Motifs and interface amino acid–mediated regulation of amyloid biogenesis in microbes to humans: potential targets for intervention

Abstract

Amyloids are linked to many debilitating diseases in mammals. Some organisms produce amyloids that have a functional role in the maintenance of their biological processes. Microbes utilize functional bacterial amyloids (FuBA) for pathogenicity and infections. Amyloid biogenesis is regulated differentially in various systems to avoid its toxic accumulation. A familiar feature in the process of amyloid biogenesis from humans to microbes is its regulation by protein-protein interactions (PPI). The spatial arrangement of amino acid residues in proteins generates topologies like flat interface and linear motif, which participate in protein interactions. Motifs and interface residue–mediated interactions have a direct or an indirect impact on amyloid secretion and assembly. Some motifs undergo post-translational modifications (PTM), which effects interactions and dynamics of the amyloid biogenesis cascade. Interaction-induced local changes stimulate global conformational transitions in the PPI complex, which indirectly affects amyloid formation. Perturbation of such motifs and interface residues results in amyloid abolishment. Interface residues, motifs and their respective interactive protein partners could serve as potential targets for intervention to inhibit amyloid biogenesis.

Introduction

Amyloid is a fibrous cross-β sheet protein structure associated with toxicity and many debilitating diseases in humans (Chiti and Dobson 2006). Reportedly, many organisms produce amyloid in nontoxic amounts that have functional roles in biological processes (Fowler et al. 2007). In lower organisms, amyloids have functional properties such as biofilm formation, pathogenicity, host-microbe interactions and maintenance of infections (Shanmugam et al. 2019).Moreover, recent studies have suggested that functional amyloids are involved in the exacerbation or onset of amyloidosis in higher organisms (Friedland and Chapman 2017; Osorio et al. 2019).

Focusing on the molecular level, amyloid fibrillation ultimately requires “aggregation prone” residues to serve as a template for hydrophobic interaction–mediated β-sheet stacking (Chiti and Dobson 2006). Contrary to fibrillation, the process of amyloid biogenesis is complex, multistep and depends on the interplay of environmental conditions, chaperones, protein trafficking, sorting and precise localization to prevent failure or premature toxic aggregation (Del Prete et al. 2017; Haass et al. 2012; Lazarev et al. 2017; Taglialegna et al. 2016; Evans et al. 2015). A prelude to understanding any biogenesis is to acknowledge that cellular pathways depend on protein’s signalling, translocation, compartmentalization and PTM (Bourret et al. 1991; Levine et al. 2005). Almost every aforementioned process requires PPIs, which in turn depend on selectiveness of amino acid residues and the topography of interacting section of protein (Destainville et al. 2008; Hegde and Kang 2008; Takayama and Clore 2012; Westermarck et al. 2013). For instance, linear motifs of amino acids at the N and C termini of proteins are recognized for signalling-based translocation, serve as a substrate for PTM and guide compartmentalized retention of protein (Filloux 2010; Owji et al. 2018). Likewise, the interfacial arrangement of selective amino acid residues at the protein’s domain contributes to PPI-mediated execution of a functional step (Nyfeler et al. 2005). Hence, structural parameters of proteins are crucial for intracellular interactions, and they categorize PPI as MMI (motif-mediated interaction) and DMI (domain-mediated interaction). MMI requires short linear motifs, commonly residing within IDRs (intrinsically disordered regions) of proteins to form interaction complex (Seo and Kim 2018). MMI usually forms a transient complex as interactions have smaller interfaces and moderate affinities (Perkins et al. 2010). DMI involves the arrangement of hotspot residues in the protein’s domain that supports compact flat interfaces with a partner domain (Moreira et al. 2007). DMI uses interfaces that are rich in hydrophobic and non-polar surface areas (Moreira et al. 2007). Apart from the direct consequences of motifs and domain-interface interactions, PPI stimulates global conformational changes in complexed proteins, which indirectly effects the function of various other domains (Tobi and Bahar 2005). In summary, the conservation of motifs and interface residues regulates protein interactions that ensure proper cellular functioning. Acknowledging the importance of motifs and interface residues, their roles in different amyloid biogenesis systems are discussed here. Exploitation of such protein topographies as drug targets in mitigating the detrimental effects of amyloids is briefly discussed.

