Human gut symbionts secrete a ubiquitin homologue that inactivates an essential periplasmic chaperone for interbacterial competition

    Bacteroidetes are the most abundant Gram-negative bacterial phylum in the human gastrointestinal tract. They play a critical role in maintaining intestinal homeostasis, but are also implicated in diseases such as inflammatory bowel disease (IBD) and colorectal cancer (CRC). To establish and maintain themselves in the limited intestinal tract, Bacteroides species have evolved several specific proteins that act as antibacterial weapons against other bacteria, such as the Bacteroidales secreted antimicrobial proteins, BSAP ¹.

Ubiquitin is a highly conserved protein modifier found in all eukaryotes. It is most often covalently attached to target proteins by the coordinated activity of enzymes referred to as E1s, E2s, and E3s, resulting in an isopeptide bond formed between the C-terminal glycine residue of ubiquitin and a lysine residue of targets ². Signaling by ubiquitination regulates virtually every cellular process in eukaryotes. Although some ubiquitin-like proteins (UBLs) have been identified as small covalent modifiers in prokaryotic cells with diverse functions, such as sulfur-transfer proteins ThiS and MoaD of E. coli ^3,4 and small archaeal modifier proteins (SAMPs) in Haloferax volcanii ⁵, covalently attach to the protein targets via its C-terminal glycine. However, none of them share sequence similarities to human ubiquitin (HmUbb) except for two C-terminal glycine residues.

In 2011, Sheila Patrick and Garry W. Blakely et al. found that Bacteroides fragilis encodes a ubiquitin homologue (termed BfUbb) ⁶, shares 63% sequence identity and more than 80% sequence similarity to HmUbb (Fig. 1). It is estimated that there are 10⁹ to 10¹⁴ cell/g of Bacteroides species in cultured human feces, of which B. fragilis ranges from 4 to 14%. Around 12% of B. fragilis strains in the human gut metagenomes encode a ubiquitin homologue, suggesting an important function for BfUbb.

Fig.1 Sequence alignment of BfUbb and human ubiquitin.

Initially, Sheila Patrick et al. showed that BfUbb can interact with the ubiquitin-activating enzyme (E1) in vitro and speculate that BfUbb can act as a suicide substrate, leading to the inactivation of E1 and blockade of the ubiquitination cascade ⁶. However, in vivo experiments to prove this hypothesis are still lacking. The following study suggests that there is a potential for antibodies against HmUbb in human serum due to the antigenic mimicry by BfUbb, which may be a trigger for autoimmune disease ⁷. Nevertheless, it remains unknown whether the immune response to BfUbb is a consequence or a driver of disease development. Therefore, the question of whether it interferes with the host's ubiquitination process in vivo, and relates to host health remains unclear.

Recently, Laurie E. Comstock group found that BfUbb exhibited potent toxicity against a subset of B. fragilis strains in vitro ⁸. However, homologues of E1, E2, or E3 enzymes are undetectable in B. fragilis, indicating a unique mechanism of action for BfUbb, which was of great interest to us. Unlike HmUbb, BfUbb has a secretion signal peptide and is secreted outside the bacteria. In addition, the C-terminus of BfUbb is quite different from that of HmUbb. The di-glycine motif in the C-terminus of HmUbb has been replaced by a cysteine residue in BfUbb (Fig. 1). Our structural studies further showed that although BfUbb shares a similar fold to HmUbb, it exhibits an essential C-terminal disulfide bond formed by the uniquely positioned cysteine residues (Cys70 and Cys76), which confers BfUbb its distinct mode of action.

