An Innate Pathogen Sensing Strategy Involving Ubiquitination Of Bacterial Surface Proteins Part 1

INTRODUCTION

Pathogenic invasion triggers a battery of immune responses stimulated by the surveillance mechanisms of the host. This is generally initiated by recognizing conserved microbial molecular structures known as pathogen-associated molecular patterns (PAMPs). Effective sensing of these PAMPs by pattern recognition receptors (PRRs) rapidly induces various host immune responses via the activation of complex signaling pathways triggering pathogen clearance. 

In recent years, with the development of biotechnology and the continuous improvement of research technology, people have a deeper understanding of the relationship between microbial molecules and immunity. Conserved microbial molecules are a class of microbial molecules that can be recognized by organisms and trigger immune responses. They mainly include molecular components of microorganisms such as bacteria, viruses, and fungi, and are highly conserved and specific. Studies have shown that conserved microbial molecules can resist various pathogens and enhance immunity by activating the body's immune system.

Conserved microbial molecules trigger inflammatory and immune responses by activating important molecules in the immune system, such as T cells, B cells, and macrophages. After recognizing conserved microbial molecules, these cells will release various immune mediators, including cytokines and chemokines, thereby attracting other immune cells to join the immune response. At the same time, conserved microbial molecules can also serve as antigens to activate antibody responses in the body, thereby forming antibody protection. The interaction of these cells and immune factors can better protect the body from pathogens.

Many studies have shown that conserved microbial molecules can improve the body's immunity, thereby preventing and treating a variety of diseases, such as infectious diseases, allergic diseases, cancer, etc. For example, the commonly used recombinant hepatitis B vaccine is made based on the conserved microbial molecule on the surface antigen of the hepatitis B virus. The immune response of the body can be induced by inoculation of the vaccine, to generate the surface antibody of the hepatitis B virus, to achieve the purpose of preventing the infection of the hepatitis B virus.

In conclusion, there is a strong link between conserved microbial molecules and immunity. By making full use of the immunomodulatory function of conserved microbial molecules, the body's immunity can be improved, and various diseases can be better prevented and treated. Therefore, we should actively pay attention to the applied research of conserved microbial molecules and promote their application in disease prevention and treatment. From this point of view, we need to improve immunity. Cistanche can significantly improve immunity, because meat ash contains a variety of biologically active ingredients, such as polysaccharides, two mushrooms, and Huangli, etc. These ingredients can stimulate the immune system. Various types of cells, increase their immune activity.

Click health benefits of cistanche

To date, several classes of PRRs, such as Toll-like receptors, Retinoic acid-inducible gene I (RIG-I)–like receptors, NODlike receptors, and DNA receptors (cytosolic sensors for DNA), have been discovered and characterized (1). These PRRs are at the forefront of both extracellular and intracellular pathogen recognition and sense various classes of molecules in microbes including proteins, lipids, carbohydrates, and nucleic acids (2). This is pivotal to halt disease progression and promote host survival.

Surveillance of the intracellular milieu for restriction of pathogen proliferation is critical to preserve cytosolic sterility. A breach of such defense mechanisms provides the pathogen a refuge from extracellular innate immunity and offers an opportunity for rapid multiplication and dissemination within the host (3). Therefore, potent pathogen sensing mechanisms and cell-autonomous defense systems are critical to restrict invasive pathogens. Ubiquitination is one strategy that plays a pivotal role in pathogen recognition and elimination (4). 

The degradative pathway dictated by ubiquitination acts as a final frontier against cytosol-dwelling bacteria that often evade the classical endocytic killing by rupturing pathogen-containing vacuoles to invade host cytosol. Several host E3 ubiquitin ligases have been identified to decorate cargos including intracellular pathogens with poly-ubiquitin (Ub) chains (5), and although few bacterial targets such as outer membrane proteins have been detected (6), extensive knowledge regarding substrate identification strategy remains limited. 

A recent study has demonstrated secreted effector proteins to contain ubiquitin-associated domain (UBA) in Mycobacterium tuberculosis (Mtb) that passively recruit ubiquitin moieties, ultimately delivering pathogen to Microtubule-associated protein 1A/1B-light chain 3 (LC3) -associated autophagosomes (7). Alongside, unusual ubiquitin substrates like lipopolysaccharide (LPS) and glycan have been elegantly illustrated in a couple of bacterial pathogens, showcasing the versatility of ubiquitination substrates (8, 9). 

