Introduction
Interest in the study of dormant forms of non-sporulating bacteria is caused, on one hand, by the desire to understand long-term bacterial survival mechanisms under conditions that are not conducive to growth that occur in natural ecosystems (El-Registan et al., 2006; Lennon and Jones, 2011). On the other hand, the occurrence of dormancy is associated with pathogen persistence via the formation of antibiotic-resistant, persistent cells (Keren et al., 2004; Zhang, 2004; Lewis, 2007). The problem of antibiotic resistance and the emergence of antimicrobial resistant strains has become especially important in connection with current infectious disease expansion. For example, resistance has been observed in Staphylococcus aureus, which is able to form cells with increased antibiotic resistance, which enhances persistence in the host organism and causes chronic or recurrent infections (Chambers and Deleo, 2009). A similar phenomenon has also been described in Mycobacterium tuberculosis (Kaprelyants et al., 2018).
The bacterial transition to a dormant state has been documented in experimental studies that described the process in a number of non-sporulating bacteria (Kaprelyants et al., 1993). Further, dormant bacterial cells may develop “non-culturability” (NC) (a term which reflects the inability to form colonies on agar plates), which prevents the cells from being detected in vivo using routine plating procedures (Roszak and Colwell, 1987; Cellini et al., 1994; Porter et al., 1995; Kell et al., 1998; Barer and Harwood, 1999; Tholozan et al., 1999; Biketov et al., 2000; Currás et al., 2002; Shleeva et al., 2002; Mukamolova et al., 2003; Wood et al., 2005; Lee et al., 2007). For many non-sporulating gram-negative and gram-positive bacteria, including non-pathogenic relatives of pathogens, morphologically differentiated dormant forms have previously been described (Cellini et al., 1994; Currás et al., 2002; Shleeva et al., 2002, 2004; Lee et al., 2010; Mulyukin et al., 2010). Studying the possible persistence and development of antibiotic resistance of other pathogenic and conditionally pathogenic bacteria, including representatives of the genus Corynebacterium is also relevant (Blokpoel et al., 2005; Tauch et al., 2005; Soriano et al., 2009; Olender, 2012).
Many corynebacteria are pathogenic (Bernard, 2012). Corynebacteria of the species Corynebacterium diphtheriae, which produce a very strong exotoxin, are the causative agents of one of the most famous human infections, diphtheria. So-called non-diphtheria corynebacteria (Corynebacteria non-diphtheriae) C. ulcerans and C. pseudotuberculosis cause diphtheria-like diseases that include pseudomembranous pharyngitis, moderate pharyngitis, otitis, lymphadenitis and skin ulcers. C. minutissimum is a causative agent of erythrasma and chronic pseudomycosis. C. amycolatum, C. urealyticum, and C. striatum are particularly resistant to penicillins, aminoglycosides and quinolones. C. jeikeium strains are causative agents of hospital infections. C. jeikeium is responsible for a number of nosocomial infections such as endocarditis, device-connected infection, osteomyelitis (van der Lelie et al., 1995; Mookadam et al., 2006). The bacteria has often found in cancer patients with compromised immune system, inserted in medical instruments, skin lesions, and after antibiotic therapy (Funke et al., 1997). A high mortality rate was documented for C. jeikeium sepsis in hematological patients (van der Lelie et al., 1995), and immunocompromised patients carrying prosthetic valves or catheters are particularly susceptible to infection. Further, the curing of C. jeikeium frequently limited by developing of multidrug-resistance of the bacteria (Olson et al., 2009; Ifantidou et al., 2010).
Whilst the phenomenon of dormancy in non-sporulating bacteria is extensively studied for many years, corynebacteria dormancy in vivo and in vitro did not attract much attention. Only one model has been used to assess dormant forms of C. pseudodiphtheriticum, which is based on the fivefold limit of the nitrogen source provided in growth media (Mulyukin et al., 2014).
