Antibiotics Against Gram-Negative Bacteria

Researchers all over the world work hard in the discovery of antibiotics, one such researcher is Dr. Kim Lewis, leader of the Lewis Lab at Northeastern University where he is a distinguished professor with expertise in molecular microbiology. Dr. Kim Lewis’ lab consists of the Lewis Lab which studies persister cells responsible for tolerance to antibiotics, uncultured bacteria of the environment and the microbiome. It also includes the antimicrobial discovery center which translates basic discoveries into novel antimicrobial therapies to combat conventional pathogen threats and biowarfare. Working under Dr. Kim Lewis are Dr. Yu Imai and Dr. Kirsten Meyer, who together studied the effects of many potential compounds and their antimicrobial effectiveness in hopes of discovering a viable antibiotic. Dr. Yu Imai started as postdoctoral research associate and became an associate research scientist and now works as an assistant professor at Shinshu University. He received his bachelor’s degree in the faculty of agriculture, then a master’s degree from the graduate school of agriculture to then receive a doctor of philosophy, PhD, at the Interdisciplinary Graduate School of Science and Technology, all at Shinshu university. Dr. Kirsten Meyer was a postdoctoral research associate in Lewis lab at Northeastern University and is now a postdoctoral fellow at the University of Toronto. She got a bachelor’s degree from Victoria University of Wellington in honours biomedical science, majoring in molecular pharmacology and medicinal chemistry to then go on and achieve a PhD in pharmacology from Johns Hopkins University School of Medicine. Her work now at the University of Toronto involves optimized antimicrobial combinations and their packaging in extracellular vesicles from soil bacteria, towards the development of new antimicrobials to counter resistance. Together at Northeastern University, along with the collaborative efforts of other researchers, they worked to find new antibiotics that selectively kill Gram-negative pathogens as an effort against the emergence of antibacterial resistance.

Figure 1: Stages in the formation of antibiotic resistant bacteria (BioMérieux, 2016).

Infectious diseases are a significant cause of death worldwide, accounting for approximately one third of all deaths among the top ten leading causes of death (Center for Disease Control (CDC), 1999). Currently, the main treatment for infectious diseases is prescribing antibiotics; however, the rate of resistance to antibiotics by bacteria is increasing whilst the production of new antibiotics is slowing down (Lewis, 2013). Bacteria can mutate and evolve which, over time, allows them to develop resistance to currently prescribed antibiotics. A random mutation can occur in bacteria which protects the bacterial cell from the effects of the antibiotic. As a result, bacteria without the mutation die when the antibiotic is administered. The antibiotic resistant bacteria survive and reproduce with less competition from non-resistant bacteria. Antibiotic resistant genes are passed onto subsequent generations and over time the whole bacterial population becomes resistant to the antibiotic (Mak et al., 2014).

There are several factors that contribute to antibiotic resistance in both developing nations as well as developed nations. Developing nations may not effectively track resistance development in bacterial strains which contributes to the antibiotic resistance problem as novel resistant strains of bacteria are not documented (Chokshi et al., 2019). Additionally, these countries may have lower quality antibiotics because they lack quality assurance mechanisms. Developing nations also have a greater ease of availability of antibiotics due to limited regulation on retail of pharmaceuticals. The ease of availability combined with clinical misdiagnoses and antibiotic misuse contributes to antibiotic resistance as they are used when not needed.

On the other hand, developed nations also contribute to the antibiotic resistance issue. In these countries, antibiotics are used prophylactically in medical settings to reduce the risk of infections during surgeries and procedures (Chokshi et al., 2019). There is also a wide use of antibiotics in agriculture for livestock to treat diseases, prevent infections, and increase animal growth. Due to the interactions between humans and animals, antibiotic resistance is gradually conferred to human pathogens.

Despite the range of factors contributing to antibiotic resistance and its rise as a global threat, there is a lack of research on new antibiotics.

