The Effect of Antibiotic-Resistant and Sensitive Escherichia coli on the Production of Pro-Inflammatory Cytokine Response by Human Peripheral Blood Mononuclear Cells

Copyright: © 2017 Volodina GV, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. The Effect of Antibiotic-Resistant and Sensitive Escherichia coli on the Production of Pro-Inflammatory Cytokine Response by Human Peripheral Blood Mononuclear Cells Galina V. Volodina1, Tigran K. Davtyan2*, Murat E. Kulmanov1, Ardak B. Dzhumagazieva1, Sholpan K. Tursunova1, Assima O. Abekova1, Indira E. Bishimova1, Zhansaya S. Abramova1, Roza T. Kenzhebekova1, Sabina G. Murzageldinova1, Ilya S. Korotetskiy1 and Alexander I. Ilin1 1Scientific Center for Anti-Infectious Drugs, 75B Al-Farabi Ave, Almaty, Kazakhstan 2Analytical Laboratory of the Scientific Center for Drug and Medical Technology Expertise. MH RA, 49/4 Komitas Ave, Yerevan, Armenia


Introduction
Drug resistance of microorganisms remains an urgent and global problem of modern medicine and microbiology [1]. The present-day antibiotic therapy and arsenal of existing courses of antimicrobial treatment for infectious diseases caused by pathogens with drug resistance are currently ineffective and promote the wide spread of resistance genes in the strains of microorganisms in both the human population and environment. The search and development of new approaches and a fundamentally new generation of drugs that could increase the sensitivity of bacteria to traditional antibiotics is a perspective and promising trend in the fight against drug resistance [2]. It was shown previously that the ionic nanostructured complex formed by proteins and carbohydrates, metal salts and iodine intercalated into them, which is a new drug FS-1 [3], is capable of inducing reversion to sensitivity or reversion of resistance to antibiotics during prolonged culturing with antibiotic-resistant bacteria [4].
It should also be noted that a comparative study of the effect of resistant and sensitive bacteria on the induction and regulation of the adaptive and innate immune response can open new prospects for developing more effective approaches to complex immunotherapy of infectious diseases caused by pathogens with drug resistance. It is known that the activation of the mechanisms involved in innate immunity and induction of the production of pro-inflammatory cytokines IL-1β, IL-6, TNF-α is caused by interaction of both PAMPs (Pathogen-Associated Molecular Patterns), a complex of canonical, evolutionarily conservative microbial molecules, associated with antigenicity, and DAMPs (Danger-Associated Molecular Patterns), an integral complex of canonical, evolutionarily conservative microbial molecules associated with antigenicity, as well as pathogen-associated or virulence-associated molecules, with pattern recognition receptors PRRs (Pattern Recognition Receptors) on cells of the innate immune system, resulting in inducing a protective inflammatory response for effective control of infections [5][6][7].
The purpose of this study was to compare the effect of Escherichia coli bacteria, differing in sensitivity to gentamicin, on the production of proinflammatory cytokines IL-1β, IL-6, TNF-α by human peripheral blood mononuclear cells and to determine the changes in the induction threshold of the cytokine response.

