key: cord-0959621-fsmuqktd
authors: Dyakov, Ilya N.; Mavletova, Dilara A.; Chernyshova, Irina N.; Snegireva, Nadezda A.; Gavrilova, Marina V.; Bushkova, Kristina K.; Dyachkova, Marina S.; Alekseeva, Maria G.; Danilenko, Valery N.
title: FN3 protein fragment containing two type III fibronectin domains from B. longum GT15 binds to human tumor necrosis factor alpha in vitro
date: 2020-08-06
journal: Anaerobe
DOI: 10.1016/j.anaerobe.2020.102247
sha: 3973dee0a10700bcbb31bf7506462d254bf4aaf6
doc_id: 959621
cord_uid: fsmuqktd

Most species of the genus Bifidobacterium contain the gene cluster PFNA, which is presumably involved in the species-specific communication between bacteria and their hosts. The gene cluster PFNA consists of five genes including fn3, which codes for a protein containing two fibronectin type III domains. Each fibronectin domain contains sites similar to cytokine-binding sites of human receptors. Based on this finding we assumed that this protein would bind specifically to human cytokines in vitro. We cloned a fragment of the fn3 gene (1503 bp; 501 aa) containing two fibronectin domains, from the strain B. longum subsp. longum GT15. After cloning the fragment into the expression vector pET16b and expressing it in E. coli, the protein product was purified to a homogenous state for further analysis. Using the immunoferment method, we tested the purified fragment's ability to bind the following human cytokines: IL-1β, IL-6, IL-10, TNFα. We developed a sandwich ELISA system to detect any specific interactions between the purified protein and any of the studied cytokines. We found that the purified protein fragment only binds to TNFα.

The biomass of BL21(DE3) strain containing the pET16b:fn3 plasmid was 126 resuspended in lysis buffer (50 mM NaH2PO4, 5 mM Tris-HCl, 300 mM NaCl, 10 mM 127 imidazole, 1 mM PMSF, 5 mM DTT, pH 8.0) containing lysozyme (1 mg/ml), 0.5% 128 Triton X-100 and 20 mM 2-mercaptoethanol and was sonicated. Insoluble particles were 129 removed by centrifugation (7500 g, 30 min, 4°C). The resulting lysate was filtered using 130 a 0.22 µm pore size Millex-GP. 131

Protein isolation and purification were performed using a BioLogic LP 132 chromatography system equipped with a fraction collector (Bio-Rad, USA). The 133 clarified lysate was loaded onto a Bio-Scale TM Mini Profinity TM IMAC Cartridge (5 ml, 134 Bio-Rad, USA), then equilibrated and washed with the lysis buffer, followed by another 135 washing with the buffer (50 mM NaH2PO4, 5 mM Tris-HCl, 300 mM NaCl , 50 mM 136 imidazole, 1 mM PMSF, 5 mM DTT, pH 8.0). The bound protein was eluted from the 137 columns with buffer containing 300 mM imidazole. For further purification of the 138 protein, dialysis was performed in PBS buffer containing 10% glycerol and 1 mM PMSF. 139

The concentration of the isolated protein was measured by Qubit fluorimeter 140 (Invitrogen, USA). The average yield of purified protein was approximately 20 mg per 141 liter of E. coli culture. The purified protein was stored at -80°C. 142

The FN3-specific polyclonal antibodies were raised by immunization of rabbits, 144 which was carried out via subcutaneous immunisation ( Freund's adjuvant. Two weeks after the last immunization, 30-50 ml of blood was 151 collected from each immunized rabbit. Blood serum was collected and stored at 4°C 152 after the addition of 0.1% sodium azide (NaN3). 153

The FN3-specific antibodies were isolated by affinity chromatography. The serum 154 from immunized rabbits was diluted 4-fold in PBS, centrifuged and passed through 4B 155 TMB reagent were performed as described above. 195

In the second stage, we used 96-well polystyrene flat bottom plates coated with 197 cytokine-specific antibodies of from the human cytokine determination kits (VECTOR-198 BEST, Russia). We added in each well one at a time, first 100 µl of the cytokines IL6, 199 IL10 or TNFα at a concentration of 250 pg/ml, then 100 µl of the FN3 protein at a 200 concentration of 2 µg/ml and lastly 100 µl of the FN3-specific polyclonal antibodies, 201 obtained as described above, at a concentration of 20 ng/ml. In each case, the plates 202 were incubated for 1 hour at 37°C, followed by washing (4 times) with PBST. Finally, incubation in carbonate/bicarbonate buffer for 1 hour at 37°C followed by washing with 210 PBST four times. To avoid nonspecific adsorption, the wells were incubated in a 1% 211 casein solution for 1 hour at 37°C. After washing, 100 µl of the FN3 protein solution at a 212 concentration of 2 µg/ml was added into the wells, incubated for 1 hour at 37°C and 213 washed four times with PBST. At the next stage, solutions of the cytokines at a 214 concentration of 250 pg/ml were added, incubated for 1 hour at 37°C and washed as 215 To reduce nonspecific interaction with various substrates of the protein FN3, which is 223 classified as an adhesive protein, it was introduced into wells in PBST containing 1% 224 casein (Invitrogen, USA). 225

To search for nucleotide and amino acid sequences we used the databases available 227 on NCBI (http://www.ncbi.nlm.nih.gov/) and UniProt (http://www.uniprot.org/). To 228 calculate the molecular weight and isoelectric point of proteins, we used the ProtParam 229 program (http://web.expasy.org/protparam/).

