Dynamic Effect of β-Lactam Antibiotic Inactivation Due to the Inter- and Intraspecies Interaction of Drug-Resistant Microbes

Abstract

The presence of β-lactamase positive microorganisms imparts a pharmacological effect on a variety of organisms that can impact drug efficacy by influencing the function or composition of bacteria. Although studies to assess dynamic intra- and interspecies communication with bacterial communities exist, the efficacy of drug treatment and quantitative assessment of multiorganism response is not well understood due to the lack of technological advances that can be used to study coculture interactions in a dynamic format. In this study, we investigate how β-lactamase positive microorganisms can neutralize the effect of β-lactam antibiotics in a dynamic format at the inter- and intraspecies level using microbial bead technology. Three interactive models for the biological compartmentalization of organisms were demonstrated to evaluate the effect of β-lactam antibiotics on coculture systems. Our model at the intraspecies level attempts to mimic the biofilm matrix more closely as a community-level feature of microorganisms, which acknowledges the impact of nondrug-resistant species in shaping the dynamic response. In particular, the results of intraspecies studies are highly supportive of the biofilm mode of bacterial growth, which can provide structural support and protect the bacteria from an assault on host or environmental factors. Our findings also indicate that β-lactamase positive bacteria can neutralize the cytotoxic effect of β-lactam antibiotics at the interspecies level when cocultured with cancer cells. Results were validated using β-lactamase positive bacteria isolated from environmental niches, which can trigger phenotypical alteration of β-lactams when cocultured with other organisms. Our compartmentalization strategy acts as an independent ecosystem and provides a new avenue for multiscale studies to assess intra- and interspecies interactions.

Introduction

Antimicrobial resistance is the ability of microorganisms to survive when exposed to antibiotics/drugs due to either inherent or acquired genetic traits. Since bacteria have acquired or selected genes of resistance, microorganisms with conferred antibiotic resistance properties exert a substantial influence directly or indirectly on other organisms.¹⁻⁴ For example, many horizontal gene transfer processes have the potential to alter the composition of microbial consortia.^5,6 Antibiotic resistome, which comprises of the collection of all antibiotic resistance genes could be reshaped in an isolated ecosystem such as the human gut.^7,8 Herein, the dynamics of bacteria colonies (i.e., perturbations of bacterial composition or genetic alterations) could potentially influence mammalian cell physiology and immunity, which is common in patients with chronic inflammatory disorders.^9,10 Further, excessive and repetitive use of antibiotics has the potential to disrupt biological networks, trigger an imbalance in microbiota, and lead to the selection of highly resistant bacteria.¹¹ Growing evidence suggests a high prevalence of antimicrobial resistance in clinical patients and the widespread dissemination of antibiotic-resistant bacteria from environmental sources into the human body. This interaction was thought to be responsible for adverse clinical outcomes, such as a decrease in treatment efficacy, among other complications.¹²⁻¹⁵

The β-lactam class of antibiotics encompasses a broad spectrum of antimicrobials, including ampicillin, penicillin, and cephalosporins, comprising 65% of the world market for antibiotics.¹⁶ To date, β-lactamases have been considered as one of the major mechanisms of resistance in the fight against β-lactam class of antibiotics by acting on the bacterial cell wall synthesis.^17,18 This mechanism of resistance is even more common in microorganisms that reside in the human gut and play a protective role in minimizing the damage caused by antibiotics.¹⁹⁻²² Conceptualized approaches to β-lactamase administration were shown to impart changes to the gut resistome through antibiotic inactivation.^23,24 These studies emphasized the potential of the inactivation of β-lactam antibiotics to protect the gut microbiome in the gastrointestinal tract (GI) tract. However, the role of β-lactamases on various organisms is not fully understood due to the lack of model systems that can facilitate the evaluation of specific organisms’ interaction in a dynamic manner. Beyond monitoring the presence of β-lactamases and their organisms, studying the interactions between β-lactamase expressing bacteria and other organisms is highly important as it provides information about antibiotic resistance, including influence, selection, and dissemination.

The present study demonstrates three interactive models to assess the compartmentalization of drug-resistant bacteria and their nullifying effect on ampicillin. Bacterial enzymes are crucial in imparting antibiotic resistance even though multiple drug-resistance factors (e.g., altered target sites, limiting uptake, active efflux, and/or inactivation of a drug) are involved at the same time.^18,25,26 Conventional approaches (e.g., liquid culture, agar plate culture, etc.)²⁷⁻³⁰ to study bacterial enzymatic factors are limited in their ability to discern the dynamic interaction of microorganisms. A pertinent model is biofilm-like compartmentalization considering factors such as biocompatibility, diffusion of secreted molecules, and stability.^31,32 Technological efforts to assess dynamic interactions utilizing microfluidic devices^33,34 exist, but multiscale studies are limited and cannot be generalized. Further studies that demonstrate intra- and interspecies communication of bacterial communities^35,36 are rare due to the lack of model systems that can provide a conducive environment for such studies. To this end, we utilized a soft hydrogel-based entrapment model to study the interaction of β-lactamase positive bacteria and its role in enzyme-mediated alteration of drug efficacy at the inter- and intraspecies level. Hydrogel core–shell compartments with precisely controllable features (e.g., size, shell thickness, bacterial density, etc.) have the potential to serve as an independent ecosystem and afford coculture studies in vitro. Further, β-lactamase positive and negative bacterial strains (Bacillus toyonensis, Micrococcus luteus, Escherichia coli, and Klebsiella oxytoca) isolated from human stool and soil samples were examined for the validation of our biomimetic compartmentalization models in relation to antibiotic resistance.

Experimental Section

Bacteria Culture and Antibiotic Resistance

All of the strains used in this study were listed in Table S1. E. coli DH5α (EC0111) was purchased from Invitrogen. E. coli DH5α pAAV-amp (E. coliamp; β-lactamase expressing bacteria) and E. coli DH5α pSmart-tet-kan-sfGFP (E. coligfp-kan) were used. E. coli DH5α were grown in Luria–Bertani (LB) broth medium (BD Difco, Miller) or LB agar plates (BD Difco, Miller) at 37 °C. E. coligfp-kan was grown in LB broth medium supplemented with 50 μg/mL of kanamycin (Sigma) at 37 °C. E. coliamp was grown in LB broth medium supplemented with 50 μg/mL of ampicillin (Sigma) at 37 °C. The agar plate was prepared by the addition of 1.5% agar (w/v). Bacteria were cultivated overnight on the agar plate, and a single colony was picked and inoculated in the media with shaking. Antibiotic resistance was verified either in liquid media or on agar plate-supplemented antibiotics.

