Discovering novel class anti-SARS-CoV-2 compounds with novel backbones is essential for preventing and controlling SARS-CoV-2 transmission, which poses a substantial threat to the health and social sustainable development of the global population because of its high pathogenicity and high transmissibility. Although the potential mutation of SARS-CoV-2 might diminish the therapeutic efficacy of drugs, 3CL Mpro is the target highly conservative in contrast with other targets. It is an essential enzyme for coronavirus replication. Based on this, this study utilized the drug discovery strategy of Knime molecular filtering framework, ROC-guided virtual screening, clustering analysis, binding mode analysis, and activity evaluation approaches to identify compound Z195914464 (IC: 7.19 μM) is a novel class inhibitor of anti-SARS-CoV-2 against the 3CL Mpro target. In addition, based on molecular dynamics simulations and MMPBSA analyses, discovered that compound Z195914464 can interact with more key residues and lower bonding energies, which explains why it exhibited more activity than the other three compounds. In summary, this study developed a method for the rapid and accurate discovery of active compounds and can also be applied in the discovery of active compounds in other targets.

The rapid spread of antibiotic-resistant strains of bacteria has created an urgent need for new alternative antibiotic agents. Membrane disrupting antimicrobial peptides (AMPs): short amino acid sequences with bactericidal and fungicidal activity that kill pathogens by permeabilizing their plasma membrane may offer a solution for this global health crisis. Magainin 2 is an AMP secreted by the African clawed frog (Xenopus laevis) that is described as a toroidal pore former membrane disrupting AMP. Magainin 2 is one of the most thoroughly studied AMPs, yet its mechanism of action is still largely hypothetical: visual evidence of the pore formation is lacking, and the molecular mechanism leading to pore formation is still debated. In the present study, quartz crystal microbalance (QCM) based viscoelastic fingerprinting analysis supported by dye leakage experiments and atomic force microscopy (AFM) imaging was used to glean deeper insights into the mechanism of action. The effect of membrane charge, acyl chain unsaturation and cholesterol concentration were also investigated. The results show lipid specific disruptive mechanism of magainin 2. QCM nano-viscometry measurements revealed the presence of distinct stages in the mechanism of magainin 2 action that, with dye leakage data, confirm the existence of an initial transient pore stage that may result in peptide flip-flop between the outer and inner membrane leaflets. There is evidence of a further mechanistic stage at high peptide concentrations that is consistent with membrane collapse into a peptide-lipid mixed phase that is distinct from the transient pore formation. The results confirm some of the earliest hypotheses about magainin 2 action, while also highlighting the membrane modulating effect of this peptide.

Quantitative characterization of protein conformational landscapes is a computationally challenging task due to their high dimensionality and inherent complexity. In this study, we systematically benchmark several widely used dimensionality reduction and clustering methods to analyze the conformational states of the Trp-Cage mini-protein, a model system with well-documented folding dynamics. Dimensionality reduction techniques, including Principal Component Analysis (PCA), Time-lagged Independent Component Analysis (TICA), and Variational Autoencoders (VAE), were employed to project the high-dimensional free energy landscape onto 2D spaces for visualization. Additionally, clustering methods such as K-means, hierarchical clustering, HDBSCAN, and Gaussian Mixture Models (GMM) were used to identify discrete conformational states directly in the high-dimensional space. Our findings reveal that density-based clustering approaches, particularly HDBSCAN, provide physically meaningful representations of free energy minima. While highlighting the strengths and limitations of each method, our study underscores that no single technique is universally optimal for capturing the complex folding pathways, emphasizing the necessity for careful selection and interpretation of computational methods in biomolecular simulations. These insights will contribute to refining the available tools for analyzing protein conformational landscapes, enabling a deeper understanding of folding mechanisms and intermediate states.

The development of small molecule drugs that target protein binders is the central goal in medicinal chemistry. During the lead compound development process, hundreds or even thousands of compounds are synthesized, with the primary focus on their binding affinity to protein targets. Typically, IC or EC values are used to rank these compounds. While thermodynamic values, such as the dissociation constant (KD), would be more informative, they are experimentally less accessible. In this study, we compare isothermal calorimetry (ITC) with surface plasmon resonance (SPR) using human STING, a key protein of innate immunity, and several cyclic dinucleotides (CDNs) that serve as its ligands. We demonstrate that SPR, with recent technological advancements, provides KDs that are sufficiently accurate for drug development purposes. To illustrate the versatility of our approach, we also used SPR to estimate the KD of poxin binding to cyclic GMP-AMP (cGAMP) that serves as a second messenger in the innate immune system. In conclusion, SPR offers a high benefit-to-cost ratio, making it an effective tool in the drug design process.

