Exploring Intrinsic Disorder in Stress Granules Proteome: Implications for Structure-Function Dynamics and Pathological Aggregation

doi:10.21203/rs.3.rs-4930444/v1

Download PDF

Research Article

Exploring Intrinsic Disorder in Stress Granules Proteome: Implications for Structure-Function Dynamics and Pathological Aggregation

https://doi.org/10.21203/rs.3.rs-4930444/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Prolonged cellular stress causes the formation of small physiological stress granules (SGs) that gradually accumulate due to chronic stress or mutations in RNA-binding proteins and form aggregations in diseased brains with progressive, age-related impairment of proteasomal and autophagic systems. These aggregations are a consequence of multivalent interactions mediated by the intrinsically disordered regions (IDRs) among the SG proteins. Although several studies have focused on identifying the SGs implicated in various pathologies, their structure-function relationship remains largely unexplored. In this study, we characterized the 460 SG protein components from the mammalian stress granule proteome (MSGP) database and analysed them for the presence of IDRs and molecular recognition feature (MoRF) sites. Our study identified abundant intrinsic disorder in the SG proteome, especially among its RNA-binding proteins. Contrastingly, the intrinsic disorder is not very abundant among the autophagy-linked SG proteome with the presence of mostly structured/ordered proteins. These results have provided newer insights into their molecular interactions, functions, and transition from physiological to pathological entities. The findings will provide an important direction for further research and experimental characterization of these disordered proteins to decipher their pathological role in various human diseases.

Stress granules

intrinsically disordered regions

RNA-binding proteins

autophagic systems

protein-protein interaction

Stress granules (SGs) are membrane-less, cytosolic messenger ribonucleoproteins (mRNPs) that rapidly aggregate in response to various environmental stresses- oxidative stress, heat shock, viral infection, UV radiation, hyperosmolarity and glucose deprivation and represent localized zones of impeded translation initiations within the cytoplasm [1]. SGs constitute stalled 48S pre-initiation complex with translationally arrested mRNAs, small ribosomal subunits, translation initiation factors, RNA binding proteins (RBPs). Depletion of these mRNA translation initiation factors or overexpression of certain RBPs is known to induce formation of SGs. They form by liquid-liquid phase separation (LLPS) via co-assembly or condensation of translationally inactive mRNAs (bimolecular condensates) [2]. SGs primarily function to protect against environmental insults and promote cell survival by condensing translationally stalled translational initiation factors, RBPs, and serve as sites of mRNA triage wherein these sequestered mRNAs are individually sorted for storage, degradation, and translation during stress and recovery [3]. Although SGs have a pro-survival adaptive response to external stress, abnormal persistent aggregation of SGs or aberrant assembly of SGs and their impaired disassembly contributes to the pathology of cancer, viral infections, and neurodegeneration [4, 5].

RNA-binding proteins (RBPs) aggregate and limit mRNA utilization under stress conditions partly via their formation of SGs [4]. RBPs, collectively known as heterogeneous nuclear ribonucleoproteins (hnRNPs) or mRNA-protein complex proteins (mRNP proteins) bind to single or double-stranded RNA and are involved in the formation of cytoplasmic and nuclear ribonucleoprotein complexes to regulate essential aspects of RNA transcription, splicing, modification, intracellular trafficking, translation, and degradation [6]. In response to stress or external cues, RBPs play a role in the stability or destabilization of mRNAs [7]. SGs form by aggregation of glycine-rich domain of these RBPs and strive to sequester, silence and/or degrade RNA transcripts while adjusting to a pattern of local RNA translation [8]. These RBPs respond to a myriad of biological and environmental stimuli and therefore, in light of this, it is evident that a variety of diseases including cancer, metabolic diseases, neuropathies, and muscular atrophy can be associated with repressed RBP expression and function [9, 10].

Autophagy is a self-degradation process responsible for the removal of misfolded and/or aggregated proteins, RNP granules [11] and damaged organelles in addition to destroying intracellular pathogens [12]. Prolonged cellular stress causes the formation of small physiological SGs that gradually accumulate due to chronic stress or mutations in RBPs and form aggregations in diseased brains with progressive, age-related impairment of their proteasomal and autophagic systems [13].

Intrinsically disordered proteins (IDPs) are biologically active, hybrid proteins that lack a stable 3D structure but contain ordered domains and intrinsically disordered regions (IDRs) and are essential to any proteome [14]. These IDRs confer the function of recognition and control of various signalling pathways. Characterization of these intrinsically disordered proteins is crucial to decipher the molecular mechanisms underpinning various human diseases. Individual proteins of the SG proteome contribute to distinct condensates due to interactions of IDRs and RNAs [15]. Previous RNA interactome analysis have suggested the presence of abundant IDRs [16]. Therefore, in this study we aim to extensively characterize 460 proteins established as SGs components and identify the occurrence of IDPs/IDRs in these components. Presence of IDPs/IDRs can provide insights into their molecular functions and interactions and may uncover its probable link to various neurodegenerative diseases.

Collection and classification of SG proteins

The 464 proteins identified as SG components were obtained from the Mammalian Stress Granule Proteome (MSGP) database (https://msgp.pt/region/all-proteins/).The MSGP database (https://msgp.pt/) is an online open-access database that provides a description of all the protein components of SGs in mammalian cells [17]. The database is structured in a way that classifies the SG component proteins into three classes: All proteins (464 proteins), RBP (253 proteins), and autophagy (32 proteins). For a better understanding and quantification of results, we will independently evaluate and analyse the intrinsic disorder propensity in the SG proteins as classified in these three sections. The “All proteins” section consists of all 464 SG component proteins. The subsequent sections narrow the search to SG proteins that are either RNA-binding or linked to the autophagy pathway. After the removal of duplicate and redundant proteins, we analysed 460 proteins for the presence of intrinsic disorder. Protein sequences of the 460 proteins were obtained in FASTA format from UniProt database [18].

Evaluation of intrinsic disorder propensity of SG proteins

Further, we quantified the intrinsic disorder in these proteins by employing several members of the Predictors of Natural Disordered Regions (PONDR) namely- VL-XT[19], VSL2[20], VL3[21] and FIT[22]. These members are distinguished on the basis of various machine learning models and training sets that have been used to develop them [23, 24]. Moreover, two forms of IUPred2A tool [25] has also been utilized for the prediction of long (≥ 30 residues) and short (< 30 residues) IDRs in query proteins. Per-residue disorder score is obtained from these platforms, the residues with score between 0.2 and 0.5 are considered flexible while those with score greater than 0.5 are considered disordered residues. The value of predicted percent of disorder (PPID) was calculated for each protein from the corresponding per-residue predicted disorder scores generated. The MobiDB database [26], which provides predictions of intrinsic disorder propensity in proteins, further validated these outcomes

Using combined CH-CDF approach to predict disorder in the entire SG proteome

To characterize the intrinsic disorder in the entire proteome and distinguish the proteins into highly ordered and highly disordered, we used the combined CH-CDF analysis approach [27] (available at http://www.pondr.com/) which uses two binary predictors of disorder- Charge hydropathy (CH) and cumulative distribution frequency (CDF) and is based on charge, hydropathy, and PONDR scores of the query proteins. The coordinates of each point in the resulting CH-CDF plot is calculated as the distance of corresponding protein from the boundary (Y-axis) and average distance of respective CDF curve from the CDF boundary (X-axis) [28].