Colony PCR was performed to confirm drug-resistant genes in the plasmid of E. coliamp and E. coligfp-kan. Bacterial colony was picked from the agar plate using a sterile tip and resuspended in the PCR master mix. Cell lysis was performed by keeping the PCR master mix at 94 °C for 3 min and then followed by the standard PCR protocol. PCR products were run on a 1% agarose gel prepared and run in 1× TBE (Tris-borate-EDTA) buffer for 1 h and visualized using the Chemidoc imaging system (Biorad) and compared with a 100 bp ladder (New England Biolabs).

Bacterial Strain Isolation

Bacterial strains were isolated from two different microbial niches: soil (collected from the location coordinates: 40.1132974, −88.2177687 Urbana, IL) and human stool (sampled from Crohn’s disease patient). 100 mg of soil or stool were resuspended in 1× phosphate buffer saline (PBS) and then serially diluted up to 10^–6 folds. The last 2 dilutions (10^–5 and 10^–6) were streaked on Luria–Bertani (LB) agar medium supplemented with/without 100 μg/mL ampicillin. These plates were incubated at 37 °C in aerobic conditions. After 24 h of incubation, pure colonies were picked up and transferred to an LB medium containing 100 μg/mL ampicillin. These single colony cultures in broth were preserved as frozen stocks and stored at −80 °C for downstream experimentation.

Bacterial DNA Extraction and PCR

Genomic DNA (gDNA) of bacterial isolates was extracted using the PureLink Genomic DNA mini kit (Invitrogen) according to the instructional manual provided by the company. The concentration and purity of DNA was evaluated by the NanoDrop 1000 spectrophotometer (Thermo Fisher Scientific). PCR was performed using the Phusion High-Fidelity PCR Master Mix with high-fidelity (HF) buffer (Thermo Fisher Scientific) with a total reaction volume of 30 μL. Primers listed in Table S2 were used to amplify the 16S rRNA gene for sequencing and detect different β-lactamase genes in the genome of these isolates. After the identification of these isolates based on the 16S rRNA gene, sequences of the β-lactamase genes encoded in the genome of these isolates were acquired, and primers were designed using the NCBI Primer-Blast tool and ordered from IDT (Coralville, IA). The PCR reaction mixture contained 15 μL of PCR master mix, 50 ng of gDNA as a template, and the final concentration of 0.5 μM primers. Molecular-grade water was added to constitute a total volume of 30 μL.

Characterization of Bacterial Isolates

Bacterial isolates were identified by sequencing the full-length 16S rRNA gene using 27F and 1492R primers.³⁷ After PCR, the reaction mixture was cleaned using the Genjet PCR purification kit (Thermo Scientific). Purified PCR products were sequenced using Sanger sequencing (Applied Biosystems 3730xl DNA Analyzers–Thermo Fisher). The resulting sequences were analyzed and processed using the software Sequencher version 5.4. Low-quality 5′ and 3′ ends were trimmed, and the forward and reverse sequences were assembled automatically using standard parameters in the Sequencher. The minimum overlapping region was 20 base pairs for the forward and reverse sequences. The processed sequences were blasted against the NCBI nucleotide database for species identification. Sequence similarities criteria were conclusive if being equal or greater than 99.5% with known sequences in GenBank. These 16S rRNA gene sequences (>1300 bp) were deposited in GenBank, and the following accession numbers were obtained: OP340990, OP340991, OP340992, and OP340993.

Hydrogel Core–Shell Encapsulation of Bacteria

All of the bacterial strains were prepared without further genetic transformation in encapsulation studies. Encapsulation processes were followed by the reported method.^32,38 Briefly, bacteria were prepared and cultured overnight on agar plates with the appropriate antibiotics. A single colony was picked and inoculated in the medium with shaking to reach an exponential growth phase (optical density 600: OD₆₀₀ = 0.5, approximately in arbitrary units). A ultraviolet/visible (UV/vis) spectrophotometer (Eppendorf Biophotomer) was used for the measurement of bacteria density in the media. Bacteria aliquots at the midexponential phase were collected for the estimation of bacteria concentration by the standard agar plate count. Bacterial aliquots at the midexponential phase (OD₆₀₀ = 0.5) were resuspended in fresh media prior to use in the encapsulation processes and then thoroughly mixed with 3 wt % (w/v) of alginate solution. Alginate core beads were fabricated using a dropwise process into a 0.1 M CaCl₂ solution. The gelled core beads containing bacteria at a desired density were transferred to a low concentration of alginate solution (less than 0.1 wt % due to viscosity) to fabricate a hydrogel-shell layer. Upon contact, the concentration of the alginate solution increased very quickly by adding 1 wt % of alginate solution (final concentration was up to 0.5 wt %). The reaction vessel was then vigorously shaken for the prevention of core-bead aggregation. The process of a hydrogel-shell layer can be stopped by adding an excessive amount of deionized (DI) water. The formed core–shell beads were transferred into a 0.01 M CaCl₂ solution under mild stirring for stabilization and stored in refrigerated conditions at 4 °C until further use.

Hydrogel Stability and Leakage Evaluation

The stability of formed hydrogel beads with bacteria growth was observed using an inverted microscope (Leica DMI3000B) equipped with a CCD camera (Qimaging EXi Blue). Bacteria leakage was evaluated in nutritionally complete LB media. Two groups of E. coli DH5α beads with/without the hydrogel shell were cultured for 48 h in 10 mL of LB media at 37 °C. The supernatant in each group was measured by spectrophotometry every 6 h to test for the leaked bacteria. Bacterial leakage from the hydrogel samples fabricated in this study was verified prior to experiments in a similar manner.