This investigation aims to immobilize peroxidase onto 3-aminopropyltriethoxysilane (APTES)-functionalized MgFeO magnetic nanoparticles to increase enzyme stability, efficiency, and recyclability. The synthesized samples were analyzed using X-ray diffraction, Fourier transform infrared spectroscopy, Thermogravimetric analysis, Vibrating sample magnetometer, and Scanning electron microscopy. The free and immobilized peroxidase were examined against different pH and temperatures as well as storage time and reuse. The immobilized peroxidase maintained 52 % of its initial activity after 10 consecutive measurements thanks to easy magnetic separation. In addition, antioxidant activity was increased by immobilizing the peroxidase to the MgFeO magnetic nanoparticles. Congo red dye removal for peroxidase immobilized MgFeO-APTES was found to be 98.6 % for 240 min. Also, it showed approximately two times more dye decolorization efficiency compared to MgFeO and APTES modified MgFeO. Finally, the immobilized peroxidase was studied by a decolorization study of congo red. So, we believe that the immobilized peroxidase on magnetic nanoparticles reported in this study may be utilized in diverse biotechnology, industrial, and environmental practices.

KIA peptides were designed as a series of cationic antimicrobial agents of different lengths, based on the repetitive motif [KIAGKIA]. As amphiphilic helices, they tend to bind initially to the surface of lipid membranes. Depending on the conditions, they are proposed to flip, insert and form toroidal pores, such that the peptides are aligned in a transmembrane orientation. Tryptophan residues are often found near the ends of transmembrane helices, anchoring them to the amphiphilic bilayer interfaces. Hence, we introduced Trp residues near one or both termini of KIA peptides with lengths of 14-24 amino acids. Our hypothesis was that if Trp residues can stabilize the transmembrane orientation, then these KIA peptides will exhibit an increased propensity to form pores, with increased membranolytic activity. Using solid-state N NMR, we found that peptides with Trp near the ends are indeed more likely to be flipped into a transmembrane orientation, especially short peptides. Short KIA peptides also exhibited higher antimicrobial activity when modified with Trp, while longer peptides showed similar activities with and without Trp. The hemolytic activity of KIA peptides of all lengths was higher with Trp near the ends. Vesicle leakage was also increased (sometimes more than 10-fold) for the Trp-mutants, especially in thicker membranes. Higher functionality of amphiphilic helices may thus be achieved in general by exploiting the anchoring effect of Trp. These results demonstrate that the incorporation of Trp increases membranolytic activities (vesicle leakage, hemolysis and antimicrobial activity), in a way compatible with a transmembrane pore model of peptide activity.

Self-assembling peptide nanoparticles (SAPN) based delivery systems, including virus-like particles (VLP), have shown great potential for becoming prominent in next-generation vaccine and drug development. The VLP can mimic properties of natural viral capsid in terms of size (20-200 nm), geometry (i.e., icosahedral structures), and the ability to generate a robust immune response (with multivalent epitopes) through activation of innate and/or adaptive immune signals. In this regard, coiled-coil (CC) domains are suitable building blocks for designing VLP because of their programmable interaction specificity, affinity, and well-established sequence-to-structure relationships. Generally, two CC domains with different oligomeric states (trimer and pentamer) are fused to form a monomeric protein through a short, flexible spacer sequence. By using combinations of symmetry axes (2-, 3- and 5- folds) that are unique to the geometry of the desired protein cage, it is possible, in principle, to assemble well-defined protein cages like VLP. In this review, we have discussed the crystallographic rules and the basic principles involved in the design of CC-based VLP. It also explored the functions of numerous noncovalent interactions in generating stable VLP structures, which play a crucial role in improving the properties of vaccine immunogenicity, drug delivery, and 3D cell culturing.