Generation of protein–protein interaction network

The Search Tool for Retrieval of Interacting Genes (STRING) database (available at https://string-db.org/) has been utilized to generate a comprehensive functional-association network of the 460 proteins of the stress granule proteome. STRING uses predicted and experimentally validated information on the interactivity of the proteins of interest to generate a network of predicted associations [29].

Characterization of intrinsic disorder in the most disordered SG proteins

We further narrow our study to the 10 most disordered proteins in each category of SG protein components. The AlphaFold database, a novel deep learning algorithm that generates highly accurate predicted protein structures based on evolutionary, physical, and geometric constraints protein structures, was used to obtain high-resolution 3D crystal structures of these disordered proteins [30]. The D²P² database [31] provided complementary disorder evaluations along with significant disorder-related functional data and identified MoRF sites, a distinct class of intrinsically disordered protein regions, exhibiting molecular function and interaction features. Molecular Recognition Feature (MoRF) sites are short segments within the disordered regions that undergo disorder-to-order transition upon interaction with specific partners [32]. Furthermore, the gene expression value of these highly disordered proteins in the context of various neurodegenerative disorders in the form of a graph based on differential expression analysis has also been taken from the MSGP database [17].

All proteins

SGs are known to play a prominent role in neurodegenerative diseases and myopathies either through their accumulation in neurodegenerative brains or via mutations in genes encoding RBPs for SG response (such as those linked with motor neuron diseases) [13]. Several bioinformatics studies have linked intrinsically disordered proteins with various diseases such as diabetes, cancer, and cardiovascular and neurodegenerative diseases [10, 33–35]. IDPs can fold upon interaction with the corresponding binding partner and have precise functional control over post-translational modifications and alternative splicing in addition to the promiscuous and plastic nature of binding which makes them key players in various pathological conditions. Furthermore, protein-protein interactions involving intrinsically disordered proteins are a promising target to develop new strategies for drug development wherein, the drug can be designed that either mimics the disordered motif or the IDP and targets the binding site [36].

A dense protein-protein interaction (PPI) network generated for highly interconnected 453 proteins of the SG proteome were generated using the STRING database on medium confidence of 0.4 (Supplementary Figure S1). The 453 nodes (proteins) in this PPI network have 6153 edges (connections), and this is significantly higher than 2289 expected number of edges for a random set of proteins of the same size and degree distribution drawn from the genome characterized by PPI enrichment p-value of < 10^− 16. Therefore, such an enrichment indicates that the proteins are at least partially biologically connected as a group. The average node degree, defined as the number of connections of a particular node to other nodes is 27.2 for this network (i.e. on an average, each protein interacts with 27 other proteins). The average local clustering coefficient defines how close the neighbours of a node are to being a complete clique, and is equal to 1, if every neighbour connected to a given node N_i is also connected to every other node in the neighbourhood and is equal to 0 if none of the nodes connected to a given node N_i connects to any other node connected to N_i. The average local clustering coefficient for this network is 0.384. The most important gene-ontology (GO) assigned biological processes to this network are post-transcriptional regulation of gene expression, regulation of translation, regulation of mRNA metabolic process, mRNA metabolic process, nucleic acid metabolic process, regulation of gene expression and cellular nitrogen compound metabolic process. The prominent molecular functions associated with this network are RNA binding, nucleic acid binding, heterocyclic compound binding, organic cyclic compound binding, binding, mRNA binding, translation regulator activity, mRNA-3’UTR binding, protein binding and cadherin binding. Among the cellular components of this network are ribonucleoprotein granule, cytoplasmic ribonucleoprotein granule, cytoplasmic stress granule, supramolecular complex, nucleus, cytosol, intracellular, non-membrane-bounded organelle, ribonucleoprotein complex and intracellular non-membrane-bounded organelle.

The proteins of the SG proteome have been classified on the basis of PPID_Mean, calculated by averaging the PPID values obtained across six disorder predictors (PONDR- VLXT, VSL2, VL3, FIT and IUPred2A- long and short) (Supplementary Table 1). These proteins have been classified as highly ordered (PPID_Mean <10%), moderately disordered (10% ≤ PPID_Mean <30%) and highly disordered (PPID_Mean ≥30%) [37]. From this PPID_Mean -based classification, 39 proteins out of 460 are predicted as highly ordered, 161 are moderately disordered and 260 are highly disordered. Figure 1(A) shows a two-dimensional PPID_PONDR−FIT vs. PPID_Mean plot representing the predisposition of 460 SG component proteins to intrinsic disorder.

To gain further insight into the nature of intrinsic disorder in the entire proteome, we employed the combined CH-CDF plot analysis (Supplementary Table 2). The combined CH-CDF plot analysis provides information about its structural and physical characteristics based on the quadrant in which it is located. The four quadrants of the CH-CDF plot are classified as- Q1 (protein is predicted to be ordered by both CH and CDF predictors), Q2 (protein predicted to be ordered by CH and disordered by CDF plot), Q3 (protein predicted to be disordered by both CH and CDF plot), Q4 (protein predicted to be disordered by CH plot and ordered by CDF plot). Figure 1(B) shows the combined CH-CDF plot analysis employed for 460 stress granule protein components. A total of 189 proteins are found in Q1 and are predicted to be ordered by both predictors, while 163 proteins in Q2 are molten globular proteins and/or hybrid proteins, predicted as ordered by CH plot and disordered by CDF. The ordered and disordered residues in these proteins are comparable and these proteins may lack unique stable 3D structures. The Q3 quadrant has 99 highly disordered proteins with presence of intrinsic disorder predicted by both the predictors. The Q4 quadrant constitutes 9 proteins predicted as disordered by CH plot and ordered by CDF plot. Out of the 460 SG component proteins, 271 (58.9%) proteins are predicted to be significantly disordered further validating that the intrinsic disorder is abundant in the SG proteome.

The 10 most disordered proteins in “All proteins” category of MSGP, based on PPID_Mean are Non-histone chromosomal protein HMG-14 (HMGN1; PPID_Mean: 100; MobiDB: 100%), High mobility group protein HMG-I/HMG-Y (HMGA1; 100; 100%), CASC3 (94.38; 82.5%), PRRC2A (PRRC2A; 93.21; 86.6%), Phostensin (PPP1R18; 89.01; 80.4%), Tankyrase-1-binding protein (TNKS1BP1; 88.85; 79.1%), PRKC apoptosis WT1 regulator protein (PAWR; 87.4; No entry in MobiDB), Plasminogen activator inhibitor 1 RNA-binding protein (SERBP1; 86.68; 83.6%), Intracellular hyaluronan-binding protein 4 (HABP4; 86.44; 75.8%) and RNA-binding protein EWS (EWS; 85.87; 64.9%) (Table 1).