Roles of motifs and interface residues in amyloid biogenesis systems

Curli system in Escherichia coli

Escherichia coli is a common pathogen involved in gut illness and urinary tract infections (Yang et al. 2017; Hooton and Stamm 1997). Escherichia coli and Salmonella enterica produce curli, a functional amyloid, which is one of the major constituents of biofilm (Hung et al. 2013). Curli utilizes a type VIII bacterial secretion system to assemble curli subunits into adhesive amyloid fibrils (Cao et al. 2014). The curli system has CsgA-curli forming unit, CsgB-nucleator, CsgC-inhibitor to premature amyloid formation and toxicity, CsgD-transcriptional regulator, CsgE/F-chaperone and CsgG-membrane channel (Hammer et al. 2012; Barnhart and Chapman 2006; Robinson et al. 2006; Evans et al. 2015). Curli secretion utilizes conserved amino acids in its micro-interactome (Fig. 1a). At the periplasmic site, CsgAN₂₂ (1–22 amino acids at the N-terminus) interfaces with CsgG, this sets the substrate recognition step for transport (Wang and Chapman 2008). The linear motif- N6 of CsgAN₂₂ interacts with the small hydrophobic pocket of CsgG while CsgAC_R1–5 (repeats 1–5 at C-terminus) interacts with CsgE, which prevents repeats from premature aggregation (Fig. 1a, b) (Evans et al. 2015). In the periplasm, CsgE nonamers bind to CsgG and form a capping adapter to close the opening of CsgG channel (Yan et al. 2020; Goyal et al. 2014). The binding and unbinding of CsgE nonamers dictate translocation events through CsgG (Van Gerven et al. 2015). The interactive relationship of CsgE, CsgG and CsgA facilitates CsgE-mediated efficient translocation of CsgA through CsgG channel (Yan et al. 2020; Goyal et al. 2014). At the extracellular site, three phenylalanine residues (F₅, F₇ and F₁₂) of CsgFN are critical and promote CsgFG pore complex formation (Fig. 1b) (Yan et al. 2020). Furthermore, Asn17 of CsgFN protrudes to form larger pore than CsgG as its side chain controls pore size and selective biophysical contacts for CsgA passage (Fig. 1a) (Yan et al. 2020). CsgBN domain interfaces with the CsgFN and CsgBC domain interact with CsgA; these interactions establish scaffold for curli fibrillation at the extracellular site (Fig. 1a) (Hammer et al. 2012; Yan et al. 2020). Residues like CsgFN_Asn17 or CsgFN_{F5, 7, 12} have potential to be used for drug design because of their accessible location and critical role in curli assembly. Moreover, decoding multiple CsgB interfaces could potentiate anti-biofilm drug discovery.

Fig. 1.

a Schematic representation of curli secretion system where protein interactions are highlighted, adapted from Yan et al. (2020). b Six critical residues of CsgA responsible for transport recognition, adapted from Yan et al. (2020). Residues in N-terminus domain of CsgF engaged in assembly of CsgGF complex and transport of CsgA, adapted from Van Gerven et al. (2015). c TapA-SipW-TasA secretion system. d Representation of TapA motif responsible TasA assembly, adapted from Romero et al. (2014). e Modulation of Bap auto-aggregation by environmental signals, adapted from Taglialegna et al. (2016). f Diagram representing various domains of Bap where amino acid composition of EF-hand-like-motifs is highlighted, adapted from Arrizubieta et al. (2004). g Als structure with highlighted key interface residues of PBC involved in ligand binding. Changes in PBC site indirectly affecting AFR conformational transition leading to subsequent loss in aggregation, based on Hoyer and Cota (2016); Lipke et al. (2012); Lin et al. (2014). h Representation of various domains in Als3 protein, adapted from Hoyer et al. (2008). i Hypothesised role of MSP-2 amyloids in malarial pathology. j Diagram of different regions in MSP-2 protein. Sequence of amyloid-prone signal region, adapted from Boyle et al. (2013); Lozano et al. (2013). k Schematic representation of APP processing by different secretase, adapted from Haass et al. (2012). Important motifs and residues in APP subjected to PTM-mediated control of amyloid production are represented at the bottom. EM: extracellular membrane, N: N terminal, C: C terminal, S: signal peptide, W: cell wall anchor domain, AICD: amyloid intracellular domain, C1-C5: cysteine residues, R1-R5: repeats

