Machine Learning Meets Cancer.

The role of machine learning (a part of artificial intelligence-AI) in the diagnosis and treatment of various types of oncology is steadily increasing. It is expected that the use of AI in oncology will speed up both diagnostic and treatment planning processes. This review describes recent applications of machine learning in oncology, including medical image analysis, treatment planning, patient survival prognosis, and the synthesis of drugs at the point of care. The fast and reliable analysis of medical images is of great importance in the case of fast-flowing forms of cancer. The introduction of ML for the analysis of constantly growing volumes of big data makes it possible to improve the quality of prescribed treatment and patient care. Thus, ML is expected to become an essential technology for medical specialists. The ML model has already improved prognostic prediction for patients compared to traditional staging algorithms. The direct synthesis of the necessary medical substances (small molecule mixtures) at the point of care could also seriously benefit from the application of ML. We further review the main trends in the use of artificial intelligence-based technologies in modern oncology. This review demonstrates the future prospects of using ML tools to make progress in cancer research, as well as in other areas of medicine. Despite growing interest in the use of modern computer technologies in medical practice, a number of unresolved ethical and legal problems remain. In this review, we also discuss the most relevant issues among them.

Alkylresorcinols as New Modulators of the Metabolic Activity of the Gut Microbiota.

Alkylresorcinols (ARs) are polyphenolic compounds with a wide spectrum of biological activities and are potentially involved in the regulation of host metabolism. The present study aims to establish whether ARs can be produced by the human gut microbiota and to evaluate alterations in content in stool samples as well as metabolic activity of the gut microbiota of C57BL, db/db, and LDLR (-/-) mice according to diet specifications and olivetol (5-n-pentylresorcinol) supplementation to estimate the regulatory potential of ARs. Gas chromatography with mass spectrometric detection was used to quantitatively analyse AR levels in mouse stool samples; faecal microbiota transplantation (FMT) from human donors to germ-free mice was performed to determine whether the intestinal microbiota could produce AR molecules; metagenome sequencing analysis of the mouse gut microbiota followed by reconstruction of its metabolic activity was performed to investigate olivetol's regulatory potential. A significant increase in the amounts of individual members of AR homologues in stool samples was revealed 14 days after FMT. Supplementation of 5-n-Pentylresorcinol to a regular diet influences the amounts of several ARs in the stool of C57BL/6 and LDLR (-/-) but not db/db mice, and caused a significant change in the predicted metabolic activity of the intestinal microbiota of C57BL/6 and LDLR (-/-) but not db/db mice. For the first time, we have shown that several ARs can be produced by the intestinal microbiota. Taking into account the dependence of AR levels in the gut on olivetol supplementation and microbiota metabolic activity, AR can be assumed to be potential quorum-sensing molecules, which also influence gut microbiota composition and host metabolism.

Elevation of truncated (48 kDa) form of unconventional myosin 1C in blood serum correlates with severe Covid-19.

In Covid-19 and autoimmune patients, there are several similarities revealed in the immune responses (Liu et al., 2021; Woodruff et al., 2020). Earlier, we firstly detected a truncated (48 kDa) form of the unconventional Myosin 1C (48/Myo1C) in a fraction of proteins soluble in 10% 2,2,2-trichloroacetic acid (TCA). These proteins were obtained from blood serum of patients with autoimmune diseases, such as multiple sclerosis, systemic lupus erythematosus, and rheumatoid arthritis (Kit et al., 2018). Here, we demonstrated that content of 48/Myo1C was also elevated in blood serum of the severe Covid-19 patients. Whereas in blood of 28 clinically healthy human individuals regularly tested for Covid-19 infection, the amount of this protein was undetectable or very low, in blood of 16 of 28 patients hospitalized with severe course of this disease, its amount was significantly increased. Dexamethasone, steroid hormone which is widely used for treatment of severe Covid-19 patients, induced time-dependent elevation of the 48/Myo1C in blood of such patients. The 48/Myo1C dose-dependently suppressed the viability of anti-CD3-activated lymphocytes of human peripheral blood. Recently, we used affinity chromatography on the magnetic poly(glycidyl-methacrylate) (mag-PGMA-NH2) microparticles functionalized with Myo1C and MALDI-TOF mass spectrometry with molecular modeling in silico in order to identify potential molecular partners of the 48/Myo1C. It was found that 48/Myo1C might bind to component 3 of the complement system and the anti-thrombin-III (Starykovych et al., 2021). Thus, the mechanisms of the pathogenic action of truncated form of Myo1C in severe COVID-19 patients may involve a suppression of the immune cells, as well as modulation of complement and coagulation cascades.

