Insight into the protein solubility driving forces with neural attention

Protein solubility is a key aspect for many biotechnological, biomedical and industrial processes, such as the production of active proteins and antibodies. In addition, understanding the molecular determinants of the solubility of proteins may be crucial to shed light on the molecular mechanisms of diseases caused by aggregation processes such as amyloidosis. Here we present SKADE, a novel Neural Network protein solubility predictor and we show how it can provide novel insight into the protein solubility mechanisms, thanks to its neural attention architecture. First, we show that SKADE positively compares with state of the art tools while using just the protein sequence as input. Then, thanks to the neural attention mechanism, we use SKADE to investigate the patterns learned during training and we analyse its decision process. We use this peculiarity to show that, while the attention profiles do not correlate with obvious sequence aspects such as biophysical properties of the aminoacids, they suggest that N- and C-termini are the most relevant regions for solubility prediction and are predictive for complex emergent properties such as aggregation-prone regions involved in beta-amyloidosis and contact density. Moreover, SKADE is able to identify mutations that increase or decrease the overall solubility of the protein, allowing it to be used to perform large scale in-silico mutagenesis of proteins in order to maximize their solubility.

The solubility of proteins is a crucial biophysical aspect when it comes to understanding many human diseases and to improve the industrial processes for protein production. Due to its relevance, computational methods have been devised in order to study and possibly optimize the solubility of proteins. In this work we apply a deep-learning technique, called neural attention to predict protein solubility while “opening” the model itself to interpretability, even though Machine Learning models are usually considered black boxes. Thank to the attention mechanism, we show that i) our model implicitly learns complex patterns related to emergent, protein folding-related, aspects such as to recognize β-amyloidosis regions and that ii) the N-and C-termini are the regions with the highes signal fro solubility prediction. When it comes to enhancing the solubility of proteins, we, for the first time, propose to investigate the synergistic effects of tandem mutations instead of “single” mutations, suggesting that this could minimize the number of required proposed mutations.

When a common biological role does not imply common disease outcomes: Disparate pathology linked to human mitochondrial aminoacyl-tRNA synthetases

Mitochondrial aminoacyl-tRNA synthetases (mt-aaRSs) are essential components of the mitochondrial translation machinery. The correlation of mitochondrial disorders with mutations in these enzymes has raised the interest of the scientific community over the past several years. Most surprising has been the wide-ranging presentation of clinical manifestations in patients with mt-aaRS mutations, despite the enzymes' common biochemical role. Even among cases where a common physiological system is affected, phenotypes, severity, and age of onset varies depending on which mt-aaRS is mutated. Here, we review work done thus far and propose a categorization of diseases based on tissue specificity that highlights emerging patterns. We further discuss multiple in vitro and in cellulo efforts to characterize the behavior of WT and mutant mt-aaRSs that have shaped hypotheses about the molecular causes of these pathologies. Much remains to do in order to complete our understanding of these proteins. We expect that futher work is likely to result in the discovery of new roles for the mt-aaRSs in addition to their fundamental function in mitochondrial translation, informing the development of treatment strategies and diagnoses.

Role of N-terminal region of Escherichia coli maltodextrin glucosidase in folding and function of the protein

Maltodextrin glucosidase (MalZ) hydrolyses short malto-oligosaccharides from the reducing end releasing glucose and maltose in Escherichia coli. MalZ is a highly aggregation prone protein and molecular chaperonins GroEL and GroES assist in the folding of this protein to a substantial level. The N-terminal region of this enzyme appears to be a unique domain as seen in sequence comparison studies with other amylases as well as through homology modelling. The sequence and homology model analysis show a probability of disorder in the N-Terminal region of MalZ. The crystal structure of this enzyme has been reported in the present communication. Based on the crystallographic structure, it has been interpreted that the N-terminal region of the enzyme (Met1-Phe131) might be unstructured or flexible. To understand the role of the N-terminal region of MalZ in its enzymatic activity, and overall stability, a truncated version (Ala111-His616) of MalZ was created. The truncated version failed to fold into an active enzyme both in E. coli cytosol and in vitro even with the assistance of chaperonins GroEL and GroES. Furthermore, the refolding effort of N-truncated MalZ in the presence of isolated N-terminal domain didn't succeed. Our studies suggest that while the structural rigidity or orientation of the N-terminal region of the MalZ protein may not be essential for its stability and function, but the said domain is likely to play an important role in the formation of the native structure of the protein when present as an integral part of the protein.