Bilirubin, a yellow bile pigment, plays an important role in the body, being a potent antioxidant and having anti-inflammatory, immunomodulatory, cytoprotective, and neuroprotective functions. This makes bilirubin promising as a therapeutic and diagnostic agent in biomedicine. However, excess bilirubin is toxic and should be removed from the body. Bilirubin exhibits photochemical activity, which has been the subject of numerous studies up to now. Such studies are relevant because the bilirubin photochemistry provides the basis for bilirubin removing in phototherapy of neonatal jaundice (neonatal hyperbilirubinemia) and for some therapeutic applications. Furthermore, it can model several elementary processes of molecular photonics. In particular, the bilirubin molecule is capable of ultrafast Z-E photoisomerization and contains two almost identical dipyrromethenone chromophores capable of exciton coupling. The present review considers the data on the photophysical and photochemical properties of bilirubin and ultrafast routes of its phototransformations, as well as its photochemical reactions in phototherapy of neonatal hyperbilirubinemia and the ways to decrease the possible adverse effects of the phototherapy. The main analytical methods of bilirubin measurement in biological systems are also viewed.

Serotonin-receptor binding plays a key role in several neurological and biological processes, including mood, sleep, hunger, cognition, learning, and memory. In this article, we performed molecular dynamics simulation to examine the key residues that play an essential role in the binding of serotonin to the G-protein-coupled 5-HT receptor (5HTR) via electrostatic interactions. Key residues for electrostatic interactions were identified via bond distance analysis and frustration analysis methods. An end-point free energy calculation method determines the stability of the 5-HTR due to serotonin binding. The single-point mutation of the polar/charged amino acid residues (Asp129, Thr134) on the binding sites and the calculation of binding free energy validate the quantitative contribution of these residues to the stability of the serotonin-receptor complex. The principal component analysis reflects that the serotonin-bound 5-HTR is more stabilized than the apo-receptor regarding dynamical changes. The difference dynamic cross-correlations map shows the correlation between the transmembranes and mini-G, which indicates that the signal transduction happens between mini-G and the receptor. Allosteric pathway analysis reveals the key nodes and key pathways for signal transduction in 5-HTR. These results provide useful insights into the study of signal transduction pathways and mutagenesis to regulate the binding and functionality of the complex. The developed protocols can be applied to study local non-covalent interactions and long-range allosteric communications in any protein-ligand system for computer-aided drug design.

Glucose isomerase is generally used in the industrial production of high-fructose corn syrup, and a heat- and acid-resistant glucose isomerase is preferred. However, most glucose isomerases exhibit low activity or inactivation at low pH. In this study, we demonstrated that two combination mutants formed by introducing positive and negative charges near the active site and on the surface of the enzyme demonstrated a successful reduction in the optimal pH and increase in the specific activity of glucose isomerase from Thermotoga maritima (TMGI). Thirteen residues, eight surface amino acids and five near the vicinity of active sites, were selected by introducing positively charged residues near the active site (mutant E237R/N298K/N337R) and negatively charged residues at the enzyme surface (mutant R112E/K220E) and were site-mutated on the basis of computational analysis. In mutants E237R/N298K/N337R and R112E/K220E, there was a decrease in the optimal pH of the glucose isomerase from 7.0 to 6.0 and 5.5, respectively, and an increase in the optimal temperature of E237R/N298K/N337R from 95 °C to 100 °C. At pH 5.5 and pH 6.0, the specific activities of R112E/K220E and E237R/N298K/N337R were 2.81 and 1.79 times greater than that of the wild-type enzyme, respectively, and their thermostabilities were greater than that of TMGI. Therefore, these two mutants (E237R/N298K/N337R and R112E/K220E) have great potential for use in the industrial production of high-fructose corn syrup. Moreover, glucose isomerase was expressed in Pichia pastoris, which demonstrated that the high expression and secretion capacity of Pichia pastoris could be used to reduce the production cost of high-fructose corn syrup.

The inhibition of the TIGIT/PVR interaction demonstrates considerable anticancer properties by enhancing the cytotoxic activity of natural killer (NK) and CD8+ T cells. However, the development of small molecule inhibitors that target TIGIT is currently limited. In this study, small molecules with the capacity to bind TIGIT and block the TIGIT/PVR interaction were screened through an advanced computational process, subsequently confirmed by blocking assays. Combined machine learning model XGBOOST and centroid-based molecular docking were employed to expeditiously exclude negative molecules, thereby reducing the chemical space. Subsequently, a blockade assay targeting the TIGIT/PVR interaction was conducted on 14 candidate molecules along with positive control, wherein compound MCULE-5547257859 exhibited the most potent inhibitory effect. Molecular dynamics simulations and binding free energy analyses revealed that compound MCULE-5547257859 possesses a thermodynamically stable conformation, indicative of a stronger binding affinity to TIGIT. In conclusion, our investigation has delineated that compound MCULE-5547257859 effectively impedes the TIGIT/PVR interaction, thereby offering a novel therapeutic modality for oncology.

