1. Introduction
The proteasome, a key regulator of protein homeostasis, accounts for as much as 1–2% of the proteome in healthy cells
[1]. The proteasome not only degrades misfolded or otherwise damaged proteins
[2], but also works together with the autophagy machinery to dispose of larger protein aggregates, intracellular bacteria, or even organelles
[3]. By maintaining intracellular protein quality control, it regulates many aspects of cellular existence, including the cell cycle, apoptosis
[4], and antigen processing
[5]. While the proteasome’s age-related reduction in activity may lead to neurodegeneration
[6], its role in inflammation-related diseases and cancer
[7] is well-characterized, and has prompted the development of proteasome-specific drugs
[8].
Initially, proteasomal degradation was intimately linked to the post-translational modification of ubiquitination. This specific modification relies on the attachment of one ubiquitin (Ub) or several ubiquitin residues to the designated substrate. The reaction is orchestrated by a set of three enzymes: E1 responsible for ATP-dependent Ub activation, followed by the transfer of a Ub thioester to a Ub-conjugating enzyme, E2, and finally, E3, catalyzing an isopeptide bond formation. The Ub chain can be elongated through the successive attachment of Ub moieties to the N-terminal methionine residue or, more frequently, to a lysine side-chain present in Ub (K6, K11, K27, K29, K33, K48, K63), resulting in different forms of linear or branched polyUb chains
[9]. In fact, the type of assembled polyUb chain determines the final fate of the substrate. This varies from the modification of its activities to the modulation of protein localization or interactions
[10], or the regulation of its half-life (which ranges from seconds to hours). Consequently, the Ub system is very diverse.
2. Human Pathogens Hijacking the UPS
Given the sheer number of functions that UPS plays in cells, it comes as no surprise that its dysregulation presents a danger to cells. Whereas both the non-proteolytic ubiquitination in signaling and in human disease
[11] and proteolytic ubiquitination in tumorigenesis and cell cycle control
[12] are most often reviewed, the researchers report here that host–pathogen interactions that occur through the UPS have been gaining attention in recent years. Remarkably, it is widely accepted that pathogenic bacteria and viruses lack their own Ub system, but instead can designate proteins with a prokaryotic ubiquitin-like protein (Pup) for proteasomal degradation, at least in some cases
[13]. Nevertheless, they have mastered the ability to hijack host ubiquitination machinery for their own benefit, including disabling key targets, evading the cell defense system, and promoting their replication and pathogenicity
[14]. Although bacteria have evolved several different sophisticated methods to promote their proliferation through the host Ub system, including modifying the dynamics of the actin cytoskeleton
[15] or impairing JNK’s activation
[16], here the researchers focus on the strategies related directly to protein stability (summarized in
Table 1).
All cellular compartments are vulnerable to pathogen infection, starting from the very first step of entering the cell. Some viruses rely upon ubiquitination, which promotes general internalization and sorting to the late endosomes and lysosomes
[17]. The influenza A virus (IAV) utilizes a combination of viral and cellular mechanisms to coordinate the transport of its proteins and gene segments in and, when fully assembled, out of the cell. During uptake, the viral ribonucleoprotein complex is released from the endosome by taking advantage of a host E3 ligase, Itch. Itch ubiquitinates the viral protein M1, triggering viral egress into the cytosol and eventually transporting it to the nucleus
[18], where it replicates.
There has been much research surrounding the concept of pathogens compromising the host immune response. This problem has always attracted considerable attention and the discovery of pathogens’ ability to hijack the UPS has shed even more light on this topic.
Antigen presentation is crucial for triggering T cell immune responses. Two classes of major histocompatibility complexes (MHC), MHC-I and MHC-II, are involved in this process, and both of them are attractive targets for pathogens. Upon ubiquitination, MHC-II is internalized and directed towards endolysosomal degradation while resulting peptides are presented at the cells surface as antigens. The MHC-II -related antigen presentation is downregulated by the Salmonella effector SteD
[19]. This depletes mature MHC-II molecules from the surface of infected antigen-presenting cells through their E3 ligase activity, which results in MHC class II ubiquitination and degradation
[19]. To enhance its effect, SteD is also ubiquitinated, leading to even more reduced levels of surface MHC-II
[19].