Neurodegenerative disease-associated mutants of a human mitochondrial aminoacyl-tRNA synthetase present individual molecular signatures

Mutations in human mitochondrial aminoacyl-tRNA synthetases are associated with a variety of neurodegenerative disorders. The effects of these mutations on the structure and function of the enzymes remain to be established. Here, we investigate six mutants of the aspartyl-tRNA synthetase correlated with leukoencephalopathies. Our integrated strategy, combining an ensemble of biochemical and biophysical approaches, reveals that mutants are diversely affected with respect to their solubility in cellular extracts and stability in solution, but not in architecture. Mutations with mild effects on solubility occur in patients as allelic combinations whereas those with strong effects on solubility or on aminoacylation are necessarily associated with a partially functional allele. The fact that all mutations show individual molecular and cellular signatures and affect amino acids only conserved in mammals, points towards an alternative function besides aminoacylation.

Pathogenic mutations causing LBSL affect mitochondrial aspartyl-tRNA synthetase in diverse ways

The autosomal recessive white matter disorder LBSL (leukoencephalopathy with brain stem and spinal cord involvement and lactate elevation) is caused by mutations in DARS2, coding for mtAspRS (mitochondrial aspartyl-tRNA synthetase). Generally, patients are compound heterozygous for mutations in DARS2. Many different mutations have been identified in patients, including several missense mutations. In the present study, we have examined the effects of missense mutations found in LBSL patients on the expression, enzyme activity, localization and dimerization of mtAspRS, which is important for understanding the cellular defect underlying the pathogenesis of the disease. Nine different missense mutations were analysed and were shown to have various effects on mtAspRS properties. Several mutations have a direct effect on the catalytic activity of the enzyme; others have an effect on protein expression or dimerization. Most mutations have a clear impact on at least one of the properties of mtAspRS studied, probably resulting in a small contribution of the missense variants to the mitochondrial aspartylation activity in the cell.

Pathogenic implications of human mitochondrial aminoacyl-tRNA synthetases

Mitochondria are considered as the powerhouse of eukaryotic cells. They host several central metabolic processes fueling the oxidative phosphorylation pathway (OXPHOS) that produces ATP from its precursors ADP and inorganic phosphate Pi (PPi). The respiratory chain complexes responsible for the OXPHOS pathway are formed from complementary sets of protein subunits encoded by the nuclear genome and the mitochondrial genome, respectively. The expression of the mitochondrial genome requires a specific and fully active translation machinery from which aminoacyl-tRNA synthetases (aaRSs) are key actors. Whilst the macromolecules involved in mammalian mitochondrial translation have been under investigation for many years, there has been an explosion of interest in human mitochondrial aaRSs (mt-aaRSs) since the discovery of a large (and growing) number of mutations in these genes that are linked to a variety of neurodegenerative disorders. Herein we will review the present knowledge on mt-aaRSs in terms of their biogenesis, their connection to mitochondrial respiration, i.e., the respiratory chain (RC) complexes, and to the mitochondrial translation machinery. The pathology-related mutations detected so far are described, with special attention given to their impact on mt-aaRSs biogenesis, functioning, and/or subsequent activities. The collected data to date shed light on the diverse routes that are linking primary molecular possible impact of a mutation to its phenotypic expression. It is envisioned that a variety of mechanisms, inside and outside the translation machinery, would play a role on the heterogeneous manifestations of mitochondrial disorders.

Thermodynamic properties distinguish human mitochondrial aspartyl-tRNA synthetase from bacterial homolog with same 3D architecture

In the mammalian mitochondrial translation apparatus, the proteins and their partner RNAs are coded by two genomes. The proteins are nuclear-encoded and resemble their homologs, whereas the RNAs coming from the rapidly evolving mitochondrial genome have lost critical structural information. This raises the question of molecular adaptation of these proteins to their peculiar partner RNAs. The crystal structure of the homodimeric bacterial-type human mitochondrial aspartyl-tRNA synthetase (DRS) confirmed a 3D architecture close to that of Escherichia coli DRS. However, the mitochondrial enzyme distinguishes by an enlarged catalytic groove, a more electropositive surface potential and an alternate interaction network at the subunits interface. It also presented a thermal stability reduced by as much as 12°C. Isothermal titration calorimetry analyses revealed that the affinity of the mitochondrial enzyme for cognate and non-cognate tRNAs is one order of magnitude higher, but with different enthalpy and entropy contributions. They further indicated that both enzymes bind an adenylate analog by a cooperative allosteric mechanism with different thermodynamic contributions. The larger flexibility of the mitochondrial synthetase with respect to the bacterial enzyme, in combination with a preserved architecture, may represent an evolutionary process, allowing nuclear-encoded proteins to cooperate with degenerated organelle RNAs.