Complementarily, Rickettsia parkeri was found to actively modify surface proteins, protecting them from ubiquitination and subsequent killing (10). Together, these independent studies emphasize the significance of the surface localization of the ubiquitin substrate. However, the identity of a proteinaceous substrate in pathogen and how they could be precisely identified by host E3 ligase remain elusive.
Quite a few host ubiquitin ligases, such as leucine-rich repeat and sterile α-motif containing 1 (LRSAM1), Parkin, Ring finger protein 166 (RNF166), RNF213, Ariadne RING-BetweenRING-RING (RBR) E3-ubiquitin protein ligase 1 (ARIH1), SMAD-specific E3-ubiquitin protein ligase 1 (Smurf1), and Skip-Cullin-F-box protein 2 containing complex (SCFFBXO2) are reported to decorate pathogen or pathogen-containing vacuoles with a variety of ubiquitin chain topologies (8, 9, 11–18). Notably, in particular, LRSAM1 through auto-ubiquitination possibly generates a robust ubiquitin signal around the bacteria to recruit the autophagic machinery (19, 20). The ubiquitin ligases responsible for pathogen marking are also involved in maintaining cellular homeostasis, creating an extremely frugal system for efficient and optimal resource utilization. 

Among various ubiquitin chain topologies formed on the pathogen, M1-Ub decoration primarily drives induction of inflammation (21), while both K48- and K63-Ub chain topologies effectively target microbes toward the autophagy or proteasomal system, respectively (22). We have recently demonstrated that the K48-Ub chain has a more dominant antibacterial effect compared to K63-Ub (23). Generally, cellular proteins destined for proteasomal degradation are tagged by K48- Ub chain–specific ligases. The critical signal for substrate recognition by such ligases is principally directed by a degron motif (24).

In this study, we identify the existence of degron motifs in surface proteins of phylogenetically diverse bacteria of both Gram-positive and Gram-negative origin. The targeting of such substrates by ubiquitination machinery propels efficient pathogen elimination from the host cell. Using this, we demonstrate the conversion of a nonubiquitinable surface protein into a ubiquitin substrate by engineering degron insertion to promote bacterial clearance. This simple yet generic principle for identifying bacterial substrates potentially serves as a conserved mechanism of cytosolic pathogen recognition, promising to be efficient and multipurpose in fending off bacterial infections.

RESULTS

The K48-Ub chain promotes the sensing of cytosolic bacterial pathogens

Upon sensing cytosolic invasion by pathogens, the host marked them with poly-Ub chains to trigger their clearance (22). Since such polyUb chains are primarily composed of K48- and K63-Ub, we first explored the predominance and spatial location of these chain types on two phylogenetically distinct pathogens, Streptococcus pneumoniae (SPN) and Salmonella enterica serovar Typhimurium (STM), which cause pneumonia and gastroenteritis in humans, respectively. For these pathogens, survival and proliferation within the host cell cytosol have been documented (13, 25). Using ubiquitin linkage–specific antibodies, we observed that a significantly higher proportion of intracellular bacteria were marked with K48- Ub chain type (~26% for SPN and 37% for STm) in contrast to K63- Ub (Fig. 1A). 

Analysis of spatial location by structured illumination microscopy (SIM) indicated that cytosolic (free of vacuolar remnants) or cytosol-exposed bacteria (within damaged endosome) are primarily associated with K48-Ub, while K63-Ub signal was located on damaged endosomes, marked with Galectin-8 (Gal8; endosome damage sensing marker) (Fig. 1, B to E) (26). About 99 and ~76% of K48-ubiquitinated SPN and STm, respectively, were devoid of Gal8 (Fig. 1F). The bacterial presence in the cytosol was further validated by transmission electron microscopy (TEM) and immunostaining with a membrane marker FM4-64 (fig. S1, A to F). We found 78.4 and 80.4% K48-Ub–positive SPN and STm, respectively, devoid of any membrane association, while 77.7 and 74.4% K63-Ub–positive SPN and STm, respectively, were confined within the vacuole (Fig. 1G). Collectively, these findings suggested that coating the bacterial surface with K48-Ub chains is a major pathogen-sensing mechanism used by the host for recognizing cytosol-dwelling microbes.

Degron is a generic code for bacterial ubiquitination

We next attempted to identify the substrate for K48 ubiquitination on the bacterial surface. Critically, host E3 ubiquitin ligases, reported to be involved in bacterial ubiquitination, are also implicated in crucial cellular functions (12, 14) where K48-Ub chains act as a major signal for cellular proteostasis. We hypothesized that similar principles could be adopted by the host for the identification of the K48-Ub substrate on the bacterial surface. For host proteins, the presence of a tripartite motif (a primary degron sequence followed by a proximal lysine residue and a disordered region in between) is reported to be a prerequisite for K48 ubiquitination (24). 