Signaling pathways and therapeutic approaches in glioblastoma multiforme (Review).

Glioblastoma multiforme (GBM) is the most aggressive type of primary brain tumor and is associated with a poor clinical prognosis. Despite the progress in the understanding of the molecular and genetic changes that promote tumorigenesis, effective treatment options are limited. The present review intended to identify and summarize major signaling pathways and genetic abnormalities involved in the pathogenesis of GBM, as well as therapies that target these pathways. Glioblastoma remains a difficult to treat tumor; however, in the last two decades, significant improvements in the understanding of GBM biology have enabled advances in available therapeutics. Significant genomic events and signaling pathway disruptions (NF­κB, Wnt, PI3K/AKT/mTOR) involved in the formation of GBM were discussed. Current therapeutic options may only marginally prolong survival and the current standard of therapy cures only a small fraction of patients. As a result, there is an unmet requirement for further study into the processes of glioblastoma pathogenesis and the discovery of novel therapeutic targets in novel signaling pathways implicated in the evolution of glioblastoma.

Synthesis, Characterization and Biodistribution of GdF3:Tb3+@RB Nanocomposites.

Herein we report the development of a nanocomposite for X-ray-induced photodynamic therapy (X-PDT) and computed tomography (CT) based on PEG-capped GdF3:Tb3+ scintillating nanoparticles conjugated with Rose Bengal photosensitizer via electrostatic interactions. Scintillating GdF3:Tb3+ nanoparticles were synthesized by a facile and cost-effective wet chemical precipitation method. All synthesized nanoparticles had an elongated "spindle-like" clustered morphology with an orthorhombic structure. The structure, particle size, and morphology were determined by transmission electron microscopy (TEM), X-ray diffraction (XRD), and dynamic light scattering (DLS) analysis. The presence of a polyethylene glycol (PEG) coating and Rose Bengal conjugates was proved by Fourier-transform infrared spectroscopy (FTIR), thermogravimetric analysis (TG), and ultraviolet-visible (UV-vis) analysis. Upon X-ray irradiation of the colloidal PEG-capped GdF3:Tb3+-Rose Bengal nanocomposite solution, an efficient fluorescent resonant energy transfer between scintillating nanoparticles and Rose Bengal was detected. The biodistribution of the synthesized nanoparticles in mice after intravenous administration was studied by in vivo CT imaging.

Prognostic role and biologic features of Musashi-2 expression in colon polyps and during colorectal cancer progression.

BACKGROUND: The RNA-binding protein Musashi-2 (MSI2) controls the translation of proteins that support stem cell identity and lineage determination and is associated with progression in some cancers. We assessed MSI2 as potential clinical biomarker in colorectal cancer (CRC) and tubulovillous adenoma (TA) of colon mucosa. METHODS: We assessed 125 patients, of whom 20 had polyps of the colon (TAs), and 105 had CRC. Among 105 patients with CRC, 45 had stages I-III; among metastatic CRC (mCRC) patients, 31 had synchronous and 29 metachronous liver metastases. We used immunohistochemistry to measure MSI2 expression in matching specimens of normal tissue versus TAs, primary CRC tumors, and metastases, correlating expression to clinical outcomes. We analyzed the biological effects of depleting MSI2 expression in human CRC cells. RESULTS: MSI2 expression was significantly elevated in polyps versus primary tissue, and further significantly elevated in primary tumors and metastases. MSI2 expression correlated with decreased progression free survival (PFS) and overall survival (OS), higher tumor grade, and right-side localization (p = 0.004) of tumors. In metastases, high MSI2 expression correlated with E-cadherin expression. Knockdown of MSI2 in CRC cells suppressed proliferation, survival and clonogenic capacity, and decreased expression of TGFß1, E-cadherin, and ZO1. CONCLUSION: Elevated expression of MSI2 is associated with pre-cancerous TAs in the colonic mucosa, suggesting it is an early event in transformation. MSI2 expression is further elevated during CRC progression, and associated with poor prognosis. Depletion of MSI2 reduces CRC cell growth. These data imply a causative role of MSI2 overexpression at multiple stages of CRC formation and progression.

