Page 101 - MI-2-4

Microbes & Immunity Brachyspira pilosicoli novel outer membrane proteins

in the ECL region (Figure 2H), S105 in the β-barrel, and OMBB protein of B. hyodysenteriae, whereas the remaining
the remaining residues within the periplasmic domain three proteins matched to an uncharacterized protein of
(Table S4). Brachyspira spp. (Tables 2 and S3). Functional roles of these
five proteins remain uncertain, highlighting the need for
3.2.1.9. Eight-stranded β-barrel proteins annotated as SR experimental validation. Amino acid sequence comparison
domain-containing proteins
across nine B. pilosicoli strains revealed several variations
Thirteen B. pilosicoli proteins annotated in UniProt as (Tables 3 and S4). BP951000_RS02055 and BP951000_
“SR domain-containing proteins” were predicted in this RS02050 had variations in ECL regions (Figure 2J and 2K);
study as OMBB proteins. Five proteins—BP951000_ variations in the remaining proteins were located in TM or
RS04880, BP951000_RS02055, BP951000_RS02050, periplasmic regions (Figure 2I, L, and M, Tables 3 and S4).
BP951000_RS07540, and BP951000_RS00180—were
classified as group A (predicted as OMBB proteins by 3.2.2. Group B
all tools). Except for BP951000_RS02055 (annotated as Group B included 29 proteins with diverse β-barrel
an OMBB in NCBI), the rest are listed as hypothetical in architectures: a 26-stranded barrel (BP951000_RS09575);
NCBI. SRs, or G-protein-coupled receptors (GPCRs), are a 22-stranded barrel (BP951000_RS03215); an 18-stranded
eukaryote-specific heptahelical membrane proteins. 98,99 barrel (BP951000_RS04405); four 16-stranded barrels
However, AlphaFold 3 models of these B. pilosicoli (BP951000_RS09655, BP951000_RS04440, BP951000_
proteins displayed eight-stranded β-barrel architectures RS08285, and BP951000_RS04505); two 14-stranded
with four extracellular and three periplasmic loops barrels (BP951000_RS08455 and BP951000_RS01090);
(Figure 2I–M). These conformations were consistently one 13-stranded barrel (BP951000_RS08975); four
supported by ESMFold, SWISS-MODEL, RoseTTAFold, 12-stranded barrels (BP951000_RS06935, BP951000_
and TrRosetta tools. As GPCRs are absent in prokaryotes, RS11380, BP951000_RS03405, and BP951000_RS00185);
and given that all tools predicted TM β-barrel structures one 10-stranded barrel (BP951000_RS10320); and six
rather than α-helices, we conclude that these proteins are 8-stranded barrels (BP951000_RS05445, BP951000_
misannotated as SR proteins in UniProt. The reason for RS08300, BP951000_RS05490, BP951000_RS07500,
this misannotation is that UniProtKB/TrEMBL annotates BP951000_RS01590, and BP951000_RS08295). In
proteins automatically using computational pipelines. addition, eight 8-stranded β-barrel SR domain-containing
These annotations are not manually curated and rely on
sequence similarity and automated rule-based systems. 100 proteins (BP951000_RS06930, BP951000_RS03290,
BP951000_RS00765, BP951000_RS01280, BP951000_
DALI analysis showed that BP951000_RS02055, RS10445, BP951000_RS00365, BP951000_RS04620, and
BP951000_RS07540, and BP951000_RS00180 best BP951000_RS04220) were identified (Table 1).
matched the N-terminal β-barrel domain of E. coli
K-12 OM protein A (OmpA) (PDB ID: 9FZC), whereas 3.2.2.1. BP951000_RS09575
BP951000_RS02050 exhibited the best structural match BP951000_RS09575 is annotated as LptD in B.pilosicoli, a
20
with another E. coli K-12 OmpA structure (PDB ID: component of the LPS transport (LPT) system responsible
9FZD) (Tables 2 and S3). Unlike E. coli OmpA, which for transporting LPS from the inner OM leaflet to the Gram-
includes both β-barrel and periplasmic domains, these negative bacterial surface. It carries a predicted secretory
110
B. pilosicoli proteins lack the periplasmic domain. E. coli signal peptide. The AlphaFold 3 structural model revealed a
OmpA functions in phage recognition, colicin transport, 26-stranded β-barrel spanning the OM, with a lateral opening
conjugation, membrane integrity maintenance, solute between strands 1 and 26, and a distinctive periplasmic
diffusion, and virulence. It also contributes to the β-jelly roll domain (Figure 3A). DALI analysis showed the
virulence and pathogenicity of E. coli, making it a key best structural alignment with Yersinia pestis LptD (PDB
target in the immune response. 101-109 BP951000_RS04880 ID: 5IXM) (Tables 2 and S3). Functional annotation by
aligned best with NspA of N. meningitidis (PDB ID: 1P4T) PANNZER and eggNOG-mapper confirmed roles in cell
(Tables 2 and S3), a potential vaccine candidate involved in envelope biogenesis and LPT function, respectively, further
host adhesion and immune interaction. 86-89 This suggests validating its identity as LptD in B. pilosicoli (Tables 2 and S3).
that BP951000_RS04880 might have a role in adhesion Sequence variation analysis across nine B. pilosicoli strains
and may be explored experimentally as a potential vaccine revealed seven variations: N14, G137, I257, I382, E454, D600,
candidate. and G944 (Figure 3A, Tables 3 and S1). Structural mapping
PANNZER annotated all five proteins as OMBB domain- showed G944 in the ICL region, D600 in the ECL, I382 and
containing proteins. Foldseek identified BP951000_RS02055 E454 in the TM region, and the remaining variations in the
and BP951000_RS02050 to be structurally closest to an β-jelly roll domain (Table S4).

Volume 2 Issue 4 (2025) 93 doi: 10.36922/MI025230050