The Kinocidin Interleukin-26 Shows Immediate Antimicrobial Effects Even to Multi-resistant Isolates

The cationic proinflammatory cytokine Interleukin 26 (IL-26) shows antibacterial activity and inhibits the replication of cytomegalovirus and hepatitis C virus. This study evaluates the early microbicidal activities of IL-26 against major bacterial species including multi-resistant variants and Candida albicans. Recombinant IL-26 was bacterially expressed and studied for its microbicidal effects in culture. We show that IL-26 has strong 90% bactericidal activities against Enterococcus faecalis, Enterococcus faecium, Staphylococcus aureus, and Acinetobacter baumannii. Similarly, IL-26 sensitivity was also detectable in vancomycin-resistant Enterococcus species, methicillin-resistant S. aureus, and carbapenem-resistant A. baumannii clinical isolates. Additionally, a significant, albeit weak fungicidal effect against Candida albicans was observed. Activities against Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa were not detectable. The proinflammatory cytokine and kinocidin IL-26 shows strong bactericidal activities against A. baumannii and, almost selectively, against Gram-positive bacteria.


Production of Recombinant Protein
Induction of IL-26 protein synthesis was achieved in Escherichia coli strain XL1blue, which is capable of producing IL-26 in inclusion bodies, by adding 1 mM isopropyl β-D-1thiogalactopyranoside and cultivation for 6 h (Knappe et al., 2000;Hör et al., 2004). First, the bacteria were disrupted mechanically in 6 M guanidinium chloride with 0.1 M NaH 2 PO 4 , 0.01 M Tris, and 100 mM β-mercaptoethanol. Bacterial debris was removed by centrifugation for 30 min at 8,873 × g and by sterile filtration. Nickel-chelate affinity chromatography was used for protein purification under denaturing conditions (Ni Sepharose 6 Fast Flow, GE Healthcare Life Sciences, Buckinghamshire, UK). Particle-bound IL-26 was eluted under increasing imidazole concentrations (40 mM, 500 mM, and 600 mM). Eluate fractions were tested for 19 kDa bands in Coomassie-stained protein gels. Renaturation of relevant eluate fractions was achieved by dialysis for 24 h twice, at 4°C (20 mM HEPES, 1 mM MgCl 2 , 20 mM KCl, 0.1 mM EDTA, 1% glycerol, 1 mM oxidated glutathione, and 5 mM reduced glutathione, pH 8.0).

Testing for Antimicrobial Activity
Bacteria and C. albicans were initially grown on Columbia sheep blood agar at 37°C overnight and then for 20 h at 37°C in 10 ml tryptic soy broth (TSB) medium with 10 mM NaCl. Then, 50 μl of the culture was added to 10 ml of TSB/10 mM NaCl medium and incubated for 3 h at 37°C for reaching the logarithmic phase of growth and the optical density at 600 nm was determined. We diluted the culture with 10 mM NaCl to reach target concentrations of 10 5 CFU/ml. Next, 100 μl of the microbial suspension was added to 100 μl of different concentrated IL-26 solutions (100, 30, 10, 3, and 1 μg/ml) to achieve final IL-26 concentrations in the wells of 50, 15, 5, 1.5, and 0.5 μg/ml. We always ran negative controls with pure dialysis buffer instead of the IL-26 solution, as well as positive controls with bacteria that were previously shown to be highly sensitive to IL-26, when species with low or lacking IL-26 sensitivity were examined.
Immediately after the start of the treatment, and after 1, 2, 3, and 4 h, 20 μl from each culture was sampled and diluted in a 0.85% NaCl solution and 100 μl of that dilution was pipetted on two lysogeny broth (LB) agar plates each. The plates were incubated overnight at 37°C and the colonies were counted. The numbers of colony forming units per milliliter (CFU/ml) were calculated after averaging of the counting results of the two separate plates in consideration of the dilution factor. Totally, 15,000 LB agar plates were plated by hand for 66 independent experiments.