Table 1

The top 5 disordered proteins of the stress granule proteome. Universal protein resource (Uniprot) ID, molecular function, PPIDmean, and their OMIM information for their probable links to human genetic disorders.
Protein Name	Abbreviation	Molecular Function	Uniprot ID	PPID_Mean	OMIM Disease
Non-histone chromosomal protein HMG-14	HMGN1	DNA binding, nucleosomal DNA binding	P05114	100
High mobility group protein HMG-I/HMG-Y	HMGA1	5'-deoxyribose-5-phosphate lyase activity, DNA binding, chromatin binding, DNA-lyase activity, enzyme binding, ligand-dependent nuclear receptor transcription coactivator activity, peroxisome proliferator-activated receptor binding, retinoic acid receptor binding, transcriptional factor activity, transcription factor binding	P17096	100	Susceptibility to Diabetes mellitus
Phostensin	PPP1R18	Actin binding, phosphatase binding	Q6NYC8	89.01604133
182 kDa tankyrase-1-binding protein	TNKS1BP1	Ankyrin repeat binding, cadherin binding, enzyme binding, macromolecular complex binding	Q9C0C2	88.85601504
PRKC apoptosis WT1 regulator protein	PAWR	Actin binding, enzyme binding, leucine zipper domain binding, transcription co-repressor activity	Q96IZ0	87.94068627

Figure 2 represents functional disorder profiles of the 5 (out of 10 most disordered proteins of the “All proteins” section of MSGP database) generated by D²P² platform. D²P² provides complementary disorder evaluations along with disorder related functional information by combining the outputs of IUPred, PONDR® VLXT, PrDOS [38], PONDR® VSL2B, PV2, and ESpritz [39] (representing disordered regions by coloured bars). Figure 2 also shows the positions of mostly structured SCOP domains [40] generated by the SUPERFAMILY predictor [41], Molecular recognition features (MoRF regions) identified by the ANCHOR algorithm [42], and different post translational modifications (PTMs) obtained from the PhosphoSitePlus platform [43]. PTMs affect efficiency of protein folding, stability of their conformations and influence their biological and catalytic activity [44]. The functional disorder profiles indicate the abundance of MoRF sites in all of these 5 proteins. Moreover, phosphorylation sites are abundant in PPP1R18, TNKSBP1 and PAWR proteins. In addition to phosphorylation sites, several ubiquitination and acetylation sites for HMGN1 and ubiquitination, acetylation and methylation sites for HMGA1 can also be seen.

Per-residue disorder profile for proteins – HMGN1, HMGA1, TNKSBP1, PPP1R18 and PAWR are shown in Fig. 3. These graphs suggest the presence of many IDRs that lack well-defined structures in their native state. Remarkable agreement between outputs of these six predictors validate higher tendency of disorder.

The corresponding high resolution Alphafold structures of the five most disordered SG proteins were shown in Supplementary Figure S2, that indicates the presence of intrinsic disorder in all of these proteins (as indicated by low per-residue confidence scores suggesting unstructured domains). This is in complete agreement with the predictions by various disorder predictors employed above.

Further, we next analysed the functional associations among the most disordered proteins of the SG proteome using the STRING database. Our analysis initially showed that none of the 10 most disordered proteins are interacting with each other (Fig. 4A). This indicates that this set of the top 10 most disordered proteins of the SG proteome are not very well connected. It is noteworthy here, that it is not necessary that this set of proteins are not biologically important but rather it simply means that these proteins have not been studied very much and that their interactions might not be deposited in databases. Next, we extended the PPI network to include 50 first shell interactors of these 10 highly disordered proteins. The corresponding functional association network using STRING generated on highest confidence of 0.9 is shown in Fig. 4B. The resulting PPI network includes 60 nodes connected by 390 edges, which is significantly greater than expected 112 number of edges for a randomly selected set of proteins of comparable size. The average node degree of this network is 13 (on average, each protein is linked to 13 other proteins) and average local clustering coefficient is 0.789. The top 10 GO-annotated biological processes of this network are nuclear- transcribed mRNA catabolic process, mRNA metabolic process, nuclear-transcribed mRNA catabolic process, nonsense-mediated decay, RNA metabolic process, nucleic acid metabolic process, heterocycle metabolic process, protein targeting to ER, cellular aromatic compound metabolic process, viral transcription and SRP-dependent co-translational protein targeting to membrane. Among the molecular functions associated with this network are RNA binding, structural constituent of ribosome, nucleic acid binding, structural molecule activity, heterocyclic compound binding, organic cyclic compound binding, rRNA binding, mRNA binding, binding and ubiquitin ligase inhibitor activity. The cellular components associated with this network are ribonucleoprotein complex, cytosolic ribosome, ribosomal subunit, ribosome, cytosolic small ribosomal subunit, protein containing complex, U2-type catalytic step 1 spliceosome, nuclear lumen, nucleus and exon-exon junction complex.

The MSGP database (https://msgp.pt/) used GEO2R tool to perform a differential transcriptomic analysis of the GEO Datasets: GSE33000 [45], GSE28894 and GSE4595 [46] to assess the expression profiles of SG components. The study showed that out of the 464 SG genes, 380 genes showed the altered expression in the brain of AD patients, in which 187 genes were upregulated and 193 genes were downregulated [47]. Similarly, 395 genes have altered expression in the brain of HD patients, in which 195 genes have increased and 200 genes have decreased expression.

Here, the gene expression level of the five most disordered proteins HMGN1, HMGA1, PPP1R18, TNKS1BP1 and PAWR in context of various neurodegenerative disorders obtained from the MSGP database are shown in Supplementary Figure S3. The genes HMGN1, HMGA1, and TNKS1BP1 were upregulated in ALS while gets downregulated in both AD and PD. The genes PPP1R18 and PAWR were upregulated in AD and HD. This indicates that these genes have differentially regulation in different neurodegenerative diseases.

Five out of these 10 most disordered proteins are also RNA-binding proteins (CASC3, PRRC2A, SERBP1, HABP4 and EWS) and therefore, will be discussed in the next section.

RNA-Binding Proteins

RNA binding proteins (RBPs) regulate essential aspects of cell’s life including transcription, splicing, modification, intracellular trafficking, translation, degradation and elimination, and are critical effectors of gene expression, with their dysfunction leading to several human diseases. Certain RBPs have the tendency to coalesce into membrane-less compartments by liquid-liquid phase separation and this process in mediated by establishment of weak interactions between RNA molecules and RBPs involving low-complexity and intrinsically disordered regions. The dynamics of phase separation can be potentially linked with age as is the gradual loss of regulation of gene expression and increase in number of misfolded proteins combined with decline in mitochondrial activity leading to aggregation of amyloids [16]. Furthermore, mutations in SG components are linked to several neurodegenerative diseases and these mutations have been reported in aggregation prone Prion-like domains (PLDs), low-complexity domains (LCDs), intrinsically disordered regions (IDRs) and RNA-binding motifs within the RBPs [48].

The functional-association network of 250 RBPs of the SG proteome generated using STRING on a medium confidence of 0.4 is shown in Supplementary Figure S4. The 245 nodes (proteins) in this PPI network have 3634 edges (connections), which is significantly higher than 756 expected number of edges for a random set of proteins of same size and distribution drawn from the genome, and this is characterized by PPI enrichment p-value of < 10^− 16. Such an enrichment score indicates that a partially biological connections exist between these proteins. The average node degree is 29.7 (i.e. on average, each protein interacts with 29 other proteins) and the average local clustering coefficient is 0.464 for this network. The most important gene-ontology (GO) assigned biological functions to this network are posttranscriptional regulation of gene expression, regulation of translation, regulation of mRNA metabolic process, RNA metabolic process, mRNA metabolic process, nucleic acid metabolic process, nucleic acid metabolic process, gene expression, cellular nitrogen compound metabolic process, RNA processing and Nucleobase-containing compound metabolic process. The prominent molecular functions of this network are RNA-binding, heterocyclic compound binding, organic cyclic compound binding, mRNA binding, binding, mRNA 2’-UTR binding, translation regulator activity, double-stranded RNA binding, translation regulator activity, nucleic acid binding, and single stranded RNA binding. Among the cellular components of this network are ribonucleoprotein granule, cytoplasmic ribonucleoprotein granule, ribonucleoprotein complex, cytoplasmic stress granule, nucleus, supramolecular complex, nuclear lumen, protein-containing complex, nucleoplasm and intracellular.