After infecting humans, bacteria are typically captured by macrophages where they are influenced by a number of stresses including low pH, elevated levels of the active forms of oxygen and nitric oxide and the activity of lysosomal hydrolases (Rook et al., 2001; Hacker et al., 2016; BoseDasgupta and Pieters, 2018). However, some bacteria, such as M. tuberculosis (Deretic and Fratti, 1999) and C. ulcerans (Hacker et al., 2016), can maintain their viability within macrophages despite these harmful factors. Therefore, naturally induced stressful conditions are likely to be factors that are also useful for inducing a dormant state in vitro. M. tuberculosis (Shleeva et al., 2011) and non-pathogenic M. smegmatis (Kudykina et al., 2011) transition to dormant, non-culturable, persistent states in response to the gradual acidification of their environments, therefore, we may suggest that slow decrease in pH levels may result in induction of dormancy in corynebacteria, which, like mycobacteria, belong to the order Actinomycetales.
In order to cure chronic infections caused by dormant forms of pathogens, new ways should be established (Kaprelyants et al., 2018). In this regard, application of physical factors seems to be promising in order to destroy metabolically passive dormant bacterial forms. Recently we found that significant concentrations of the intermediates participating in protoporphyrin biosynthesis were present in dormant forms of M. smegmatis (Nikitushkin et al., 2016), a fast-growing bacterium which is genetically close to M. tuberculosis, These findings suggested that dormant bacteria may be killed by photodynamic inactivation (PDI) when fluorescent porphyrins serve as intracellular photosensitizers. Studies have also demonstrated the photoinactivation of dormant mycobacterial forms in vitro in the rapidly growing, tuberculosis-related pathogenic strain, M. smegmatis (Shleeva et al., 2019b, 2020).
The goal of this study was to find if C. jeikeium are able to form dormant cells as a result of slow decrease of pH level of growth medium in stationary phase. We also tested whether the stimulation of endogenous porphyrin production in dormant corynebacteria enhances their sensitivity to photodynamic inactivation (PDI).
Materials and Methods
Organisms and Media
The C. jeikeium K411 strain (from State collection of pathogenic microorganisms FSBI Scientific Center for Expert Evaluation of Medicinal Products of the Ministry of Health of the Russian Federation) was grown in TSB broth (Himedia, India) at 37°C for 20–24 h while stirring (200 rpm). A 0.2 mL inoculum was added to 100 mL developed by us 2AS medium (105 cells/mL) with the following composition: 20 g/L glucose; 0.125 g/L MgSO4 × 7H2O; 1.5 g/L NaCl; 2.5 g/L (NH4)2SO4; 13.6 g/L KH2PO4; 0.44 g/L histidine (Sigma); 4.0 g/L glutamic acid (Sigma); 8 mL trace element solution; and 0.1% Tween-80. The media was pH adjusted to 6.0 using NaOH. Trace element solution contained 1.0 g/L EDTA, 10.0 g/L MgCl2 × 6H2O, 0.1 g/L CaCl2 × 2H2O, 0.04 g/L CoCl2 × 6H2O, 0.1 g/L MnCl2 × 2H2O, 0.02 g/L Na2MoO4 × 2H2O, 0.2 g/L ZnSO4 × 0.02 g/L 7H2O, CuSO4 × 5H2O and 0.5 g/L FeSO4 × 7H2O. A final concentration of 0.1% Tween 80 was added. The culture was grown at 37°Ñ in a shaker (200 rpm) for 13–16 days until a pH of approximately 5.5 was established.
Estimation of Viability
Bacterial suspensions were serially diluted in fresh TSB medium, and three 10 μl samples from each dilution were spotted on TSB (Himedia, India) agar containing 0.1% Tween-80. Plates were incubated at 37°C for 5 days. After incubation, the number of the colony forming units (CFUs) present was counted. The limit of detection was 10 CFU/mL.
The same diluted suspensions (100 μL each) were also used for most probable number (MPN) assays and to evaluate cell resuscitation in 48-well Corning microplates containing appropriate medium (0.9 mL) (see below). Microplates were incubated at 37°C for 14 days without agitation. Wells with visible bacterial growth were considered positive. The number of cells that remained intact after exposure to damage-inducing effects was also determined microscopically by counting propidium iodide (PI)-negative cells in a Helber’s chamber (no less than 10 large fields were counted for each sample).
