Investigation of stress responses in microorganisms and elucidation of the mechanisms of their formation are among the most important aspects of fundamental biology. The evolutionarily developed resistance of microorganisms to stressors of various nature is considered to be the primary cause of the increasing inefficiency of antimicrobial therapy. Under natural conditions, the ability of microorganisms to resist stresses due to production of extracellular compounds plays a key role in the formation and maintenance of homeostasis of microbial communities. Our previous studies have demonstrated the ability of bacteria, archaea and lower eukaryotes (yeasts) to produce extracellular compounds of peptide nature, which exert a protective and reactivating effect on the cells of both producers and members of other domains after exposure to various stressors, including UV radiation, heating, oxidants, and bile salts (Vorobjeva et al., 2013, 2015).

The test objects exposed to lethal impacts were gram-positive and gram-negative bacteria (Vorob’yova et al., 2010; Vorob’eva et al., 2013), yeasts of different physiological and ecological groups (Vorob’eva et al., 2011), and archaea (Vorob’eva et al., 2013). The basic model organism for studying the patterns of reactivating factor (RF) formation, its physicochemical characteristics, and conditions of manifestation of its biological activity was the bacterium Luteococcus japonicus subsp. casei (L. casei). Object selection was determined by a broad range of its adaptive properties, i.e., the ability to utilize various substrates (Vorobjeva, 1999; Thorenoor et al., 2009) and by occurrence of representatives of the genus Luteococcus in a wide environmental range. Luteococcus sp. strains were isolated from soil and water (Tamura et al., 1994), dairy products and cheese (Vorob’eva et al., 1983), blood samples (Collins et al., 2003), deep-sea sediments of the Pacific Ocean (Fan et al., 2014), and activated sludge of wastewater treatment plants at cellulose factories (Thorenoor et al., 2009).

The L. japonicus subsp. casei strain was isolated from Sovetskii cheese. Molecular biological studies, ribotyping and restriction analysis of the 16S and 23S rDNA fragments have shown that the isolated Luteococcus strain should be assigned to the family Propionibacteriaceae (Van Niewholtz, 1998; Vorobjeva, 1999). The closest relative of the strain is Propioniferax innocuum. Previously, from the culture liquid of L. casei we have isolated a reactivating factor exerting protective and reactivating effects on the cells of its producer and other organisms: bacteria, archaea, and yeasts. It has been shown that RF is a peptide and a complex of two active fractions (Vorob’eva et al., 2015). RF is synthesized and excreted by undamaged cells under normal growth conditions at extremely low concentrations (Vorob’eva et al., 2013). An inverse dependence was found between stressor intensity and RF efficiency (the survival rate of microbial cells) (Vorob’eva et al., 2011; Vorob’eva et al., 2013). Thus, the function of RF is of particular importance under stresses of lethal intensity. RF has no strict species specificity as was demonstrated by the cross-domain effect of the preparation (Vorob’eva et al., 2013). The type of concentration dependence, the high rate of reactivating effect and stress resistance suggest the signaling mechanism of action of the peptide RF (Vorobjeva et al., 2009). Using the constructed tester strain Escherichia coli bearing the umuD-lacZ operon it was established that the protective effect of RF was not associated with the induction of SOS response (Loiko et al., 2013) and seemed to be provided by the signaling-type RF effect on the cytoplasmic membrane.

The goal of the present work was to identify the structure of the peptide reactivating factor from L. casei.

MATERIALS AND METHODS

Research subjects and cultivation conditions. The research subject was a gram-positive bacterium Luteococcus japonicus subsp. casei originally isolated from cheese (Vorob’eva et al., 1983). The strain was cultivated under static conditions at 32°С in flasks with 100 mL of glucose-mineral medium containing the following (%): glucose, 1.0; (NH4)2SO4, 0.3; KH2PO4, 0.1; NaH2PO4, 0.2; MgSO4, 0.002; CaCl2, 0.002; NaCl, 0.002; yeast extract, 0.1; pH 7.0. The L. casei culture in the linear growth phase was centrifuged (10 000 g, 20 min) and the culture liquid (CL) was used for RF isolation. L. casei cells were washed twice with 0.05 M Na-phosphate buffer solution, pH 7.4, and resuspended in the same buffer to the cell density of OD540 = 0.4–0.6. The cell suspension was used to determine the biological activity of RF.

RF activity assay. CL was run through a nitrocellulose membrane filter with 0.22-µm pore diameter (Millipore, United States); the components adsorbed on the filter were eluted with 3% NaCl solution and then the solution was run through a low protein affinity membrane filter (Pall, United States) to completely remove the cells.