The combined CH-CDF plot has been analysed for the characterization of intrinsic disorder in the RNA-binding proteome (Fig. 5A). In the combined CH-CDF plot for the RBPs, 64 proteins are found in Q1, indicating order by both the predictors. 93 proteins are located in Q2, predicted as ordered by CH plot and disordered by CDF. A total of 89 proteins are predicted as disordered by the both the predictors and are found in Q3. The Q4 constitutes 4 proteins predicted as disordered by CH plot and ordered by CDF plot. Out of the 250 RBPs in the SG proteome, 186 (74.4%) proteins are identified as having a considerable amount of intrinsic disorder, validating previous studies that suggested the abundance of disorder in RBP components of SGs. From the PPID_Mean- based classification in the 250 RBPs of the SG proteome, 14 proteins are highly ordered (PPID_Mean<10), 81 proteins are moderately disordered (10 ≤ PPID_Mean <30) and 155 proteins are highly disordered (PPID_Mean≥30). Figure 5B shows a two-dimensional PPID_PONDR−FIT vs. PPID_Mean plot representing the predisposition of 250 RBPs component of stress granules to intrinsic disorder.

The 10 most disordered RBPs of the mammalian SG proteome are CASC3 (CASC3; PPID_Mean: 94.38; No predicted disorder entry on MobiDB), PRRC2A (PRRC2A; 93.21; No entry on MobiDB), Plasminogen activator inhibitor 1 RNA-binding protein (SERBP1; 86.68; 42%), Intracellular hyaluronan-binding protein 4 (HABP4; 86.44; No entry on MobiDB), RNA-binding protein EWS (EWS; 85.87; No entry on MobiDB), Fused in Sarcoma (FUS; 84.53; No entry), Scaffold attachment factor B2 (SAFB2; 84.12; 49.8%), Eukaryotic translation initiation factor 4B (EIF4B; 84.06; No entry), Coiled-coil domain containing 9B (CCDC9B; 84.04; No entry) and Antigen KI-67 (MKI67; 83.88; No entry) (Table 2).

Table 2

The 10 most disordered RNA-binding proteins of the stress granule proteome name, their Universal protein resource (Uniprot) ID, molecular function, PPID_mean, and OMIM information for their probable links to human genetic disorders.
Protein Name	Abbreviation	Molecular Function	Uniprot ID	PPID_Mean	OMIM Disease
Protein CASC3	CASC3	Enzyme binding, identical protein binding, ubiquitin protein ligase binding	O15234	94.38158606
Protein PRRC2A	PRRC2A	RNA-binding	P48634	93.21509581
Plasminogen activator inhibitor 1 RNA-binding protein	SERBP1	Cadherin binding, mRNA 3'-UTR binding, RNA binding	Q8NC51	86.68094771
Intracellular hyaluronan-binding protein 4	HABP4	RNA-binding	Q5JVS0	86.44173123
RNA-binding protein EWS	EWS	Calmodulin binding, identical protein binding, metal ion binding, RNA binding	Q01844	85.875	Ewing sarcoma, Neuroepithelioma
Fused in Sarcoma	FUS	DNA binding, estrogen receptor binding, identical protein binding, ionotropic glutamate receptor binding, metal ion binding, myosin V binding, retinoid X receptor binding, RNA binding, thyroid hormone receptor binding, transcription coactivator activity	P35637	84.53581749	Amyotrophic lateral sclerosis, Tremor
Scaffold attachment factor B2	SAFB2	Double-stranded DNA binding, identical protein binding, RNA binding, sequence-specific DNA binding	Q14151	84.12063484
Eukaryotic translation initiation factor 4B	EIF4B	Helicase activity, ribosomal small subunit binding, RNA binding, RNA strand annealing and exchange activity, translation initiation factor activity	P23588	84.06964539
Coiled-coil domain containing 9B	CCDC9B	RNA-binding	Q6ZUT6	84.04989388
Antigen KI-67	MKI67	ATP binding, DNA binding, protein C-terminus binding, RNA binding	P46013	83.88072072

Figure 6 shows functional disorder profiles providing complementary disorder evaluations along with disorder related functional information of the 10 most disordered RBPs generated by D²P² platform. With higher intrinsic disorder propensity, these RBPs can be seen as enriched with MoRF sites. Per-residue disorder profile for proteins – CASC3, PRRC2A, SERBP1, HABP4, EWS, FUS, SAFB2, EIF4B, CCDC9B AND MKI67 are shown in Fig. 7. These graphs suggest the presence of multiple intrinsically disordered regions and remarkable agreement can be seen between outputs of the six predictors. The corresponding Alphafold structures of these 10 most disordered RBPs indicates the presence of intrinsic disorder in all these proteins (Supplementary Figure S5). This is in complete concordance with the predictions by various disorder predictors employed above.

Further, we focused towards analysing the functional associations among the top 10 most disordered RBPs of the SG proteome using the STRING database on highest confidence of 0.9. The 10 proteins (nodes) of this network form do not form any connection and the current set of proteins that are not very well connected (PPI enrichment p-value = 1) (Fig. 8A). This could simply mean that these proteins have not been studied very much and that their interactions might not yet be known to STRING. The average node degree and average local clustering coefficient of this network is 0. Moreover, RNA-binding and SUMO-binding are the major molecular functions associated with this network. We further extended the PPI network to include 50 first shell interactors of these 10 highly disordered proteins. The corresponding functional association network using STRING generated on highest confidence of 0.9 is shown in Fig. 8B. The resulting STRING network has 60 nodes and 282 edges which is significantly more than 99 expected number of edges, suggesting that compared to a random set of proteins of the same size and degree distribution selected from the genome, this set of proteins interacts with one another more frequently. An enrichment of this kind suggests that, collectively, the proteins have some biological connections. The average node degree and average local clustering coefficient for this network is 9.4 and 0.729, respectively. The GO assigned biological processes of this network are nuclear-transcribed mRNA catabolic process, nonsense-mediated decay, mRNA metabolic process, gene expression, RNA catabolic process, translation initiation, RNA metabolic process, cellular nitrogen compound metabolic process, posttranscriptional regulation of gene expression, nucleic acid metabolic process and translation. The most prominent molecular functions associated with this network are RNA binding, nucleic acid binding, heterocyclic compound binding, organic cyclic compound binding, structural constituent of ribosome, mRNA binding, binding, translation initiation factor activity, translation regulator activity and structural molecule activity. Among the cellular components are ribonucleoprotein complex, U-2 type catalytic step 1 spliceosome, cytosolic ribosome, protein containing complex, exon-exon junction complex, ribsosome, ribsosomal subunit, spliceosomal complex, nucleoplasm and nuclear lumen.