Measuring Levels of 3H-uracil Incorporation to Determine the Metabolic Activity of Cells
One milliliter samples from cells suspensions were incubated with 1 μl [5,6-3H] uracil (10 μCi; 0.2 μmol in 50% ethanol) and incubated for 2 h at 37°C with agitation (45–60 rpm). Cells were then harvested on glass fiber GFC filters (Whatman, United Kingdom) and washed with 3 mL 7% trichloroacetic acid. Next, cells were washed using 3 mL absolute ethanol. Air-dried filters were placed in scintillation liquid and incorporated radioactivity was measured using a scintillation counter (Beckman, United States).
Spent Medium Preparation
Supernatants (SN) with resuscitating activity were obtained from C. jeikeium cultures grown in either TSB or 2AS (initial pH 7.0) media after consecutive sub-culturing. First, a 0.2 mL stock culture (stored at −70°C) was used to inoculate 100 mL culture medium to produce an initial density of approximately 103 cells/mL. The cultures were incubated overnight with agitation (100 rpm). Cells were sub-cultured by transferring a 0.1 mL inoculum to 100 mL fresh medium and cultivating as previously described for 2–30 h. Then cultures were subjected to centrifugation (12,000 × g, 20 min) and sterilization using 0.22 μm filters (Whatman). Fifty milliliters volumes of SN produced in this manner were frozen and stored at −70°C. C. jeikeium growth stimulating activity of SN was assessed after SN was added to freshly inoculated TSB or 2AS media at a 1:1 vol/vol ratio. Stored SNs were used immediately after thawing and were not refrozen and reused.
Resuscitation of “Non-culturable” C. jeikeium Cells
Non-culturable cells were separated from spent medium (centrifugation for 20 min at 5, 000 × g), that was serially diluted and used to inoculate either TSB or 2AS medium. An equal volume of either SN was prepared as indicated above, or appropriate uninoculated medium (control), and numbers of viable resuscitated cells were determined using the MPN assay (de Man, 1974).
Sensitivity to Antibiotics and Heat Treatment
Four milliliters of an early stationary phase (ESP) culture that was grown in TSB medium (pH 7.0) for 1 day or 4 mL of a 25 days old culture incubated at reduced pH were treated with 0–10 μg vancomycin/mL or 0–100 μg rifampin/mL and incubated at 37°C or room temperature for 24 h without agitation. The number of resistant cells was determined using the MPN assay in the presence of TSB culture SN (5 h, see above). One milliliter samples of ESP cultures or 25 days long incubated suspensions containing dormant cells were heated to 60, 65, 70, 75, or 80°C for 10 min, and the number of survivors within samples were determined by evaluating the growth of bacteria in presence of SN (MPN assay).
Pigment Extraction From the Cells
Pigment was extracted from the biomass in accordance with the method described by Bligh and Dyer (Bligh and Dyer, 1959). First, 1 mL chloroform and 2 mL methanol were added to the wet biomass of cells (0.8 g). Cells were agitated for 12 h in the extraction mixture and subsequently centrifuged (2,000 × g), followed by the addition of 1 mL water and 1 mL chloroform (to separate phases). The chloroform layer was washed three times with 0.1 M NaCl. The bacterial pellet was re-suspended in ice-cold 100 mM HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) buffer (pH 8.0) containing 2% Triton × 100 (Sigma, United States) then lysed by using zirconium beads on a bead beater homogenizer (MP Biomedicals FastPrep-24) for 1 min, five times and stayed in dark place at room temperature for 5 h. The bacterial lysate was centrifuged at 13,000 rpm for 15 min at 4°C.
Absorption and Fluorescence Spectra
Absorption spectra were recorded on a Cary 300 Bio Spectrophotometer (Varian, Inc., United States). Fluorescence measurements were carried out with a Varian Cary Eclipse fluorimeter (Varian, Inc., United States).
