Protein misfolding diseases: Prospects of pharmacological treatment
Abstract
Protein misfolding has been linked to numerous inherited diseases. Loss- and gain-of-function mutations (common features of genetic diseases) may cause the destabilization of proteins, leading to alterations in their properties and/or cellular location, resulting in their incorrect functioning. Misfolded proteins can, however, be rescued via the use of proteostasis regulators and/or pharmacological chaperones, suggesting that treatments with small molecules might be developed for a range of genetic diseases. This work describes the potential of these small molecules in this respect, including for the treatment of congenital disorder of glycosylation (CDG) due to phosphomannomutase 2 deficiency (PMM2-CDG).
1 INTRODUCTION
Proteostasis requires that a delicate balance be maintained between protein synthesis, folding, trafficking and degradation. The cell is thus equipped with protein quality control systems (QCS) in which chaperone molecules, the ubiquitin proteasome pathway, and autophagy play important roles.1
In eukaryotic cells, protein folding occurs in a number of cell compartments, including the mitochondria and peroxisomes, the much more massive endoplasmic reticulum (ER, involved in the folding of membrane and secreted proteins), the cytosol, and the nucleus. Cells respond to protein folding problems following different strategies depending on the compartment in which they arise: the unfolded protein response (UPR) in the ER, the mitochondrial UPR in the mitochondria, and the heat-shock response (HSR) in the nucleus and cytosol.2, 3 These responses are made to small perturbations in protein homeostasis and play vital roles in helping misfolded proteins regain their correct conformation.4
Molecular chaperones are central elements of protein QCSs. They interact cotranslationally with nascent polypeptide chains, helping non-native proteins acquire a native conformation.5 A number of conserved families of molecular chaperones guide cytosolic proteins in efforts to prevent misfolding and aggregation. Their members are referred to as stress proteins or heat shock proteins (Hsps) and are produced in larger amounts under conditions of conformational stress. The major chaperone families are classified by molecular weight (Hsp40, Hsp60, Hsp70, Hsp90, Hsp100, and the small Hsp).6 Heat shock factor 1 (HSF1) is a key molecule in the coordination of the HSR. This transcription factor is activated during cellular stress induced by the presence of unfolded proteins, leading to the transcription of chaperones and other molecules that modulate folding.
Protein misfolding induced by missense mutations has been identified as the cause of many genetic (or in this case conformational) diseases.7-9 Loss of protein function results from early degradation, mislocalization, structural alteration or aggregation10-13 leading to pathological dysfunctions.14 When misfolded proteins cannot be properly refolded, the ubiquitin-proteasome system, autophagy and ER-associated degradation begin to degrade them.15-17
Knowledge of the proteostasis network and protein QCS is, however, paving the way for new treatment strategies for protein misfolding diseases. Influencing the proteostasis network, or directly stabilizing target proteins using proteostasis regulators (PRs) or pharmacological chaperones (PCs), offers the possibility of treating several severe diseases.12, 18
2 NEW THERAPIES FOR TREATING PROTEIN MISFOLDING DISEASES
PRs and PCs lie at the centre of a revolutionary approach for treating protein misfolding diseases. Their stabilizing effect on mutant proteins allows their correct delivery and activity. The degree of functional correction achieved depends on the absolute amount of protein rescued, its intrinsic activity, and its new stability at the appropriate functional location.
2.1 Proteostasis regulators
PRs improve proteostasis by facilitating protein folding, enhancing the degradation of non-native protein species, and minimizing misfolding. This is achieved by increasing the function and availability of molecular chaperones, and/or the activation of the protein QCS.12 Small molecule PRs, such as non-steroidal anti-inflammatory drugs,19 proteasome inhibitors,20-22 celastrol,23, 24 Hsp90 inhibitors25-27 and the benzyl pyrazole derivative Human heat shock factor protein (HSF1) activator,28 are able to modulate the HSR. Other small molecule PRs, such as rapamycine,29 inositol-lowering compounds,30 and the inhibitor of USP1431 modulate autophagy or the ubiquitin proteasome system. The capacity of the proteostasis network can be enhanced by the modulation of calcium signaling32 via the use of thapsigargin33 or curcumin,34 among others.35
The modulation of UPR and ER-associated degradation is mediated by different stress-response transmembrane proteins,36 including transcription factor 6 (ATF6), protein kinase RNA (PKR)-like ER kinase (PERK), and inositol-requiring protein 1 (IRE1). These integral membrane proteins act as ER stress transducers and respond to the accumulation of misfolded proteins in the ER lumen by leading to the activation of transcription factors that regulate the expression of UPR target genes.35
2.2 Pharmacological chaperones
PCs bind to proteins via electrostatic forces, van der Waals forces, or hydrogen bonding. They do not have the ability to refold the target protein, but they are able to shift the equilibrium toward the folded state and, therefore, are valid for pathogenic mutations that induce protein denaturation. Specific ligand-binding sites are often located at the interface between protein domains or subdomains; the corresponding ligands can therefore be particularly effective at stabilizing a protein's structure.37 PCs are protein specific, and some are mutation specific.