Despite being mostly neglected in structural biology, the C-terminal Regions (CTRs) are studied to be multifunctional in humans as well as in viruses. Previously, SARS-CoV-2 Spike and NSP1 proteins' CTRs are observed to be disordered, and experimental evidence showed a gain of structure properties in different physiological environments. In this line, we have investigated the structural dynamics of CTR (residues 38-61) of SARS-CoV-2 ORF6 protein, disrupting bidirectional transport between the nucleus and cytoplasm. ORF6-CTR is disordered in nature but doesn't gain any structure in most conditions. As per studies, residue such as Methionine at 58th position in ORF6 is critical for interaction with Rae1-Nup98. Therefore, along with M58, we have identified a few other mutations from the literature and performed extensive structure modelling and dynamics studies using computational simulations. The exciting revelations in CTR models provide evidence of its structural flexibility and possible capabilities to perform multifunctionality inside the host.

Lipid-based nanocarriers provide versatile platforms for the encapsulation and delivery of many different bioactive compounds to improve the solubility, stability and therapeutic efficacy of bioactive phyto-compounds. In this study, liposomes were used to load leaf extract of Coffea Arabica, which is known to be rich beneficial substances such as alkaloids, flavonoids, etc. The aim of this work is to optimize the valorization of agricultural wastes containing natural antioxidants. The physico-chemical properties of the liposomes loaded with chlorogenic acid or Coffea arabica leaf extract were evaluated. The average size of empty and loaded liposomes was found to range of 120-150 nm, which is consistent with the fact that the addition of chlorogenic acid or Coffea arabica leaf extract can change the average size of the vesicles without affecting the physicochemical properties of the lipid bilayer, which remain stable systems. A structural and morphological characterization as well as an evaluation of biological properties such as viability in normal human dermal fibroblasts, is also been carried out. The cationic liposomes show a good average size and low polydispersity index values, indicating that the liposomes tend to be monodisperse and therefore stable. In particular, DOPC/DOTAP liposomes generally have better properties than DOPC/DDAB liposomes for use as encapsulation systems for natural plant extracts.

SARS-CoV-2 remains a global threat with new sublineages posing challenges, particularly in the Philippines. Hesperidin (HD) is being studied as a potential prophylactic for COVID-19. However, the virus's rapid evolution could alter how HD binds to it, affecting its effectiveness. Here, we study the mutation-induced variabilities of HD dynamics and their effects on molecular energetics in SARS-CoV-2 spike receptor complex systems. We considered eight different point mutations present in the Omicron variant. Root-mean-square deviation and binding energy analysis showed that S477N and Omicron did not eject HD throughout the simulation. Hydrogen bond distribution analysis highlighted the involvement of hydrogen bonding in mutant-HD stabilization, especially for S477N and Omicron. Root-mean-square fluctuation analysis revealed evidence of Y505H destabilization on complex systems, while distal-end loop mutations increased loop flexibility for all models bearing the three mutations. Per-residue energy decomposition demonstrated that Q493R substitution increased HD interaction. Free energy landscape and essential dynamics through principal component analysis provided insights into the conformational subspace distribution of mutant model molecular dynamics trajectories. In conclusion, significant mutations contributed to the HD interaction in different ways. S477N has shown significant binding contributions through favorable ligand interaction, while other mutations contribute via conformational modifications, increased affinity due to sidechain mutations, and increased loop flexibility.