Musashi 2 (MSI2) expression as an independent prognostic biomarker in non-small cell lung cancer (NSCLC).

BACKGROUND: Musashi-2 (MSI2) is a member of RNA-binding protein family that regulates mRNA translation of numerous intracellular targets and influences maintenance of stem cell identity. This study assessed MSI2 as a potential clinical biomarker in non-small cell lung cancer (NSCLC). METHODS: The current study included 40 patients with NSCLC, of whom one presented with stage 1, 14 presented with stage II, 15 presented with stage III, and 10 patients had stage IV. All patients received standard of care treatments. All patient samples were obtained before treatment started. We used immunohistochemical (IHC) approach to measure MSI2 protein expression in matching specimens of normal lung versus tumor tissues, and primary versus metastatic tumors, followed by correlative analysis in relation to clinical outcomes. In parallel, clinical correlative analysis of MSI2 mRNA expression was performed in silico using publicly available datasets (TCGA/ICGC and KM plots). RESULTS: MSI2 protein expression in patient samples was significantly elevated in NSCLC primary tumors versus normal lung tissue (P=0.03). MSI2 elevated expression positively correlated with a decreased progression free survival (PFS) (P=0.026) combined for all stages and with overall survival (OS) at stage IV (P=0.013). Elevated MSI2 expression on RNA level was confirmed in primary tumor versus normal tissue samples in TCGA dataset (P<0.0001), and positively correlated with decreased OS (P=0.02). No correlation was observed between MSI2 expression and age, sex, smoking, and treatment type. CONCLUSIONS: Elevated MSI2 expression in primary NSCLC tumors is associated with poor prognosis and can be used as a novel potential prognostic biomarker in NSCLC patients. Future studies in an extended patient cohort are warranted.

Musashi-2 (MSI2) regulates epidermal growth factor receptor (EGFR) expression and response to EGFR inhibitors in EGFR-mutated non-small cell lung cancer (NSCLC).

Non-small cell lung cancer (NSCLC) has limited treatment options. Expression of the RNA-binding protein (RBP) Musashi-2 (MSI2) is elevated in a subset of non-small cell lung cancer (NSCLC) tumors upon progression, and drives NSCLC metastasis. We evaluated the mechanism of MSI2 action in NSCLC to gain therapeutically useful insights. Reverse phase protein array (RPPA) analysis of MSI2-depleted versus control KrasLA1/+; Trp53R172HΔG/+ NSCLC cell lines identified EGFR as a MSI2-regulated protein. MSI2 control of EGFR expression and activity in an NSCLC cell line panel was studied using RT-PCR, Western blots, and RNA immunoprecipitation. Functional consequences of MSI2 depletion were explored for cell growth and response to EGFR-targeting drugs, in vitro and in vivo. Expression relationships were validated using human tissue microarrays. MSI2 depletion significantly reduced EGFR protein expression, phosphorylation, or both. Comparison of protein and mRNA expression indicated a post-transcriptional activity of MSI2 in control of steady state levels of EGFR. RNA immunoprecipitation analysis demonstrated that MSI2 directly binds to EGFR mRNA, and sequence analysis predicted MSI2 binding sites in the murine and human EGFR mRNAs. MSI2 depletion selectively impaired cell proliferation in NSCLC cell lines with activating mutations of EGFR (EGFRmut). Further, depletion of MSI2 in combination with EGFR inhibitors such as erlotinib, afatinib, and osimertinib selectively reduced the growth of EGFRmut NSCLC cells and xenografts. EGFR and MSI2 were significantly co-expressed in EGFRmut human NSCLCs. These results define MSI2 as a direct regulator of EGFR protein expression, and suggest inhibition of MSI2 could be of clinical value in EGFRmut NSCLC.