The most commonly used PCs directed against misfolded enzymes are based on their substrates, though these suffer from the drawback of inhibiting enzyme function when in high concentration.38 However, new, non-substrate-like and therefore non-inhibitory PCs have been used in the treatment of certain genetic disorders.39 The emerging concepts of PC therapy include both inhibitory and non-inhibitory (allosteric) strategies. The former involve competitive PCs that reversibly bind to the active site of the target enzyme via hydrogen bond networks and van der Waal's interactions. These have been used in the treatment of lysosomal diseases to help form stable protein complexes in the ER that are then transported to the lysosome. Here, these enzymes remain stable but are catalytically active for some time.40 In allosteric PC therapy, however, the chaperone is non-competitive and binds to a site other than the active site (as in methylmalonic aciduria cblB type).41 For example, tafadamis is an allosteric PC that has received market approval for transthyretin-related hereditary amyloidosis.42, 43 These PCs do not inhibit the target enzyme and thus have the greatest potential as the next generation of chaperones, either alone or in combination with PCs that target the active site.44 Enzyme cofactors provide another type of PC. An increase in the amount of the natural cofactor of an enzyme might help stabilize it.45 Such is the case of (6R)-l-erythro-5,6,7,8-tetrahydrobiopterin (BH4), the natural cofactor of phenylalanine hydroxylase (PAH), the defective enzyme in phenylketonuria (PKU).46, 47 Additionally, endogenous Hsp can be used for the stabilization of mutant proteins in some diseases by using molecular chaperone inducers.6, 48
PCs hold the promise of being able to reduce the severity of several genetic diseases, but it is important to note that the number of PCs described for different targets is low. For instance, in cystic fibrosis (CF), a number of PCs and PRs that might rescue CF transmembrane conductance regulator (CFTR) have recently been identified,34, 49-51 one of which is the PR roscovitine, which is involved in calcium regulation.52 The use of therapeutic autophagy inducers has also been tested in CF. For example, the repurposed drug cysteamine—which is Food and Drug Administration (FDA)-approved for treating cystinosis53—has been shown to restore CFTR function in a clinical trial involving CF patients.54
When a PC therapy is mutation specific it cannot, of course, be prescribed for all patients with the same genetic disease.48 However, some PCs are not mutation specific. Many mutant forms of gonadotropin-releasing hormone receptor, vasopressin type 2 receptor, and rhodopsin have been successfully rescued in cell cultures by the same PCs,55 offering hope that a single PC could be used to treat patients with different mutant forms of the target enzyme.