Statistical Evaluation
Statistical evaluation was performed with GraphPad PRISM 8 (GraphPad Software, Inc., San Diego, CA, United States). Values of p were determined by using the independent two-sample t-test. We used linear regression for calculating the minimal concentration for 90% bactericidal or fungicidal effects (MBC 90 / MFC 90 ), respectively, the lethal dose or concentration for 90% killing (LD 90 /LC 90 ). All results of this study are given as MBC 90 / MFC 90 , even though the use of LD 90 /LC 90 is also common in AMP research. The use of the terms MBC/MFC may be more precise from a microbiological point of view. Results with values of p ≤ 0.05 were rated significant. As far as not given in numbers, significances are labeled as followed: *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; and ****p ≤ 0.0001, ns = not significant. Top: Detection of phospho-STAT3 using primary polyclonal rabbit-raised phospho-STAT3 antibodies. Bottom: Detection of STAT3α and STAT3β using primary polyclonal rabbit-raised STAT3 antibodies. For both analyses, horseradish peroxidase-conjugated anti-rabbit IgG antibodies were used as second reagents. Chemiluminescence detection was performed with a charge-coupled device camera (LAS-3000, Fujifilm, Japan).

IL-26 was tested for bactericidal activities against different
Gram-positive and Gram-negative strains of different antibiotic resistance phenotypes, as well as for fungicidal activities against C. albicans. All experiments were performed with a standardized protocol and were highly reproducible. The anti-infective activities were analyzed quantitatively for the minimal concentration for 90% bactericidal or fungicidal effects (MBC 90 /MFC 90 ).

Activity Against Enterococci
Directly after adding 50 μg/ml IL-26 to Enterococcus faecalis ATCC 29212 cultures, an immediate and highly significant (p ≤ 0.0001) reduction of the number of colony forming units (CFU) of approximately 98.4% was detected. After 1 h, significant effects were found from 5 μg/ml on. Complete eradication of all bacteria was reached at 50 μg/ml IL-26, after incubation for at least 1 h. The CFU numbers after 2, 3, and 4 h of incubation were reduced by 97.3, 92.3, and 96.1% at 15 μg ml IL-26 (Figure 2A).
Similar to the antibiotic-sensitive type strain, a vancomycinresistant clinical isolate of E. faecalis showed an instant significant effect. After 1 h, significant CFU number reductions were found starting from 0.5 μg/ml, whereas the strongest results were observed at 15 μg/ml (reduction of 72.5%) and 50 μg/ml (99.6%). The reduction levels stayed constant over time. After 4 h, the incubation with 50 μg/ml IL-26 resulted in a highly significant decline of CFU numbers by 99.95% ( Figure 2B).
Likewise, Enterococcus faecium ATCC 6057 showed a prompt significant drop of the CFU numbers at 15 μg/ml by 70.6% and at 50 μg/ml by 99.5%. This effect increased, when the incubation was performed for 1 h. Reduction values were 81.3% for 5 μg/ ml, 99.96% for 15 μg/ml, and a complete killing of all bacteria was achieved for 50 μg/ml. After 2 and 4 h, complete eradications were reached at 15 μg/ml. At 3 h, reductions amounted to 99.98% for 15 μg/ml and 100% for 50 μg/ml ( Figure 2C).

Activity Against Staphylococcus aureus
In the case of S. aureus ATCC 6538, we found an instant reduction of the CFU numbers of approximately 23.2% at 50 μg/ml IL-26 (p = 0.37), which increased after 1 h to a highly significant reduction by 94.6% (p < 0.0001). Further increments were achieved upon ongoing incubation. CFU reductions at 50 μg/ml reached 95.6, 98.0, and 93.4% after 2, 3, and 4 h. After 2 and 3 h, significant but weak effects were detectable at 15 μg/ml (Figure 2E).

