Fighting Resistance With Calorimetry: New Tools for Antimicrobial Drug Development

Antibiotic-resistant bacteria are a growing concern, and with good reason—an estimated 23,000 deaths and two million illnesses occur annually in the U.S.1 Major contributing factors are overuse, lack of regulation and scarcity of novel antibiotics.

Antibiotic development and monitoring

Development of antibiotic-resistant compounds has largely followed traditional Pasteur-style microbiology, i.e., culturing and plating and manual operation. Molecular-based, nonculture methods that rely on DNA typing or proteomics are often high in cost, have low specificity and are unable to differentiate between living, dead and persistent dormant bacteria. A sensitive, label-free cell-based assay was needed to measure bacterial activity in real time with minimal effort.

Calorimetry-based microbial measurements

Calorimetry-based monitoring of living systems has fallen out of use because it is perceived to be complex. But calorimetry-based cell monitoring and the data produced are uniquely suited to the development of novel antibiotics.

Calorimetry measures the power produced in a cell culture at any given time as Joules/second (W). The heat generated is a measure of the metabolic processes in the cells and, as a consequence, gives a true phenotype fingerprint of the organism measured. Different bacteria and treatments create unique heat profiles that reveal significant information about the system tested. Calorimetry provides label-free, nondestructive measurement, and thus makes postexperimental analysis possible, while being independent of sample morphology. This means that assays can be performed on bacteria in solution as well as on solid media, including three-dimensional matrices such as bone biopsies and surgical and dental implant materials.

A unique property of calorimetry-based metabolic monitoring of bacterial growth is that the pattern of energy expenditure is species, as well as strain, specific—over time, each bacterium gives rise to a specific growth pattern as heat production. This can be used to quantify the number of bacteria and determine the species (see Figure 1). The bacterial load determination is similar to a quantitative PCR measurement in which the curves are identical in shape, but different numbers of cycles are needed to reach the detection limit, and different loads of bacteria require a varying number of cell divisions to reach the detection limit. The metabolic output assay thus becomes quantitative as well as qualitative. Minor changes in growth behavior such as metabolic pathway mutations are detected, as are biofilm formation and, most importantly, antimicrobial sensitivity.

 Figure 1 – Heat flow curves of serial dilutions of S. typhimurium demonstrating the shift in lag time and consistent maximal heat output. Curve shape is strain and media specific.

By integrating the metabolic power over time to accumulated heat over time (in Joules), a growth curve is established that is the equivalent of a traditional growth curve (as measured by optical density of the culture) (see Figure 2). From this data it is possible to calculate both the lag time and the maximum growth rate of the culture. This is the basis for determining the effect of antibiotic treatment.

 Figure 2 – Integration over time for the data from Figure 1. A growth curve can be derived by plotting the accumulated energy release. Fitting the data to growth models allows the determination of lag-phase time and maximum growth rate. The maximum growth rate corresponds to the maximum slope of the heat over time curve. Duration of the lag phase is measured as the time from the start of the experiment to the interception of a line tangent to the maximum growth rate point and the baseline.

A prolonged lag phase is indicative of a bactericidal action, since the starting number of live bacteria will be less, and a decrease in growth rate will suggest a bacteriostatic effect. The starting number of bacteria will not change, but the cell division time will increase. This data can be used to qualify the mechanism of action, based on the inhibiting properties and the curve shape, compared to substances with known mechanisms of action (see Figures 3 and 4). Dose-response curves, plotting dose against lag time and dose against maximal growth rate, are easy to derive from the data. Further, the total energy release for a given time frame plotted against concentration is a measure of the total biomass formation and can be used as a measurement of antibacterial efficiency.

 Figure 3 – Heat flow curves of E. coli growth without antibiotic and two different concentrations of antibacterial compound. The inoculum of bacteria is identical for all three samples. The blue curve depicts untreated sample; red represents the low concentration of antibiotics demonstrating a prolonged lag phase but similar maximum peak power. The green curve depicts a high antibiotic concentration demonstrating a shift in lag-phase time as well as growth rate and peak output power. All samples show that there is residual metabolic activity, even for prolonged experiments, and the cells still metabolize to some extent and a cell population is viable at high initial antibacterial load.
 Figure 4 – Integration over time for the data from Figure 3. The growth curves demonstrate that the low antibiotic concentration has a shift in lag-phase time but similar maximum growth rate indicative of a bactericidal action (red curve). The green curve, high initial concentration, shows a prolonged lag phase as well as a significantly decreased growth rate, indicative of additional bacteriostatic effect at high concentration.

An important consideration is a compound’s bioavailability. Since the calorimetric measurement is a measure of the total metabolism, bioavailability is a nonissue since it is accounted for in the measurement. Compare this to culturing assays on solid media, where the diffusion of compound in the media can lead to measurement errors. The formation of biofilms can also be a concern in the potency measurements of antibiotics since the efficacy of antibiotics differs between planktonic and biofilm growth. Biofilm formation can be monitored by calorimetry because the metabolic status and the treatment efficacy are clearly different.