PRs could also be used in combination with PCs.56 PCs stabilize the pool of natively folded proteins, while PRs increase the proteostasis network capacity. Their co-administration might therefore have additive or synergistic effects.2, 18 PRs activate the UPR, which increases the pool of folded mutant proteins which PCs can then stabilize. The elucidation of the exact mechanism by which the protein QCS acts in specific diseases, might allow the design of improved treatment options for individual patients.57
3 PROTEIN MISFOLDING DISEASES
The first step before attempting any type of pharmacotherapy is to determine whether the disease under study might be responsive to this type of treatment. A candidate disease for folding therapy is one that involves destabilizing missense mutations leading to the unstable proteins formed being degraded or aggregated. The gathering of computational evidence, such as that afforded by FoldX58 or residue location analysis, together with experimental evidence regarding the oligomeric state of the mutant protein and its stability,59 are necessary steps in determining whether a missense variation alters a protein's folding. The fact that PRs act within the protein homeostasis network rather than directly on any specific protein may render them valuable against a whole group of diseases that alter a particular process in a similar fashion.45
3.1 Neurodegenerative diseases caused by misfolded functional proteins
The formation of insoluble oligomers and toxic aggregates of misfolded functional proteins can lead to a toxic gain-of-function. The accumulation of toxic aggregates in the brain is an upstream event in the pathological cascade leading to neuronal dysfunction, cell death, and eventually neurodegenerative diseases such as Alzheimer, Parkinson, Huntington and Creutzfeldt-Jakob disease.60, 61 Amyloid plaques and neurofibrillary tangles are the hallmarks of Alzheimer disease, and both arise from protein misfolding62; the amyloid-β (Aβ) peptide and tau protein suffer conformational changes that produce toxic aggregates.63 In Parkinson disease, the death of dopaminergic neurons and protein aggregation (Lewy bodies) is present.64 The use of PCs/PRs that prevent or reduce neurodegeneration in Parkinson disease models and related synucleinopathies have been described. Those most commonly used in Parkinson disease are geldanamycin, celastrol, trehalose and 4-phenylbutyrate. The most frequently used PRs in synucleinopathies are HSP90 inhibitors (geldanamycin, 17-allyamino-17-demethoxy-geldanamycin (17-AAG), and SNX compounds), and enhancers of HSF-1 (carbenoxolone). Chemical chaperones such as trehalose and mannitol may also have a clinical future.42, 65
3.2 Dominant diseases caused by misfolded structural proteins
Misfolded structural proteins that are not degraded may also form toxic aggregates in the cell. Monogenic conformational diseases with a dominant inheritance pattern include familial cardiomyopathies,66 collagen disorders67 and keratin disorders.68 In epidermolysis bullosa simplex, an inherited connective tissue disorder, mutant forms of the keratin proteins KRT5 and KRT14 lead to severe blistering of the skin in response to injury.14 Disease-associated mutations in the keratin gene cause the misfolding and aggregation of the protein, particularly in response to mechanical stress.69, 70 Depending on the nature and position of the causal mutation, these diseases may be mild or severe. Inherited Li-Fraumeni syndrome and some early onset cancers involving p53 mutations71 may be added to this negative dominant type of protein misfolding diseases.72
3.3 Recessive metabolic genetic diseases
The term “inborn errors of metabolism” (IEM) covers a large group of conditions affecting the biosynthesis or breakdown of biomolecules involved in specific pathways, recognizable by specific biochemical tests. Recently, it has become necessary to redefine IEMs as they have been found not only to involve metabolic pathways but also cellular processes.73 Most IEMs remain without effective treatment. However, in recent years, the therapeutic use of small molecules has emerged as a promising possibility in the treatment of protein misfolding disease—the mechanism behind many autosomal recessive metabolic disorders. The resulting pathologies may involve misfolded proteins in the cytosol, for example, PKU,74 mitochondrial methylmalonic aciduria cblB type,13 or organelles, for example, lysosomal storage disorders (LSDs). The degree of the misfolding determines the severity and outcome of the associated disease; a range of phenotypes are therefore usually observed.12, 72
3.4 Conformational disorders in the cytosol compartment
PKU, the first well-known misfolding disease,38 is caused by mutations affecting the PAH enzyme, which is responsible for metabolizing the essential amino acid phenylalanine. Most disease-causing mutations in the PAH gene produce misfolded proteins that are rapidly degraded. A considerable subset of patients respond to treatment with the protein's natural cofactor BH4 which, in certain mutations, can act as a PC.46, 75 The approval of sapropterin dihydrochloride as an orphan drug for BH4-responsive PAH deficiency,76-78 which in effect saw an oral pharmacological treatment substitute or combine with a traditional dietary treatment, marked a paradigm shift in the treatment of this disease. Nowadays, approximately half of all PKU patients benefit from this therapy. Efforts are being focused on the discovery of cofactor analogues with improved pharmacokinetic properties, as well as non-substrate-like analogues.79 Candidate PCs have also been identified through shape-focused virtual screening.80
Classical homocystinuria is caused by mutations affecting the cystathionine beta-synthase (CBS) protein. As 85% of these are of the missense type, it has been postulated that protein misfolding plays an important role in the pathogenesis of CBS deficiency.81 An important fraction of patients responds to pharmacological doses of pyridoxine, increasing hepatic CBS activity.82 Heme arginate also increases CBS residual activity and promotes proper enzyme assembly in vitro; although the clinical use of heme arginate may be limited by its cost and potential side effects.81 Recently, the first evidence has been reported for the use of S-adenosylmethionine as a kinetic stabilizer for CBS,83, 84 along with different molecules acting as chemical chaperones, such as betaine and taurine,85 Dimethyl sulfoxide (DMSO), glycerol, proline and Trimethylamine N-oxide (TMAO).86 PRs have also been used to rescue destabilizing CBS mutants; the proteasome inhibitors, bortezomib and ONX0912 have both been proven effective in an animal model of homocystinuria.87 In primary hyperoxaluria type 1, a disorder of glyoxylate metabolism, betaine and pyridoxine have been suggested as PCs.88
The authors' group is investigating a range of molecular therapies for congenital disorder of glycosylation (CDG) due to phosphomannomutase 2 deficiency (PMM2-CDG or CDG type Ia).59, 89 PMM2 is a homodimeric enzyme that catalyzes the conversion of mannose-1-phosphate into mannose-6-phosphate.90 This metabolic disease involves a defect in protein glycosylation, for which no effective treatment is available. PMM2-CDG is the most common form of CDG.
