Bistable Bacterial Growth Dynamics in the Presence of Antimicrobial Agents

The outcome of an antibiotic treatment on the growth capacity of bacteria is largely dependent on the initial population size (Inoculum Effect). We characterized and built a model of this effect in E. coli cultures using a large variety of antimicrobials, including conventional antibiotics, and for the first time, cationic antimicrobial peptides (CAMPs). Our results show that all classes of antimicrobial drugs induce an inoculum effect, which, as we explain, implies that the dynamic is bistable: For a range of anti-microbial densities, a very small inoculum decays whereas a larger inoculum grows, and the threshold inoculum depends on the drug concentration. We characterized three distinct classes of drug-induced bistable growth dynamics and demonstrate that in rich medium, CAMPs correspond to the simplest class, bacteriostatic antibiotics to the second class, and all other traditional antibiotics to the third, more complex class. These findings provide a unifying universal framework for describing the dynamics of the inoculum effect induced by antimicrobials with inherently different killing mechanisms.

curves (Figures 2B,D,3B,4B,5B,7C) are found by the following algorithm: a) Smooth the data (by the MATLAb program smooth which uses a moving average filter over nsmooth=3 terms), b) Cut the data points which are below the OD reader resolved region (ODcut=0.11) c) To clear the oscillation pass the maximal capacity, set all data points after the maximal value of the smoothed data to this maximal value. d) Ignore the data points which have a slope which is small (smaller than half the averaged slope of the data). e) Compute the growth function for the smoothed, cut data : ODsm(t+dt)-ODsm(t)/dt and plot it versus the corresponding mid-point value ((ODsm(t)+ODsm(t+dt))/2). Figure S1 examines the dependency of the above algorithm on the number nsmooth of terms in the smoothing filter and demonstrates that the maximal capacity hardly depends on nsmooth whereas the maximal growth rate depends on nsmooth.
Maximal capacity : If there is growth, and if the growth is complete in the experiment time frame, the value of the maximal capacity ( Figures 3C,D, 4C,D, 5C-F,7D,E) is defined to be the maximal smooth OD value (last point on the solid growth curve), normalized by the mean of the control maximal smooth OD values. The growth is called completed, either when the cutting time is prior to the last possible one (one before last reading) or if the OD value is sufficiently large (larger than the control maximal capacity mean minus 3 standard deviations of the control maximal capacity data). We notice that the Maximal capacity produces a more robust criteria than the maximal growth rate, which is defined as the maximal values of the OD growth function divided by the OD value: Rmax = Max_t {2*[ODsm(t+dt)- ((ODsm(t)+ODsm(t+dt))]}, see Figure S1D.

Peptide loss of potency leads to BIK dynamics.
Figure S2 demonstrates that the dynamics displayed by bacteria in the presence of antimicrobial peptides is due to the peptides losing their ability to damage bacterial can survive the initial encounter with the peptide and show undisturbed growth later on. Indeed, comparing fresh Polymixin B (PMB), PMB that was incubated for an hour in LB media prior to usage, or PMB that was incubated for an hour with 10^6 cfu bacteria in LB (and then centrifuged to discard the bacteria), the 3 types of PMB give different inhibitory effects over an inoculum of 10^5 bacteria. This is shown in a PMB concentration low enough for the amount that was "wasted" upon the previous encounter with bacteria to be significant. As can be seen in Figure S2, the least potent PMB is the one incubated for an hour with bacteria, and the most potent (that also inhibits bacterial growth completely in this PMB concentration) is the freshly added PMB. Figure S2. PMB potency after incubation with bacteria. Bacteria were grown in different media as indicated for 16h and their OD600 was monitored in a microplate reader every 20 min. Bacteria grown in LB with PMB that was previously incubated with bacteria grow to full capacity, bacteria grown in LB with PMB that was previously incubated in 37°C grow late, and bacteria grown in LB with fresh PMB fail to grow (for each condition duplicates are shown).

Bacteria grown in the presence of commercial bacteriostatic antibiotics exhibit bistability and A-dependent dynamics (BAD)
Comment: Experiments were performed in two distinct plate readers. Often, results from one of the readers contained a distinct growth output till an OD600 of about 0.3. This growth is probably a result of technical issues in the said plate reader and didn't change the trends of the results. Below we present growth curves, growth functions and maximal capacity of all experiments performed with bacteriostatic antibiotics (Chloramphenicol and Tetracycline - Figures S4A and S4C), and the maximal capacity of 7 experiments with Chloramphenicol together ( Figure S4B).

Bacteria grown in the presence of commercial bacteriolytic and bactericidal antibiotics exhibit bi-stability and BMFD dynamics
Growth curves, growth functions and maximal capacity of all experiments performed with bactericidal and bacteriolytic antibiotics (Gentamycin, Kanamycin, Ampicillin and Carbenicillin see Figure S5 A-D).

β -lactamase resistant β-lactams induce different dynamics than regular β-lactams dynamics
Growth curves, growth functions and maximal capacity of all experiments performed with Oxacillin ( Figure S6).