Gram-negative bacteria contain a cell envelope that consists of three layers, the outer membrane, peptidoglycan and the inner membrane. The outer membrane, unique to gram-negative bacteria, acts as a protective layer. It contains several membrane proteins such as porins, that allow small molecules such as amino acids to pass through the membrane (Breijyeh et al, 2020). The periplasmic space, the space between the two membranes, contains a thin layer of peptidoglycan. The inner membrane is a phospholipid bilayer crucial for functions such as structure, transport and biosynthesis. The outer membrane is the primary cause of antibacterial resistance in gram-negative bacteria. Antibiotics such as β-lactams, quinolones, and colistins must pass through the outer membrane of bacteria in order to access their targets. While some hydrophobic drugs can pass through diffusion, most require porins and membrane proteins(Breijyeh et al, 2020). Mutations to the outer membrane, either to the hydrophobic properties or the membrane proteins, create resistance. Compared to gram-positive bacteria that lack the outer membrane, gram-negative bacteria or more susceptible to antibiotic resistance. The increase in multidrug resistance in E.coli is becoming a growing concern for researchers. Through different mechanisms, such as lateral gene transfer, E.coli has been able to confer resistance to broad-spectrum cephalosporins, carbapenems, aminoglycosides, quinolones and polymyxins, increasing the need for novel antibiotics (Breijyeh et al, 2020).

A new potential antibiotic is darobactin which targets an outer membrane protein (OMP) of gram negative bacteria (Guo et al., 2022). The target of darobactin is an OMP called the β-barrel assembly machinery (BAM), which is a pentameric complex with BamA being the essential core and BamB, BamC, BamD, and BamE being the accessory lipoproteins to BamA (Guo et al., 2022; Noinaj et al., 2017; Pinto et al., 2018). This OMP is essential for bacterial physiology, pathogeny, and drug resistance (Pinto et al., 2018). The four accessory lipoproteins each carry a N-terminal post-translational lipid modification that signals them to attach to BamA (Noinaj et al., 2017). However, the component of interest is BamA which has been studied to be the target of darobactin (Guo et al., 2022). As the core of the BAM complex, BamA is found embedded within the outer membrane due to its amphiphilic nature, being hydrophobic on the outside to interact with the membrane and hydrophilic on the inside to interact with solvents (Noinaj et al., 2017). It has a N-terminal periplasmic domain that carries five polypeptide transport associated (POTRA) domains and a C-terminal domain that consists of a 16 stranded, anti-parallel, β-barrel domain (Noinaj et al., 2017). POTRA domains do not have a clear role in the OMP’s functions but it is suggested that they act as docking sites for BamB, BamC, BamD, and BamE onto BamA (Kim et al., 2012). In particular, β1 and β16 interact in a way that form a bend or kink towards the lumen of the domain which gives rise to a lateral gate with the ability to provide direct insertion of nascent OMPs into the outer membrane during biogenesis (Noinaj et al., 2017). The lateral gate has two proposed functions being that it is involved in β-augmentation and that it releases the substrate into the bilayer due to the kink producing a slight destabilization (Pinto et al., 2018). β-augmentation occurs when β1 and β16 strands and the substrate protein interact to fold the substrate of interest (Pinto et al., 2018).

Figure 2: Two views of the BAM complex as modeled using cryo-EM (Konovalova et al., 2017).

Other than POTRA5 which is known to be an essential domain, the other POTRA domains appear to have varying conclusions drawn in the literature (Browning et al., 2013). Some have been shown to be both essential and non-essential. Amongst the domains, there exists conservation with their folding structures but not with their sequences or function (Konovalova et al., 2017). This structure consists of three strand β-barrel covered by an antiparallel helix pair (Konovalova et al., 2017). In terms of their function, the POTRA domains allow for interactions between BamA and both other units in BAM and other proteins. POTRA1 allows for interactions with the chaperone SurA, POTRA3 allows for interactions with BamB, and POTRA5 allows for interactions with the BamCDE subcomplex through the formation of an interaction interface (Konovalova et al., 2017). Alongside the other parts of BAM, the POTRA domains form a ring-shaped structure oriented parallel to the plane of the membrane (Konovalova et al., 2017). Due to the hydrogen bonds that form between the phospholipid heads of the membrane and the POTRA domains, there is a high affinity between the two where tryptophan residues, conserved in the POTRA domains, allow the domains to partition into the membrane (Konovalova et al., 2017).