Loss-of-function mutations in patients with PMM2-CDG involve the increased susceptibility of PMM2 to degradation and/or aggregation. In certain cases, however, including the mutations characterized by our group, that is, p.L32R, p.V44A, p.D65Y, p.P113L, p.R123Q, p.R141H, p.F157S, p.R162W, p.F183S, p.P184T, p.F207S, p.T237M and p.C241S,91, 92 PMM2 activity might be rescuable via the use of synergetic PRs and/or PCs. The fact that 80% of PMM2 mutations are missense (The Human Gene Mutation Database (HGMD) professional release), and that most have been identified in a compound heterozygous state—with 1 severe null mutation and 1 destabilizing mutation allowing PMM2 to retain some residual activity89, 93—suggests that PMM2-CDG is a conformational disease59 that PCs might be able to alleviate in many patients. In our work, 8 possible PCs were selected from a 10 000 compound library screening. The compound 1-(3-chlorophenyl)-3-3-bis(pyridine-2-yl)urea (compound VIII) stood out, based on its pharmacochemical properties, its lack of an inhibitory effect on PMM2 activity, and its positive effect on the activity and stability of a number of destabilizing PMM2 mutations. Together, the results provided the first proof-of-concept that PCs can be used to treat PMM2-CDG, plus a basis for developing a new therapy based on the chemical optimization of compound VIII (Figure 1).89 We are currently exploring the effect of PRs on the activity and amount of protein produced by different PMM2-CDG mutants. The next step will be to study whether the PC compound VIII and PRs have a synergistic effect.
3.5 Mitochondrial conformational disorders
A PC for the treatment of methylmalonic aciduria cblB type was also identified by our group via the high-throughput ligand screening of more than 2000 compounds. One of these compounds, N{[(4-chlorophenyl)carbamothioyl]amino}-2-phenylacetamide, was selected based on the increase it afforded in the stability of purified ATP:cob(I)alamin adenosyltransferase (ATR). It also enhanced ATR activity in fibroblasts from patient with a destabilizing hemizygous I96T mutation. Cobalamin, when present, improved its effect. An increase in steady-state levels of the ATR protein was also observed in both the liver and brain of mice after 12 days of oral administration of the compound.41 Our group has now characterized a number of destabilizing mutations that respond to treatment with this compound (data not published).