PDI
Either dormant or active cell suspensions were used for light-inactivation experiments. C. jeikeium suspensions with OD values equal to 0.1, which corresponded to approximately 107 bacteria/mL, were used. Bacterial suspensions (100 μL) were added to the wells of a Nunc 96-well plate (ThermoFisher Scientific, United States). Samples were illuminated with a SOLIS-4C light-emitting diode at 565 nm, 3.2 W using bandpass filter MF565-24 (Thorlabs, United States). The light beam was collimated to a diameter of 5 mm, which corresponded to the diameter of the wells of the 96-well plate, and samples were illuminated for 60 min. The power density of the light was 145 mW per well (513 mW/cm2) as determined using a 2,936-c power meter (Newport, United States). Temperature was controlled an accuracy of ± 0.2°C by placing an 80BK type-K multimeter thermocouple (Fluke, Germany) directly in the microcell before and after lighting in the presence and absence of bacterial suspension. Temperatures were below 40°C in the wells of all samples. After the illumination of samples, numbers of viable bacteria were estimated using an MPN assay in presence of 5-h-old SN obtained from TSB medium (see above).
Microscopy
Cell suspensions were examined under a microscope (Eclipse E4000, Nikon, Japan) in phase-contrast and epifluorescence modes after staining with propidium iodide (PI) (3 mM) to detect injured cells, ethidium bromide (5 mM) to detect DNA-containing cells, or Nile red (4 mg/mL) to detect lipid inclusions. The excitation value was at 510 nm and emission was at > 560 nm.
Statistical Analysis
Statistical processing was carried out using the analysis of the standard deviation or relative error within the data group. MPN values were determined using de Man’s tables calculated on the base of Poisson distribution (de Man, 1974). For the MPN assay (95%) confidence limits were calculated. The MPN values were considered statistically different if low and high confidence limits were not overlapped.
Discussion
In this study, we ascribe conditions needed for corynebacterial cells to transition to a state of dormancy, which was accompanied by: (1) the formation of stress-resistant forms intended for long-term survival that have been identified in many spore-forming and non-spore-forming bacteria (Zhang, 2004; Mulyukin et al., 2010, 2014; Lennon and Jones, 2011) and (2) the acquisition of an NC state during the prolonged incubation of post-stationary cultures, which was previously described for Micrococcus luteus (Kaprelyants and Kell, 1993), Rhodococcus rhodochrous (Shleeva et al., 2002) and mycobacteria (M. tuberculosis and M. smegmatis) (Shleeva et al., 2003, 2004, 2011, 2015; Kudykina et al., 2011). The dormant forms of mycobacteria, which form when the external medium undergoes gradual acidification, can be distinguished from active bacteria based on their distinct proteomic (Trutneva et al., 2018, 2020) and metabolomic (Nikitushkin et al., 2020) profiles.
C. jeikeium cells formed in post-stationary cultures should be classified as dormant bacterial forms based on the following features: (i) the retention of viability throughout an extended incubation period (up to 5 months) (Figure 4), (ii) their low level of metabolism (Table 1), (iii) their enhanced resistance to deleterious factors such as high temperatures and antibiotic treatment (Figure 5), (iv) the acquisition of an NC state. In the NC state, cells maintain viability (the potential for proliferative activity), but could not produce colonies on agar media and, therefore, are not detected using standard tests (Colwell and Grimes, 2000; Oliver, 2010). The determination of whether non-sporulating C. jeikeium have the ability to transition to a long-lived dormant form of the bacteria has the potential to broaden our understanding of the mechanisms by which bacteria survive in natural systems when they encounter conditions not conducive for growth.