Activity Against Multi-resistant Gram-positive Bacteria
The MBC 90 for immediate effects ranged from 38.5 μg/ml (MRSA ATCC 33593) to 57.5 μg/ml (vancomycin-resistant E. faecalis), except for the methicillin-sensitive S. aureus ATCC 6538 with an unusually high MBC 90 of 136 μg/ml. Two groups of similar MBC 90 ranges were defined. In the first group, E. faecalis ATCC 29212, E. faecium ATCC 6057, vancomycin-resistant E. faecium ATCC 17050, and MRSA ATCC 33593 had average MBC 90 values between 12.5 μg/ml and 13.4 μg/ml IL-26 for 1 to 3 h and 19.3 μg/ml after 4 h. In the other group, the average MBC 90 for 1 to 4 h for vancomycin-resistant E. faecalis and methicillin-sensitive S. aureus ATCC 6538 ranged between 43.7 μg/ml and 45.5 μg/ml (Table 1). Hence, we were able to show for the first time that IL-26 functions as a highly active bactericidal agent against different Gram-positive bacteria. These effects are independent of the antibiotic resistance phenotypes since multi-resistant strains of S. aureus, E. faecalis, and E. faecium were as sensitive as or even more sensitive than their antibiotic-sensitive counterparts.

Activity Against Acinetobacter baumannii
Concerning Gram-negative species, we first tested A. baumannii ATCC 19606. Similar to all tested Gram-positive bacteria, an initial albeit weak CFU reduction was observed. After 1 h of incubation, a highly significant decline of 99.8% was detected at 50 μg/ml IL-26 and stayed constant for 4 h (±0.3%). Additionally, we found a 50% non-significant reduction at 15 μg/ml ( Figure 2G).
Moreover, two highly resistant A. baumannii CR isolates were analyzed, which solely were sensitive for colistin. The first one was isolated in 2014 from a patient from a local outbreak at the University Hospital Schleswig-Holstein in Kiel, Germany. An immediate reduction was not observed but, again, highly significant CFU reductions were measureable at 1, 2, 3, and 4 h (99.4, 99.3, 99.7, and 99.7%) at IL-26 concentrations of 50 μg/ml ( Figure 2H). The other A. baumannii CR isolate resulted from a patient from 2018 with previous hospitalization in a country with high prevalence of colonization with CR A. baumannii. Here, we detected a significant immediate Frontiers in Microbiology | www.frontiersin.org reduction of 31.7% at 50 μg/ml IL-26. Furthermore, weaker and non-significant reductions were detectable at 0.5 μg/ml, 5 μg/ml, and 15 μg/ml. After 1 h, the decline of the CFU count at 50 μg/ml IL-26 was 98.7% and increased in the course of the experiments to 99.4% (2 h), 99.8% (3 h), and 99.6% (4 h; Figure 2I). The mean MBC 90 values for the three A. baumannii   (Figures 2J-L).

Activity Against Candida albicans
Last, the sensitivity of C. albicans ATCC 24433 against IL-26 was investigated. Immediate effects were not detected at up to 50 μg/ml IL-26. The CFU reduction values at 50 μg/ml IL-26 were 29% after 1 h (non-significant), 67.5% after 2 h, 58.2% after 3 h, and 78.5% after 4 h (all significant). The MFC 90 values ranged from 54.7 to 81.41 μg/ml (Table 1). Hence, we were able to show a reproducible and significant, albeit weak fungicidal activity of IL-26 against C. albicans.
Thus, the cytokine IL-26 with its proinflammatory, bactericidal, antiviral, and fungicidal activities can be attributed to the group of kinocidins which was defined for cytokines with direct antimicrobial effects, such as human mammalian platelet factor 4 (hPF-4; Yount et al., 2004;Yeaman et al., 2007;Larochette et al., 2019).