The gene expression profile of these 10 most disordered RBPs in context of different neurodegenerative diseases, as obtained from MSGP database are shown in Supplementary Figure S6. In fact, most of the SG components are RBPs, which play a variety of crucial roles in neurons physiology. Therefore, changes in their expression could influence or be the cause of the neurodegenerative pathophysiology. In case of ALS, only EWS have decreased expression while rest all have increased expression. For AD, CASC3 and HABP4 show decreased expression while for HD, SERBP1, HABP4, and FUS show decreased expression. In cased of PD, most of the genes are upregulated also. The expression analysis results thus indicated the increased in expression of most of the RBPs in different neurodegenerative diseases.

Autophagic Proteins

Autophagy is a major catabolic pathway for degradation and recycling of cellular components, is tightly maintained to regulate cellular homeostasis and provides an alternative pathway for stress granule clearance [49]. Defective autophagy pathway has been linked to Parkinson’s disease, Huntington’s disease, amyotrophic lateral sclerosis, multiple sclerosis, frontotemporal dementia, cortical atrophy, glaucoma, epilepsy, diabetes mellitus and cancer. Moreover, recent studies have demonstrated that autophagy pathway heavily relies on dynamic and flexible protein regions and occurrence of intrinsic disorder provides suitable conformation for regulation [50–52]. Furthermore, under stress conditions neurons may interact with their intracellular SGs to degrade large amount of RNA through the autophagy pathway. When the crisis is removed, SGs get eliminated through autophagy and normal translation is restored [5]. However, with persistent cellular stress, c9orf72 mutation [53, 54] results in blocked stress granule-autophagy pathway leading to formation of TDP-43 inclusions in neuronal cytoplasm, a characteristic hallmark of ALS/FTD.

The functional association network of 32 autophagic proteins of the SG proteome is shown in Supplementary Figure S7. The 32 nodes (proteins) are connected by 62 edges (connections), which is significantly greater than 21 expected number of edges expected for a random set of proteins of the same size and degree distribution drawn from the genome. The average node degree and average local clustering coefficient for this network is 3.88 and 0.591, respectively. The GO assigned biological processes are regulation of cellular protein metabolic process, cellular response to stress, regulation of nitrogen compound metabolic process, regulation of translation, regulation of primary metabolic process, regulation of cellular biosynthetic process, regulation of cellular metabolic process, regulation of metabolic process, cellular protein modification process and regulation of cellular macromolecule biosynthetic process. The most prominent molecular functions associated with this network are translation regulator activity, protein serine/threonine kinase activity, protein kinase activity, nucleotide binding, catalytic activity, acting on a protein, purine nucleotide binding, purine ribonucleoside triphosphate binding, adenyl nucleotide binding, purine ribonucleotide binding and enzyme binding. Among the cellular components of this network are cytoplasm, cytosol, translation release factor complex, glutamatergic complex, TORC1 complex, aggresome, lysosome, supramolecular complex, autolysosome and cytoplasmic stress granule.

The PPID_Mean –based classification of 32 SG proteins linked to the autophagy pathway identified 5 proteins as highly ordered (PPID_Mean<10), 17 as moderately disordered (10 ≤ PPID_Mean <30) and 10 as highly disordered (PPID_Mean≥30). Figure 9A shows a two-dimensional PPID_PONDR−FIT vs. PPID_Mean plot representing the predisposition of 32 autophagic proteins of SGs to intrinsic disorder. The combined CH-CDF plot analysis (Fig. 9B) provided further insight into the presence of intrinsic disorder within the autophagic SG proteome. The 32 autophagic SG proteins are found only in the Q1 and Q2 quadrants of the CH-CDF phase space. Out of these, 18 proteins are found in Q1 and predicted as ordered by both predictors whereas, 14 proteins are predicted as ordered by CH and disordered by CDF, found in Q2. Therefore, only 14 proteins (43.75%) of the autophagic SG proteome are considered having some measurable amount of disorder.

These results are contrasting to the CH-CDF plot obtained for RBPs wherein, abundant intrinsic disorder was predicted in the CH-CDF plot. The CH-CDF plot obtained for the autophagic SG proteome therefore suggests that intrinsic disorder propensity among these proteins is significantly lower.

The 10 most disordered autophagy proteins of the SGs are BAG family molecular chaperone regulator 3 (BAG3; PPID_Mean: 83.76, MobiDB: 86.43%), Hamartin (TSC1; 56.30; No predicted disorder entry on MobiDB), Sequestosome-1 (SQSTM1; 54.31, 37.05%), Serine/threonine-protein kinase PAK 4 (PAK4; 51.01; 28.26%), Eukaryotic translation initiation factor 2-alpha kinase 1 (EIF2AK1; 36.69; 38.73%), Protein transport protein Sec24C (SEC24C; 36.34; 29.80%), CUGBP Elav-like family member 1 (CELF1; 34.05; 38.89%), Histone deacetylase 6 (HDAC6; 33.96; 30.29%), Interferon-induced, double-stranded RNA-activated protein kinase (EIF2AK2; 33.153; 10.71%) and Dual specificity mitogen-activated protein kinase 7 (MAP2K7; 30.83; 7.40%) (Table 3).

Table 3

The 10 most disordered autophagy-linked proteins of the stress granule proteome name, their Universal protein resource (Uniprot) ID, molecular function, PPID_mean, and OMIM information for their probable links to human genetic disorders.
Protein Name	Abbreviation	Molecular Function	Uniprot ID	PPID_Mean	OMIM Disease
BAG family molecular chaperone regulator 3	BAG3	Adenyl-nucleotide exchange factor activity, cadherin binding, chaperone binding, protein complex binding	O95817	83.76956522	Cardiomyopathy, dilated, 1HH; Myopathy, myofibrillar
Hamartin	TSC1	Chaperone binding, GTPase activating protein binding, protein N-terminus binding	Q92574	56.30111111	Focal cortical dysplasia, Lymphangioleiomyomatosis, Tuberous sclerosis
Sequestosome-1	SQSTM1	Enzyme binding, identical protein binding, ionotropic glutamate receptor binding, K63-linked polyubiquitin modification-dependent protein binding, protein homodimerization activity, protein kinase binding, protein kinase C binding, protein serine/threonine kinase activity, receptor tyrosine kinase binding, SH2 domain binding, ubiquitin binding, ubiquitin protein ligase binding, zinc ion binding	Q13501	54.31712121	Frontotemporal dementia and/or amyotrophic lateral sclerosis, Myopathy, Neurodegeneration (with ataxia, dystonia, and gaze palsy), Paget disease of bone
Serine/threonine-protein kinase PAK 4	PAK4	ATP binding, cadherin binding, cell-cell adhesion, serine/threonine kinase activity, Rac GTPase binding	O96013	51.01591653
Eukaryotic translation initiation factor 2-alpha kinase 1	EIF2AK1	ATP binding, eukaryotic translation initiation factor 2 alpha kinase activity, heme binding, protein homodimerization activity	Q9BQI3	36.69179894
Protein transport protein Sec24C	SEC24C	SNARE binding, zinc ion binding	P53992	36.34955515
CUGBP Elav-like family member 1	CELF1	BRE binding, mRNA binding, RNA binding, translation repressor activity, nucleic acid binding	Q92879	34.05430041
Histone deacetylase 6	HDAC6	histone deacetylase activity, actin binding, tubulin binding, tubulin deacetylase activity, beta-catenin binding, dynein complex binding, promoter binding, enzyme binding, Hsp90 binding, microtubule binding, misfolded protein binding, polyubiquitin binding, tau protein binding, ubiquitin protein ligase binding, zinc ion binding	Q9UBN7	33.96599451	Chondrodysplasia
Interferon-induced, double-stranded RNA-activated protein kinase	EIF2AK2	ATP binding, double-stranded RNA binding, non-membrane spanning protein tyrosine kinase activity, protein serine/threonine kinase activity, protein phosphatase regulator activity, RNA binding	P19525	33.15298246
Dual specificity mitogen-activated protein kinase 7	MAP2K7	ATP binding, enzyme binding, magnesium ion binding, protein serine/threonine kinase activity, protein tyrosine kinase activity, MAP kinase activity, protein phosphatase binding	O14733	30.83772076