Some amino acid sequences serve as stop signals for proteasomal degradation. The Gly-Ala repeat, found in the Epstein–Barr virus protein EBNA-1, prevents EBNA-1 proteasomal degradation
[20]. This is required for the generation of EBNA-1 peptides that MHC-I can use to bind, present, and activate cytotoxic T-cells. Introducing the Gly-Ala sequence into EBNA-4 chimeras
[21], as well as Gly-Ala p53 functional chimeras, allows these chimeras to efficiently escape proteasomal degradation by inhibiting their unfolding
[22], despite the fact that they are ubiquitinated
[23]. Nevertheless, the presence of these repeats alone cannot fully explain EBNA-1’s long stability in cells
[24]. The authors also noted that, when the Gly-Ala repeat is placed in close vicinity to other unfolded regions, the 26S proteasomal degradation of the protein was inhibited. Alternatively, when the repeat is far from other unfolded regions, the protein is easily degraded
[24]. In sum, the presence of specific motifs within the protein sequence regulates the proteasomal degradation in a substrate- and positional-dependent manner. In addition, these findings highlight the importance of the unfolding process for proteasomal degradation.
Another pathway, the ER-associated protein degradation (ERAD) pathway, plays an associated role in protein quality control. This involves the retrotranslocation of errant proteins from the ER into the cytosol for proteasomal degradation
[25]. Cholera toxin finds its way into the cell by retrograde trafficking and utilizing ERAD to enter the cytosol. Similarly to typical ERAD substrates, the enzymatic A1 chain of cholera toxin, which is not ubiquitinated itself
[26], must first be unfolded by retrotranslocation complex components Derlin-1 and HRD1
[27]. Although the typical substrates exported by ERAD are proteasomally processed, cholera toxin is able to escape this fate by its rapid refolding
[26]. In the cytosol, it adopts its native conformation to activate adenylate cyclase by the ADP-ribosylation of the G-protein Gs, resulting in the extreme form of diarrhea that is characteristic of cholera.
In the case of viruses, the ER is used for both entry and replication, as well as for the assembly of infectious viral particles. This can be achieved by hijacking host E3 ligases found in ER. For instance, the human cytomegalovirus (HCMV) takes control over TMEM129, recruited by viral US11, to induce the degradation of MHC-I signaling molecules in a Ub-dependent manner
[28] through Derlin1
[29]. Interestingly, rather than using host proteins, the mouse γ-herpesvirus 68 utilizes its own E3 ligase, mK3
[30] to compromise the immune response in infected organisms, exactly like HCMV.
The E3 ligase TRC8, which is essential for US2-mediated MHC-I breakdown, is usurped by US2 from HCMV. In essence, MHC-I undergoes fast polyubiquitination as a result of TRC8 binding to the cytoplasmic tail of US2
[31]. This causes MHC-I to be delivered into the cytoplasm, where it is degraded. While US11-mediated degradation is restricted to MHC-I, US2, unfortunately, also stimulates the downregulation of a number of immunoreceptors to modify cellular motility and immunological signaling
[32].
Among its many roles in HIV-1 pathogenesis, the viral protein Vpu counteracts host antiviral responses by the downregulation of the HIV-1 receptor CD4
[33] that is produced de novo in the ER by acting as an adaptor to the Skp1/Cullin1/F-box (SCF) Ub ligase complex through β-TrCP. Concomitantly, the Vpu-β-TrCP complex induces the mono- or polyubiquitination of BST-2
[34][35] to abolish its function of inhibiting the release of infectious viral particles. This complex also promotes the degradation of NF-kB and AP1
[36].