TasA-SipW-TapA system of Bacillus subtilis

Bacillus subtilis is generally classified as non-pathogenic and probiotic bacterium (Oggioni et al. 1998). There have been reports where B. subtilis is found in mammalian infections (Gu et al. 2019; Oggioni et al. 1998). Subtilis employs the TapA-SipW-TasA system to secrete amyloids, which provides resistance to biofilm from denaturants and degradation (Romero et al. 2010). SipW is a serine peptidase that processes the signal sequences of TapA and TasA (Stover and Driks 1999). TapA is an essential minor component of Tas A amyloid fibre formation and anchors TasA to the cell surface (Fig. 1c) (Stover and Driks 1999). The processed N-terminus of TapA has imperfect repeats (50–68 residues) and such longer repeats are present at the C-terminus (194–230) (Fig. 1d). TapA^∆194–230 had no effect on Tas A machinery, while the N terminal domain TapA^∆50–57 led to the abolishment of the secretion system (Romero et al. 2014). Notably, TapA^∆50–57 properly localizes to the extracellular site, but absence of 8-residue long motif disrupts interaction-mediated localization and assembly of Tas A (Romero et al. 2014). Furthermore, Tap A has five conserved cysteine residues as shown in Fig. 1d. The deletion of three residues (3, 4, 5 cysteine) showed impaired pellicle formation, possibly due to their role in the stabilization of amyloids (Romero et al. 2014). Peptides designed to have topography analogous to 8-residue motif could inhibit interaction-mediated Tas A amyloid assembly.

Bap system in S. aureus

Biofilm-associated proteins (Bap) are found in Staphylococci, Enterococci, Pseudomonas and Salmonella biofilms. Bap necessitates S. aureus to form biofilm, intercellular aggregation and adherence to abiotic surfaces (Speziale et al. 2014). Bap-mediated pathogenicity is a leading cause of mastitis infections (Speziale et al. 2014; Di Martino 2016). Bap is a 284.4 kDa multidomain protein with an N-terminal Sec signal sequence and a C-terminus LPETG motif for cell adherence (Lasa and Penades 2006). The C domain (320–2148 aa) of Bap has 13 identical repeats, which form amyloid β-sheet structure (Lasa and Penades 2006). Bap assembly is modulated by environmental signals like pH and Ca²⁺ ion concentration (Taglialegna et al. 2016). Provided that the N-terminus domain of Bap is processed, acidic pH promotes its auto-aggregation, while high concentration of calcium inhibits Bap assembly (Fig. 1e) (Taglialegna et al. 2016). Interestingly, the bap proteins have EF hand-like-motifs to which Ca²⁺ ions bind (Fig. 1f) (Arrizubieta et al. 2004). EF hand motifs consist of helix-loop-helix conformation and have characteristic 12 residues loop where 1, 3, 5, 7, 9 and 12 residues partake in Ca²⁺ chelation (Gifford et al. 2007). EF hand motifs occur in even numbers and their functionality depends on pairing of motifs (Gifford et al. 2007). Similarly, a pair of EF2 and EF3 hand-like-motif, upon Ca²⁺ binding, induces conformational change in Bap, which subsequently inhibits aggregation (Arrizubieta et al. 2004). Bap aggregation depends on Ca²⁺ levels; hence, gland conditions are suitable for S. aureus biofilm formation (Arrizubieta et al. 2004). Calcium supplement would be insufficient to combat mastitis due to poor absorption. Knowledge of Ca²⁺-bound EF hand-like-motif structure could derive lead molecules that would produce non-aggregative Bap conformation.

ALS of Candida albicans

Candida albicans is top opportunistic pathogen, one of the leading causes of vaginal thrush, nosocomial blood stream infections and subsequent fatality (Lin et al. 2014). Candida produces a protein, Als (agglutinin-like sequence), which provides abiotic and biotic adhesion power (Lipke et al. 2012). Candida produces Als 1–7 and Als9, where deletion of Als3 leads to loss of adhesion (Hoyer et al. 1998). Als proteins have four domains: NT domain (N-terminal), T domain (before TR), TR (Tandem repeat) and CT domain (C-terminal) (Hoyer et al. 1998) (Fig. 1h). Als proteins have conserved signal peptide sequences, the amyloid-forming region (AFR) in the NT and GPI anchor regions in CT (Hoyer et al. 1998). Mechanical stimulation enables AFR to form homotypic interactions that assemble Als molecules into amyloid nanodomain (Lipke et al. 2012). These amyloid domains account for mature biofilms consequential to which candidiasis further deteriorates (Gilchrist et al. 2012). The NT domain has a peptide-binding cavity (PBC) that promotes adhesion by accommodating up to six residues of the C termini of the host proteins (Salgado et al. 2011). The PBC conformation affects AFR positioning in the Als3 structure (Hoyer and Cota 2016). In peptide-free PBC, AFR is free for self-complementation and aggregation. In peptide-bound PBC, AFR associates with the Als3 surface becoming unavailable for associative aggregation (Fig. 1g) (Hoyer and Cota 2016; Lin et al. 2014). There are four hotspot residues K59, A116, Y301 and S170 in PBC that interfaces with the host peptide. Mutations in PBC-K59M, A116V and Y301F mimic peptide-bound PBC conformation and led to non-adhesive phenotypes, which also disrupted AFR-mediated aggregation (Lin et al. 2014) (Fig. 1g). The S170Y mutation inhibits peptide binding, but displays a similar loss of adhesion and aggregation (Lin et al. 2014). Understanding of PBC residues could be exploited to design inhibitors that could result imitation of ligand-bound conformation. Furthermore, peptides designed to interact PBC with higher affinity could mitigate host-pathogen interaction and aggregation (Hoyer and Cota 2016).