The NC state can be reversed by applying a resuscitation procedure in MPN assay either in fresh medium or in the presence of SN. MPN assay was used to evaluate a number of potentially viable cells in cultures that were diluted to disappearance in liquid medium. At the same time this approach permitted the resuscitation of cells that were deprived of the ability to produce colonies on agar plates. This procedure was previously developed to resuscitate mycobacteria by cultivating “non-culturable” cells in liquid medium (Shleeva et al., 2011). The increased necessity of SN-based resuscitation as C. jeikeium cultures age could be due to the existence of cells in different physiological state: those that are simply unable to form colonies on plates and those that are at the depth of dormancy. Previously, we have found that a secreted protein Rpf (resuscitation promoting factor) provided culture SN activity for Micrococcus luteus (Mukamolova et al., 1998). The protein stimulates the resuscitation of non-culturable cells and shortens the lag phase of active M. luteus cultures inoculated with a low dose of bacteria. Rpf family is comprised of peptidoglycan hydrolases (Kana and Mizrahi, 2010) and the products of their enzymatic activity stimulate the reactivation of dormant mycobacteria (Nikitushkin et al., 2013, 2015). The expression of the rpfA gene in M. tuberculosis is under control of the transcriptional regulator Rv3676 and a cAMP receptor protein (CRP) (Rickman et al., 2005). Presumably, a protein similar to the Rpf protein in M. luteus is the active factor present within the cell free C. jeikeium culture liquid. Indeed, the C. jeikeium genome contains three genes encoding proteins similar to Rpf (JK_RS02150/jk0416, RpfA; JK_RS07760/jk1512, RpfB; JK_RS00265/jk0051, RpfC), which may be released into the environment and function to resuscitate NC C. jeikeium cells.
However, other reactivation stimulants, such as phospholipids (Zhang et al., 2001), free unsaturated fatty acids (Nazarova et al., 2011), cAMP (Shleeva et al., 2013), muropeptides (Nikitushkin et al., 2013), may also be present within the supernatant of actively growing bacterial cultures. It was previously shown that unsaturated fatty acid-dependent adenylate cyclase Rv2212 and a cAMP-dependent transcription factor of the Crp family Rv3676 participate in the reactivation of dormant forms of M. tuberculosis (Shleeva et al., 2017, 2019a). There is also a link between the Crp factor Rv3676 and RpfA protein synthesis (Bai et al., 2005). The C. jeikeium genome contains a gene encoding a cAMP-dependent transcriptional regulator of Crp family, glxR (JK_RS10065/jk1972), which may have a function that is similar to Rv3676.
It is clear that dormant cells could be of significant importance for medicine and microbiology due to their ability to form normal, viable organisms after resuscitation (Kaprelyants et al., 1993; Barer, 1997; Barer et al., 1998; Kell et al., 1998; Barer and Harwood, 1999; Mukamolova et al., 2003). For example, diphtheria bacteriocarrier phenomenon is known, and is considered one of the mechanisms by which a pathogen can persist within the host organism (Deviatova, 1956; DeWinter et al., 2005; Kostyukova and Bechalo, 2018).
C. jeikeium cells are resistant to a number of many beta-lactam antibiotics, doxycycline and ciprofloxacin, but they are sensitive to vancomycin and rifampicin (Soriano et al., 1995). However, the dormant forms of C. jeikeium used in the study that were produced as a result of the acidification of the environment proved to be resistant to both rifampicin and vancomycin (Figures 4B,C). Researchers previously demonstrated that, in some cases, vancomycin may be ineffective in the treatment of C. jeikeium bacteremia and endocarditis and disease recurrence occurred despite providing an extended course of antimicrobial therapy (Vanbosterhaut et al., 1989; Clarke et al., 2019). Wang et al. (2001) “retrospectively reviewed 53 cases of C. jeikeium bacteremia in bone marrow transplant recipients who had a Hickman catheter without signs of local catheter site infection. The results showed that salvage of catheter with vancomycin therapy is successful in most patients (93%)” (Bookani et al., 2018). However, 7% of patients that underwent catheter salvage were affected by recurrent bacteremia. Whilst the genetic characterization of causative pathogen was not performed in the study, the emergence of such resistance of corynebacteria to the antibiotic could be a consequence of the transition of bacteria to a dormant state, which protects them from being killed by antibiotic treatment. The fact that C. jeikeium was resistant to both vancomycin and rifampicin allows us to suggest a model of C. jeikeium dormancy that can be used to screen for new drugs that possess activity against chronic C. jeikeium bacteremia.