Colonization in complex matrices like bone can be difficult to assay. “Normal” assays cannot provide a representative sample of bacteria colonizing three-dimensional surfaces. Consequently, large deviations may be found when using microscopy, fluorescence and molecular methods. The heat produced by bacterial metabolism in 3-D matrices can be measured regardless of the sample properties, permitting new areas of investigation.

The bacterial growth assay in calorimetry can be performed in both liquid and solid media so that different properties can be studied during colonization of, for example, dental and surgical implant materials. A wide range of possible growth conditions allow for the study of both aerobic and difficult-to-grow anaerobic systems, as well as for monitoring tuberculosis and other slow-growing mycobacteria.

Correct quantification of the number of cells, as well as the number of living cells, can be a challenge in antibiotic development. Since many bacteria give rise to clustered cells, biofilms, etc., there will likely be a misrepresentation when using standard plating/growth analysis. A single colony may originate from a cluster of living bacteria, thus giving false numbers for efficacy. Calorimeter-based assays account only for the actual number of metabolic active, live cells. This also has implications for the comparison to DNA or protein-based assays, where it can be difficult to distinguish between the number of live active cells and DNA/protein remaining in inactive/dead cells protected by biofilm. It is easy to monitor the metabolic activity for prolonged times using calorimetry—a typical assay runs from a few hours up to days or weeks if needed. This permits monitoring of persister cells or cells derived with antibiotic resistance from biofilm formation; they produce metabolic activity at a lower but constant rate during a prolonged time and can be distinguished in the assay. Up to 80% of all infections are complicated by bacteria forming biofilms,2 and antibiotics typically developed using bacteria in planktonic growth may be largely ineffective for treating biofilm-derived infections.

Another use of a calorimetric assay is in antibiotic development in native conditions such as a serum-supplemented growth media, mimicking the conditions in the human body. The possible degradation and instability of tested compounds can allow regrowth of persister cells; this can easily be monitored by following the total metabolic state of the culture for a prolonged time.

Increasing efficacy of antibiotics

Potentiating treatments are being used more frequently to increase antibiotic efficacy. Multiple modes of action of combined therapies and the use of potentiating compounds with no inherent antibiotic properties can be monitored using a calorimetric assay. Since there is no need to know the mechanism of action prior to the experiment, unbiased phenotype screening is achieved.

Conclusion

The 32-channel calScreener (Symcel Sverige AB, Kista, Sweden) enables efficient screening of lead compounds and dose response measurements, performing calorimetric assays in a microtiter plate-based format. Small sample volumes and multiple parallel channels increase throughput, with presterilized single-use consumables suited for bacterial growth.

Calorimetry-based monitoring of the spread of resistant strains is rapid, sensitive and cost-effective. Combining detection with an indication-based panel of antibiotics will make it possible to identify the presence of an infecting agent and determine the correct antibiotic treatment in hours rather than days.

References

  1. Antibiotic resistance threats in the United States, Centers for Disease Control and Prevention, 2013; http://www.cdc.gov/ drugresistance/pdf/ar-threats-2013-508.pdf.
  2. Hay, M.; Thomas, D.W. et al. Clinical development success rates for investigational drugs. Nature Biotechnol. 2014, 32(1), 40–51; doi:10.1038/nbt.2786.)

Additional reading

  1. Von Ah, U.; Wirz, D. et al. Rapid differentiation of methicillin-susceptible Staphy-lococcus aureus from methicillin-resistant S. aureus and MIC determinations by isothermal microcalorimetry. J. Clin. Microbiol. 2008 Jun, 46(6), 2083–7.
  2. Braissant, O.; Bonkat, G. et al. Microbial growth and isothermal microcalorimetry: growth models and their application to microcalorimetric data. Thermochim. Acta Mar 2013, 555(10), 64–71.
  3. Howell, M.; Wirz, D. et al. Application of a microcalorimetric method for determining drug susceptibility in mycobacterium species. J. Clin. Microbiol. 2012, 50(1), 16.
  4. Braissant, O.; Keiser, J. et al. Isothermal microcalorimetry accurately detects bacteria, tumorous microtissues, and parasitic worms in a label-free well-plate assay. Biotechnol. J. 2015 Mar, 10(3), 460–8.
  5. Costerton, J.W.; Stewart, P.S. et al. Bacterial biofilms: a common cause of persistent infections. Science 1999, 284, 1318–22.

Magnus Jansson, Ph.D., is chief scientific officer, SymCel Sverige AB, Isafjordsgatan 39B, SE-164 40 Kista, Sweden; tel.: +46 8 5000 49 26; e-mail: [email protected]www.symcel.se