Merozite surface protein of protozoan

Malarial infection caused by Plasmodium spp. exhibits a multiphasic cell cycle (Beeson et al. 2016). P. falciparum asexual blood-stage merozoites abundantly produce GPI-anchored merozoite surface protein (MSP-2) (Beeson et al. 2016). MSP-2 aids merozoite to invade RBCs and is later degraded inside RBC (Beeson et al. 2016). The signal sequences at N and C termini of MSP-2 are amyloid-prone regions (Fig. 1j) (Low et al. 2007). Amyloids formed by the N-terminus signal motif (MSP-2_1-25) show toxicity similar to that of Aβ and tau proteins (Lu et al. 2018). MSP-2 oligomeric aggregates disrupt membrane mimetics by pore formation, similar to the aforementioned clinically relevant amyloid proteins (Lu et al. 2018). It is speculated that MSP-2_1-25-mediated oligomers have a role in RBC egression or maintenance of the life cycle of pathogen (Moles et al. 2005). In case of cerebral malaria, integrity of blood-brain barrier (BBB) is compromised particularly in vessels inhabitant to merozoite-infected RBCs (Idro et al. 2005). It is hypothesised that BBB’s disruption could be due to egression of merozoites from RBCs, consequentially releasing MSP-2 amyloids (Moles et al. 2005) (Fig. 1i). These hypotheses are open to exploration, but the MSP-2_1-25 amyloidogenic motif could serve as a target for future antimalarial drugs.

Amyloid precursor protein in humans

Alzheimer disease is a consequence of insoluble accumulation of amyloid β peptide (Aβ_1–42) (Haass et al. 2012). Amyloid β peptide is a proteolytic product of amyloid precursor protein (APP) (Haass and Selkoe 1993). Aβ generation localized to the ER lumen involves precise interactive recognition between APP and secretase (Placido et al. 2014). APP processing by different sets of proteases generates an amyloidogenic and non-amyloidogenic pathway. The APP ectodomain processed by β-secretase generates a carboxy terminus fragment that acts as a substrate for γ-secretase to produce an amyloidogenic product (Fig. 1k) (Haass 2004). Alternatively, the APP ectodomain processed by α-secretase generates substrate for γ-secretase that produces non-amyloidogenic product (Fig. 1k) (Haass 2004). Mutations in or around the Aβ domain consequentially lead to improper processing by secretase, which causes many familial Alzheimer disease (FAD) (Hunter and Brayne 2018). APP processing gets more complicated and produces varying products by the introduction of secretases: δ-secretase, η-secretase and meprin-β-secretase (Andrew et al. 2016). Furthermore, PTM of APP provides regulated interactions with partner proteins, which dictate the outcome of Aβ peptides. For instance, an APP residue Ser⁶⁷⁵ phosphorylation favours the interaction and processing by meprin-β-secretase instead of α-secretase, which generates truncated Aβ2–40/42 peptides that have higher aggregation propensity (Menon et al. 2019). Likewise, APP has a conserved Y₆₈₂ENPTY₆₈₇ motif at C-terminus, and phosphorylation at Tyr682 modulates its trafficking and sorting (Slomnicki and Lesniak 2008). Phosphorylation of Tyr682 produces conformational changes in C-terminus of APP which disfavours interactions with proteins that prevent the Aβ pathway and instead reinforces PPIs that propagate pathophysiology (Ando et al. 2001). Another APP motif, ⁶⁵³YTSI⁶⁵⁶, has a phosphorylation tendency at Ser⁶⁵⁵ (Slomnicki and Lesniak 2008). Phosphorylated APP at Ser⁶⁵⁵ is unable to interact with an adapter protein that directs the cargo to the lysosomal compartment (Tam et al. 2016). This reduced trafficking to lysosome substantially decreases Aβ toxicity (Tam et al. 2016). In context to disease alleviation, PTM-mediated APP regulation could be exploited without disrupting the global interactome. Advent of phosphonate and phosphate prodrugs with proper delivery strategies could synthetically control phosphorylation, favourable for inhibition of disease (Wiemer and Wiemer 2015). Furthermore, Aβ toxicity could be combated by targeting secretase and APP interactions.

