Page 40 - Read Online

Alekseeva et al. Microbiome Res Rep 2023;2:10 https://dx.doi.org/10.20517/mrr.2023.06 Page 7 of 15

TNFα
The FN3 protein encoded by the fn3 gene of the PFNA operon of B. longum subsp. longum GT15 consists of
1,994 amino acid residues and harbors two FN3 domains, which stretch from amino acid 1,494 to amino
acid 1,581 and from amino acid 1,586 to amino acid 1,671, respectively. The region following the second
FN3 domain (from amino acid 1,672 to amino acid 1,994) is the C-terminal region of the protein.
Previously, we cloned in E. coli a fragment of the fn3 gene encoding the part of the protein stretching from
amino acid 1,494 to amino acid 1,994 and encompassing both the FN3 domains and the C-terminal region.
After obtaining the recombinant protein (designated by us as ΔFN3.1), we were able to demonstrate that
this fragment of the FN3 protein can selectively bind to the tumor necrosis factor TNFα .
[24]

In this study, we tested the ability of the fragments that make up the ΔFN3.1 protein: 2D FN3 (from amino
acids 1,494 to 1,671; containing only two FN3 domains) and CD FN3 (from amino acids 1,672 to 1,994;
containing only the C-terminal region of the protein) to binding to TNFα. The molecular weights of the
recombinant proteins corresponded to the calculated molecular weights (including the pET16b plasmid
linker containing His-Tag) of the 2D FN3 protein fragments stretching from amino acid 1494 to amino acid
1,671 (21 kDa) and CD FN3 stretching from amino acid 1,672 to amino acid 1,994 (36 kDa).

At the next stage of the work, the possibility of purification of recombinant 2D FN3 and CD FN3 proteins
[24]
was tested under conditions that were previously optimized for the ∆FN3.1 protein . The recombinant 2D
FN3 and CD FN3 proteins [Supplementary Figure 1] were isolated and purified in sufficient quantities to
study their binding to TNFα.

At the first stage of the study, cross-linking between polyclonal rabbit IgG antibodies and the FN3 protein
harboring 2D FN3 and CD FN3 fragments was established. The process of obtaining antibodies was
[24]
described in our previous paper . We established that polyclonal rabbit IgG antibodies were able to bind to
the mentioned fragments at high dilutions - a positive reaction was detected for both 2D FN3 and CD FN3
at a concentration of 0.008 μg/mL. For the sake of comparison, the reaction between ΔFN3.1 and antibodies
was detected previously at a concentration of 0.004 μg/mL. The addition of nonspecific γ-globulins instead
of polyclonal antibodies did not lead to a cross-reaction with FN3, nor with its fragments.

In order to assess the binding specificity between 2D FN3 and CD FN3 to TNFα, we employed the ELISA
[24]
scheme described above and in our previous publication . First, FN3-specific polyclonal rabbit IgG
antibodies were sensitized on the solid phase. Second, 2D FN3 or CD FN3 fragments or the whole FN3
protein was added as a “second layer”. Next, a solution of TNFα conjugated to HRP from the CYTOKINE
commercial kit (Russia) was added. The devised ELISA scheme is a well-validated model for the detection of
interaction between FN3 and TNFα. However, the results showed that neither 2D FN3 nor CD FN3 bound
TNFα [Figure 3]. Our data indicate that for binding to TNFα, the complete fragment ΔFN3.1 containing
both 2D FN3 and CD FN3 fragment is necessary.

Prediction of 3D structures of FN3 protein fragments of B. longum subsp. longum GT15
Since the 3D structures of bifidobacteria proteins containing FN3 domains have not yet been determined,
we used the trRosetta servers to predict the spatial structures of the studied proteins. In this study, using
trRosetta software, we built structural models of fragments of FN3 proteins of the strain B. longum subsp.
longum GT15: 2D FN3 [Figure 4A], CD FN3 [Figure 4B], and ΔFN3.1 [Figure 4C] proteins, with a
prediction confidence of 0.710, 0.506, and 0.527, respectively.