Figure 10 shows functional disorder profiles of the 10 most disordered autophagy-linked SG proteins generated by D²P² platform. The functional disorder profiles indicate that with a few IDRs among the autophagic proteins, only a few MoRF sites can be seen. Per-residue disorder profile for proteins –BAG3, TSC1, SQSTM1, PAK4, EIF2AK1, SEC24C, CELF1, HDAC6, EIF2AK2 and MAP2K7 are shown in Fig. 11. These graphs suggest the presence of multiple intrinsically disordered regions and remarkable agreement can be seen between outputs of these six predictors.

The corresponding Alphafold structures of the 10 most disordered autophagic SG proteins indicates the presence of intrinsic disorder (Supplementary Figure S8). These structures show highly structured regions (in blue) and that is in concordance with disorder predictions by tools employed above and the unstructured regions are few (represented by yellow and orange).

A PPI network of top 10 most disordered autophagic SG proteins generated on highest confidence of 0.9 is shown in Fig. 12A. This functional association network has 10 nodes and 2 edges (connections) which is higher than 0 expected edges. The average node degree and average local clustering coefficient of this network are 0.4 and 0.2, respectively. The GO assigned biological processes are aggresome assembly, positive regulation of macroautophagy, negative regulation of translation, regulation of cellular response to stress, negative regulation of cellular protein metabolic process, regulation of protein phosphorylation, and negative regulation of cellular macromolecule biosynthetic process, cellular response to stress, negative regulation of gene expression and regulation of cellular protein metabolic process. The molecular functions associated with this network are protein serine/threonine kinase activity, eukaryotic translation initiation factor 2alpha kinase activity and translation regulator activity. Among the cellular components are aggresome and chaperone complex.

Next, we extended our PPI network to include 50 first shell interactors of the top 10 most disordered autophagic SG proteins (Fig. 12B). The 60 nodes (proteins) form 199 edges (connections) which is higher than 91 expected number of edges. The average node degree of this network is 6.63 and the average local clustering coefficient is 0.656. The GO assigned biological processes of this network are cellular response to stress, response to stress, positive regulation of molecular function, regulation of protein metabolic process, regulation of cellular protein metabolic process, regulation of cellular response to stress, regulation of protein modification process, positive regulation of catalytic activity, COPII-coated vesicle cargo loading and regulation of phosphorylation. The most prominent molecular functions are ubiquitin protein ligase binding, enzyme binding, protein binding, protein serine/threonine kinase activity, binding, protein kinase activity, purine nucleotide binding, nucleotide binding, purine ribonucleoside triphosphate binding and protein ribonucleotide binding. Among the cellular components are cytosol, COPII vesicle coat, cytoplasm, ER to Golgi transport vesicle membrane, perinuclear region of cytoplasm, whole membrane, intracellular membrane-bounded organelle, cytoplasmic vesicle, intracellular organelle and bounding membrane of organelle.

The gene expression analysis of the 10 most disordered autophagy components of the SG proteome in context of different neurodegenerative diseases, as obtained from MSGP database are shown in Supplementary Figure S9. All of the genes show increased expression in case of AD except EIF2AK2. The expression of genes BAG3 and CELF1 were repressed in case of ALS while for HD, HDAC6 and CELF1 have decreased expression. In case of PD, PAK4, EIF2AK1, SEC24C and MAP2K7 showed downregulation suggesting different roles of autophagic proteins in context of neurodegenerative diseases.

Stress granules associated inclusions are known to be associated with pathogenesis of several neurodegenerative disorders and RNA-binding proteins contribute to disease progressions by forming larger SG aggregates. These aggregations are known to be mediated by intrinsically disordered regions within the protein. Our analysis suggested that the intrinsic disorder is abundant among the stress granule proteins. Within the stress granule proteome, RBPs are majorly found to be disordered whereas, autophagic proteins are majorly ordered. At the same time, most of the disordered RBPs have increased expression in the brain of patients suffering from neurodegenerative diseases. However, autophagic components of SG have different expression regulation in PD patients. Our study has provided significant insights into the presence of intrinsic disorder among the stress granule proteome and their link to pathogenesis of various maladies can be corroborated by additional and more experimental research.

Ethics Approval:

Not applicable.

Conflicts of Interest:

The authors declare that they have no conflicts of interest to report regarding the present study.

Author Contribution

S.A. performed the experiments and calculations, wrote the draft, and prepared all the figures. V.K conceived the idea, wrote the final manuscript and supervision. All authors reviewed the manuscript.

Acknowledgments

VK sincerely thanks Indian Council of Medical Research (ICMR) New Delhi, India (Grant No. DHR/Neuro/2020-NCD-I). The authors sincerely thank Amity University, Noida for providing facilities.

Availability of Data and Materials:

The datasets generated during and/or analysed during the current study are available within the manuscript and supplementary materials.