We screened surface proteins of SPN for the presence of similar features (Fig. 1H), identifying BgaA and PspA as putative targets for ubiquitination (Fig. 1H and Fig. S2, A, and B). BgaA is a β-galactosidase reported to function as an adhesin for SPN, while PspA is a choline-binding protein that binds lactoferrin and is required for complement evasion (27, 28). We observed a ~50 to 53% reduction in association of K48-Ub chain type for both ΔbgaA and ΔpspA mutants, without any change in K63-Ub levels (Fig. 1I). This reduction was pronounced (~75%) in a double-knockout strain (ΔbgaAΔpspA), suggesting the nonredundant nature of these ubiquitin substrates (Fig. 1I). 

Moreover, expression of BgaA-T (a truncated version of the protein, consisting of amino acids 1 to 1049) and PspA in host cells leads to their ubiquitination with K48-Ub topology (Fig. 1J). The validity of the predicted targets was confirmed by complementation, and the model was strengthened by using ΔhysA (SPN surface protein that does not fulfill tripartite degron criteria) mutant as a control to score K48-Ub association levels (Fig. 1I).

We next explored the effect of K48-Ub decoration on bacterial clearance. 

The absence of K48-Ub substrates impeded bacterial clearance, resulting in significantly improved intracellular persistence for both mutant SPN strains (~1.8-fold for ΔbgaA and ~2-fold for ΔpspA) (Fig. 1K). The universality of our substrate prediction approach was validated by the identification of several surface-exposed proteins in various other pathogens as putative substrates for ubiquitination (table S1). One such putative candidate, an outer membrane protein RlpA on STm was confirmed as a target for the host ubiquitination machinery, as the ΔrlpA mutant exhibited ~1.5- fold reduced association with K48-Ub compared to wild-type (WT) STm (Fig. 1L). This finding established the broad applicability of our substrate selection strategy. To the best of our knowledge, these are the first bacterial surface proteins reported to be recognized by host ubiquitination machinery for pathogen sensing and clearance.

A tripartite motif is a prerequisite for precise ubiquitin tagging of bacterial surface proteins

Following substrate identification, we aimed to test the critical features of the tripartite motif which formed the backbone of our screen (Fig. 2A and Fig. S2A). Deletion of the degron sequence ( 102VTPKEE107) in BgaA-T resulted in a ~50% drop in the association of K48-Ub chain type, compared to WT SPN regardless of similar growth kinetics and cell adherence ability (Fig. 2C and Fig. S3, A, and B). This reduction was comparable to the ΔbgaA knockout strain, confirming the degron-specific phenotype. Apart from the degron sequence, a lysine residue in close vicinity is crucial for the attachment of the ubiquitin moiety to the substrate. In BgaA, the degron sequence is accompanied by two proximal considerably reduced K48 ubiquitination compared to BgaA-T (Fig. 2D). Notably, the possibility of severe conformational change in BgaA-TΔDegron protein to affect ubiquitination was nullified by in silico prediction and circular dichroism (CD) spectroscopy of purified BgaA-TΔDegron protein, which showed similar structural signatures to BgaA-T (fig. S3, C and D). Like BgaA, degron sequence ( 327PETPAPE333) deletion and lysine mutation (K315R) in PspA also led to significantly reduced K48-Ub association, coupled with prolonged surviving ability (fig. S4, A to D).

We next engineered the SPN surface protein HysA, originally lacking a primary degron sequence, by addition of a degron sequence within a structurally disordered region of the protein that contains a lysine residue (Fig. 2, F and G, and fig. S2, C to E). This modification conferred recognition and K48 ubiquitination of the previously nonubiquitinable HysA protein. K48-Ub association levels of SPN strains carrying the engineered HysA protein (ΔbgaA:pHysADegron-BgaA and ΔpspA:pHysADegron-PspA) were 2.2- and 3.2-fold higher compared to either ΔbgaA and ΔbgaA: pHysA or ΔpspA strains, respectively (Fig. 2H). 

The enhanced decoration of ΔbgaA:pHysADegron-BgaA and ΔpspA:pHysADegron-PspA strains with K48-Ub robbed SPN of the survival gain provided by the absence of degron in ΔbgaA and ΔbgaA: pHysA or ΔpspA (Fig. 2I). All the engineered SPN strains produced similar levels of the pore-forming toxin pneumolysin (Ply) (Fig. 2, B and G), which is a prerequisite for endomembrane damage and subsequent ubiquitination (25). This nullifies the possible contribution of low or extensive membrane damage promoting a marked change of ubiquitination levels in mutant SPN strains. Collectively, these suggest that the artificial addition of the degron sequence promotes ubiquitin-mediated detection and elimination of pathogens. Notably, the degron sequence in BgaA was highly conserved across different pneumococcal serotypes (fig. S5A). 