Biomass and Protein Quantification

A single hydrogel bead containing bacterial colonies was harvested at each time point for the evaluation of the biomass and proteins. The collected beads were degraded by adding 0.1 M sodium citrate (Sigma/S4641), and then fresh LB was added. Measurement of the OD₆₀₀ of the samples was by spectroscopy to estimate bacterial densities and biomass in the bead. The supernatant of the sample was discarded after centrifugation at 15,000 rpm for 5 min. The biomass pellets were used in the quantification of intracellular proteins. Degraded biomass samples were solubilized by incubation in 0.1 M NaOH and 0.1% sodium dodecyl sulfate (SDS) at 95 °C for 15 min. Quantification of total proteins was performed with the bicinchoninic acid (BCA) assay (Thermo Scientific) per manufacturer’s instructions. Data were normalized against the total protein content determined by the BCA assay. Measurements from three biological samples were repeated in triplicate.

Metabolic Activities

Hydrogel-shell beads containing bacteria colonies were collected at each time point after incubation. Metabolic activity was estimated based on an MTT colorimetric assay (Invitrogen). Briefly, to determine the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) reduction activity of microbial colonies entrapped in beads, beads were incubated with 100 μL of MTT solution (25 μg/mL) mixed with LB media at 30 °C in a mini thermal incubator (Thermo Fisher Scientific). After 2 h of incubation, the media was removed, and then the water-insoluble formazan crystal trapped in the bead was dissolved by adding dimethyl sulfoxide (DMSO). After 8 h, the formazan crystal solution dissolved from the bead was measured by spectroscopy using a Synergy H1 plate reader (BioTek) at 560 nm.

β-Lactam Degradation Test

Disc diffusion susceptibility and planktonic cell growth test were performed for the evaluation of hydrolyzed ampicillin using a negative control strain, E. coli DH5α. β-lactamase positive (E. coliamp) beads with hydrogel shell were prepared and incubated for 24 h in 5 mL of LB media containing ampicillin at 100 μg/mL. Susceptibility testing for hydrolyzed ampicillin was based on the agar overlay disc diffusion method. The standardized inoculum of E. coli DH5α was spread on the LB agar surface, followed by the disc diffusion protocol.³⁹ The plate was left for 1 h at room temperature to allow the diffusion of the media containing hydrolyzed ampicillin and then inverted and incubated overnight at 37 °C. The diameter of the inhibition zone was measured to evaluate ampicillin degradation. In addition to verifying ampicillin degradation in the media, β-lactamase negative E. coli DH5α were inoculated in the LB media cultivated with β-lactamase positive hydrogel beads. Planktonic bacterial growth of E. coli DH5α was measured by a UV/vis spectrophotometer.

Nitrocefin Assay

To test β-lactamase-catalyzed reactions, a nitrocefin colorimetric assay was performed. Briefly, hydrogel-shell beads containing β-lactamase positive bacteria were prepared and incubated for 24 h. The colonized beads were washed once with DI water and transferred to 48-well plates (a single bead per well). The β-lactamase-catalyzed activity was determined in 500 μL of a reaction mixture with 0.1 mM nitrocefin substrates (Sigma). The mixture with hydrogel-shell bead encoding β-lactamase enzyme was incubated at room temperature, in which the yellow substrate nitrocefin (λ_max = 390 nm) was converted to a red product (λ_max = 486 nm). The reactive solution of nitrocefin assays was transferred into 96-well microtiter plates (Corning Costar) and measured using a microplate reader (Synergy H1, BioTek Instruments) fitted with a 486 nm filter in the absorption mode. Hydrolysis rates were calculated from plots of the linear range of the increasing absorbance at 486 nm. Sulbactam (Sigma) at a concentration of 10 μg/mL was used for the β-lactamase inhibitor to verify the competitive inhibition of β-lactamase hydrolysis of β-lactamase positive hydrogel beads. Nitrocefin colorimetry assay of all bacteria and hydrogel beads was performed in the same manner prior to experiments and after encapsulation.

LC-ESI-MS/MS Analysis

Antibiotic degradation was analyzed by ultraperformance liquid chromatography–mass spectrometry (UPLC-MS) procedures. Briefly, the supernatant from the bacteria beads cultured in the media at 100 μg/mL of ampicillin was collected to evaluate ampicillin degradation. Supernatants were analyzed with UPLC-MS using the SYNAPT G2-Si system (Waters Co.) at the University of Illinois, Urbana–Champaign School of Chemical Sciences Mass Spectrometry Laboratory.

The LC separation was performed using the Acquity BEH C18 column (1.7 μm, 2.1 mm × 50 mm). For the gradient elution, 2 mobile phases were used: mobile phase A contains 95% water, 5% acetonitrile, 0.1% formic acid, and mobile phase B contains 95% Acetonitrile, 5% water, 0.1% formic acid. The flow rate was kept at 0.5 mL/min. The injected volume of the sample was 0.1 μL. The following elution conditions were used: initially, mobile phase A was 100% and kept for half a min. Over the next 3.5 min, mobile phase B was linearly increased, reaching up to 30% at time T = 4 min. For the next 2 min, mobile phase B was linearly increased to 100% and kept for 2 min. Then, a steep reversal to the initial conditions was done from time 8 min and maintained for up to 10 min.

The LC eluents were introduced into the mass spectrometer equipped with electrospray ionization (ESI) and were run in positive ion mode for ampicillin detection. The following optimized conditions were used: capillary voltage of 3 kV, desolvation temperature of 350 °C, cone voltage of 30 V, collision energy of 4 eV, collision gas helium, source temperature of 120 °C, cone gas flow of 5 L/h, and desolvation gas flow of 800 L/h. The mass range was 50–2000 Da. Mass Lynx v4.1 (Waters) was used to run the instrument, and then the chromatogram was analyzed for ampicillin peaks.