The cell number in the suspensions of L. casei and in the strains of yeasts, lactic acid bacteria, and propionic acid bacteria was determined by the titer of colony-forming units (CFU/mL) after plating of 10N-fold diluted suspensions on the relevant solid (2% agar). Micro inoculation with 5-µL aliquots of suspensions was performed with six replicate samples per dilution.

Reactivating factor isolation and structure determination. In this study we used the previously developed scheme of fractionation and purification of polypeptide components secreted by L. casei into the culture liquid (Vorob’eva et al., 2015). The culture liquid was desalted, concentrated, and fractionated by affine chromatography in a Heparin HiTrap Sepharose column (GE HealhCare, Sweden). The anionic total fraction not bound to the sorbent was rechromatographed by the analytical reversed-phase (RP) HPLC in a Vydac C18 Protein & Peptide 4.6 × 250 mm column (GraceVydac, United States). Chromatography was carried out with the linear gradient of solvent B (80% CH3CN, 0.1% TFA) relative to solvent A (0.1% TFA) from 0 to 50% for 50 min at a flow rate of the mobile phase of 1 mL/min. The peptides were detected by absorption at 214 nm. Molecular masses of the active components were measured on an Ultraflex MALDI-mass spectrometer (Bruker Daltonics, Germany) equipped with an UV laser (337 nm) in the positive ion mode. The matrix used was 2,5-dihydroxybenzoic acid. Equal volumes (1 µL) of the samples and the matrix (20 mg matrix/mL in 80% aqueous CH3CN, 0.1% TFA in MQ deionized water) were mixed on the target and the resultant mixture was air-dried. The mass spectra were analyzed using FlexAnalysis for TOF. The error of measurement was 2 ppm. The primary structure of the biologically active peptide was determined by the Edman degradation with a PPSQ-33A automated protein and peptide sequencer (Shimadzu Corp., Japan) according to the manufacturer’s protocol. The experimental data were interpreted using the LabSolutions software package (Shimadzu Corp., Japan). The homologies of amino acid sequences were searched in the UniProt/ SwissProt and trEMBL databases using the BLASTP algorithm; the translated sequences were searched on the basis of the L. casei annotated genome kindly provided by V. Loux (Collective Genomes Gbk 17320. contigs final gbk).

Determination of the RF biological activity. The stress factors used were UV radiation, bile salts, and oxidative stress. The source of UV radiation consisted of two 30-W BUV-15 lamps (Russia) mounted in parallel, with the major part of emission at 253.7 nm. Preliminary experiments showed the “dose–response” dependence. At a radiation dose of 81 J/m2, the survival rate of irradiated cells was 0.01–0.03%. The target activity was determined in the cell suspension incubated with RF for 10 and 15 min before and after stress exposure, respectively. The reactivating effect was assessed by comparing viable cell numbers (CFU) emerging on a solid medium (2% agar) in experimental and control variants; in the latter there was no reactivating factor. The bacterial survival rates in experimental and control variants were expressed in percentage relative to non-irradiated control. The test objects were the cells of RF producer Luteococcus casei, as well as the cells of lactic acid bacteria from the collection of the Department of Microbiology (Moscow State University, the Faculty of Biology): Lactobacillus acidophilus (Lb. acidophilus), Lactobacillus casei (Lb. casei). The propionic acid bacteria Propionibacterium freudenreichii (no. 1857) and P. acidipropionici (no. 1859) were obtained from the Czech Collection of Microorganisms (ССM). The yeast Saccharomyces cerevisiae VKPM Y-1200 was obtained from the All-Russian Collection of Industrial Microorganisms. The cell suspensions of tester strains were obtained as described above.

RESULTS AND DISCUSSION

Biological characteristics of RF. The bacterium L. casei is a facultative anaerobe represented by large, nonmotile non-spore-forming spherical cells producing orange colonies on agar surface. The major fermentation products are propionic and acetic acids; caprylic, isocaproic, butyric, hexanoic (caproic), and formic acids are produced in minor amounts (Vorob’eva et al., 2014). It was previously shown that the number of surviving bacteria exposed to UV radiation and heating increased in the case of short-term (10–15 min) pre- or postincubation of L. casei cells with the fraction of extracellular peptide metabolites of different degrees of purity (Vorob’eva et al., 2003).

The extracellular peptide fraction referred to as RF (reactivating factor) isolated from L. casei CL had a protective effect on stress-exposed cells of not only the producer but also of other prokaryotic and eukaryotic objects.

Peptide exometabolites with antistressor properties are formed by the cells and excreted into the medium under the conditions favorable for normal growth of microorganisms and are accumulated in the medium by the beginning of the stationary growth phase (Table 1).