However, in serotype 19F which is often associated with an increased risk of death from bacteremic pneumonia and sepsis (29–31), the primary degron was found to be mutated (P104Q). We observed that mimicking this mutation in BgaA (fig. S5B) imparted poor ubiquitination and improved survival ability to ΔbgaA:pBgaA-TP104Q compared to ΔbgaA:pBgaA-T (fig. S5, C and D). This highlights degron recognition as a strategy used by the host to protect itself against severe bacterial infections.

SCFFBW7 is an antimicrobial E3 ubiquitin ligase

The canonical degron sequence present in the selected ubiquitin substrates is predicted to be identified by the SCFFBW7 E3 ubiquitin ligase complex (24), which is involved in the regulation of cell cycle and growth (32). It is composed of two conserved proteins, S-phase kinase-associated protein 1 (SKP1) and a member of the Cullin protein family, along with a variable F-box protein that provides substrate specificity (33). To verify the involvement of SCFFBW7 in SPN ubiquitination, we first assessed the association of FBXW7 with SPN. We found ~31% of intracellular SPN to be associated with FBXW7 upon immunofluorescence analysis (Fig. 3A and Fig. S6A). Expectedly, FBXW7-positive SPN also colocalized with K48 ubiquitin (fig. S6B). To prove the involvement of SCFFBW7, labeling of the bacteria with K48-Ub chains was examined by immunofluorescence, following down-regulation of the expression of Cullin1, SKP1, and FBXW7 genes using targeted small interfering RNAs (siRNAs; fig. S7, A to C). 

In particular, FBXW7 silencing was validated by the cyclin E1 accumulation level in siFBXW7-treated cells (fig. S7F). We observed ~45 to 60% reduction in SPN association with K48-Ub in Cullin1, SKP1, and FBXW7 knockdown cells (Fig. 3B), which, in turn, led to ~1.6- to 1.75-fold increase in SPN persistence within host cells (Fig. 3E). The specific targeting of degron motif by SCFFBW7 was proved by unaltered differences in K48-Ub colocalization and survival ability of ΔpspAΔbgaA and ΔbgaA:pBgaA-TΔDegron strains in siFBXW7-treated cells (fig. S8, A to D). These findings were substantiated by notable reductions in K48 ubiquitination of BgaA-T in host cells following knockdown of FBXW7 (Fig. 3C). 

Further, in vitro, ubiquitination with purified BgaA-T (fig. S9, A to D) and the SCF complex components unambiguously demonstrates SCFFBW7 as the bona fide E3 ligase responsible for ubiquitination of BgaA. Recombinant SCFFBW7 was capable of ubiquitinating purified BgaA-T but failed to ubiquitinate the degron-deleted variant BgaA-TΔDegron or the lysine-to-arginine substitution variant BgaA-TK97R (Fig. 3D). Moreover, host cells expressing FBXW7R505C variant, which exhibits impaired recognition ability for cyclin E1 (a substrate of FBXW7) (fig. S7E), showed reduced (~50%) K48 ubiquitination of SPN, as well as ~2-fold higher survival of SPN compared to WT cells (Fig. 3, F and G). These experiments prove the key role of the SCFFBW7 E3 ligase in the detection of cytosol-dwelling pathogens and targeting them toward killing pathways.

GSK3β-mediated phosphorylation of degron motif potentiates the antimicrobial activity of SCFFBW7

In general, F-box proteins recognize phosphorylated substrates to promote their ubiquitination (34). We, therefore, investigated the likelihood and impact of phosphorylation of bacterial substrates on the K48-Ub coating of the pathogen. Bioinformatics analysis revealed the presence of a putative phosphorylable threonine residue ( 102VT*PKEE107) within the degron sequence in BgaA. We observed that the SPN strain harboring a BgaA-TT103A mutation (ΔbgaA:pBgaA-TT103A) (Fig. 4A) manifested 71% reduced colocalization with K48-Ub compared to WT (Fig. 4B), revealing the relevance of phosphorylation in substrate recognition by the SCF complex. 

Critically, the decreased propensity of BgaA phosphorylation in ΔbgaA:pBgaA-TT103A abrogated the host’s ability to eliminate intracellular bacterial loads (~1.8-fold) (Fig. 4C). In parallel to BgaA, a PspA degron variant (ΔpspA:pPspAT329A) also showed a 51% drop in K48-Ub colocalization that was associated with prolonged intracellular persistence (fig. S10, A to C). In general, SCFFBW7 target substrates have a threonine/serine (T/S*) next to a proline residue, which is phosphorylated by a proline-directed protein kinase, GSK3β (35–37). We, therefore, attempted to unravel the involvement of GSK3β in augmenting substrate recognition. 