Bacterial strains and cultivation
The following strains were used in the study: a) the resistant Escherichia coli ATCC-BAA-196 J53 pMG223p (ATCC) strain with multiple drug resistance; b) the resistant Escherichia coli subculture (E. coli R) obtained by culturing E. coli ATCC-BAA-196 strain for 50 generations in a medium with ampicillin and gentamicin at a concentration of 40 μg/ml, which retained resistance to antibiotics; c) the gentamicin-sensitive Escherichia coli subculture (E. coli Rev) obtained under experimental conditions by culturing E. coli ATCC-BAA-196 strain for 50 generations in a medium with a constant content of the drug FS-1 at a ½ concentration of the minimum bactericidal concentration MBC (500 μg/ml).
For long-term serial cultivation of E. coli in the presence of antibiotics and FS-1, 1.5 х 10 6 CFU/mL of cells in the Mueller Hinton broth (Sigma Chemical Co., St. Louis, MO) were incubated into separate tubes with total volume of 1.0 ml containing 40 μg/ml concentration of ampicillin and gentamicin or 250 μg/ml concentration of FS-1 and incubated at 37 ± 1°C for 24 h for obtaining one generation. The selected concentrations for antibiotics and FS-1 represented the 1/50 and ½ concentrations of MBC respectively, which caused no cytotoxicity or growth suppression of initial E. coli ATCC-BAA-196 J53 pMG223p (ATCC) strain. The bacterial strains were propagated in the presence of selecting drugs for 50 generations and every fifth generation of each strain was tested for morphology, enzymes profiling and MIC value and cell growth in solid agar plates at 37 ± 1°C for 24 h were estimated.
The bacteria were cultured in a solid nutrient medium until the middle of the log phase of growth at the temperature of 37 ± 1°C. The bacteria were fixed with 37% formalin at 37 ± 1°C for 20 min, followed by triple washing with PBS (phosphate-buffered saline, pH 7.2) by centrifuging for 10 min at 3,000 rpm. In the control experiments the survival of the formalin-fixed cultures was monitored. The concentrations of live and fixed bacteria from 10 1 CFU/ml to 10 7 CFU/ ml were used for the co-cultivation with PMBC.
The determination of MBC and minimum inhibitory concentration (MIC) of antibiotics against the strains under study was carried out by the serial dilution method [8]. The antibiotic sensitivity of the E. coli Rev strain was evaluated using the disc diffusion method according to recommendations and requirements of CLSI (Performance Standards for Antimicrobial Disc Susceptibility Tests) [9].

PBMC isolation and activation
Heparinized peripheral blood samples were obtained from 8 (4 male, 4 female) age-matched normal donors. Peripheral blood mononuclear cells (PBM) were separated from heparinized whole blood by Ficoll-Hypaque (histopaque) (Sigma Chemical Co., St. Louis, MO) density gradient centrifugation. Polyglucin (6% dextran) was added to the blood to precipitate the red blood cells. The supernatant was washed, RPMI-1640 re-suspended, and fractionated by the density gradient corresponding to the floating density of human PBMC (ρ=1.076 g/ml), at 4°C and 3,000 rpm for 20 min. The PBMC fraction was washed by centrifugation and re-suspended in the RPMI-1640 culture medium containing 10% FCS, 2 mM L-glutamine, 100 U/mL penicillin, 100 µg/mL streptomycin and 0.025 µg/mL amphotericin B (Sigma, USA). The percentage of viable cells was assessed by the trypan blue exclusion (Sigma, USA). Cell suspensions with viability, greater than 90%, were used in the experiments.

Determination of cytokine minimum inducing concentration (cMIC)
As the cMIC, a lowest concentration of bacteria was taken that was able to induce statistically significant cytokine production in the threshold level above 25-30% uncertainty range of the un-stimulated PBMC negative control level [10,11]. To determine cMIC, the PBMC suspension was dispensed into the 96-well flat-bottom plates (BD Falcon, USA) at the concentration of 1 × 10 5 cells per well, and an appropriate E. coli strain was added at concentrations from 10 1 to 10 7 CFU/ml in a volume of 100 μl. Physiological saline was used as a control. The bacteria and PBMC were cocultured during 4 h or 12 h incubation at 37°C in an atmosphere containing 5% CO 2 . At the end of the incubation, the plates were centrifuged at 300-500 g for 10-15 min, and the culture supernatant was collected. Quantitative analysis of cytokines was carried out using the commercial reagent kits IL-1 beta-IFA-BEST, IL-6-IFA-BEST, alpha-TNF-IFA-BEST (Vector-Best, Novosibirsk, Russia). Optical density measurement and calculation of cytokine concentrations were performed on the Sunrise RC.4 microtiter plate reader (Tecan, Austria) with the Magelan 2.0 software (Tecan, Austria).

Statistics
The method of dispersion analysis with parametric and nonparametric procedures was used in this study. The data reliability was assessed with the GraphPad Prism version 6.00 for Windows (GraphPad Software, La Jolla California USA), using the one-way ANOVA and unpaired t-test. The values of P>0.05 were considered insignificant.

