Has resistance to chlorhexidine increased among clinically-relevant bacteria? A systematic review of time course and subpopulation data.

Chlorhexidine (CHX) was introduced for use as an antimicrobial more than 70 years ago. CHX has been and continues to be used broadly for disinfecting surfaces in medical and food service facilities as well as directly on skin of humans and animals. Considering its widespread use over many decades, questions of resistance to CHX have been raised. Additionally, questions of possible coincident resistance to the biocide and resistance to clinically relevant antibiotics have also been raised. A number of important questions remain, including is there consistent evidence of resistance, what is the degree of resistance, especially among clinically isolated microbial strains, and what is the degree of resistance compared to the typical concentrations of the biocide used? Data for microbial species isolated over the last 70+ years were compiled to construct as complete a picture as practical regarding possible resistance, especially among species in which resistance to commonly used antibiotics has been noted to be increasing. This is a compilation and analysis of individual MIC values for CHX reported in the literature, not a compilation of the conclusions individual authors reached. The data were analyzed using straight-forward and robust statistical procedures to detect changes in susceptibility to CHX over time, i.e. linear regression. Linear regression was supplemented with the use of nonlinear least squares regression analysis to detect the presence of population parameters associated with subpopulations of microbial strains which exhibit increased resistance to CHX. Pseudomonas aeruginosa, Klebsiella pneumoniae, and Acinetobacter baumannii were all found to have an increased resistance to CHX over time with the most profound change detected in A. baumannii. Additionally, subpopulations with log-normal distributions were found consistent with the presence of a baseline subpopulation of susceptible strains and a subpopulation with increased resistance to CHX. However, the CHX-resistant subpopulations did not correlate exactly with antibiotic resistance, so details of the relationship remain to be addressed. Increased resistance over time was not detected for Escherichia coli, Enterobacter faecalis, Staphylococcus aureus, or Candida albicans, although a subpopulation with greater than baseline resistance to CHX was detected among strains of E. faecalis and C. albicans. A difference in susceptibility to CHX was also detected between methicillin-resistant (MRSA) and methicillin-sensitive (MSSA) S. aureus strains. The levels of resistance to CHX detected were all markedly lower than concentrations routinely used in medical and food service applications. Reaching conclusions regarding the relationship between antibiotic and CHX resistance was complicated by the limited overlap between tests of CHX and antibiotic resistance for several species. The results compiled here may serve as a foundation for monitoring changes in resistance to CHX and possible relationships between the use of CHX and resistance to antibiotics commonly used in clinical medicine.

Use of mixture distributions to deconvolute the behavior of "hits" and controls in high-throughput screening data.

The stochastic nature of high-throughput screening (HTS) data indicates that information may be gleaned by applying statistical methods to HTS data. A foundation of parametric statistics is the study and elucidation of population distributions, which can be modeled using modern spreadsheet software. The methods and results described here use fundamental concepts of statistical population distributions analyzed using a spreadsheet to provide tools in a developing armamentarium for extracting information from HTS data. Specific examples using two HTS kinase assays are analyzed. The analyses use normal and gamma distributions, which combine to form mixture distributions. HTS data were found to be described well using such mixture distributions, and deconvolution of the mixtures to the constituent gamma and normal parts provided insight into how the assays performed. In particular, the proportion of hits confirmed was predicted from the original HTS data and used to assess screening assay performance. The analyses also provide a method for determining how hit thresholds--values used to separate active from inactive compounds--affect the proportion of compounds verified as active and how the threshold can be chosen to optimize the selection process.

Calculating the probability of detection for inhibitors in enzymatic or binding reactions in high-throughput screening.

In high-throughput screening (HTS) for drug candidates from a library containing tens of thousands to millions of chemical compounds, one problem is assessing the sensitivity of an assay for detecting compounds with a particular potency. For example, when looking for inhibitors of an enzyme, what is the potency of an inhibitor that will be readily detected by an enzyme inhibition assay? Similarly, when assessing compounds that inhibit binding between receptors and ligands or similar molecule-to-molecule interactions, what potency of an inhibitor will be readily detected? In this article, the well-established concepts of Michaelis-Menten kinetics and Langmuir binding isotherms are combined with fundamental statistical principles to yield a measure of assay sensitivity. The approach is general and can be modified to accommodate situations where the reaction kinetics is known to be more complicated than situations described by the Michaelis-Menten and Langmuir equations. The calculations presented take into account the concentration of inhibitor used, the variability of the assay, the relationship between the K(m) or K(d) of the reaction and the substrate or ligand concentration used, the threshold or cutoff value used for determining "hits," and the number of replicates used in screening.

Effects of membrane partitioning and other physical chemical properties on the apparent potency of "membrane active" compounds evaluated using red blood cell lysis assays.

The membrane-destabilizing properties of Amphotericin B and Zwittergent were used as benchmark compounds for examining in detail their membrane-altering effects in a series of human red blood cell lysis assays. The procedures included examining dose responses and the effects of different cell concentrations on potency in rbc lysis assays. In order to enhance detection of subtle membrane effects, we also used a range of NaCl concentrations to osmotically stress the rbc's. Using the benchmark compounds, a set of conditions was developed for examination of subtle membrane effects that may be applied to series of compounds with suspected membrane-perturbation activity. A group of experiments was defined that allow detection of the most important membrane-modifying behaviors among a diverse group of compounds. From an initial screen of bacterial growth inhibition over 150 compounds were examined for membrane-altering properties using the limited experimental protocols developed from the benchmark compounds. Several dose-response patterns were observed as useful for classifying compounds based on their tendency to alter membrane integrity and to partition into the lipids of membranes, as well as their propensity to form aggregates or precipitates. The methods may prove generally useful for distinguishing compounds whose primary activity is membrane destabilization from more interesting and useful pharmacological mechanisms of action.