Another HIV-1 accessory protein, Vif
[37], also functions as an adaptor of cellular ubiquitination pathways. To enhance the probability of successful infection, this virulence factor takes control of the CUL5 E3 ligase by targeting APOBEC3G/F and STAT1/3 to counteract the former
[38], and inhibit INF-α-signaling in the case of the latter
[39]. The third HIV-1 accessory protein, Vpr, as well as its paralog from HIV-2 and a subset of simian lentiviruses, hijacks the CRL4A (DCAF1) E3 ligase to degrade SAMHD1, a nucleotide triphosphohydrolase that interferes with viral infection
[40] and thus enhances viral reverse transcription.
Some HIV proteins compromise the host UPS to mask themselves from recognition and subsequent proteasomal degradation. For example, the integrase (IN) protein, named after its function of inserting the proviral dsDNA into the host genome, is readily proteolyzed in vitro. However, its lifetime in infected cells is significantly prolonged as it hijacks cellular components such as p75
[41] and Ku70
[42] to shield it from recognition by host E3 ligases.
During budding and when leaving the trans-Golgi network, some viruses must ultimately be released from the vesicle upon reaching their final destination. To accomplish this, the herpes simplex virus, HSV-2, hijacks the activity of the Nedd4 family of E3 ligases. By acting as an adaptor protein, the UL56 tegument protein from HSV-2 enhances the ubiquitination of Nedd4 without being ubiquitinated itself
[43]. In infected cells, this results in the degradation of Nedd4 in a strictly UL56-dependent manner. More recently, the ORF0 of the varicella–zoster virus, UL42 of HCMV, and U24 of human herpesvirus 6A were shown to bind to Itch, a member of the Nedd4 family, through their PPxY motif, and modulate its activity and reduce its protein level
[44].
The HSV-1 E3 ligase ICP0 disrupts components of the so-called promyelocytic leukemia nuclear bodies (PML-NB), which consist of an assembly of about 70 different proteins that prevent the silencing of the viral genome
[45]. Specifically, ICP0 targets the degradation of the promyelocytic leukemia protein, which is a critical scaffolding protein required for the recruitment of other associated proteins and the assembly of PML-NB
[45]. This E3 ligase and its critical role in the infectious cycle of HSV-1 have recently been comprehensively reviewed
[46].
Interestingly, to promote infection, γ-herpesviruses, such as murid herpesvirus-4 (MuHV-4), target Myc, which is essential to the formation and maintenance of germinal center B-cells. In order to stabilize Myc, the viral E3 ligase mLANA catalyzes the attachment of non-canonical Ub chains onto c-Myc, independently of its phosphorylation
[47].
In addition to viral E3 ligases and adaptors, the HSV-1 arsenal also encompasses one more kind of effector that affects the host UPS, a DUB UL36. UL36 removes both the K63- and K48-linked polyUb chains of TRAF3 and abrogates the TRAF3 mediation of IFN-β production
[48]. Similarly to other effectors, the target list of UL36 is not limited to just a single protein. By deubiquitinating IκBα and thus enhancing its lifetime, the NF-κB activation is inhibited to further dampen host antiviral responses in the DNA sensing pathway
[49].
In fact, host proteins involved in antiviral mechanisms are often targeted by different kinds of DUBs. For instance, two papain-like proteases from the severe acute respiratory syndrome coronavirus (SARS-CoV) and the notorious SARS-CoV-2 viruses preferentially target K48-linked polyUb and the ubiquitin-like interferon-stimulated gene 15 protein, respectively
[50], both of which are known to act as regulators of the host innate immune pathways. Similarly, pathogenic DUBs expressed by bacteria and viruses (recently reviewed in
[51]) target host Ub pathways to compromise the immune response.
Given the important role of the mitochondrial antiviral signaling (MAVS) protein in the antiviral immune response, it is hardly surprising that MAVS represents an attractive target for pathogens. The Orf9b of SARS-CoV usurps the HECT domain E3 ligase AIP4 to trigger the degradation of MAVS, as well as TRAF3 and TRAF 6
[52], which are crucial signaling intermediaries in the antiviral defense. In addition, it also triggers the Ub-dependent proteasomal degradation of dynamin-like protein 1, a host protein involved in mitochondrial fission.