Evidently, dormant bacteria with little metabolic activity are capable of avoiding effects of antibiotics even though they remain suitable targets. It can be assumed that the effect of complete elimination of dormant, “non-culturable” bacteria can be accomplished by factors which produce indirect, harmful effects on bacteria (Kaprelyants et al., 2018). In this regard, we employed a previously discovered phenomenon of the accumulation of free and methylated porphyrins in dormant mycobacteria (Nikitushkin et al., 2016). This finding made it possible to induce the PDI of mycobacteria (Shleeva et al., 2019b, 2020). In the present study we found that dormant C. jekeium cells accumulate a pigment which according to the spectral properties (Figures 6A,B) evidently belong to a class of porphyrins (Gouterman, 1961).
Of the currently available photosensitizers (PSs) available, porphyrin and its analogs have the following advantages: (1) many are effective producers of singlet oxygen; (2) they usually do not generate singlet oxygen in the absence of light; (3) they absorb in the red region of the optical spectrum; and (4) they are relatively stable. As an alternative to antibiotic treatment for killing dormant C. jeikeium, PDI was a promising approach that was used for killing dormant bacteria (Figure 7). In a previous study, an elevated accumulation of coproporphyrin III was observed in C. diphtheriae culture filtrates (Gray, 1948). The overproduction of porphyrins was also observed in 15 days old C. acnes cultures (Cornelius and Ludwig, 1967). However, the link between porphyrin accumulation and C. jeikeium dormancy was not established previously.
The functional role for the porphyrins that accumulate in dormant bacteria is not clear. Likely that newly found, hydrophobic porphyrins may stabilize and protect dormant cells against unfavorable conditions or destructive factors (Patel and Day, 1999; Castello et al., 2008; Antonova et al., 2010). In addition, stored porphyrins may be utilized as metabolites in biochemical pathways during dormant bacterial cell reactivation.
Bacterial killing in association with PDI seems to be due to the production of reactive oxygen compounds, which participate in the oxidation of such important molecules as enzymes, proteins, lipids and nucleic acids, are lethal to the bacterial cell (Demidova and Hamblin, 2004; Jori et al., 2006). PDI is known as a robust approach for killing bacterial cell including multi-resistant bacteria (Grinholc et al., 2008; Cassidy et al., 2010). PDI was found to be efficient for eliminating of a number of bacteria (Plavskii et al., 2018), including virulent E.coli O157H7, Listeria monocytogenes (Romanova et al., 2003), Neisseria gonorrhoeae (Wang et al., 2019) and Legionella rubrilucens (Schmid et al., 2019). Commonly, exogenously added photosensitizers, in particular porphyrins, were applied in PDI experiments (Liu et al., 2015; Guterres et al., 2020; Rossi et al., 2020). For example, cyclodextrin with bound porphyrins has been applied for the treatment of Staphylococcus aureus (Hanakova et al., 2014). Endogenous porphyrins were utilized for the killing of pathogenic periodontal bacteria (Cieplik et al., 2014; Hope et al., 2016), skin pathogens St. aureus (Ashkenazi et al., 2003; Lipovsky et al., 2009), Staphylococcus carnosus (Hoenes et al., 2020), Hemophilis parainfluenzae (van der Meulen et al., 1997) and Helicobacter pylori (Hamblin et al., 2005; Morici et al., 2020).
Apart of photosensitizer, application of 5-aminolevulinic acid (ALA) was employed as another approach. ALA being photodynamically inactive stimulates biosynthesis of endogenous PSs in cells that enhances harmful effect upon illumination (Harris and Pierpoint, 2008; Bohm et al., 2020). In bacteria, close to eukaryotic cells, the biosynthesis of porphyrins from ALA took place, and ALA administrated to the cell culture enhances porphyrin accumulation in the cell (Nitzan et al., 2004).
Application of ALA-PDI in clinic revealed positive results notably for the skin diseases such as acne, chronic folliculitis, rosacea, lichen sclerosis and skin lesions caused by Mycobacterium marinum. According to the clinical trials, the usage of ALA-PDI look promising for the treatment of ulcers caused by Helicobacter pylori (Wilder-Smith et al., 2002; Nitzan et al., 2004). However, the application of PDI for eliminating both active and dormant corynebacteria was demonstrated for the first time in the present study. This work suggested a new approach that has the potential to be used to eradicate dormant corynebacteria in clinical settings.