Coculture Study with Neighboring Beads

E. coliamp (β-lactamase positive) was grown in LB medium containing ampicillin at 50 μg/mL, and E. coligfp-kan (β-lactamase negative) was grown at 37 °C in LB medium containing kanamycin at 50 μg/mL. The agar plate was prepared by the addition of 1.5% agar (w/v). Bacteria were cultivated overnight on agar plates, and a single colony was picked and inoculated in the media with shaking. E. coliamp beads and E. coligfp-kan beads were prepared, respectively, and incubated together in 35 mm Petri dishes at 37 °C for 24 h in 10 mL LB media containing ampicillin at 50 μg/mL. β-lactamase negative (E. coligfp-kan) beads containing bacteria colonies were collected at each time point after incubation, and gfp fluorescence was measured using a Synergy H1 plate reader (BioTek). After incubation for 48 h, fluorescence images of hydrogel samples were characterized using a Chemidoc XRS system (Biorad). Data are shown as relative fluorescence (arbitrary) units (RFU).

Coculture Study in a Stratified Bead

Hydrogel beads with a hierarchical structure were fabricated by alternating the procedures of multilayered hydrogel fabrication based on previous reports.³² Briefly, E. coligfp-kan core beads were prepared as described above. Before fabricating the hydrogel-shell layers, an aliquot of E. coliamp in the midexponential phase was mixed with an alginate solution. The subsequent procedures were the same for the fabrication of the hydrogel-shell beads. By changing the concentration of E. coliamp bacterial aliquots, three types of stratified hydrogel architecture could be fabricated in the layered geometries. The stratified hydrogel beads were incubated in 35 mm Petri dishes at 37 °C for 24 h in 10 mL of LB media containing ampicillin at 50 μg/mL and observed to verify the growth of β-lactamase negative (E. coligfp-kan) bacteria in the core structure using a fluorescent microscope. The dynamic fluorescent signal emitted from E. coligfp-kan entrapped in the core structure was measured at an emission wavelength of 521 nm every 5 min over 20 h. Data are shown as relative fluorescence (arbitrary) units (RFU).

Coculture Study with Human Cancer Cells

Human renal carcinoma A498, colorectal carcinoma HCT-116, and breast adenocarcinoma MCF-7 supplied by ATCC were used for in vitro coculture studies with bacteria beads. A498 cells were grown in Eagle’s MEM with l-glutamine (EMEM; ATCC) with 10% fetal bovine serum (FBS; Gibco), supplemented with 1% penicillin/streptomycin (Lonza). HCT-116 and MCF-7 cells were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM, Corning) supplemented with 10% fetal bovine serum (FBS, Gibco) and 1% penicillin/streptomycin (Lonza). The cells were maintained in a CO₂-humidified incubator (at 37 °C and 5% CO₂). Prior to coculture experiments with bacteria beads, the WST-1 (Roche Applied Science) assay for cytotoxicity was performed to determine the 50% inhibition concentration (IC₅₀) value of ampicillin. Sigmoidal dose–response curves were generated using Origin 2022 software (Origin Lab), and the IC₅₀ value was obtained.

Before initiation of cocultivation experiments utilizing hydrogel-shell beads containing bacteria, biomass production and metabolic status of bacteria colonized in a hydrogel-shell structure was determined in EMEM with l-glutamine with 10% fetal bovine serum (FBS; Gibco) with/without ampicillin (at a concentration of 100 μg/mL). To verify β-lactam inactivation in cocultures at IC₅₀ concentration of ampicillin, A498 cells were seeded on standard 96-well plates at a density of 3 × 10³ per well and preincubated with β-lactamase bacteria beads for 12 h. The cells were subsequently incubated in an IC₅₀ concentration of ampicillin for 24 h. The inactivation effect of β-lactam antibiotics with bacteria beads on cell proliferation was measured by the WST-1 assay (Roche Applied Science). Morphological features were observed using an inverted fluorescence microscope (Leica DMI3000B) after coculture.

Statistical Analysis

All assays and experiments were performed in triplicate and repeated at least three independent times, otherwise noted. The data for the experiments was expressed as mean ± standard deviation (s.d.) and analyzed using unpaired Student’s t-test and one-way ANOVA with a posthoc test (Graphpad prism 9.0.), where posthoc comparisons were conducted using Tukey’s method. P values <0.05 were considered statistically significant, and the details of the statistical tests are presented in each Figure legend. No statistical analysis was performed to predetermine a required effect size.

Ethical Statement

Human stool samples were collected under an approved IRB (IRB# 19DHI2003) by the Carle Foundation Hospital (Urbana, IL). All participants provided written informed consent prior to participating.

Results

β-Lactam Biodegradation System

Bacteria have been exploited in the transformation of raw materials through various routes as biocatalysts.^40,41 Experimental concepts of a β-lactam inactivation system are depicted in Figure 1. β-lactamase positive bacteria produce β-lactamase enzymes to inactivate β-lactam antibiotics when exposed to β-lactam antibiotics for survival. Exploiting the capabilities of microorganisms involved in β-lactam transformation is an important step in the development of individualized compartments (independent hydrogel encapsulation systems) for the degradation of β-lactam antibiotics.

Figure 1.

Schematic of the experimental design for the β-lactam enzymatic degradation system. Three coculture models were developed to evaluate the effect of β-lactam antibiotic degradation. (i–iii, upper panel) from left to right: neighboring communities, stratified communities, and other organisms. β-lactam antibiotics can be inactivated through bacteria-based functional hydrogel bead encapsulating β-lactamase expressing bacteria. Experimental validation of the β-lactam enzymatic degradation system (lower panel). Various bacteria samples (stool and soil) were collected from humans and soil. After the isolation of uncharacterized bacteria, β-lactamase expression was tested prior to encapsulation. In the last step, coculture studies with encapsulated bacteria isolated from various sources were performed to validate the effect of β-lactam degradation on the viability of cancer cells.