Remarkably, bacterial effectors, such as
Salmonella enterica AvrA
[53] and SSeL
[54], also deubiquitinate and impair IκBα degradation in vivo. Another pathogenic function of the latter relies on its ability to remove Ub chains from specific aggregated structures required for
Salmonella replication, thus preventing their recognition and protecting them from autophagic degradation
[55]. More recently, it was suggested that AvrA also decreases Beclin-1 ubiquitination to suppress autophagy
[56]. In addition to these DUBs,
Salmonella produces its own HECT-like E3 ligase, SopA, that targets at least two host E3 ligases, TRIM56 and TRIM65, to promote their degradation and inhibit interferon production
[57].
Another human pathogen,
Shigella flexneri, is known to secrete a range of effectors hijacking host UPS. To date, several
Shigella E3 ligases have been identified that target several different agents involved in the immune response. For example, a secreted E3 ligase IpaH1.4 decorates the linear Ub chain assembly complex (LUBAC) with K48-chains
[58], destining it for degradation. In this way, IpaH1.4 antagonizes the LUBAC-mediated accumulation of M1-linked Ub chains on bacterial surfaces, as well as the recruitment of Optn and Nemo, and abolishes LUBAC-dependent xenophagy
[59]. Two other
Shigella E3 ligases, IpaH7.8 and IpaH0722, have different functions. The former destines the inflammasome inhibitor glomulin for Ub-depended degradation through the activation of caspase-1-mediated cell death
[60]; the latter promotes the degradation of TRAF2 to inhibit NF-κB activity in invaded epithelial cells
[61]. Host cells have mechanisms that prevent actin-dependent cell-to-cell infection to thwart intracellular bacteria by coating the surface of the invading pathogen with interferon-induced guanylate-binding proteins. The
Shigella effector, IpaH9.8, interferes with this process through the targeted proteasomal degradation of critical proteins
[62]. This clever strategy has been demonstrated for both human cell lines
[62] and murine models
[63].
In the case of
Legionella pneumophila, a bacterium causing severe pneumonia known as Legionnaires’ disease, an active host Ub system is required and both E3 ligases and DUBs have been identified among
Legionella’s ca. 300 effector molecules
[64]. AnkB, a F-box domain-containing Ub ligase, directs K48-chain attachment to proteins coating the
Legionella-containing vacuole (LCV) to induce host proteasomal-degradation of -modified proteins, thus providing an amino acid supply to enhance bacterial proliferation
[65].
Fascinatingly, SidE family effectors, including SdeA (Lpg2157), SdeB (Lpg2156), SdeC (Lpg2153), and SidE (Lpg0234), represent a novel type of Ub modification with the ability to modify substrates by means of phosphoribosylated ubiquitination, which is independent of E1 and E2 enzymes
[66]. Upon phosphoribosylation on a specific arginine residue, a modified Ub is conjugated by SdeA to serine residues of protein substrates, and this impairs mitophagy, TNF signaling, proteasomal degradation, and other cellular processes
[67].
Interestingly, the half-life of pathogenic effectors can be regulated by other effectors, called metaeffectors, in a Ub-dependent manner.
Legionella LubX and SidH exemplify probably the best-known example of this kind of temporal regulation. Although initially discovered to target host Cdc2-like kinase 1
[68], LubX, a U-box-type E3 ligase, also efficiently promotes the ubiquitination and degradation of SidH
[69]. It has been proposed that while SidH is required for the very first phases of infection, it must eventually be degraded by the metaeffector to prevent the death of the host cell
[69].
The discussed impact of pathogens on the animal UPS is presentenced in Figure 1 (top panel).
Table 1. Vertebrate pathogens that exploit protein degradation or stabilization by the UPS.
Figure 1. Different modes of action of animal (top) and plant (bottom) pathogens in the process of hijacking host UPS by effector E3 ligases, DUBs and adaptor proteins.