Antibiotic resistance of selected bacterial strains
In this study, the Escherichia coli ATCC-BAA-196 strains were used which differ in their sensitivity to gentamicin: the resistant E. coli R subculture obtained by culturing the stock E. coli ATCC-BAA-196 strain for 50 generations in a medium with ampicillin and gentamycin, and the gentamicin-sensitive E. coli Rev subculture obtained by culturing the E. coli ATCC-BAA-196 strain for 50 generations in a medium with a constant content of the drug FS-1. The results from the MIC studies showed that for E. coli ATCC-BAA-196 the MBC of gentamicin was>2000 μg/ml, and MIC of gentamicin was equal to 1800 μg/ml. With respect to the sensitive E. coli Rev strain, there was a decrease in the MBC for gentamicin to 950 μg/ml and a decrease in MIC for gentamicin to 800 μg/m. The gentamicin MBC for E. coli R strain was equal to 2000 μg/ml and MIC to 1700 μg/ml respectively. These data indicates the acquisition of the sensitivity to gentamicin in E. coli Rev. The reversion to antibiotic sensitivity in the E. coli Rev strain was also determined by the disc-diffusion method, according to which the growth retardation zone to the ampicillin increased up to 15 mm and to ceftriaxone-up to 21 mm. Alongside with that, the growth retardation zone of E. coli R to the ampicillin was 10 mm, to ceftriaxone-17 mm.

DAMP-mediated differences in cytokines production by PBMC
To assess the threshold for induction of cytokine production as a result of the exposure to DAMPs, we have selected a model for co-cultivation of PBMC with live E. coli Rev and E. coli R. First, in preliminary experiments we calibrated the live bacteria-PBMC cocultivation system by studying the time-dependence of cytokine response and found the best-fit linearity of response during 4 h incubation time. Dose-dependence preliminary studies revealed that IL-1β, IL-6 and TNF-α production by PBMC induced by live E. coli Rev at the concentration of 10 1 -10 4 CFU/ml had an exponential character; and at concentrations of bacteria 10 5 CFU/ml and higher, the production of cytokines reached to a plateau (data not shown). At 10 1 -10 3 CFU/ml concentration stock E. coli BAA-196 and resistant E. coli R strains exponentially increased the production of cytokines by PBMC, whereas at 10 4 CFU/ml and higher bacterial load caused starvation of PBMC cytokine response.
We observed that co-cultivation of stock E. coli BAA-196 strain and the gentamicin-resistant E. coli R strain in concentrations of 10 1 -10 3 CFU/ ml with PBMC for 4 h caused a dose-dependent increase in IL-1β, IL-6, TNF-α production. As it shown on figures 1, 2 and 3, the DAMP-mediated cytokine response induced antibiotic-resistant bacteria characterized by high similarity. From the data presented in Figure 1, it follows that the resistant E. coli R strain and the stock E. coli ATCC-BAA-196 strain in concentrations of 10 1 -10 2 CFU/ml induced the production of IL-1β at the un-stimulated control level (0 CFU/ml). When the concentration of both E. coli R and E. coli ATCC-BAA-196 strains was raised up to 10 3 CFU/ml, the induction of IL-1β production increased (P<0.05) as compared to the un-stimulated PBMC (Figures 1a and 1b).
In contrast to the tested resistant strains, the gentamicin-susceptible E. coli Rev strain in concentrations of 10 1 -10 3 CFU/ml induced low production of IL-1β, which did not differ from the un-stimulated control level (Figure 1c), however, E. coli Rev at the concentration of 10 4 CFU/ml induced an increase (P<0.05) in the level of IL-1β production relative to the untreated control (0 CFU/ml).
The production of IL-6 by PBMC as a result of the action of resistant strains E. coli R and the E. coli ATCC-BAA-196 at concentrations of 10 1 -10 2 CFU/ml was insignificant and did not differ from the untreated control level. With an increase in the concentration of E. coli R to 10 3 CFU/ml, increase in IL-6 production was observed in comparison with the control level (P<0.05). The strain E. coli ATCC-BAA-196 at the same concentration also induced (P<0.05) production of IL-6 (Figures 2a and 2b). ml, and significant IL-6 production was recorded at the concentration of 10 4 CFU/ml (P<0.05) in relation to the untreated control level (Figure 2c).
As it follows from the Figure 3, the production of TNF-α in response to resistant E. coli R and E. coli ATCC-BAA-196 strains at concentrations of 10 1 -10 2 CFU/ml was insignificantly different from the untreated control level. Both, E. coli R and E. coli ATCC-BAA-196 at the concentration of 10 3 CFU/ml increased TNF-α production compared to the control level (P<0.05). The gentamicin-sensitive E. coli Rev strain at low concentrations of 10 1 -10 3 CFU/ml induced a baseline production of cytokine TNF-α at the control level, and significant induction of TNF-α (P<0.05) took place at 10 4 CFU/ml (Figure 3).
Thus, the obtained results revealed a DAMP-mediated difference in the induction threshold of IL-1β, IL-6, TNF-α production by PMBC by live E. coli, differing in the gentamicin sensitivity phenotype. The cMIC of the resistant E. coli BAA-196 and the gentamicin resistant E. coli R strain (10 3 CFU/ml) was at least ten times lower than cMIC of the gentamicin-sensitive E. coli Rev strain (10 4 CFU/ml).