Isolation and identification in human blood serum of the proteins possessing the ability to bind with 48 kDa form of unconventional myosin 1c and their possible diagnostic and prognostic value.

We firstly identified 48 kDa molecular form of the unconventional myosin 1c (p48/Myo1C), and isolated it from blood serum of multiple sclerosis patients. The amount of p48/Myo1C in human blood serum correlated with some autoimmune, hemato-oncological and neurodegenerative diseases and thus may serve as a potential molecular biomarker. The biological functions of this protein in human blood remain unknown. Previously, we used the monodisperse magnetic poly (glycidyl methacrylate)(mag-PGMA-NH2 ) microspheres with immobilized 48/Myo1C and western-blot analysis, which allowed us to identify IgM and IgG immunoglobulins presenting an affinity to this protein. Here, we used mass spectrometry followed by the western blotting in order to identify other blood serum proteins with affinity to 48/Myo1C. The obtained data demonstrate that 48/Myo1C binds to component 3 of the complement and the antithrombin-III proteins. A combination of magnetic microparticle-based affinity chromatography with MALDI-TOF mass spectrometry and an in silico analysis provided an opportunity to identify the partners of interaction of 48/Myo1C with other proteins, in particular those participating in complement and coagulation cascades.

Plant Defensins from a Structural Perspective.

Plant defensins form a family of proteins with a broad spectrum of protective activities against fungi, bacteria, and insects. Furthermore, some plant defensins have revealed anticancer activity. In general, plant defensins are non-toxic to plant and mammalian cells, and interest in using them for biotechnological and medicinal purposes is growing. Recent studies provided significant insights into the mechanisms of action of plant defensins. In this review, we focus on structural and dynamics aspects and discuss structure-dynamics-function relations of plant defensins.

Nanocomposites for X-Ray Photodynamic Therapy.

Photodynamic therapy (PDT) has long been known as an effective method for treating surface cancer tissues. Although this technique is widely used in modern medicine, some novel approaches for deep lying tumors have to be developed. Recently, deeper penetration of X-rays into tissues has been implemented, which is now known as X-ray photodynamic therapy (XPDT). The two methods differ in the photon energy used, thus requiring the use of different types of scintillating nanoparticles. These nanoparticles are known to convert the incident energy into the activation energy of a photosensitizer, which leads to the generation of reactive oxygen species. Since not all photosensitizers are found to be suitable for the currently used scintillating nanoparticles, it is necessary to find the most effective biocompatible combination of these two agents. The most successful combinations of nanoparticles for XPDT are presented. Nanomaterials such as metal-organic frameworks having properties of photosensitizers and scintillation nanoparticles are reported to have been used as XPDT agents. The role of metal-organic frameworks for applying XPDT as well as the mechanism underlying the generation of reactive oxygen species are discussed.

Identification of new candidate genes and signalling pathways associated with the development of neuroendocrine pancreatic tumours based on next generation sequencing data.