Table 1. Reactivating effect of the RF from L. casei on stress-exposed microbial cells
Full size table

RF exhibited a universal protective effect on the cells exposed to unrelated stresses: UV radiation, heating, oxidants and bile salts (Table 1).

There is an inverse dependence between the efficiency of protective and reactivating effects of RF and the survival rate of stress-exposed cells. The maximum antistress effect of the exometabolite was observed at an extremely low number of survivor cells after lethal impact (0.01–0.1%). The protective effect (preincubation with RF) was several times more pronounced than the reactivating effect (postincubation with RF) exhibited after lethal impact (Vorob’eva et al., 2003). It has also been noted that the efficiency of reactivation of L. casei cells in the stationary phase was 2 times less compared to the cells in the logarithmic growth phase (Vorob’eva et al., 2009).

RF showed a universal biological effect against microorganisms from different taxa (including probiotic) exposed to stress factors of diverse nature (Table 1).

The above-mentioned patterns in manifestation of the biological activity of RF suggest that its effects are associated with formation of a small subpopulation of stress-resistant persister cells. RF probably induces the phenotypic transition of regular vegetative cells into persisters, which accounts for its much (up to 10-fold) greater effect in the variants with culture preincubation before the stress compared to the variant with postincubation. However, the possible effect of RF at the stage of induction of reversed transition of persisters to vegetative growth cannot be ruled out either. In this case, the higher efficiency of RF during preincubation may be explained by an increase in its biological activity due to UV radiation. This explanation is not contrary to the data on enhanced efficiency of RF under exposure to radiation (as part of the purified fraction) before the effect on the cells in the postincubation variant (Vorob’eva and Khodzhaev, 2010).

Determination of the structure of the extracellular RF from L. casei. Extracellular RF was isolated from the culture liquid of L. casei by the optimized protocol including solid-phase extraction (desalting) of concentrated CL on a hydrophobic sorbent with the C8 phase, followed by evaporation, lyophilization and separation by affine chromatography with a stepwise gradient of increasing NaCl concentration from 0 to 1 M. The total unbound fraction, without the stage of separation by analytical gradient anion-exchange chromatography, was separated by RP-HPLC; as a result, 15 predominant fractions were collected and designated as I–XV (Fig. 1). These fractions were lyophilized to remove residual amounts of TFA. The results of testing the fractions for the sought-for reactivating effect showed its presence only in two components (I and IV). For determining the structural characteristics, both fractions were analyzed by MALDI-TOF mass spectrometry. We failed to obtain effective ionization and, accordingly, mass spectrum signals for fraction I, while the mass-to-charge ratios for component IV were 773.394 and 788.102 Da. Further identification of this active fraction was performed by the Edman’s method of stepwise automated sequencing; as a result, the complete amino acid sequences of two peptides within component IV were established: 1Ala-Pro-Asn-Glu-Asn-Gln-Gly8 (peptide IV-1) and 1Ala-Pro-Asn-Glu-Glu-Gln-Gly8 (peptide IV-2). Consequently, both peptides were highly homologous, differing in only one amino acid substitution: Asn5Glu. Their formation apparently was not associated with template synthesis, because such sequences have not been found in the genome of L. casei (V. Loux, Collective Genomes Gbk). The search of homologies in the databases of the primary structures of polypeptides revealed no similarity with any of the known full-size functional peptide molecules. However, the data obtained make it possible to state with certain probability that this peptide is a fragment of partial proteolysis of a larger protein. It is characteristic that, according to the genomic data, this protein may contain a unique repeating motif of the APNENQCAPNENQCAPNENQC type, which later on will undergo limited proteolysis with formation of short sequences with the signaling function. Moreover, an additional argument in favor of proteolytic origin of RF may be partial coincidence of the given short sequence with certain parts of some bacterial proteins according to the data (GenBank reference sequence IDs: WP_012104305.1, WP_007615369.1, WP_020569836.1). The identified peptides have no homology with the previously determined RF possessing the functions of a stress “reactivator” from baker’s yeast (S. cerevisiae) (Vorob’eva et al., 2017); however, the similar physicochemical properties, as well as the similar predicted elements of the random coil-type secondary structure, are evidence in favor of the similar mechanisms of signal peptide generation in response to stresses in phylogenetically distant living organisms belonging to different kingdoms.

Fig. 1.
figure 1

The RP-HPLC profile of fractionation of the unbound fraction after separation of desalted concentrated extracellular liquid of L. japonicus by affine chromatography with a stepwise gradient. The collected fractions are denoted by Roman numerals (I‒XV); their retention times on the column are shown. Black arrows indicate the peaks showing the presence of a reactivating effect.

Full size image