PAMP-mediated differences in cytokines production by PBMC
To assess the induction threshold of cytokine production as a result of the exposure to PAMPs, we next selected a model for co-cultivation of PBMC with the formalin-fixed bacteria E. coli Rev and E. coli R and studied how E. coli strains induced cytokine production by PBMC.
IL-1β cytokine production in response to resistant E. coli R ( Figure  4a . The bacteria were cultured in a solid nutrient medium until the middle of the log phase of growth and were fixed with 37% formalin at 37 ± 1°C for 20 min, followed by triple washing with PBS. PBMC were incubated with medium alone (0 CFU/ml) or in the presence of 10 1 -10 4 CFU/ml formalin-fixed bacteria for 4 hours and the level of IL-1β production was measured by ELISA. All data represent means ± SEM (error bars) and are significantly different comparing treated PBMC with the untreated control (*P<0.05). of 10 1 -10 2 CFU/ml was insignificant relative to the untreated control level. Fixed E. coli R and stock E. coli ATCC-BAA-196 strains at the concentration of 10 3 CFU/ml induced increase in IL-1β production (P<0.05). In contrast, the gentamicin-sensitive E. coli Rev strain ( Figure  4c) induced IL-1β production at the control level at concentrations of 10 1 -10 3 CFU/ml, and only at the concentration of 10 4 CFU/ml caused significant increase in IL-1β (P<0.05).
Co-cultivation of PBMC with resistant E. coli R (Figure 5a) and E. coli ATCC-BAA-196 (Figure 5b) strains at concentrations of 10 1 -10 2 CFU/ ml induced IL-6 production at the control level. At the concentration of 10 3 CFU/ml, the E. coli R strain and stock resistant E. coli ATCC-BAA-196 strain induced increase in IL-6 production (P<0.05). In comparison with this, the E. coli Rev strain (Figure 5c) induced IL-6 production at the untreated control level at concentrations of 10 1 -10 3 CFU/ml, and at a concentration of 10 4 CFU/ml caused significant increase in IL-6 production (P<0.05).
As for TNF-α production, the same tendency was observed. Low concentrations of 10 1 -10 2 CFU/ml of the resistant E. coli R (Figure 6a) and stock resistant E. coli ATCC-BAA-196 (Figure 6b) strains induced insignificant production of TNF-α relative to the control level. The increase in the induction of TNF-α production (P<0.05) was observed in E. coli R at the concentration of 10 3 CFU/ml, the stock resistant E. coli ATCC-BAA-196 strain at the same concentration also induced (P<0.05) production of TNF-α (Figures 6a and 6b). E. coli Rev strain (Figure 6c) at concentrations of 10 1 -10 3 CFU/ml induced a slight production of TNF-α relative to the untreated control level, while at the concentration of 10 4 CFU/ml it caused the significant increase in level of TNF-α production (P<0.05). Thereby, the results of the present studies revealed similar differences between the formalin-fixed resistant and gentamicin-susceptible E. coli strains in relation to the induction threshold of cytokine production in response to the evolutionarily conservative molecular structures of bacteria. The cMIC of the formalin-fixed resistant E. coli  and E. coli R strains for IL-1β, IL-6, and TNF-α, as well as for the alive strains, was equal to 10 3 CFU/ml, and the cMIC of the fixed revertant E. coli Rev strain, as well as for the live strain, reached 10 4 CFU/ml.