It is also reported that, in maple syrup urine disease (MSUD)94 and pyruvate dehydrogenase (PDH) deficiency, the small molecule phenylbutyrate acts by increasing the residual activity of enzymes via the inhibition of their inactivating enzymes.94, 95
Certainly, protein misfolding is a hallmark of fatty acid oxidation defects, and small molecules able to increase the functional level of the affected protein are valuable candidates in drug design. Mitochondrial cofactors and metabolites have been identified as potential stabilizers of 2 β-oxidation acyl-CoA dehydrogenases—short chain acyl-CoA dehydrogenase and the medium chain acyl-CoA dehydrogenase—as well as of glutaryl-CoA dehydrogenase, an enzyme involved in lysine and tryptophan metabolism. It is also reported that physiological concentrations of FAD result in a spectacular improvement in the thermal stabilities of these enzymes, and preventing their loss of activity.96 Riboflavin (vitamin B2) has also been reported a potential therapeutic agent for disorders affecting mitochondrial energy metabolism.97
3.6 Lysosomal storage disorders
Most LSDs are caused by the malfunction of one of the lysosome hydrolases responsible for the catabolism of glycogen, peptides, glycoproteins, mucopolysaccharides, oligosaccharides or cholesterol, leading to pathological substrate accumulation.98 LSDs represent a pharmacological therapy success story; those now treated with PCs include Fabry, Pompe and Gaucher disease, GM1-gangliosidosis (Tay-Sachs disease), GM2-gangliosidosis (Sandhoff disease), Krabbe disease, Batten disease, and Sanfilippo syndrome type C. Most of the PCs used are substrate analogues.38 The breakthrough in this kind of therapy was first reported in Fabry disease, for which a small-molecule inhibitor was found that selectively binds to and stabilizes the target protein.99 The first substrate inhibitor that received market approval was Miglustat—for Gaucher disease type I in 2002,100 and for Niemann-Pick disease type C in 2009.101 Miglustat, a glucosylceramide synthase inhibitor, prevents the formation of the substrate that would normally build up in the presence of an enzyme deficient in its breakdown. It is used to treat mild-to-moderate type I Gaucher disease when enzyme replacement therapy does not return the hoped-for results. In the first clinical trial with Miglustat, a significant reduction in disease biomarkers was reported.102 Another substrate inhibitor, Eliglustat was granted marketing authorization in 2015. This is also a glucosylceramide synthase inhibitor, but more specific and potent than Miglustat, and represents an emerging alternative to enzyme replacement therapy for the long-term treatment of adults with Gaucher disease type I.103, 104 It is important to note that Miglustat has poor central nervous system penetration. New substrate inhibitors able to cross the brain blood barrier have, however, been developed and have returned promising results in Gaucher and Fabry disease.105 As it is possible that all diseases caused by a deficiency in 1 enzyme involved in the same degradation pathway might be responsive to the same substrate inhibitor, Miglustat has also been tested in Tay-Sachs106 and Sandhoff107 patients. However, the results obtained have been controversial.
Additionally, the proteasome inhibitor bortezomib has been described to significantly improve the activity of Niemann-Pick type C mutant proteins caused by specific missense variants,108 and in Gaucher disease the proteasome inhibitor MG132, together with the heat-shock transcription factor 1 activator celastrol, have been shown to enhance the folding, trafficking and activity of misfolded glucosylceramidase proteins caused by different missense mutations.18 In Gaucher disease, calcium regulation with lacidipine is also being investigated with promising results,109 and TRMPL1 ligands are showing promise in mucolipidosis type IV.110 A curcumin derivative, BCM95, along with an autophagy inducer, hydroxypropyl-β-cyclodextrin, may also be effective in the treatment of saposin C deficiency via the improvement of lysosome function.111 Finally, autophagy inducers have been proven of potential use via the treatment of cystinosis with cysteamine.112
In total, around 40 PCs have been described as effective in the treatment of different IEMs, most of them for LSDs,12, 45 but of these only few are used clinically or are under pharmaceutical development (Table 1).
