(720c) Intracellular Absorption Underlies Collective Bacterial Tolerance Towards an Antimicrobial Peptide

Intracellular absorption underlies collective
bacterial tolerance towards an antimicrobial peptide

Fan Wu and Cheemeng Tan

Department of Biomedical Engineering,
University of California Davis, Davis, CA, 95616 USA

Statement of Purpose: Antimicrobial Peptides
(APs) are being developed as alternatives to classical antibiotics. However,
the efficacy of APs as antimicrobials in treatment is still limited by unknown
mechanisms beyond their direct interaction with the target bacteria. Previous
study and our own test show that bacteria that are typically sensitive to APs
can recover from inhibition by the APs without any known genetic mutations.
Here, we show that an AP can permeabilize cytoplasmic membranes of a subpopulation
of bacteria, which then absorbs the AP into its intracellular space, leading to
the recovery of other bacteria in the same population. To reveal quantitative
insights into this highly dynamic LL37-absorption mechanism, we combine
mathematical modeling with population and single-cell experiments in this
study.

Methods: We investigated
LL37, a cathelicidin family AP from human and Escherichia coli BL21PRO strain in our study. The
strain expressed lux genes, leading
to luminescence that was tracked as a surrogate of bacterial viability. The recovery
of the bacterial population under LL37 treatment was tracked using a
platereader (Fig. 1a). To track the
activity of LL37 molecules in bacterial culture, we utilized rhodamine-labeled LL37 (Rh-LL37), or his-tag labeled
LL37 (his-LL37) in our treatment. The activity, quantity, and location of the drug molecules were assessed
using wide-field single-cell microscopy (Fig.
1b), western blotting, flow cytometry,
and structured illuminated microscopy (Fig.
1d). We developed a mathematical model using ODEs to investigate the kinetics
of LL37-absorption. The parameters of the model were estimated by tracking the
status of different bacterial subpopulations during Rh-LL37 treatment using
flow cytometry. To further explore the LL37-absorption mechanism, we extended
our model to predict the effects of a second strain with different
LL37-absorption kinetics (E. coli
MG1655) or a synthetic peptide (LBP) on BL21PRO recovery (Fig. 1e&f).   

Results: Through a
series of qualitative assays, we show that the bacterial recovery during LL37
treatment is not due to any known resistant mechanisms, such as genetic
mutations or the degradation of the AP molecules. Using single-cell fluorescent
microscopy, we show that the bacterial cells absorb LL37 at a timing that
coincides with the permeabilization of their cytoplasmic membranes (Fig. 1b&c). Combining flow
cytometry and structured illuminated microscopy, we find that bacterial
population can be further characterized by three subpopulations during Rh-LL37
treatment: living subpopulation, binding
subpopulation (¢Ù in Fig.
1d), and absorbing subpopulation (¢Ú in Fig. 1d). Based on the finding, we
propose a mechanistic model where Rh-LL37 can bind to a subpopulation of E. coli (transition from living
subpopulation to binding subpopulation) and permeabilize it (transition from
binding subpopulation to absorbing subpopulation), which then absorbs the
remaining Rh-LL37 into its intracellular space (Fig. 1d left schematic), leading to the recovery of other bacteria
in the same population. Facilitated by a mathematical model, we further show
that bacterial strains have different kinetic parameters of absorption and, as
a result, one bacterial strain can protect another strain against killing by LL37
(Fig. 1e). Moreover, we show that a synthetic
peptide can delay the bacterial recovery by slowing down the absorption of LL37 (Fig. 1a&f).

Conclusions: Combining qualitative and quantitative approaches, we discover a unique bacterial tolerance
mechanism towards an antimicrobial peptide: LL37 permeabilizes a subpopulation
of E. coli, which then absorbs the
remaining LL37 into its intracellular space, and in turn allows other bacteria
to recover from the LL37 treatment. Leveraging the fundamental discovery, we
demonstrate new synthetic-biology strategies to enhance the efficacy of antimicrobial
peptides by controlling bacterial population dynamics during the treatment. We
are combining machine learning, genome-wide screening, and in vivo model systems to study multi-scale dynamics of the antimicrobial-peptide
tolerance mechanism.