DISCUSSION
Whereas all tested Gram-positive bacterial strains were highly sensitive to IL-26, we observed a remarkable difference between A. baumannii strains and all other Gram-negatives (enterobacteria and P. aeruginosa). Thus, the questions arise why all other Gram-negative species are non-sensitive for IL-26 and what the essential factor is for the sensitivity of A. baumannii. The lipopolysaccharides (LPS) and especially the O-antigen, which is the outer chain of the LPS, are two known factors for AMP resistance in Gram-negative bacteria (Silhavy et al., 2010;Joo et al., 2016). Acinetobacter species are unable to produce complete LPS due to the absence of O-antigen-ligase activity and the lipooligosaccharide (LOS) core, the lipid A, is remaining (Weber et al., 2016). Thus, O-antigen might be responsible for the IL-26 resistance of Gram-negative species.
Concerning the mode of IL-26 action, the direct interaction with the bacterial cell membrane including pore formation seems likely, similarly to other AMP (Patel and Akhtar, 2017). Due to its high cationic charge, IL-26 binds to glycosaminoglycans of the surface of eukaryotic cells (Hör et al., 2004), as well as to LPS and lipoteichoic acid of the surface of bacteria (Meller et al., 2015). Based on electron micrographs of P. aeruginosa ATCC 27853, bleb-formation followed by membrane disruption was described as the mode of IL-26 action (Meller et al., 2015). However, this needs to be interpreted with caution, since exactly the same P. aeruginosa strain has been classified as IL-26 resistant in this study.
The initial publication concerning antimicrobial activities of IL-26 described bacteriostatic effects at 50% level for 5 to 10 μM IL-26 against P. aeruginosa ATCC 27853, E. coli ATCC 11775, K. pneumoniae O1:K2, and S. aureus ATCC 6538 but no detectable effects against E. faecalis ATCC 29212 and C. albicans ATCC 24433 (Meller et al., 2015). In contrast, our study was able to demonstrate strong bactericidal activities at 90% level against the Gram-positive strains E.    (Agak et al., 2018). In a third publication, an activity against biofilm formation of S. aureus was detected which was more pronounced for IL-26 than for the AMP LL-37 (Woetmann et al., 2018). Regarding Mycobacteria (M.) IL-26 has been shown to inhibit growth and reduce viability of M. leprae and M. tuberculosis in axenic cultures as well as within macrophages, probably by inducing lysis by bleb-formation after interaction with lipoarabinomannan (Dang et al., 2019;Hawerkamp et al., 2020). The reason for these functional differences might be due to the different sources and qualities of commercially available IL-26 (Knappe et al., 2000;Hör et al., 2004). In order to ensure high quality and functionality, we used self-produced IL-26 which was functionally tested for STAT3 activation in a colonic carcinoma cell line. Moreover, molar concentrations are difficult to interpret since IL-26 occurs as monomers, dimers, oligomers, and even multimers (Knappe et al., 2000;Fickenscher et al., 2002;Meller et al., 2015).
Concerning the raising relevance of antibiotic-resistant bacteria, AMP have come under scrutiny, especially due to the lack of resistance development, even though they have always been present during evolution (Zasloff, 2002;Gallo and Hooper, 2012). Besides potential pharmacotechnological and galenic challenges in the production of this instable kinocidin, additional immunologic effects of the proinflammatory IL-26 need to be considered in the case of a possible therapeutic application. In published experiments, IL-26 or LL-37 was applied nasally in mice after nasal application of bacteria and a certain reduction of the CFU number (factor 10-100) was seen (Meller et al., 2015). However, these experiments were performed with K. pneumoniae, which was classified as IL-26 resistant in this study.
Compared to other AMP, which usually have a broad activity spectrum against Gram-positive and Gram-negative bacteria, protozoa, and fungi (Ebbensgaard et al., 2015), IL-26 has its main target in Gram-positive bacteria, which has not yet been described for other AMP. Thus, IL-26 is a unique member of the family of cationic AMP. In summary, we showed for the first time that IL-26 is a proinflammatory kinocidin with bactericidal and fungicidal activities, which is also active against A. baumannii and C. albicans and kills Gram-positive bacteria almost selectively.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article, further inquiries can be directed to the corresponding author.

AUTHOR CONTRIBUTIONS
BTH wrote the manuscript, designed and performed the experiments, and prepared the figures and tables. GM contributed to the experiments, reviewed and edited the manuscript, and supervised the project. RP gave advice for the experimental design and reviewed and edited the manuscript. HF wrote the manuscript, gave advice for the experimental design, and supervised the project. All authors contributed to the article and approved the submitted version.

FUNDING
This project was supported in part by the Excellence Clusters Inflammation at Interfaces and Precision Medicine at Kiel and funded in part by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) -project number 413490537 within the Kiel Clinician Scientist Programme in Evolutionary Medicine. We acknowledge financial support by the state Schleswig-Holstein within the funding programme Open Access Publikationsfonds.

ACKNOWLEDGMENTS
We thank Andrea Hölzgen and Petra Krüger for expert technical assistance. Frontiers in Microbiology | www.frontiersin.org