The goal of BAM is to catalyze the proper folding and insertion of new β-barrel proteins such as BAM itself through a process called bacterial OMP biogenesis (Imai et al., 2019; Kim et al., 2012). Nascent OMPs are produced within the cytosol with a N-terminal signal sequence that directs it to the inner membrane to be transported to the periplasm through the Sec translocation system (Kim et al., 2012). During this time, the signal sequence is cleaved by signal peptidase I (SPase I) in order for the OMP to travel to the periplasm (Kim et al., 2012). Then, it will travel to the outer membrane through the SurA pathway or Skp/DegP pathway and finally integrate itself into the outer membrane as a β-barrel protein (Kim et al., 2012).

Coming back to the β-barrel at the C-terminus of BamA, there are various loops (L) and turns (T) connecting the 16 β-strands (Browning et al., 2013). All loops are extracellular while all turns are periplasmic. The loops are especially important in the function of BamA where insertions in some loops may cause large issues in protein function (Browning et al., 2013). Specifically, loops L2, L5, and L8 appear to be conserved quite well amongst orthologues while contrarily L3, L4, and L7 are not (Browning et al., 2013). L1 is conserved well enough that mutations are quite rare to observe (Browning et al., 2013). This implies the possibility of functional/structural importance being linked to the amino acid sequences in L2, L5, and L8 (Browning et al., 2013). Experimentation has been that shows that L4, L6, L7, and L8 cause major issues upon deletion, but deletion of L3 appears to have no major problems in the protein (Browning et al., 2013). Due to this, it was inferred that L3 instead has steric effects on the protein (Browning et al., 2013).

Figure 3: A models BamA using a collection of information from various studies; B is a topological model of the β-barrel constructed through predictions using bioinformatics (Browning et al., 2013).

Based on a homology with FhaC, another outer membrane protein, L6 shows importance being an essential structure in the functioning of FhaC (Browning et al., 2013). There exists a conserved RGF motif on L6, being the notable portion that allows for L6 to be essential (Browning et al., 2013). It is notable that the RGF motif cannot be deleted while retaining normal BamA function (Browning et al., 2013). Unlike the other loops which exist outside of the β-barrel, L6 instead is found inside the β-barrel (Browning et al., 2013). It has been hypothesized that the strands of the β-barrel and L6 undergo some conformational change during the protein folding process where the other loops aid via stabilization through the stages of movement (Browning et al., 2013).

Darobactin:

Imani et al. (2019) investigated the secretions of Gram-negative bacteria, Photorhabdus and Xenorhabdu, that live symbiotically in the microbiome of nematode worms. They selected these bacteria due to their ability to produce antimicrobials that are non-toxic to the nematode and have good pharmacokinetics when protecting against invading Gram-negative pathogens. Through screening and plating strains of Photorhabdus and Xenorhabdu, the researchers found that a concentrated extract of Photorhabdus khanii HGB1456 inhibited the growth of Escherichia coli (Imani et al., 2019). Imani et al. (2019) isolated the extract and identified the active compound as darobactin, a potential antibiotic for Gram-negative bacteria. Further experimentation with different species and strains of Photorhabdus indicate that darobactin production is very low, offering an explanation as to why it has not been previously discovered as an antibiotic.

It was determined that darobactin directly binds to the BamA region that is accessible from the extracellular space (Imani et al., 2019). This binding stabilizes BamA into a potentially inactive conformation which prevents outer membrane proteins from exiting into the outer membrane. The ability of darobactin to target BamA and inhibit the BAM complex is promising in the development of novel antibiotics.

Murepavadin Analogue:

Luther et al. (2019) screened cyclic peptides related to Murepavadin - a macrocyclic beta-hairpin peptidomimetic antibiotic that targets outer membrane proteins (OMPs) of Gram-negative bacteria - against a panel of Gram-negative pathogens. Using information gathered from the screen, Luther et al. (2019), synthesized chimeric molecules with an increased ability of the beta-hairpin to target OMPs. This led to the discovery of a family of chimeric peptidomimetic antibiotics that are promising therapeutics against Gram-negative bacteria. Specifically, four compounds (chimera peptides 3, 4, 7 and 8) demonstrated low toxicity towards mammalian cells, high potency in human serum, low chance of resistance, good pharmacokinetics as well as efficacy in mouse models (Luther et al., 2019).

The chimeras act similarly to darobactin by targeting BamA and locking it into a closed state. However, they are also able to permeabilize both the inner and outer membrane which is indicative of the compound acting directly on the membrane (Luther et al., 2019).

