Member Research and Reports

Member Research and Reports

Texas A&M Researcher’s Cautionary Note on the Mutation Frequency in Microbial Research

Bacterial mutation is a recurring theme in biomedical and public health research. Not all bacterial strains mutate at the same rate. If a strain mutates at an exceptionally high rate (e.g., a 10-fold higher rate) relative to a reference strain, scientists call that strain a mutator strain. Identifying mutator strains is a major topic of interest in microbial research, with applications in public health research like better understanding of the dynamics of drug-resistant bacteria. Mutation frequency has long been a commonly used tool in microbial mutation research and could soon become a standard measurement in scientists’ relentless hunt for mutator strains. However, this could be problematic as mutation frequency is frequently misunderstood and often used inappropriately.

In an article in the journal Mutation Research – Fundamental and Molecular Mechanisms of Mutagenesis, Dr. Qi Zheng, associate professor in the department of epidemiology and biostatistics at the Texas A&M School of Public Health, outlines some of the shortcomings of using mutation frequency in microbial research and focuses on why researchers should resist the temptation to use mutation frequency instead of  the mutation rate to identify microbial mutator strains.

Dr. Zheng notes that some researchers have used the terms mutation frequency and mutation rate interchangeably; however, the two concepts have seemingly subtle differences that affect their usefulness as measurement tools. The mutation frequency is simply a measure of what proportion of cells in a given cell population are mutants. On the other hand, mutation rate describes the probability of a mutation happening when a cell divides.

[Photo: Dr. Qi Zheng]

The mutation frequency can be measured in one of two forms in the laboratory: the mean mutation frequency and the median mutation frequency. The former is the mean proportion of mutants across cultures, while the latter is the median proportion of mutants across cultures. Dr. Zheng noted that the mean mutation frequency can yield inconsistent results even between cultures of identical bacteria because the mean mutation frequency increases in line with the number of generations that have passed since the experiment was initiated. If two cultures had experienced different numbers of cell divisions they would have different mean mutation frequencies, despite having the same mutation rate. Some researchers have attempted to use the median mutation frequency in hope of circumventing that problem, because the median mutation frequency shows less variation than the mean mutation frequency. However, the correlation between median mutation frequency and mutation rate is disappointing. Citing simulation results, Dr. Zheng encouraged researchers to eschew both forms of the mutation frequency.

Another problem Dr. Zheng noted is that mutation frequency can be significantly affected by cell death. One example the paper cites is recent studies of Beijing strains of Mycobacterium tuberculosis, which are believed to be more likely to become drug resistant than other M. tuberculosis strains. Drug resistant bacteria are a growing public health threat and major topic of interest for public health researchers; however, in these cases, mutation frequency may be painting an inaccurate picture of drug resistance. When exposed to antibiotics, many of the non-mutated bacteria die, leaving mutants that are resistant to the drug. After multiple generations this would leave a population heavily skewed toward mutants. In Dr. Zheng’s hypothetical example this yields a mutation frequency that was about 64,000 times greater than the actual mutation rate.

Dr. Zheng noted that the widespread use of mutation frequency had been a long-overlooked obstacle to improving efforts to identify microbial mutators, and he emphasizes the importance of dropping the use of mutation frequency and embracing mutation rate as the preferred measure of microbial mutation. This change would add to researchers’ computational workload; however, new software tools such as “rSalvador” stand to reduce this.

“More research is still needed to improve the ability to identify bacterial mutators,” Dr. Zheng said. “However, moving from mutation frequency-based methods to techniques based on mutation rate is the first step toward improving search for mutators, and toward enhancing understanding of microbial drug resistance.”