Li YR, King OD, Shorter J and Gitler AD (2013) Stress granules as crucibles of ALS pathogenesis. J Cell Biol 201:361-72. doi: 10.1083/jcb.201302044
Youn JY, Dyakov BJA, Zhang J, Knight JDR, Vernon RM, Forman-Kay JD and Gingras AC (2019) Properties of Stress Granule and P-Body Proteomes. Mol Cell 76:286-294. doi: 10.1016/j.molcel.2019.09.014
Wolozin B (2012) Regulated protein aggregation: stress granules and neurodegeneration. Mol Neurodegener 7:56. doi: 10.1186/1750-1326-7-56
Jeon P and Lee JA (2021) Dr. Jekyll and Mr. Hyde? Physiology and Pathology of Neuronal Stress Granules. Frontiers in Cell and Developmental Biology 9. doi: 10.3389/fcell.2021.609698
Hofmann S, Kedersha N, Anderson P and Ivanov P (2021) Molecular mechanisms of stress granule assembly and disassembly. Biochim Biophys Acta Mol Cell Res 1868:118876. doi: 10.1016/j.bbamcr.2020.118876
Dreyfuss G, Kim VN and Kataoka N (2002) Messenger-RNA-binding proteins and the messages they carry. Nat Rev Mol Cell Biol 3:195-205. doi: 10.1038/nrm760
Alves LR and Goldenberg S (2016) RNA-binding proteins related to stress response and differentiation in protozoa. World J Biol Chem 7:78-87. doi: 10.4331/wjbc.v7.i1.78
Wolozin B and Ivanov P (2019) Stress granules and neurodegeneration. Nat Rev Neurosci 20:649-666. doi: 10.1038/s41583-019-0222-5 10.1038/s41583-019-0222-5 [pii]
Castello A, Fischer B, Eichelbaum K, Horos R, Beckmann BM, Strein C, Davey NE, Humphreys DT, Preiss T, Steinmetz LM, Krijgsveld J and Hentze MW (2012) Insights into RNA biology from an atlas of mammalian mRNA-binding proteins. Cell 149:1393-406. doi: 10.1016/j.cell.2012.04.031
Ali N, Prasad K, AlAsmari AF, Alharbi M, Rashid S and Kumar V (2021) Genomics-guided targeting of stress granule proteins G3BP1/2 to inhibit SARS-CoV-2 propagation. Int J Biol Macromol 190:636-648. doi: S0141-8130(21)01930-9 [pii] 10.1016/j.ijbiomac.2021.09.018
Monahan Z, Shewmaker F and Pandey UB (2016) Stress granules at the intersection of autophagy and ALS. Brain Res 1649:189-200. doi: 10.1016/j.brainres.2016.05.022
Glick D, Barth S and Macleod KF (2010) Autophagy: cellular and molecular mechanisms. J Pathol 221:3-12. doi: 10.1002/path.2697
Wolozin B and Ivanov P (2019) Stress granules and neurodegeneration. Nature Reviews Neuroscience 20:649-666. doi: 10.1038/s41583-019-0222-5
Uversky VN (2019) Intrinsically Disordered Proteins and Their “Mysterious” (Meta)Physics. Frontiers in Physics 7. doi: 10.3389/fphy.2019.00010
Protter DSW and Parker R (2016) Principles and Properties of Stress Granules. Trends Cell Biol 26:668-679. doi: 10.1016/j.tcb.2016.05.004
Gebauer F, Schwarzl T, Valcárcel J and Hentze MW (2021) RNA-binding proteins in human genetic disease. Nat Rev Genet 22:185-198. doi: 10.1038/s41576-020-00302-y
Nunes C, Mestre I, Marcelo A, Koppenol R, Matos CA and Nóbrega C (2019) MSGP: the first database of the protein components of the mammalian stress granules. Database 2019. doi: 10.1093/database/baz031
UniProt: the universal protein knowledgebase in 2021. Nucleic Acids Res 49:D480-d489. doi: 10.1093/nar/gkaa1100
Romero P, Obradovic Z, Li X, Garner EC, Brown CJ and Dunker AK (2001) Sequence complexity of disordered protein. Proteins 42:38-48. doi: 10.1002/1097-0134(20010101)42:1<38::aid-prot50>3.0.co;2-3
Peng K, Radivojac P, Vucetic S, Dunker AK and Obradovic Z (2006) Length-dependent prediction of protein intrinsic disorder. BMC Bioinformatics 7:208. doi: 10.1186/1471-2105-7-208
Peng K, Vucetic S, Radivojac P, Brown CJ, Dunker AK and Obradovic Z (2005) Optimizing long intrinsic disorder predictors with protein evolutionary information. Journal of bioinformatics and computational biology 3:35-60. doi: 10.1142/s0219720005000886
Xue B, Dunbrack RL, Williams RW, Dunker AK and Uversky VN (2010) PONDR-FIT: A meta-predictor of intrinsically disordered amino acids. Biochimica et Biophysica Acta (BBA) - Proteins and Proteomics 1804:996-1010. doi: https://doi.org/10.1016/j.bbapap.2010.01.011
Dunker AK, Brown CJ, Lawson JD, Iakoucheva LM and Obradović Z (2002) Intrinsic Disorder and Protein Function. Biochemistry 41:6573-6582. doi: 10.1021/bi012159+
He B, Wang K, Liu Y, Xue B, Uversky VN and Dunker AK (2009) Predicting intrinsic disorder in proteins: an overview. Cell Research 19:929-949. doi: 10.1038/cr.2009.87
Dosztányi Z, Csizmok V, Tompa P and Simon I (2005) IUPred: web server for the prediction of intrinsically unstructured regions of proteins based on estimated energy content. Bioinformatics 21:3433-3434. doi: 10.1093/bioinformatics/bti541
Piovesan D, Necci M, Escobedo N, Monzon AM, Hatos A, Mičetić I, Quaglia F, Paladin L, Ramasamy P, Dosztányi Z, Vranken WF, Davey NE, Parisi G, Fuxreiter M and Tosatto SCE (2021) MobiDB: intrinsically disordered proteins in 2021. Nucleic Acids Res 49:D361-d367. doi: 10.1093/nar/gkaa1058
Huang F, Oldfield C, Meng J, Hsu WL, Xue B, Uversky VN, Romero P and Dunker AK (2012) Subclassifying disordered proteins by the CH-CDF plot method. Pac Symp Biocomput:128-39.
Xue B, Oldfield CJ, Dunker AK and Uversky VN (2009) CDF it all: Consensus prediction of intrinsically disordered proteins based on various cumulative distribution functions. FEBS Letters 583:1469-1474. doi: https://doi.org/10.1016/j.febslet.2009.03.070
Szklarczyk D, Gable AL, Nastou KC, Lyon D, Kirsch R, Pyysalo S, Doncheva NT, Legeay M, Fang T, Bork P, Jensen LJ and von Mering C (2021) The STRING database in 2021: customizable protein-protein networks, and functional characterization of user-uploaded gene/measurement sets. Nucleic Acids Res 49:D605-d612. doi: 10.1093/nar/gkaa1074
Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, Tunyasuvunakool K, Bates R, Zidek A, Potapenko A, Bridgland A, Meyer C, Kohl SAA, Ballard AJ, Cowie A, Romera-Paredes B, Nikolov S, Jain R, Adler J, Back T, Petersen S, Reiman D, Clancy E, Zielinski M, Steinegger M, Pacholska M, Berghammer T, Bodenstein S, Silver D, Vinyals O, Senior AW, Kavukcuoglu K, Kohli P and Hassabis D (2021) Highly accurate protein structure prediction with AlphaFold. Nature 596:583-589. doi: 10.1038/s41586-021-03819-2 10.1038/s41586-021-03819-2 [pii] 3819 [pii]
Oates ME, Romero P, Ishida T, Ghalwash M, Mizianty MJ, Xue B, Dosztányi Z, Uversky VN, Obradovic Z, Kurgan L, Dunker AK and Gough J (2013) D²P²: database of disordered protein predictions. Nucleic Acids Res 41:D508-16. doi: 10.1093/nar/gks1226
Vacic V, Oldfield CJ, Mohan A, Radivojac P, Cortese MS, Uversky VN and Dunker AK (2007) Characterization of molecular recognition features, MoRFs, and their binding partners. J Proteome Res 6:2351-66. doi: 10.1021/pr0701411