Interface residues and motifs as drug target

PPI mediated by interface residues and motifs in amyloid-producing systems could serve as drug targets with few considerations. Most PPIs are transient, promiscuous (where a single motif/interface interacts with multiple proteins) and complex (Pelay-Gimeno et al. 2015; Corbi-Verge and Kim 2016). Moreover, drugs targeting PPI are preferentially large compounds that cover a large interface area with specificity (Pelay-Gimeno et al. 2015; Corbi-Verge and Kim 2016). Large compounds pose problems in reaching intracellular targets with substantial concentration and encounter cellular degradation (Corbi-Verge et al. 2017). There have been notable developments in drug design and delivery strategies to increase pharmacological response, as discussed in Table 1. Despite these efforts, drug design against interface/motif is still underdeveloped and requires improvement in various fields. There are thousands of unresolved protein structures; information on interface and protein-protein contacts in complexes is also deficit. Structural decoding in combination with machine learning, coevolution studies and scoring methods could contribute towards the development of PPI-based pharmacology (Torrisi et al. 2020). Interaction assays coupled with computational biology can resolve the cellular localization of PPI for drug targeting (Ding and Kihara 2018). Proteo-genomics could aid in the construction of new druggable motifs library. DNA nanotech, nanomotors and upgraded biologics would potentiate the bioavailability of PPI-based therapeutics (Jorge and Eritja 2018; Venugopalan et al. 2020).

Table 1.

Drug design and delivery strategies to inhibit PPI

Drug design and optimization	Characteristics	References
High throughput screening (HTS)	Could be used when less information is available on protein targets. Exploits multi-protein complexes. Lack of availability of physiologically relevant models to screen compounds	Taylor et al. (2018)
Fragment-based drug design	Fragments or low molecular compounds targeted against PPI and their library is screened by various biophysical techniques. Limitations with targets like large macromolecular and membrane complex	Robson-Tull (2018)
Peptidomimetics	Molecules designed to mimic key motifs involved in PPI interface. Variations in formulation of mimetics in order to increase its efficiency.	Akram et al. (2014).
Backbone alteration	Substitution or addition of atoms in peptides.	Mabonga and Kappo (2020)
Non-canonical amino acids	Non-canonical functional groups added to side chains in a peptide. Good Packing at interface. Higher price and large space utilization in libraries.	Corbi-Verge and Kim (2016); Corbi-Verge et al. (2017)
D amino acids	D amino acids peptides. Magnitude of potency increased. Non-immunogenic.	Corbi-Verge and Kim (2016)
Stapled peptides	Hydrocarbon α helical peptide having synthetic chemical staple linker which enables to lock peptide at specific site.	Ali et al. (2019)
Nanoswitchable peptides	Combination of nanoparticles peptides and photo modulation targeting α helical motif in PPI.	Nevola et al. (2018)
Drug delivery strategy
Pegylation and lipidation	Addition of PEG or lipid molecule to increase permeability.	Corbi-Verge et al. (2017)
Cell-penetrating peptides	Therapeutics conjugated to 3–50 amino acid peptide which has endogenous property to permeate cell.	Guidotti et al. (2017)
Liposomes, nanoparticles and microsomes	These colloidal carriers ensure prolonged and controlled release in cells.	Corbi-Verge et al. (2017)
Carbon nanotubes	Crosses BBB, high drug loading and increased pharmacological efficiency.	Guo et al. (2017)

Conclusion

Amyloids were primarily linked to mammalian diseases but their occurrence in lower organisms made the clinical situation worse. Amyloids in lower biological systems provide pathogenicity and reinforce infections in mammals. In order to combat detrimental effect of amyloid, understanding of its biogenesis is important. Process of amyloid biogenesis depends on multiple critical steps where slight interventions disrupt amyloid formation. Protein interactions, mediated by motifs and interfaces, predominantly control amyloid pathways. These protein topographies serve as an alternative interference platform to mitigate the effects of amyloids. Structural information, localization and strategic delivery are prerequisite for PPI-based therapeutics. Even though PPI-based pharmacology has advanced, inclusion of computer biology, proteo-genomics and biologics can assure its clinical implementation.

Funding

The Council of Scientific Indian Research (CSIR) provided funding in the form of CSIR-NET/SRF to AZB.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.

Ethical approval

No studies have been performed on humans and animals by any authors and not required.

Research resource identifiers (RRID)

Software: Inkscape, RRID:SCR_014479.