β-lactamase positive bacteria can be recapitulated in alginate architectures enclosed by hydrogel-shell layers, which provides a stable environment for the colonization of bacteria at a high cell density due to its biofilm-like characteristics. Figure 1 depicts the encapsulation of drug-resistant bacteria and its ability to neutralize the effect of an antibiotic on the neighboring bacterial communities (Figure 1, panel (i)), stratified bacterial communities (Figure 1, panel (ii)), and its interaction with human carcinoma cells (Figure 1, panel (iii)). In our experimental design to evaluate antibiotic inactivation, we present three model systems for coculture studies by recapitulating E. coli Dh5α, E. coli Dh5α amp, and E. coli Dh5α gfp-kan without further genetic transformation. Two of the strains have the common drug-resistance gene for ampicillin and kanamycin in the plasmid. (Figures S1 and S2) In particular, the ampicillin resistance gene (AmpC) renders a β-lactam resistant phenotype. (Figure S3) In addition, microorganisms, including bacteria, are abundant in the environment, and some of these organisms have natural resistance against β-lactam antibiotics due to their genetic traits. Representative bacterial strains isolated from both human stool and soil samples were examined to ascertain how wild-type β-lactamase positive bacteria can trigger phenotypical alteration of antitumor activity of β-lactams when cocultured with cancer cells. These interactive models of enzymatic inactivation can simulate physiological conditions and distinct biofilm-like characteristics for the bacteria to evaluate interactions.

Biocompartmentalization and Antibiotic Inactivation

Natural polysaccharide-based hydrogel can entrap bacteria and serve as a biomimetic container to impart specific functional properties.^31,42 We encapsulated E. coliamp in alginate hydrogel-shell beads (Figure 2A), where the constituents in the liquid media could permeate the hydrogel shell due to the permeable nature of alginate hydrogel structures. A single hydrogel-shell bead can serve as a biocatalytic system and mimic the properties specific to the encapsulated bacteria. The microscopic images show bacteria colonization after incubation in a biological medium. (Figure 2B) The environment in the hydrogel shell can sustain microbial activity to balance proliferation and functionality upon complete confinement. Conventional approaches for encapsulation or entrapment systems are not suitable for inter- or intraspecies studies due to planktonic (free-swimming) bacteria leaking out of the encapsulant. The leaked-out bacteria due to overgrowth can contaminate and/or distort the entire system.^32,43 The bacteria leakage test conducted for 48 h showed that bacteria leakage occurred in the absence of the hydrogel-shell layers due to hydrogel swelling and bacteria proliferation; however, the construction of hydrogel-shell layers could effectively inhibit leakage to contain the organisms and allow these to thrive within the bead in a biofilm-like environment (Figure 2C).

Figure 2.

Characterization of the hydrogel-shell encapsulation and ampicillin degradation system. (A) Schematic depiction of hydrogel-shell encapsulation. (B) Representative images of the hydrogel-shell bead containing β-lactamase expressing E. coliamp (β-lactamase positive; βL +). The scale bar = 250 μm. (C) Leakage test of the supernatant culture by the optical density (OD₆₀₀) measurement. (D) Quantification of total protein accumulated in the hydrogel structure every 12 h. (*p < 0.05, **<0.01, data are the mean ± s.d.; n = 3 biological replicates). (E) Schematic depiction of the ampicillin (concentration 100 μg/mL) degradation test. (F) Quantification of the inhibition zone formed around the diffusion disc after 24 h of incubation with hydrogel-shell beads containing bacterial colonies. E. coli (β-lactamase negative; βL −) and E. coliamp (β-lactamase positive; βL +). (***<0.001, data are the mean ± s.d.; n = 3 biological replicates). (G) Planktonic growth of E. coli (β-lactamase negative; βL −) in LB media for evaluating ampicillin degradation.

The confined bacteria will grow to occupy the core of the bead within 2–3 days in the presence of nutrients. The growth of bacterial colonies can be arrested due to the physical constraint of the hydrogel environment as time progresses. The entrapped bacteria would then gradually reach a metabolically inactive state for their long-term survival.⁴⁴ During the process, continuous production of various metabolic byproducts including proteins is possible, where most of the large-size molecules including enzymes could accumulate within the hydrogel structures (Figure 2D). We show an accumulation of β-lactamase enzymes synthesized by the entrapped E. coliamp within the structure. Further examination shows that the hydrogel bead entrapping E. coliamp can function as an artificial bioreactor while maintaining their viability and metabolic activity upon confinement without bacteria leakage.

Next, to demonstrate antibiotic inactivation due to enzymatic action, we examined β-lactam degradation after incubation with E. coliamp encapsulated in hydrogel beads in ampicillin solution at a concentration of 100 μg/mL. As shown in Figure 2E, ampicillin can be converted into an inactive form with the loss of antibiotic activity. β-lactamase enzymes synthesized by E. coliamp in the hydrogel structure inactivate β-lactam antibiotics by hydrolyzing the peptide bond of the β-lactam ring, rendering the antibiotic ineffective. It might further be metabolized under the catalysis of β-lactamase with the ring-opening reaction. The change in ampicillin concentration can be used as the basis for determining the susceptibility of β-lactam antibiotics. The inhibition zone observed in a disc diffusion test clearly decreased in the solution of ampicillin after 24 h of incubation compared with E. coli (βL −) and E. coliamp (βL +) bead due to the catalytic degradation of β-lactamase. (Figure 2F) The planktonic growth of E. coli in the LB media with ampicillin at a concentration of 100 μg/mL was also evaluated for 24 h. Bacteria that are not drug-resistant could not survive in the presence of antibiotics. As shown in Figure 2G, the growth of E. coli incubated in the ampicillin solution with E. coliamp (βL +) bead was not different compared with that of E. coli inoculated in LB medium without antibiotics. These results suggest that the expressed β-lactamase enzyme within the hydrogel structures significantly influences the hydrolysis of β-lactam antibiotics. As a result, the antimicrobial activity of ampicillin in the media diminished or was nonexistent when hydrogel beads containing E. coliamp were cultured in it, signifying the function of the developed hydrogel system for the catalytic degradation of β-lactamase.