Membrane potential is essential in biological signaling and homeostasis maintained by voltage-sensitive membrane proteins. Molecular dynamics (MD) simulations incorporating membrane potentials have been extensively used to study the structures and functions of ion channels and protein pores. They can also be beneficial in designing and characterizing artificial ion channels and pores, which will guide further amino acid sequence optimization through comparison between the predicted models and experimental data. In this study, we implemented a uniform external electric field function in the GENESIS MD simulation package to investigate the conformational dynamics of de novo-designed peptide pores. Our simulations and single-channel current recording experiments demonstrate that both charged amino acid residues in the N-terminal sequence of the peptide and the membrane potential are crucial for the structural stability and dynamics of the peptide pores. This suggests that MD simulations with an external electric field enable more accurate screening of designed proteins functioning under membrane potentials, which will ultimately contribute to a deeper understanding of voltage-sensitive membrane proteins from a bottom-up synthetic biology perspective.

With the rapid development and using of electromagnetic technology, artificial electromagnetic fields (EMFs) have become an emerging environmental factor in our daily life. Extremely-low-frequency (ELF) magnetic fields (MFs), generally generated by power lines and various electric equipment, is one of the most common EMFs in the environment which were concerned for the potential impact on human health. Base on limited evidence, ELF-MFs have been classified as possible carcinogen to human by International Agency for Research on Cancer (IARC), but the mechanisms have not been fully elucidated. Senescent cells are a group of special cells, characterized by cell cycle arrest, senescence-associated secretory phenotype (SASP), accumulation of macromolecular damage, and metabolic disturbance, play important role in fetal development, tissue aging, and even carcinogenesis. Thus, EMFs may promote carcinogenesis by affecting senescent cells, however, there are few studies. In this study, we found that exposure to 50 Hz MFs at 1.0 mT for 24 h could induce significant DNA damage in senescent but not non-senescent human fetal lung fibroblast suggested that senescent cells are more sensitive to 50 Hz MFs on DNA damage, and further results revealed that reactive oxygen species (ROS) generation mediated by mitochondrial permeability transition pore (mPTP) activation play critical role in this process. Our results indicated that cellular senescence can lead to cell sensitivity to the DNA damage effect of 50 Hz MFs, however, whether this play important role in mediating the carcinogenesis of EMFs await further study.

Parkinson's disease (PD) is a neurodegenerative disorder involving the progressive loss of dopaminergic neurons in the substantia nigra pars compacta triggered by the accumulation of amyloid aggregates of α-synuclein protein. This study investigates the potential of Salvianolic Acid B (SalB), a water-soluble polyphenol derived from Salvia miltiorrhiza Bunge, in modulating the aggregation of the A53T mutant of α-synuclein (A53T Syn). This mutation is associated with rapid aggregation and a higher rate of protofibril formation in early-onset familial PD. Computational and experimental approaches demonstrated Sal-B effectively prevents the amyloid fibrillation of A53T Syn by interacting with the N-terminal region and NAC domain. Sal-B particularly associates with the KTKEGV motif and NACore segment of A53T Syn by hydrophobic and hydrogen bonding interactions. Replica exchange molecular dynamics (REMD) simulations indicated that Sal-B reduces intramolecular hydrogen bonding and structural transitions into β-sheet rich conformations, thereby lowering the aggregation propensity of A53T Syn. Systematic analysis conducted using biophysical techniques and high-end microscopy has demonstrated significant inhibition in the amyloid transformation of A53T Syn corroborated by a 92 % decrease in ThT maxima at 100 μM Sal-B concentration and microscopic techniques validated the absence of mature fibrillar amyloids. DLS data revealed heterogeneous particle sizes, supporting the formation of smaller unstructured aggregates. These findings underscore Sal-B as a promising therapeutic candidate for PD and related synucleinopathies, warranting further investigation in cellular and animal models to advance potential treatments and early intervention strategies.