Initially, we tried several methods to elucidate the mechanism of action of BfUbb, including transposon screening, long-term evolution, and transcriptome sequencing, but were unable to identify the substrate of BfUbb. To my delight, through affinity purification coupled to LC-MS/MS, Kun found that a peptidyl-prolyl cis-trans isomerase (PPIase) emerged as a prominent BfUbb-interacting protein in the BfUbb-treated group. Through bioinformatics analysis and the toxicity changes of BfUbb to genetically engineered B. fragilis strains, we unexpectedly found that the sensitivity of B. fragilis strains to BfUbb could be identified only by analyzing the difference of glutamic acid (insensitive) and tyrosine (sensitive) at position 119 on PPIase. Crystal structure analysis also well verified that the unique disulfide bond at the C-terminus of BfUbb facilitates its binding to PPIase by orienting of Asp75 in BfUbb to interact with Tyr119 in PPIase from the BfUbb-sensitive strains. However, PPIases from BfUbb-insensitive strains bear a glutamic acid in place of the tyrosine, which likely exhibits significant electrostatic repulsion to Asp75 in BfUbb, and represents a substantial barrier to the formation of the BfUbb-PPIase complex, rendering the strains resistant to BfUbb. Bacterial co-colonization investigations in mice combined with human gut metagenomic analyses further suggest that BfUbb facilitates its encoding strains to gain significant fitness advantages in the complex human gut ecosystem through competitive intraspecies interactions
(Fig.2).

Fig. 2 Schematic of mechanism of BfUbb-encoding strains antagonizing other B. fragilis strains via BfUbb.

In this work, after identifying the substrate PPIase of BfUbb, our team elucidated the molecular mechanism of the interaction between BfUbb and PPIase and found that BfUbb is a potential ecological force in regulating the bacterial composition of the human gut in less than a year. We have overcome numerous difficulties, including shortages of laboratory supplies and staff due to the Covid-19 pandemic, difficulties in protein structure determination, and difficulties in genetic manipulation of strains, etc. These are attributed to the tremendous endogenous motivation generated by our great interest in BfUbb and the close cooperation among the team members.

In addition, many questions about BfUbb remain unanswered, including the exact secret pathway of BfUbb, how BfUbb enters the periplasm of recipient bacteria, why not all B. fragilis strains possess the insensitive PPIase gene, and the origin and evolution of BfUbb, etc. These are questions that we will continue to investigate.

References

1          García-Bayona, L. & Comstock, L. E. Bacterial antagonism in host-associated microbial communities. Science 361, doi:10.1126/science.aat2456 (2018).

2          Cappadocia, L. & Lima, C. D. Ubiquitin-like protein conjugation: structures, chemistry, and mechanism. Chemical Reviews 118, 889-918, doi:10.1021/acs.chemrev.6b00737 (2018).

3          Schmitz, J., Wuebbens, M. M., Rajagopalan, K. V. & Leimkuehler, S. Role of the C-terminal Gly-Gly motif of Escherichia coli MoaD, a molybdenum cofactor biosynthesis protein with a ubiquitin fold. Biochemistry 46, 909-916, doi:10.1021/bi062011w (2007).

4          Xu, X., Wang, T., Niu, Y., Liang, K. & Yang, Y. The ubiquitin-like modification by ThiS and ThiF in Escherichia coli. International Journal of Biological Macromolecules 141, 351-357, doi:10.1016/j.ijbiomac.2019.08.172 (2019).

5          Humbard, M. A. et al. Ubiquitin-like small archaeal modifier proteins (SAMPs) in Haloferax volcanii. Nature 463, 54-U60, doi:10.1038/nature08659 (2010).

6          Patrick, S. et al. A unique homologue of the eukaryotic protein-modifier ubiquitin present in the bacterium Bacteroides fragilis, a predominant resident of the human gastrointestinal tract. Microbiology-Sgm 157, 3071-3078, doi:10.1099/mic.0.049940-0 (2011).

7          Stewart, L., Edgar, J. D. M., Blakely, G. & Patrick, S. Antigenic mimicry of ubiquitin by the gut bacterium Bacteroides fragilis: a potential link with autoimmune disease. Clinical and Experimental Immunology 194, 153-165, doi:10.1111/cei.13195 (2018).

8          Chatzidaki-Livanis, M. et al. Gut symbiont Bacteroides fragilis secretes a eukaryotic-like ubiquitin protein that mediates intraspecies antagonism. mBio 8, doi:10.1128/mBio.01902-17 (2017).