Nitrocefin-Based Test for β-Lactam Degradation

β-lactam antibiotic-resistant bacteria have been visually detected based on β-lactamase hydrolysis indicators with a visual color change.⁴⁵ Nitrocefin, a chromogenic substrate, is highly reactive with an β-lactam ring and can be used to test for the presence of β-lactamase enzymes. In our preliminary investigation to test for the specificity of β-lactamase, nitrocefin substrate was rapidly degraded in the presence of planktonic E. coliamp. (Figure 3A) To evaluate the enzymatic reaction of β-lactamase, we monitored E. coliamp beads via a color change based on the kinetics of nitrocefin cleavage. (Figure 3B) The β-lactam ring structure of the nitrocefin substrate was hydrolyzed due to the accumulation of β-lactamase in the bead, leading to a color change discernible by the naked eye as time progressed. Its characteristic absorbance shifted from 390 to 486 nm within a few hours. While the bead containing β-lactamase negative E. coli Dh5α displayed no peak shift even after 12 h of incubation. (Figures 3C and S4) Further, the addition of sulbactam at a concentration of 10 μg/mL was tested for its neutralizing activity in the bead system. Sulbactam acts primarily by inactivating β-lactamase in most β-lactamase-producing organisms as an irreversible β-lactamase inhibitor.⁴⁶ As shown in Figure 3D, the inhibitory potency of sulbactam against the β-lactamase enzyme was apparent. The nitrocefin hydrolysis reaction decreased in the presence of sulbactam, indicating a competitive inhibition of β-lactam hydrolysis by E. coliamp in the hydrogel environment. The β-lactam hydrolysis through E. coliamp beads was further confirmed by liquid chromatography-tandem mass spectrometry (LC/MS) (Figure S5). Thus, the loss of activity of β-lactam antibiotics and its chemical degradation by bacteria-based functional hydrogel beads were verified.

Figure 3.

Nitrocefin-based colorimetric assay for evaluating the β-lactam antibiotics degradation system. (A) Color change of nitrocefin solution. (Left) hydrolysis of nitrocefin by β-lactamase on a chromogenic nitrocefin substrate. (Right) (B) hydrolysis of nitrocefin by the E. coliamp (β-lactamase positive; βL +) bead and UV–vis absorption spectra of nitrocefin solution. (C) Comparative color change of the nitrocefin solution (E. coliamp (βL +) bead vs E. coli (βL −) bead). Normalized UV–vis absorption spectra of the nitrocefin solution incubated with E. coliamp (βL +) bead vs E. coli (βL −) bead for 12 h. (D) Absorbance at 490 nm of hydrolysis of nitrocefin using four different groups. Sulbactam was used as a β-lactamase inhibitor. (**p < 0.01, ***<0.001, data are the mean ± s.d.; n = 3 biological replicates).

β-Lactam Inactivation to Near-Bacterial Communities

To study the influence of enzymatic degradation between microorganisms, we examined how β-lactamase positive bacteria can counteract the effects of ampicillin at the intraspecies level (i.e., nonresistant populations). First, we generated two types of hydrogel-shell beads encapsulating β-lactamase positive E. coliamp and β-lactamase negative E. coligfp-kan, respectively (Figure S6). E. coligfp-kan incubated with the E. coliamp beads together could survive and grow as colonies in the liquid media of the ampicillin solution at a concentration of 50 μg/mL (Figure 4A,B). The signal of gfp expressed by E. coligfp-kan in hydrogel beads was enhanced as the concentration of the ampicillin antibiotics decreased under the catalytic degradation action of β-lactamase produced by E. coliamp (Figure 4C). In microscopic images, the colony formation of E. coligfp-kan was clearly distinguishable after the coincubation of E. coliamp beads (Figures 4D and S7). The resultant growth patterns of the E. coligfp-kan encapsulated in hydrogel beads could be affected by coculturing with E. coliamp beads (Figure 4E).

Figure 4.

(A) Schematic depiction of the β-lactam antibiotic degradation system to evaluate the growth of nondrug-resistant bacteria (βL −) with the E. coliamp (βL +) bead. The bead coculture model for the growth of E. coligfp-kan (βL −) bead cocultured with the E. coliamp (βL +) bead. (B) Images of coculture for the E. coliamp (βL +) bead and E. coligfp-kan (βL −) bead after 36 h (red arrows indicate E. coligfp-kan bead). The ratio of βL + and βL – beads was 2:1. The scale bar = 2 mm. (C) Fluorescence intensity analysis of the E. coligfp-kan (βL −) bead. The plots show the fluorescent intensity line profile following the yellow line (upper right) in the respective fluorescent images. (D) Microscopic images of the E. coligfp-kan (βL −) bead, cultivated for 36 h without the E. coliamp (βL +) bead (left) and with the E. coliamp (βL +) bead (right). The scale bar = 250 μm. (E) Measurement of the GFP signal of the E. coligfp-kan (βL −) bead (relative fluorescent units, RFU) with/without the E. coliamp (βL +) bead.

Next, we hypothesized that drug-resistant bacteria protect nondrug-resistant bacteria in stratified structures similar to the structure of biofilm formation.⁴⁷ To emulate biofilm-like stratified structures during the process of growth and colonization, we utilized hierarchical encapsulation techniques to entrap different microorganisms sequentially in a layered architecture.³² As depicted in Figure 5A, nondrug-resistant bacteria can be positioned within the deep core of the structure, and drug-resistant bacteria were entrapped in the outer layer. The growth of E. coligfp-kan in the core structure was observed after 24 h of incubation in the liquid media at an ampicillin concentration of 50 μg/mL. (Figure 5B) To verify the protective role of drug-resistant bacteria, three types of hydrogel beads were fabricated by changing the density of E. coliamp confined in the layered region. (Figure 5C) The growth of nondrug-resistant E. coligfp-kan surrounded by a high density of drug-resistant E. coliamp in the layered region exhibited strong signal enhancement in real-time. On the other hand, E. coligfp-kan entrapped in the hydrogel core could not proliferate when the density of E. coliamp in the layered shell region was low. Furthermore, the growth of E. coligfp-kan surrounded by a high density of nondrug-resistant E. coli in the layered region was inhibited. (Figure S8) Therefore, this stratified model indicates that nondrug-resistant E. coligfp-kan in the stratified structure can survive with drug-resistant E. coliamp from external insults in the hydrogel structure.

Figure 5.

(A) Schematic depiction of the stratified hydrogel bead model (core: E. coligfp-kan (βL −), shell region: E. coliamp (βL +)). (B) Microscopic images of the stratified hydrogel bead at different time points at 0 and 24 h. The scale bar = 250 μm. (C) Schematic depiction of stratified beads by changing the concentration of E. coliamp (βL +) immobilized in the hydrogel-layered (shell) region. (Core: E. coligfp-kan 500 cfu/bead and shell: E. coliamp for bead 1:2000 cfu/bead, bead 2:200 cfu/bead, bead 3:0 cfu/bead). The dynamic measurement of the gfp signal from stratified beads every 5 min over 20 h. (Data are the mean ± s.d.; n = 3. The error band represents the standard deviation of three measurements).