The cannabinoid receptor 1 (CB1) is an essential component of the endocannabinoid system, responsible for regulating various physiological processes such as pain, mood, and appetite. Despite increasing interest in the therapeutic potential of CB1 modulators, the precise mechanisms by which small molecules modulate receptor activity-particularly without fully transitioning between active and inactive states-remain partially understood. In this study, the complexity of CB1-ligand interactions was evaluated for the inactive CB1 state. A comprehensive pipeline, integrating ligand-based similarity search, 2D fingerprint-based reverse virtual screening and molecular dynamics (MD) simulations, identified compounds with core scaffolds commonly found in bioactive natural products, such as stilbenoids and polyphenolic compounds. Arachidin-2 (AR2) and a polyphenolic derivative were subjected to extended MD simulations, revealing their ability to stabilize the inactive CB1 state across key helices. The distinct stability differences observed in the helices HI, HIV, and HVI of the active CB1 state further highlighted ligand-specific conformational dynamics. A comparative analysis with co-crystallized synthetic ligands AM6538 and AM841 demonstrated the distinct binding behaviors of natural and synthetic ligands. AR2 showed more favorable binding to the inactive form (-22.0 kcal/mol) than to the active state. Similarly, the polyphenolic compound exhibited a greater binding difference (∼6 kcal/mol) between the inactive and active states. Notably, AM6538 and AM841 demonstrated the strongest binding (∼30 kcal/mol) to the inactive and active state, respectively. Key residues stabilizing the identified compounds in CB1-inactive state included PHE102, GLY166, PHE170, VAL196, LEU359, SER383, and CIS386. These findings underscore the utility of computational methods in the discovery and development of novel CB1 modulators for potential biomedical applications.

Bacteriocins, a class of molecules produced by bacteria, exhibit potent antimicrobial properties, including antiviral activities. The urgent need for treatments against SARS-CoV-2 has proposed bacteriocins such as enterocin DD14 (EntDD14) as potential therapeutic agents. However, the mechanism of macromolecular interaction of EntDD14 for the inhibition of SARS-CoV-2 is not yet fully understood, and its efficacy against variants like JN.1 has not been completely established. To address these knowledge gaps, biocomputational analyses were employed using a diverse set of tools, including Markov state models and volumetric analyses. This analysis revealed a favorable interaction between EntDD14 and the receptor-binding domain (RBD) of SARS-CoV-2. Furthermore, it was found that EntDD14 induces changes in the flexibility of the RBD and alters the distribution and size of its internal cavities, particularly in the JN.1 variant. These findings align with experimental observations and support the inhibitory mechanism of EntDD14 against SARS-CoV-2. Additionally, they suggest that EntDD14 may possess a broader spectrum of action, encompassing the JN.1 variant. This study paves the way for future investigations and therapeutic applications, encouraging further exploration of the antiviral activity of bacteriocins like EntDD14 against variants of concern like JN.1. However, additional experimental demonstrations are warranted to substantiate these findings.

Solid-state NMR (ssNMR) methods have continued to be developed in recent years for the efficient assignment of signals and 3D structure modeling of biomacromolecules. Consequently, we are approaching an era in which vigorous applications of these methods are more widespread in research, including functional elucidation of biomacromolecules and drug discovery. However, multidimensional ssNMR methods are not as advanced as solution NMR methods, especially for automated data analysis. This article describes how a newly developed Graph_Robot module, implemented in MagRO-NMRViewJ, evolved from integrated tools for NMR data analysis named Kujira (developed by Kobayashi et al. [1]). These packaged tools systematically utilize flexible, sophisticated, yet simple libraries that facilitate only for solution-NMR data analysis, offering an intuitive interface accessible even to novice users. In this study, semi-automated assignments of backbone and side chain signals of ssNMR datasets for uniformly C/N labeled aquaporin Z and 42-residue amyloid-β fibril were examined as examples to demonstrate how Graph_Robot can expedite the visual inspection and handling of multidimensional ssNMR spectral data. In addition, the functionality of the Graph_Robot system enables a computer to interpret the behavior of magnetization transfer based on a finite automaton model.

Filament of human hair is formed from α-keratin protein and its physical property is predominantly dominated by the structure of microfibril (also known as intermediate filaments (IF)). It is known that human hair is swollen by permanent waving (pw) treatment which consists of the reducing process and following oxidizing process, but a detail in the swelling behaviour remains still unclarified. The present work was devoted to the analysis of the swelling behaviour of hair through the structural change of IF during pw treatment, where 1.0 mol/L ammonium thioglycolate solution (pH 9.25) was employed as reducing reagent. The structure of IF was represented in terms of its microstructure, which is given by α-helix content in keratin chain, and its macrostructure, which is given by alignment of IF along human hair axis, and the structures were studied by SAXS and CP/MAS C NMR and others. It is shown that the microstructure and macrostructure of IF simultaneously start to change at an initial stage of the pw treatment without any induction period and both structures are sufficiently swollen at that stage. Furthermore, it is shown that the microstructure and macrostructure of IF is partly destructed by reducing treatment, but the destructed structures are considerably restored by following oxidizing treatment.