MRL-494:

Hart et al. (2019) discovered MRL-494, a compound that targets BamA to inhibit the insertion of OMPs into the outer membrane. MRL-494 was identified through a screen that looked for compounds able to disrupt the outer membrane barrier and whose antimicrobial activity was not decreased by efflux pumps (used by bacteria to gain resistance and protect against antibiotics). Furthermore, they observed that MRL-494 was able to interrupt a critical outer membrane pathway and allow antibiotics to move across the barrier (Hart et al., 2019).

The researchers were able to conclude that the MRL-494 compound works as an antibiotic by inhibiting the biogenesis of outer membrane proteins. In addition to this, they were also able to determine that MRL-494 is able to inhibit BamA and prevents the assembly of outer membrane proteins (Hart et al., 2019).

Darobactin:

Figure 4. Titration NMR Spectral Positioning of Darobactin to BamA-β. Image from Imai et al. 2019.

Investigation of similarities between the three novel antibiotics and their action mechanism may potentially provide insight into key regions of BamA. The first antibiotic darobactin was found to form darobactin-stabilized conformations of BamA with most of the residues closely resembling that of the closed-gate conformation. These findings were backed by the substantial NMR spectral positioning changes (Figure 4) during a titration experiment introducing darobactin to BamA-β. These findings along with isothermal titration calorimetry measurements strongly suggest that darobactin and BamA physically interact with each other by stabilizing the proteins closed lateral gate in a potentially inactive conformation. This stabilized closed lateral gate conformation would prevent the exiting of substrates into the outer membrane inhibiting the ability of BAM complex to perform its OMP-folding function. These results are further backed by mutations that confer resistance to darobactin being located at the lateral gate of BamA (Imai et al. 2019).

Murepavadin Analogue:

These results coincide with the chimera antibiotic findings as it was also found through NMR experiments that the interaction of the chimera peptide 3 binding stabilizes BamA in a potentially inactive conformation also resembling that of the closed form of BamA. NMR spectroscopy showed that peptide 3 interacted with the BamA β-barrel domain through the external loops L4, L6, and L7 which changed the conformation of the β-barrel lateral gate between open or closed states and locked BamA into its closed state (Luther et al., 2019).

MRL-494:

However, MRL-494 antibiotic did not share such similarities with the rest displaying varying results. MRL-494 underwent cellular thermal shift assay experiments indicating direct or proximal target engagement able to stabilized BamA against thermally induced protein aggregation. Although the exact binding region of MRL-494 to BamA have not been determined, BamAE470 residue has been highlighted as key region with mutations suppressing the effects of MRL-494 on BamA despite being in its presence. MRL-494 also displays similar potency against gram-positive bacteria lacking BamA indicating a separate mechanism apart from darobactin and chimera peptide 3 (Hart et al., 2019).

Figure 5. The predicted chemical structure of the antibiotic darobactin. Image from Imai et al. 2019.

Nematodes have similar requirements for antibacterials as humans (Tobias et al. 2018). The digestive system of the worms is lined with bacteria similar to E. coli; these bacteria are released by a nematode as it invades an insect larva to feed (Crawford and Clardy 2011). The antimicrobial substances prevent other microorganisms from feeding on the larvae (Tobias et al. 2018). Some of these environmental microorganisms are Gram-negative bacteria (Tambong, 2013). The antibiotics must harm the bacteria without harming the nematode as they are in close proximity (Imai et al. 2019). The fact that the nematode microbiome can kill these bacteria so effectively means there could be a candidate human antibiotic substance produced by one of the commensal bacteria (Imai et al. 2019).

28 species of bacteria from the nematode microbiome were screened for production of antimicrobial against E. coli (Imai et al. 2019). Very few candidates produced zones of inhibition, the antibacterials are not produced in high enough quantities in vitro (Imai et al. 2019). Concentrated extracts of the bacteria were prepared and an extract from ​​Photorhabdus khanii HGB1456 effectively produced a zone of inhibition and served as a potential antibiotic candidate (Imai et al. 2019). The compound was isolated and identified to have a molar mass of 966.41047, suggesting that the molecular formula would be C47H56O12N11+ (Imai et al. 2019). This compound was checked against a natural products database to ensure that it was novel (Imai et al. 2019). Once it was confirmed to be a new substance, the research group named it darobactin (Imai et al 2019).

Figure 6. The predicted chemical structure of the antibiotic darobactin. Image from Imai et al. 2019.