β-Lactam Inactivation with Mammalian Cells

β-lactam antibiotics and their anticancer properties have been the subject of study in the past years.^48,49 Several in vitro studies have supported the fact that ampicillin, in combination with different types of antibiotics, can confer anticancer activity due to the dysregulation of mitochondrial function in various cancer cells.^50,51 Many factors, including bacterial enzymes, could be involved in the interaction between various organisms when considering the influence of microorganisms on drug response. We utilized the dynamic coculture system wherein antibiotic inactivation (i.e., β-lactam degradation) could affect the antitumor activity of β-lactam drugs on cancer cells caused by microorganisms and/or their enzymatic actions. To assess this hypothetical situation, we first estimated the half-maximum concentration (IC₅₀) of β-lactam on cancer cells. (Figure S9) A498 renal carcinoma cells that are more sensitive to β-lactam compared to other malignant cells were used.^50,52 The IC₅₀ of ampicillin for A498 renal cell carcinoma cells was estimated to be approximately 185.23 μg/mL at relatively high doses of ampicillin.

Confined bacteria should reach a high cell density as colonies to synthesize intracellular proteins, including enzymes. A498 cells were treated at the high dose of ampicillin, IC₅₀ concentration after 12 h of preincubation with bacteria beads. (Figure 6A) Time-dependent microscopic images of coculture studies showed continual proliferation of both E. coliamp in hydrogel beads and A498 cells in the presence of high-dose ampicillin (180 μg/mL) (Figure 6B). Reduced cytotoxicity of A498 cells was apparent upon coculturing with E. coliamp beads compared to negative control groups when treated at IC₅₀ of ampicillin. (Figure 6C) In the end, this result implies β-lactamase positive bacteria have the potential to trigger the alteration of antitumor activity of β-lactam antibiotics in cancer cells due to the hydrolysis of β-lactam antibiotics.

Figure 6.

Coculture study to demonstrate the species interaction (bacteria bead and cancer cells). (A) Depiction of coculture experiment. The hydrogel bead was cocultured for 12 h, and ampicillin (IC₅₀) was treated in the medium. (B) Images of E. coliamp (βL +) bead cocultured with A498 cells at different time points (6, 12, 18, and 24 h). The scale bar = 250 μm. (C) Percentage of A498 cell viability treated with the IC₅₀ of ampicillin to evaluate the effect of the presence of the E. coliamp (βL +) bead (****p < 0.0001; n = 6 biological replicates).

Validation with Bacteria Isolated from Feces and Soil

Antibiotic resistance can be inherent or acquired depending on the genetic pathways. β-lactamase positive bacteria resistant to high concentrations of β-lactam antibiotics are prevalent in the environment as well as in humans. We asked whether wild-type β-lactamase positive bacteria in a hydrogel system can be used as a biomimetic to neutralize the antitumor effect (i.e., drug inactivation) on cancer cells (Figure 7A). β-lactamase positive and negative bacteria from soil and fecal samples were utilized (Figure S10). The ability to express β-lactamase enzyme for drug resistance varies with the characteristics of bacteria. Three β-lactamase positive strains (resistant to high concentrations of ampicillin, >100 μg/mL) were screened and isolated: E. coli and K. oxytoca, commonly found in the gut microbiota, and B. toyonensis from soil. In addition, M. luteus from soil samples was selected as the negative control.

Figure 7.

(A) Experimental validation of the β-lactam enzymatic degradation system using bacteria isolated from stool samples (E. coli and K. oxytoca) and environmental sources (B. toyonensis and M. luteus). Coculture tests with encapsulated bacteria were used for validation to show how β-lactam degradation can affect cancer cell viability. (B) Measurements of biomass per bead produced from the encapsulated bacteria after 48 h, cultivated in EMEM media with ampicillin at a concentration of 100 μg/mL. (C) Measurements of metabolic activity of bacteria population per bead. (D) Absorbance at 490 nm of hydrolysis of nitrocefin from different bead groups encapsulating isolated bacteria. (E) Percentage of A498 cell viability treated with the IC₅₀ of ampicillin to evaluate the effect of the presence of bacteria bead. (βL +: E. coliDh5α amp, βL −: E. coliDh5α) (*p < 0.05, **p < 0.01, ***<0.001, ****<0.0001. Data are the mean ± s.d.; (B–D) n = 3, (E) n = 6 biological replicates).

Not all bacteria could grow in human cell culture conditions. Prior to demonstrating coculture interactions, we characterized the production of biomass and metabolic status of the encapsulated bacteria in the hydrogel-shell beads incubated in EMEM media with high concentrations of ampicillin (>100 μg/mL). Three β-lactamase positive groups (E. coli, K oxytoca, and B. toyonensis) were culturable under these conditions, while β-lactamase negative M. luteus could not proliferate after encapsulation due to the lack of drug-resistant gene (Figures 7B and S11). Evaluation of metabolic activities in the beads was found to be proportional to the increase in bacterial communities in the confined environment (Figure 7C). Further, the reaction of nitrocefin hydrolysis increased in the presence of bacteria isolated from these samples, indicating the existence of β-lactam resistance (Figures 7D and S12).

Reduced antitumor effect on the A498 renal carcinoma cells was observed as hypothesized when coincubated with the β-lactamase positive bacteria isolated from various sources (Figure 7E). Only β-lactamase positive bacteria could exhibit enhanced viability of A498 cells in the presence of IC₅₀ of ampicillin (∼185 μg/mL). Further, there is a possibility that certain bacteria could produce various secondary metabolites such as bacteriocin for recognizing inter or intraspecies under highly specific circumstances. (Figure S13) The trigger points of these responses depicting how bacteria can recognize mammalian cells in vitro and vice versa are incompletely understood. However, our demonstration shows that specific microorganisms have the potential to trigger an indirect effect on other organisms through the chemical transformation of certain drugs. This phenomenon may be associated with natural defense mechanisms that potentially compromise treatment efficacy.