Uversky VN (2010) Targeting intrinsically disordered proteins in neurodegenerative and protein dysfunction diseases: another illustration of the D(2) concept. Expert Rev Proteomics 7:543-64. doi: 10.1586/epr.10.36
Uversky VN, Dave V, Iakoucheva LM, Malaney P, Metallo SJ, Pathak RR and Joerger AC (2014) Pathological unfoldomics of uncontrolled chaos: intrinsically disordered proteins and human diseases. Chem Rev 114:6844-79. doi: 10.1021/cr400713r
Uversky VN, Oldfield CJ, Midic U, Xie H, Xue B, Vucetic S, Iakoucheva LM, Obradovic Z and Dunker AK (2009) Unfoldomics of human diseases: linking protein intrinsic disorder with diseases. BMC Genomics 10 Suppl 1:S7. doi: 10.1186/1471-2164-10-S1-S7 1471-2164-10-S1-S7 [pii]
Habchi J, Tompa P, Longhi S and Uversky VN (2014) Introducing protein intrinsic disorder. Chem Rev 114:6561-88. doi: 10.1021/cr400514h
Rajagopalan K, Mooney SM, Parekh N, Getzenberg RH and Kulkarni P (2011) A majority of the cancer/testis antigens are intrinsically disordered proteins. Journal of Cellular Biochemistry 112:3256-3267. doi: https://doi.org/10.1002/jcb.23252
Ishida T and Kinoshita K (2007) PrDOS: prediction of disordered protein regions from amino acid sequence. Nucleic Acids Research 35:W460-W464. doi: 10.1093/nar/gkm363
Walsh I, Martin AJM, Di Domenico T and Tosatto SCE (2011) ESpritz: accurate and fast prediction of protein disorder. Bioinformatics 28:503-509. doi: 10.1093/bioinformatics/btr682
Murzin AG, Brenner SE, Hubbard T and Chothia C (1995) SCOP: A structural classification of proteins database for the investigation of sequences and structures. Journal of Molecular Biology 247:536-540. doi: https://doi.org/10.1016/S0022-2836(05)80134-2
de Lima Morais DA, Fang H, Rackham OJL, Wilson D, Pethica R, Chothia C and Gough J (2010) SUPERFAMILY 1.75 including a domain-centric gene ontology method. Nucleic Acids Research 39:D427-D434. doi: 10.1093/nar/gkq1130
Mészáros B, Simon I and Dosztányi Z (2009) Prediction of Protein Binding Regions in Disordered Proteins. PLOS Computational Biology 5:e1000376. doi: 10.1371/journal.pcbi.1000376
Hornbeck PV, Kornhauser JM, Tkachev S, Zhang B, Skrzypek E, Murray B, Latham V and Sullivan M (2011) PhosphoSitePlus: a comprehensive resource for investigating the structure and function of experimentally determined post-translational modifications in man and mouse. Nucleic Acids Research 40:D261-D270. doi: 10.1093/nar/gkr1122
Walsh CT, Garneau-Tsodikova S and Gatto GJ, Jr. (2005) Protein posttranslational modifications: the chemistry of proteome diversifications. Angew Chem Int Ed Engl 44:7342-72. doi: 10.1002/anie.200501023
Narayanan M, Huynh JL, Wang K, Yang X, Yoo S, McElwee J, Zhang B, Zhang C, Lamb JR, Xie T, Suver C, Molony C, Melquist S, Johnson AD, Fan G, Stone DJ, Schadt EE, Casaccia P, Emilsson V and Zhu J (2014) Common dysregulation network in the human prefrontal cortex underlies two neurodegenerative diseases. Mol Syst Biol 10:743. doi: 10.15252/msb.20145304
Lederer CW, Torrisi A, Pantelidou M, Santama N and Cavallaro S (2007) Pathways and genes differentially expressed in the motor cortex of patients with sporadic amyotrophic lateral sclerosis. BMC Genomics 8:26. doi: 1471-2164-8-26 [pii] 10.1186/1471-2164-8-26
Nunes C, Mestre I, Marcelo A, Koppenol R, Matos CA and Nobrega C (2019) MSGP: the first database of the protein components of the mammalian stress granules. Database (Oxford) 2019. doi: 10.1093/database/baz031 baz031 5367298 [pii] baz031 [pii]
Asadi MR, Sadat Moslehian M, Sabaie H, Jalaiei A, Ghafouri-Fard S, Taheri M and Rezazadeh M (2021) Stress Granules and Neurodegenerative Disorders: A Scoping Review. Front Aging Neurosci 13:650740. doi: 10.3389/fnagi.2021.650740
Popelka H (2020) Dancing while self-eating: Protein intrinsic disorder in autophagy. Prog Mol Biol Transl Sci 174:263-305. doi: 10.1016/bs.pmbts.2020.03.002
Babu MM, van der Lee R, de Groot NS and Gsponer J (2011) Intrinsically disordered proteins: regulation and disease. Curr Opin Struct Biol 21:432-40. doi: 10.1016/j.sbi.2011.03.011
Kulkarni P and Uversky VN (2019) Intrinsically Disordered Proteins in Chronic Diseases. Biomolecules 9. doi: 10.3390/biom9040147
Uversky VN, Oldfield CJ and Dunker AK (2008) Intrinsically disordered proteins in human diseases: introducing the D2 concept. Annu Rev Biophys 37:215-46. doi: 10.1146/annurev.biophys.37.032807.125924
DeJesus-Hernandez M, Mackenzie IR, Boeve BF, Boxer AL, Baker M, Rutherford NJ, Nicholson AM, Finch NA, Flynn H, Adamson J, Kouri N, Wojtas A, Sengdy P, Hsiung GY, Karydas A, Seeley WW, Josephs KA, Coppola G, Geschwind DH, Wszolek ZK, Feldman H, Knopman DS, Petersen RC, Miller BL, Dickson DW, Boylan KB, Graff-Radford NR and Rademakers R (2011) Expanded GGGGCC hexanucleotide repeat in noncoding region of C9ORF72 causes chromosome 9p-linked FTD and ALS. Neuron 72:245-56. doi: 10.1016/j.neuron.2011.09.011
Renton AE, Majounie E, Waite A, Simón-Sánchez J, Rollinson S, Gibbs JR, Schymick JC, Laaksovirta H, van Swieten JC, Myllykangas L, Kalimo H, Paetau A, Abramzon Y, Remes AM, Kaganovich A, Scholz SW, Duckworth J, Ding J, Harmer DW, Hernandez DG, Johnson JO, Mok K, Ryten M, Trabzuni D, Guerreiro RJ, Orrell RW, Neal J, Murray A, Pearson J, Jansen IE, Sondervan D, Seelaar H, Blake D, Young K, Halliwell N, Callister JB, Toulson G, Richardson A, Gerhard A, Snowden J, Mann D, Neary D, Nalls MA, Peuralinna T, Jansson L, Isoviita VM, Kaivorinne AL, Hölttä-Vuori M, Ikonen E, Sulkava R, Benatar M, Wuu J, Chiò A, Restagno G, Borghero G, Sabatelli M, Heckerman D, Rogaeva E, Zinman L, Rothstein JD, Sendtner M, Drepper C, Eichler EE, Alkan C, Abdullaev Z, Pack SD, Dutra A, Pak E, Hardy J, Singleton A, Williams NM, Heutink P, Pickering-Brown S, Morris HR, Tienari PJ and Traynor BJ (2011) A hexanucleotide repeat expansion in C9ORF72 is the cause of chromosome 9p21-linked ALS-FTD. Neuron 72:257-68. doi: 10.1016/j.neuron.2011.09.010

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Exploring Intrinsic Disorder in Stress Granules Proteome: Implications for Structure-Function Dynamics and Pathological Aggregation

Status:

Version 1

Abstract

Figures

Introduction

Materials and Methods

Collection and classification of SG proteins

Evaluation of intrinsic disorder propensity of SG proteins

Using combined CH-CDF approach to predict disorder in the entire SG proteome

Generation of protein–protein interaction network

Characterization of intrinsic disorder in the most disordered SG proteins

Results and Discussion

All proteins

RNA-Binding Proteins

Autophagic Proteins

Conclusions

Declarations

Ethics Approval:

Conflicts of Interest:

Author Contribution

Acknowledgments

Availability of Data and Materials:

References

Additional Declarations

Supplementary Files

Status:

Version 1