Discussion
The data we have obtained due to comparability of PAMPand DAMP-mediated differences in the threshold of induction of pro-inflammatory cytokine production by PBMC seem to reflect the similarity or unidirectionality of a complex of biochemical, morphological or functional changes and coordinated expression of bacterial genes during the acquisition of antibiotic sensitivity or the formation of a reversion phenotype under the effect of a new iodine nanostructured complex FS-1. As a result of prolonged interaction of FS-1 with bacteria, a "new or acquired" phenotype is formed, which probably, due to changes in genomics, proteomics, and metabolomics of bacteria, leads to both a conjugated change in the expression and architectonics of immunogenic molecular structures and reversion to antibiotic sensitivity.
It is interesting to note that bacterial PAMPs and DAMPs differ from each other only in the availability of "additional" pathogen-associated or virulence-associated molecules (DAMP), while evolutionarily conservative molecular structures of microorganisms (PAMP) are common among E. coli strains in whole [12][13][14]. This is evidenced by the data of our comparative statistical analysis of the production level of proinflammatory cytokines IL-1β, IL-6 and TNF-α by PBMC stimulated with formalin-fixed and live E. coli ATCC-BAA-196, E. coli R, and E. coli Rev. Accordingly, the production of IL-1β, IL-6, and TNF-α by PBMC stimulated with formalin-fixed E. coli ATCC-BAA-196, E. coli R, and E. coli Rev was significantly (P<0.0005) up to ten times lower as compared with live bacteria of the same strains, which indicates the invariance in functioning of the pattern-recognition receptors (PRR) in this experimental system. Consequently, the decrease in the potential of the immune system pro-inflammatory reaction in response to the sensitive strain of bacteria is associated with changes in the immunogenicity of both virulence-associate molecules and immunogenic factors but is not a consequence of a change in the interactions of PAMPs and DAMPs with recognizing PRR receptors in the cells of the immune system.
Summarizing the foregoing, it can be concluded that we have identified comparable PAMP-and DAMP-mediated differences in the inductance of the production of pro-inflammatory cytokines IL-1β, IL-6, TNF-α by PBMC when interacting with bacteria that differ in the sensitivity phenotype. Inflammation is a defence response of immune system to hazardous stimuli such as infections or tissue injury. However, prolonged or uncontrolled inflammation is the main cause of healthy tissue destruction and chronic pathologies. An increase in the immune system sensitivity associated with a decrease in the threshold for induction of the pro-inflammatory reaction in response to resistant strains of bacteria can be considered as the immune system compensatory response to the changed fitness cost of the bacterial population compared to the population of the susceptible strain. Therefore, the extent of increase in the threshold for cytokine response induction by an integrated complex of canonical, evolutionarily conservative, associated with antigenicity, as well as pathogenassociated or virulence-associated molecules, indirectly determined by the cMIC value, may serve as an immunological marker of antibiotic resistance reversion.
The data we have obtained provide evidence that when a resistant stock E. coli BAA-196 strain was exposed to a drug FS-1, the gentamicin-sensitive strain was obtained with the microbiologically confirmed property to reverse antibiotic sensitivity. The ability of the sensitive strain to increase the threshold for pro-inflammatory cytokine induction relative to that of resistant strains appears to be an immunological evidence of resistance reversion towards reversing sensitivity to antibiotics.