this requires a whole different diagnostic procedure). Sometimes clusters of antibiotic resistance genes are carried on mobile genetic elements that can be easily traded with other bacteria. This can lead to rapid spread of antibiotic resistance.

Current diagnostic tools are slow, cumbersome and impractical in large-scale, commercial feedlots. New tests to identify bacteria carrying antimicrobial resistance or virulence genes that make them more likely to cause clinical BRD would provide an important tool in the management of BRD. New diagnostic methods are critical to provide timely, practical information to guide management of BRD and antimicrobial stewardship practices to help combat AMR.

Objectives

• Design recombinase polymerase amplification assays to detect Histophilus somni strains capable of causing bovine respiratory disease, and identify antimicrobial resistance genes (ARGs) associated with these and other bacteria from deep nasopharyngeal swabs obtained from feedlot calves.
• Use recombinase polymerase amplification to detect four main bacteria associated with BRD, their antimicrobial resistance genes, associated mobile genetic elements, and H. somni ibpA genes in samples from calves on arrival and at 13- and 36-days post-arrival at a research feedlot, as well as from calves that are sick with BRD; evaluate the performance of RPA as a diagnostic test.
• Use recombinase polymerase amplification to detect four main bacteria associated with BRD, their antimicrobial resistance genes, associated mobile genetic elements, and H. somni ibpA genes in samples collected in the field from commercial feedlot cattle near the time of arrival; evaluate the performance of RPA as a diagnostic test under field conditions.

What they Did

This project used a technology called “recombinase polymerase amplification” (RPA), an experimental test that runs at a constant, low temperature, requires minimal equipment or sample preparation, and produces results from extracted DNA in as little as 30 minutes. A new RPA test was developed to detect whether feedlot calves were carrying strains of Histophilus somni with an increased risk of causing disease. Another RPA test was developed that can detect up to three important ARGs simultaneously in deep nasopharyngeal swab (DNPS) samples.

Samples collected from high-risk, auction-mart calves were subjected to these new RPA tests and other recently developed RPA tests that identify bacteria associated with BRD and their mobile genetic elements. Calves were sampled on arrival at a research feedlot and again 13 and 36 days later, and at the time of diagnosis for any calves that became sick with BRD. RPA results were compared to those of bacterial culture and antimicrobial susceptibility testing to determine their sensitivity (i.e., can RPA identify samples with a specific gene?) and specificity (i.e., can RPA discern samples without a gene target?).

The same tests were applied to samples collected from commercial feedlot calves to determine the performance of RPA under field sampling conditions.

What They Learned

This is the first identified report to develop a real time RPA assay for the detection of multiple important macrolide resistance genes, as well as the first use of a conserved genetic region coding for a specialized H. somni protein (IbpA) as diagnostic target for RPA. The latter RPA test could detect as few as 226 copies of the ibpA gene per sample.

The macrolide AMR RPA assay was able to detect ARGs (msrE, mphE, erm42) in DNPS samples from high-risk, experimental calves, with a clinical sensitivity of 95% and specificity of 57% for detecting the group of assayed macrolide resistance genes under the conditions of the pilot test using selected samples from a single study. The presence of erm42 was low in our samples, making it difficult to evaluate the ability of RPA to detect it accurately.

Using these and previously published RPA tests for important BRD bacteria and mobile genetic elements that allow transmission of groups of AMR genes between these bacteria, RPA was used to test DNPS samples from cattle at different timepoints near the time of arrival at a research feedlot. When compared alongside bacterial culture, RPA was highly specific and able to accurately identify which samples did not contain BRD-associated bacterial pathogens, meaning that a positive test result can be trusted to rule in the presence of these bacterial targets.

The samples obtained from experimental and commercial feedlot cattle were 25-63% positive for one or more mobile genetic element targets and 47-90% positive for any of the macrolide gene targets, while 67% of the H. somni-positive DNPS samples collected from commercial feedlot calves contained the H. somni-disease-associated IbpA protein.

What It Means

In this study, new RPA tests were created to identify ARGs specific to macrolide drugs (e.g. tulathromycin) and a protein linked to H. somni disease in cattle. These new tests show promise for identifying genes in BRD-associated bacteria that can contribute to the development of more severe disease and challenges in BRD management.

RPA is fast, easy to use, and requires little equipment to identify ARGs and other genes important to BRD. RPA shows promise as a potential diagnostic method for smaller groups of cattle. However, as a target-based test, it is restricted by the number of known targets we can test for. As more genes are identified, RPA testing related to BRD needs to be optimized, as it could become limited by the cost of testing for additional targets.

This means three things. An RPA test cannot find things it is not told to look for. So if there are 5 genes involved in resistance to a particular antimicrobial, but the RPA test is only looking for three of the gene targets, the test may not predict antimicrobial resistance with 100% accuracy. Secondly, bacteria can share their antimicrobial resistance genes with other species. So the three genes targets the RPA test may not all be come from the same species of bacteria. Finally, an RPA test may look for both BRD bacteria and antimicrobial resistance, but the test only finds the gene targets, not where they’re from. So there’s no way to know whether the BRD pathogens themselves are antimicrobial resistant or not.

RPA tests to diagnose BRD pathogens or antimicrobial resistance are much faster and simpler than traditional diagnostic tests. But a high “false negative” rate means they may simply be generating the wrong answer quickly. More refinement is needed before they can be deployed with confidence.