Previously, based on the analysis of the genomes of 34 species of bifidobacteria, we 233 identified and characterized a species-specific cluster of PFNA genes, which consists of To develop a sandwich ELISA assay that would allow us to check the ability of the 298 recombinant FN3 protein to bind human cytokines, we raised polyclonal antibodies 299 specific to the recombinant FN3 protein. We obtained the FN3-specific polyclonal 300

antibodies from the serum of immunized rabbits. The process of antibody production 301 and isolation is described in the materials and methods section. The resulting affinity 302 purified and concentrated preparation contained 1 mg of FN3-specific rabbit polyclonal 303 antibodies per 1 mL and was highly active -specific interaction with the FN3 protein 304 was detected even after 1:820000 dilution. We used the bacterial aminoglycoside 305 phosphotransferase APHVIII, purified previously at our laboratory, as a negative 306

control. The produced rabbit polyclonal antibodies interacted with APHVIII only in the 307 smallest dilutions (1: 400), which is usually the case for nonspecific binding. Adding 308 rabbit nonspecific γ-globulins instead of the polyclonal antibodies did not cross-react 309 either with the FN3 protein adsorbed to the wells of a 96-well polystyrene plate (see. 310

Supplementary appendix Table S1-S3). 311

To conduct an initial evaluation of the cytokine-binding capacity of the FN3 protein, 313 we first proceeded with ELISA scheme 1, described in the materials and methods. 314

When conducting this study, we had to carry out a number of preliminary experiments, 315 the results of which are not shown. The studied fragment of the FN3 protein reacted 316 with cytokines in the same way. Assuming that the capacity of the constructed system 317 with different cytokines may vary, we had to evaluate the effect of dose dependence. It 318 turned out that scheme 1 was not suitable for our task since the FN3 protein generated a 319 strong positive reaction with all the cytokines, and the reaction intensity did not 320 correlate with the concentration of cytokines bound to the solid phase (Fig. 1) . Further 321 testing showed (scheme 2 ELISA) that the FN3 protein binds nonspecifically to virtually 322 any substrate attached to a solid surface, including the antibody-cytokine complex, 323 antibodies to cytokines, and even to non-treated polystyrene (Fig. 2) . All while 324 nonspecific interaction of other reaction components with each other and with the solid 325 surface was excluded experimentally (see materials and methods). 326

To reduce nonspecific binding to substrates, 1% casein was added to the solution of 327 the FN3 protein. The FN3 protein was not able to bind specifically to immobilized 328 cytokines (either to directly adsorbed cytokines or to cytokines attached to a trap of 329 antibodies). This could be explained by the fact that adsorption onto a solid surface can 330 lead to the blockage of the binding sites of FN3-molecule. Moreover, a strong 331 nonspecific interaction between the FN3 protein and the solid surface (see Fig. 2 ) could 332 hinder any specific interaction. 333

To exclude the possibility of blockage of the cytokine-binding motif of the FN3 335 protein and its nonspecific adsorption onto the solid surface, we employed ELISA 336 scheme 3 described in the materials and methods section: The FN3-specific rabbit 337 polyclonal antibodies were first adsorbed onto the solid surface after which the FN3 338 protein was added as a "second layer". Then, we added into the wells the cytokine 339 solutions and the peroxidase conjugates from the commercial kits VECTOR-BEST for 340 detecting cytokines. This design revealed a specific interaction between the FN3 protein 341 and the cytokine TNFα, but not with IL-6, IL-1, or IL-10 (Fig. 3) . This observation 342

suggests the possibility that the FN3 protein of B. longum subsp. longum GT15 binds 343 human cytokines, namely, the pro-inflammatory cytokine TNFα. As a control for TNFα, 344

we used a cytokine that does not bind specifically the FN3 protein in our ELISA 345 scheme, IL-6. 346

Bifidobacteria as strictly anaerobic bacteria are considered to be some of the oldest 348 representatives of Actinobacteria [36]. As they coevolved alongside their hosts (insects, 349 birds, and mammals), they appear to have developed a species-specific gene cluster 350 (PFNA operon), which allowed them to adapt better to their niches and interact with 351 the host's immune system [19, 20] . The PFNA cluster is unique to the genus the genes of the PFNA cluster exhibit unusually high level of interspecific divergence 354