Darobactin can effectively inhibit the growth of many Gram-negative bacteria and it has little effect against Gram-positive bacteria (Imai et al. 2019). There is also little activity against the Bacteroides that compose much of our commensal gut microbiome (Imai et al. 2019), Zafar et al. 2021). This led the research group to believe that the compound might inhibit lipopolysaccharide, the main component of the outer membrane of bacteria (Bertani et al. 2019). Tests of darobactin against purified LPS found that there was no inhibition of activity, meaning darobactin does not target LPS directly (Imai et al. 2019). It must target a different component of the outer membrane. A potential candidate would be BamA which is a large, crucial outer membrane protein (Kaur et al. 2021). To identify the target, a ligand-protection thermal proteome analysis was performed, this analysis showed that chaperone proteins were significantly upregulated, whereas outer membrane proteins were decreased (Imai et al. 2019). Mutant E. coli cultured in high quantities of darobactin developed resistance to the drug (Imai et al. 2019). The sequences of the mutant bacteria were observed to determine which mutated genes would confer darobactin resistance, each mutant bacteria had multiple mutations in the bamA which codes for the BamA protein (Imai et al. 2019). This proved that BamA was the likely target of darobactin (Imai et al. 2019).

Further studies were performed to determine how resistance affected the virulence and infectivity of the mutated bacteria (Imai et al. 2019). Mice injected with wild-type E. coli had a mortality rate of 60% within 24 hours of infection, whereas the bacteria injected with the resistant mutants had no deaths within 24h (Imai et al. 2019) The mutations that confer darobactin resistance significantly reduced the virulence of the bacteria (Imai et al. 2019).

Darobactin is an exciting drug candidate that targets the outer membrane protein BamA (Imai et al 2019). If approved for humans, it would be the first drug developed from the microbiome of another animal species (Candanosa, 2022).

Bertani, B., & Ruiz, N. (2018). Function and biogenesis of lipopolysaccharides. EcoSal Plus, 8(1). https://doi.org/10.1128/ecosalplus.esp-0001-2018

BioMérieux. (2016). What is antibiotic resistance ? | BioMérieux, Antimicrobial Resistance. https://www.yourgenome.org/facts/what-is-antibiotic-resistance/

Breijyeh, Z., Jubeh, B., & Karaman, R. (2020). Resistance of gram-negative bacteria to current antibacterial agents and approaches to resolve it. Molecules, 25(6), 1340. https://doi.org/10.3390/molecules25061340

Browning, D. F., Matthews, S. A., Rossiter, A. E., Sevastsyanovich, Y. R., Jeeves, M., Mason, J. L., Wells, T. J., Wardius, C. A., Knowles, T. J., Cunningham, A. F., Bavro, V. N., Overduin, M., & Henderson, I. R. (2013). Mutational and topological analysis of the Escherichia coli BAMA protein. PLoS ONE, 8(12). https://doi.org/10.1371/journal.pone.0084512

Candanosa, R. M. (2022, March 29). A new antibiotic has been hiding in the gut of a tiny worm. it may be our best weapon against drug-resistant bacteria. Northeastern Global News. Retrieved February 1, 2023, from https://news.northeastern.edu/2019/11/20/can-darobactin-a-new-antibiotic-found-in-a-tiny-worm-become-our-best-weapon-against-drug-resistant-bacteria/

Centers for Disease Control and Prevention. (n.d.). Achievements in public health, 1900-1999: Control of Infectious Diseases. Centers for Disease Control and Prevention. https://www.cdc.gov/mmwr/preview/mmwrhtml/mm4829a1.htm#fig2

Crawford, J. M., & Clardy, J. (2011). Bacterial symbionts and natural products. Chemical Communications, 47(27), 7559. https://doi.org/10.1039/c1cc11574j

Hart, E. M., Mitchell, A. M., Konovalova, A., Grabowicz, M., Sheng, J., Han, X., Rodriguez-Rivera, F. P., Schwaid, A. G., Malinverni, J. C., Balibar, C. J., Bodea, S., Si, Q., Wang, H., Homsher, M. F., Painter, R. E., Ogawa, A. K., Sutterlin, H., Roemer, T., Black, T. A., … Silhavy, T. J. (2019). A small-molecule inhibitor of bama impervious to efflux and the outer membrane permeability barrier. Proceedings of the National Academy of Sciences, 116(43), 21748–21757. https://doi.org/10.1073/pnas.1912345116