| Disease | Target | PC | Company/institution | References |
|---|---|---|---|---|
| Batten disease | Palmitoyl:Protein thioesterase | CS38 | The University of Chicago | 119 |
| Cystic fibrosis | Cystic fibrosis transmembrane conductance regulator | Lumacaftor-ivacaftor | Vertex Pharmaceuticals | 120 |
| Cystic fibrosis transmembrane conductance regulator | Cysteamine | NovaBiotics | 54 | |
| Cystinosis | Cystine | Cysteamine | Horizon Pharma | 112 |
| Fabry disease | α-galactosidase A | Migalastat hydrochloride | Amicus Therapeutics | 121-123 |
| Gaucher disease | β-glucosidase | Isofagomine | Amicus Therapeutics | 124 |
| Glucosylceramide | Miglustat | Actelion | 104 | |
| Glucosylceramide | Eliglustat | Sanofi-Genzyme | 104 | |
| GM1 gangliosidosis | Acid β-galactosidase | N-Octyl 4-Epi-beta-valienamine | Seikagaku Corporation | 125 |
| GM2 gangliosidosis | Acid β-hexosaminidase | Pyrimethamine | The Hospital for Sick Children | 126 |
| Krabbe disease | Galactosylceramidase | Lobeline | Mayo Foundation for Medical Education and Research | 127 |
| Niemann-Pick type C | Glucosylceramide | Miglustat | Actelion | 101 |
| Phenylketonuria | Phenylalanine hydroxylase | Sapropterin dihydrochloride | BioMarin; Merck Serono | 75 |
| Pompe disease | Acid α-glucosidase | 1-deoxynojirimycin | Amicus Therapeutics | 128 |
| Sanfilippo syndrome type C | Heparan sulfate acetyl-CoA | Glucosamine | The University of Montreal | 129 |
| Transthyretin type familial amyloid polyneuropathy | Transthyretin | Tafamidis meglumine | Pfizer | 42 |
4 PRs AND PCs IN THE CLINIC
The first step in developing pharmacotherapies involves the identification of potential therapeutic molecules. Two strategies can be followed: high-throughput screenings (HTSs), exploring libraries of thousands of existing chemicals and drugs, or hypothesis-driven searching. By monitoring thermal protein stability using differential scanning fluorimetry, HTS are particularly useful for identifying novel chemical structures.79 Hypothesis-driven strategies are usually based on screening for low molecular weight compounds structurally related to key moieties of the natural substrate, its cofactors or inhibitors.12, 113
The development of PRs or PCs by virtual screening is based on computational investigations. This has the major advantage of allowing large numbers of compounds to be investigated at the same time, increasing the probability of finding hits and improving the likelihood of being able to select allosteric binding sites. One of the most used virtual screening techniques is molecular docking. Another strategy is the development of a pharmacophore model—an ensemble of steric and electronic features that ensure optimal supramolecular interactions with a specific biological target.79 Multicomponent drug discovery can also be undertaken using computational tools to perform repositioning analysis. In silico predictions have been shown powerful enough to lead to new and efficient potential therapies, for example, for amyotrophic lateral sclerosis.114
When a hit is selected, it needs to pass a variety of experimental and computational filters to ensure its viability as a therapeutic agent. Once preclinical assays succeed, the most promising candidate for a given disease is selected, and is then used in animal experiments prior to entering clinical trials with human patients. The last stage of the drug discovery process is the derivatization of ligands. Lead optimization improves some of the properties of molecules, such as lipophilicity, synthetic accessibility, absorption, distribution, metabolism, toxicity and excretion.115 It should be noted that many hit compounds do not make it to the clinical stage of development as they are, in fact, artefacts, that is, their activity does not depend on a specific, drug-like interaction with the target protein. Such molecules are termed pan-assay interference compounds (PAINS).116 To avoid the problems caused by PAINS, the selection and characterization methods followed during the first steps of drug discovery need to be improved.117 Table 1 shows the diseases for which molecular therapies are under clinical trial in humans or are already on the market.
5 PERSPECTIVES
The discovery of PCs can be accelerated by contemplating drug repositioning. Several small molecule drugs have been successfully repositioned as PCs for rare disorders, including doxorubicin, an anti-neoplastic anthracycline now used in CF; diltiazem, an antihypertensive now used in Gaucher disease; ambroxol, a mucolytic agent now used for Gaucher and Fabry disease; acetylcysteine, another mucolytic agent now used in Pompe disease; pyrimethamine, an anti-parasitic compound now used in GM2 gangliosidosis; carbamazepine, a dibenzazepine now used in hyperinsulinaemic hypoglycaemia; and salycylate which has found a use in the treatment of Pendred syndrome.118
Combining PCs and PRs will likely prove the most effective treatments for protein misfolding diseases, 1 molecule acting as a stress-responsive signalling pathway activator, the other binding to and stabilizing the misfolded protein.56
ACKNOWLEDGEMENTS
This work was funded by the Fondo Investigación Sanitaria, grants PI13/01239 and PI16/00573; the Fundación Isabel Gemio, an institutional grant from the Fundación Ramón Areces and by the European Regional Development Fund. The authors thank all the PMM2-CDG families involved in the research undertaken at our laboratory.
Conflict of interest
The authors have declared no conflicts of interest.