[20], suggesting that the functions of this operon are species-specific. 355

Today, the functional role of most of the genes that make up the PFNA operon and 356 their products remains unknown. Nevertheless, at least two genes of this operon (pkb2 357 and tgm) are potential components of a signal system that accounts for a bidirectional 358 communication between the host cells and different species of bifidobacteria [19] . The presence of motifs in FN3-containing proteins similar in structure to cytokine 392 binding sites of mammalian receptors [23, 35], encouraged us to opt for FN3-containing 393 proteins as suitable candidates for this role. As mentioned before, one of these proteins 394 is encoded by one of the genes of the cluster PFNA of B. longum. Proceeding from these 395 premises, we aimed at detecting specific interaction between the FN3 protein and 396 human cytokines. 397

We set out to assess the ability of a fragment of the FN3 protein containing two FN 398 type III domains to bind in vitro to various human cytokines. However, the protein's 399 strong adhesive properties rendered the task of revealing its specific interaction with 400 cytokines immobilized on a solid surface, impossible. 401

The fact that it was impossible to detect any specific interaction between the FN3 402 protein and cytokines attached to a solid surface suggests one of two things: 403

1) The FN3 protein undergoes conformational changes when adsorbed onto plastic, 404 which blocks its active site responsible for adhesion and cytokine-binding ability; 405

2) The cytokine-binding motif of the protein is directed inward or is located very 406 close to the adhesion site, which leads to spatial blockage during the adhesion of the 407 FN3 protein to a solid surface. To find out which of these two hypotheses is true, it is 408 necessary to evaluate the crystal structure of the FN3 protein. We managed to overcome the non-specific interaction between the FN3 protein and 410 solid surfaces by constructing an ELISA sandwich system, in which the FN3 protein, 411 attached to rabbit polyclonal IgG, acted as a cytokine trap. This construction allowed us course of evolution, bacteria of the genus Bifidobacterium used proteins containing FN3 418 domains as an instrument for the identification of cytokines of the host organism. 419

It is well known that the components of the immune system can affect the growth of 420 bacteria [10-13], i.e. microorganisms not only affect the host organism, but are also able, 421

in turn, to perceive signals from it and accordingly respond to them. Until recently, 422 such data for bifidobacteria were absent from the literature. 423

It is self-evident that for bifidobacteria to be affected by the host organism, they 424 need to be able to receive a feedback signal from the human body. Table S1 . Optical density of serum samples of rabbits immunized with the FN3 464 protein fragment bound to a specific antigen. Table S2 Optical density of affinity-465 purified rabbit antibodies specific to the FN3 protein fragment bound to specific and 466 non-specific antigens. Table S3 Optical density of affinity-purified rabbit antibodies 467 specific to the FN3 protein fragment and total IgG of non immunized rabbits against the 468 protein FN3 of B. longum GT15. 

Bifidobacteria and Their Role as Members of the

Bifidobacteria and the infant gut: an example of co 483 evolution and natural selection

Mechanisms of Action of

Antioxidant 488 Properties of Probiotic Bacteria

Tumor necrosis 519 factor alpha binding to bacteria: evidence for a high-affinity receptor and alteration 520 of bacterial virulence properties

Recognition of host immune activation by Pseudomonas 524 aeruginosa

Specific high affinity binding of 527 human interleukin 1 beta by Caf1A usher protein of Yersinia pestis

Molecular and Structural 531 Characterization of a Novel Escherichia coli Interleukin Receptor Mimic Protein

Functional and structural characteristics of bacterial proteins 534 that bind host cytokines

Positive Selection in 20

longum: Genetic environment and 542 substrate specificity

Signature proteins that are distinctive 583 characteristics of Actinobacteria and their subgroups

Functional Analysis 586 of an S-Layer-Associated Fibronectin-Binding Protein in Lactobacillus acidophilus

Ihalin R. A novel intrinsically disordered outer 590 membrane lipoprotein of Aggregatibacter actinomycetemcomitans binds various 591 cytokines and plays a role in biofilm response to interleukin-1β and interleukin-8

Influence of probiotics on cytokine production in the in vitro and in vivo systems

Structure, function, and antigenicity of the SARS-CoV-2 spike glycoprotein

Pathogenic human coronavirus infections: causes 607 and consequences of cytokine storm and immunopathology

The use of anti-inflammatory drugs in the treatment 611 of people with severe coronavirus disease 2019 (COVID-19): The Perspectives of 612 clinical immunologists from China

Specific interaction of FN3 protein with cytokines. CC -«conjugate» 632 control (no cytokines; secondary andibodies added after the FN3 protein 633 fragment). The values yielded by secondary antibodies against the cytokines IL-634 6, IL-10, IL-1 and TNFα were relatively similar and grouped together