Imai, Y., Meyer, K. J., Iinishi, A., Favre-Godal, Q., Green, R., Manuse, S., Caboni, M., Mori, M., Niles, S., Ghiglieri, M., Honrao, C., Ma, X., Guo, J. J., Makriyannis, A., Linares-Otoya, L., Böhringer, N., Wuisan, Z. G., Kaur, H., Wu, R., … Lewis, K. (2019). A new antibiotic selectively kills gram-negative pathogens. Nature, 576(7787), 459–464. https://doi.org/10.1038/s41586-019-1791-1

Kaur, H., Jakob, R. P., Marzinek, J. K., Green, R., Imai, Y., Bolla, J. R., Agustoni, E., Robinson, C. V., Bond, P. J., Lewis, K., Maier, T., & Hiller, S. (2021). The antibiotic darobactin mimics a β-strand to inhibit outer membrane insertase. Nature, 593(7857), 125–129. https://doi.org/10.1038/s41586-021-03455-w

Kim, K. H., Aulakh, S., & Paetzel, M. (2012). The bacterial outer membrane β-barrel assembly machinery. Protein Science, 21(6), 751–768. https://doi.org/10.1002/pro.2069

Kirsten Meyer. (n.d.). Home [LinkedIn page]. LinkedIn. Retrieved February 1, 2023, from https://ca.linkedin.com/in/kirstenjoymeyer

Konovalova, A., Kahne, D. E., & Silhavy, T. J. (2017). Outer Membrane Biogenesis. Annual Review of Microbiology, 71(1), 539–556. https://doi.org/10.1146/annurev-micro-090816-093754

Lewis, K. (2013). Platforms for antibiotic discovery. Nature Reviews Drug Discovery, 12(5), 371–387. https://doi.org/10.1038/nrd3975

Luther, A., Urfer, M., Zahn, M., Müller, M., Wang, S.-Y., Mondal, M., Vitale, A., Hartmann, J.-B., Sharpe, T., Monte, F. L., Kocherla, H., Cline, E., Pessi, G., Rath, P., Modaresi, S. M., Chiquet, P., Stiegeler, S., Verbree, C., Remus, T., … Obrecht, D. (2019). Chimeric peptidomimetic antibiotics against gram-negative bacteria. Nature, 576(7787), 452–458. https://doi.org/10.1038/s41586-019-1665-6

Mak, S., Xu, Y., & Nodwell, J. R. (2014). The expression of antibiotic resistance genes in antibiotic-producing bacteria. Molecular Microbiology, 93(3), 391–402. https://doi.org/10.1111/mmi.12689

Ni, D., Wang, Y., Yang, X., Zhou, H., Hou, X., Cao, B., Lu, Z., Zhao, X., Yang, K., & Huang, Y. (2014). Structural and functional analysis of the β‐barrel domain of Bama from Escherichia coli. The FASEB Journal, 28(6), 2677–2685. https://doi.org/10.1096/fj.13-248450

Noinaj, N., Gumbart, J. C., & Buchanan, S. K. (2017). The β-barrel assembly machinery in motion. Nature Reviews Microbiology, 15(4), 197–204. https://doi.org/10.1038/nrmicro.2016.191

Northeastern University College of Science. (2023, February 3). Kim Lewis. https://cos.northeastern.edu/people/kim-lewis/

Tambong, J. T. (2012). Phylogeny of bacteria isolated from rhabditis sp. (Nematoda) and identification of novel entomopathogenic serratia marcescens strains. Current Microbiology, 66(2), 138–144. https://doi.org/10.1007/s00284-012-0250-0

Tobias, N. J., Shi, Y.-M., & Bode, H. B. (2018). Refining the natural product repertoire in entomopathogenic bacteria. Trends in Microbiology, 26(10), 833–840. https://doi.org/10.1016/j.tim.2018.04.007

Yu Imai. (n.d.). Home [LinkedIn page]. LinkedIn. Retrieved February 1, 2023, from https://www.linkedin.com/in/yu-imai-9a5774123

Zafar, H., & Saier, M. H. (2021). Gut Bacteroides species in health and disease. Gut Microbes, 13(1). https://doi.org/10.1080/19490976.2020.1848158