We first demonstrated that GSK3β is closely affiliated with ubiquitinated SPN, marked with FBXW7 (Fig. 4, D, and E). Subsequently, by performing an in vitro kinase assay, we observed that GSK3β could phosphorylate recombinant BgaA-T. At the same time, the BgaA-TT103A variant remained nonphosphorylated (Fig. 4F). This validated the identity of the threonine residue within the degron sequence of BgaA-T as a target for GSK3β-mediated phosphorylation. Targeted knockdown of GSK3β by siRNA (fig. S7D) led to ~58% reduction in K48 ubiquitination of SPN (Fig. 4G). This reduced ubiquitination, following down-regulation of GSK3β expression, resulted in a diminished ability of the host to clear cell-invaded pathogens (~1.5-fold) (Fig. 4H) but did not show any effect on ΔbgaA: pBgaA-TT103A (fig. S8, E, and F). Collectively, this provides the first evidence of a host kinase, specifically GSK3β, regulating the ubiquitination of bacterial surface proteins for efficient clearance of pathogens (Fig. 4I).

Ubiquitination of cytosolic pathogens imparts distinct fates for their elimination. In particular, K48 ubiquitination promotes the targeting of substrates toward proteasomes (22). Similarly, our results suggest an association of ubiquitinated SPN with proteasomal subunit, β7 (fig. S11, A, and C). Moreover, proteasomal inhibition by MG132 treatment improves the persistence of WT SPN but does not alter the survival ability of ΔpspAΔbgaA. Similar phenotypes were observed in the case of STm and ΔrlpA mutant (fig. S11, B, and D).

Pathogen surveillance guided by degron protects the host from sepsis

We then sought to determine the impact of SPN recognition via the cellular ubiquitination machinery on outcomes of infection. Using an established model of SPN sepsis (38), we compared the virulence of the ΔbgaA mutant with that of the WT SPN as well as strains complemented with either BgaA-T (ΔbgaA:pBgaA-T) or a version lacking the degron sequence (ΔbgaA:pBgaA-TΔDegron). Consistent with previous reports (39), the bgaA deletion strain showed attenuated virulence, while mice infected with WT, ΔbgaA:pBgaA-T, or ΔbgaA: BgaA-TΔDegron succumbed to infection (Fig. 5A and fig. S12, A to D). However, the group of mice infected with the SPN strain lacking the degron sequence showed a higher proportion of deaths but with delayed mortality compared to ΔbgaA:pBgaA-T–infected group (P = 0.0492, log-rank test) (Fig. 5A). Comparison of bacterial burdens in blood (Fig. 5B) and spleen (Fig. 5C) and the time course of visible disease signs in infected mice (Fig. 5D) confirmed the trend toward increased virulence in the ΔbgaA:pBgaATΔDegron strain.

Previous studies have demonstrated that SPN sepsis is established from a reservoir of bacteria in the spleen (38). While the first wave of invading bacteria in the circulation is rapidly cleared by host innate immune mechanisms, a proportion of SPN survive and proliferate within splenic macrophages, before reseeding into blood. We hypothesized that delayed onset of severe disease in mice infected with ΔbgaA:pBgaA-TΔDegron might be the result of prolonged survival of SPN within splenic macrophages, due to reduced intracellular recognition of bacteria by the host ubiquitination machinery. In support of this, we observed delayed onset of the second wave of bacteremia in mice infected with ΔbgaA:pBgaA TΔDegron strain compared to ΔbgaA:pBgaA-T (24 hours versus 12 hours), following the early clearance phase (Fig. 5E). 

However, in the eclipse phase, during which bacteria are cleared from the blood, splenic bacterial numbers were consistently higher in ΔbgaA: pBgaA-TΔDegron–infected mice (Fig. 5F). These findings suggest that the phase of SPN propagation within splenic macrophages is extended in the absence of intracellular recognition of infection via the ubiquitination machinery. As a result, increased bacterial densities can accumulate in the spleen (Fig. 5F), subsequently seeding into the blood in higher numbers, which may account for the delayed but increased mortality of ΔbgaA:pBgaA-TΔDegron–infected mice. Together, these data demonstrate that recognition and ubiquitination of intracellular SPN contribute to host control of pathogens during sepsis.

For more information:1950477648nn@gmail.com