Discussion

Microbial communities thrive symbiotically and have the ability to act as a community.⁴ Conventional culture approaches in planktonic lifestyles have a technical limitation in predicting the efficacy of antibiotic responses between multicommunities’ interactions, even if in vitro liquid culture has been standardized for determining antibiotic susceptibility (e.g., minimum inhibitory concentration (MIC)).⁵³ Intraspecies interactions with two or more different types of microorganisms can change the environment, resulting in reduced antimicrobial susceptibility, as noted. Further, it has been emphasized that multidrug-resistant bacteria (Stenotrophomonas maltophilia,⁵⁴Moraxella catarrhalis,⁵⁵etc.) can provide high levels of β-lactam antibiotic protection to other microorganisms in host niches. Protection from host antimicrobial molecules is also dependent upon either abiotic or biotic factors or both (e.g., bacteria density, temperature, pH, nutrition level, and other environmental cues).^14,18,56 Similarly, a key consideration of intraspecies interactions for multimicroorganisms is the biofilm-like lifestyles, which can alter the antibiotic susceptibility of a specific strain due to an increase in tolerance through reduced metabolic status, stress responses, diffusion gradients, and molecule diffusion.⁴⁷ Therefore, in light of the distinctive lifestyle of microorganisms in biofilms, the two-model systems demonstrated for intraspecies interactions could serve as an effective and tunable tool for studies in measuring the level of antibiotic efficacy and exposure protection against antibiotics with other environmental insults.

Recent clinical studies have demonstrated the pharmacological effect of β-lactam drugs on the gut microbiome composition.⁵⁷⁻⁵⁹ As the epicenter of bacterial resistance, the gut plays a prominent role in the development of antibiotic resistance. The gut receives a continuous intake of a range of bacteria from external sources due to food intake, wherein the emergence of drug-resistant bacteria might be continuously regulated by antibiotic-induced alterations in the gut microbiota. The drug-resistant bacteria exhibit their presence in the environment via excretions and cross-transmission of the resistant strains or genes. These can easily occur due to the spectrum of genetic pathways, allowing for the hidden selection and multiplication of resistant microorganisms. Such complex intra- and interspecies interactions could result in a broad spectrum of outcomes on the host metabolism or treatment efficacy by regulating bacterial composition or function. Since the advent of metagenomics, our understanding of the gut microbiome has noticeably increased, but the available data on the gut in relation to antibiotic resistance are still incomplete. A significant gap exists in the pharmacological effect of β-lactams on other organisms. There is much to be learned about the gut microbiome, particularly its stability and genetic population alterations. Studies to determine whether drug-resistant bacteria, along with other factors, can influence the role or effect of antibiotics on other organisms in the gut are warranted. The model systems conceived in this work can be a cornerstone in the screening for a range of organisms for antibiotic resistance both from a disease and environment perspective.

Conclusions

Taken together, this study demonstrates how drug-resistant bacteria can affect the viability of other organisms at the intra- and interspecies level when provided a microgrowth environment demonstrated by the compartmentalized hydrogel model developed in this work. We focused on the enzymatic hydrolysis of β-lactam antibiotics, although other bacterial enzymes relevant to specific biochemical mechanisms (e.g., enzymatic transformation, modification of antibiotic targets, or cellular metabolism reactions) could be similarly tested for drug resistance. To support our hypothesis of organism interaction, we presented three interactive models utilizing β-lactamase positive bacteria confined in a β-lactam degradation system for the evaluation of antibiotic resistance due to species interaction. Our findings show that β-lactamase positive bacteria can neutralize the cytotoxic effect of antibiotics not only in the neighboring communities of nondrug-resistant bacteria but also between different organisms, including cancer cells. The results from these coculture studies are highly supportive of the biofilm mode of bacterial growth, which can provide structural support and protect the bacteria from an assault on host or environmental factors. Finally, the coculture system was validated with A498 renal carcinoma cells utilizing β-lactamase positive bacteria isolated from soil and human stool samples. Our studies show that the hydrogel environment can serve as an independent ecosystem for biological compartmentalization and has excellent potential in coculture studies to discern microbial function and response in relation to changes in environmental cues.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsbiomaterials.3c01678.

• Strains and additional description of bacterial isolates and identification (Table S1); primer sequences for PCR amplification of 16S rRNA and different β lactamase genes (Table S2); plasmid map of E. coli DH5α pAAV-Amp (Amp, Ampicillin antibiotics marker) and β-lactamase expressing E. coli (Figure S1); plasmid map of E. coli DH5α pSmart-kan-sfGFP (Kan, Kanamycin antibiotics marker) (Figure S2); bacteria growth in LB broth with/without antibiotics (ampicillin 100 μg/mL and kanamycin 100 μg/mL) (Figure S3); nitrocefin-based colorimetric assay for evaluating the β-lactam antibiotic degradation system at different time points (Figure S4); mass spectra of ampicillin degradation for evaluating β-lactam antibiotics degradation system (Figure S5); results of colony PCR for bacterial strains used in this study (Figure S6); evaluation of the β-lactam antibiotic degradation system using E. coligfp-kan (Kanamycin resistance) (Figure S7); schematic depiction of two stratified hydrogel bead models and the dynamic measurement of the gfp signal from stratified beads (Figure S8); cell death index determination of ampicillin (Figure S9); screening methods of β-lactamase positive bacteria from environmental niches (Figure S10); gel electrophoresis (1% agarose): β-lactamase gene PCR product from soil and fecal bacterial isolates run against 100 bp ladder (Figure S11); bacteria isolated from environmental sources (B. toyonensis and M. luteus) and clinical patients (E. coli and K. oxytoca) (Figure S12); and percentage cell viability (A498 cells) cocultured with hydrogel-shell beads containing bacteria colonies (Figure S13) (PDF)

Author Contributions

Y.J. designed the research; Y.J. and S.A. performed the research; Y.J. and S.A. analyzed the data; Y.J. and J.I. generated the main concepts, interpreted the experiments, and wrote the manuscript; and J.I. provided overall supervision and funding.

The authors declare no competing financial interest.