Despite advances in classification, treatment, and imaging, neuroendocrine tumours remain a clinically complex subject. In this work, we studied the genetic profile of well-differentiated pancreatic neuroendocrine tumours (PanNETs) in a cohort of Caucasian patients and analysed the signalling pathways and candidate genes potentially associated with the development of this oncological disease. Twenty-four formalin-fixed paraffin-embedded (FFPE) samples of well-differentiated PanNETs were subjected to massive parallel sequencing using the targeted gene panel (409 genes) of the Illumina NextSeq 550 platform (San Diego, USA). In 24 patients, 119 variants were identified in 54 genes. The median mutation rate per patient was 5 (2.8-7). The detected genetic changes were dominated by missense mutations (67%) and nonsense mutations (29%). 18% of the mutations were activating, 35% of the variants led to a loss of function of the encoded protein, and 52% were not classified. Twenty-six variants were described as new. Functionally significant changes in the tertiary structure and activity of the protein molecules in an in silico assay were predicted for 5 new genetic variants. The 5 highest priority candidate genes were selected: CREB1, TCF12, PRKAR1A, BCL11A, and BUB1B. Genes carrying the identified mutations participate in signalling pathways known to be involved in PanNETs; in addition, 38% of the cases showed genetic changes in the regulation of the SMAD2/3 signalling pathway. Well-differentiated PanNETs in a Russian cohort demonstrate various molecular genetic features, including new genetic variations and potential driver genes. The highlighted molecular genetic changes in the SMAD2/3 signalling pathway suggest new prospects for targeted therapy.

A Proteomics Analysis Reveals 9 Up-Regulated Proteins Associated with Altered Cell Signaling in Colon Cancer Patients.

Colorectal cancer is the second most common cancer in women and third most common cancer in men. Cell signaling alterations in colon cancer, especially in aggressive metastatic tumors, require further investigations. The present study aims to compare the expression pattern of proteins associated with cell signaling in paired tumor and non-tumor samples of patients with colon cancer, as well as to define the cluster of proteins to differentiate patients with non-metastatic (Dukes' grade B) and metastatic (Dukes' grade C&D) colon cancer. Frozen tumor and non-tumor samples were collected after tumor resection from 19 patients with colon cancer. The Panorama™ Antibody Microarray-Cell Signaling kits were used for the analyses. The expression ratios of paired tumor/non-tumor samples were calculated for the each protein. We employed R packages 'samr', 'gplots', 'supclust' (pelora, wilma algorithms), 'glmnet' for the differential expression analysis, supervised clustering and penalized logistic regression. Significance analysis of microarrays revealed 9 significantly up-regulated proteins, including protein kinase C gamma, c-Myc, MDM2, pan cytokeratin, and 1 significantly down-regulated protein (GAP1) in tumoral mucosa. Pan-cytokeratin and APP were up-regulated in tumor versus non-tumor tissue, and were selected in the predictive cluster to discriminate colon cancer type. Higher levels of S-100b and phospho-Tau-pSer199/202 were confirmed as the predictors of non-metastatic colon cancer by all employed regression/clustering methods. Deregulated proteins in colon cancer are involved in oncogenic signal transduction, cell cycle control, and regulation of cytoskeleton/transport. Further studies are needed to validate potential protein markers of colon cancer development and metastatic progression.

EGFR mutations and tumor metastases in patients with nonsmall cell lung cancer in the South of Russia.

PURPOSE: To assess the frequencies of somatic EGFR mutations in the tumor tissues of patients with non-small cell lung cancer (NSCLC) residing in the South of Russia (SR), and to define the relationship between genetic subtypes of NSCLC and the emergence of different types of metastases. METHODS: DNA was extracted from formalin-fixed parrafin embedded (FFPE) samples of 721 patients. A total of 29 somatic EGFR mutations were detected using commercial Therascreen EGFR RGQ PCR Kit. RESULTS: EGFR mutations were significantly more frequent in females and non-smokers even when considering the combination of both factors. The frequency of activating EGFR mutations across three age groups (<51, 51-61, >61 years) of women with NSCLC was significantly different (x2=10.94, p=0.004) and became higher with increasing age. Both activating and resistance mutations of EGFR were not associated with the frequency of regional or distant metastases. The frequencies of both regional and distant metastases were associated with higher disease stage (odds ratio/OR)=16.71; 95% confidence interval (CI): 9.5-29.38; p<0.0001, and OR=2.94; 95% CI: 2.22-3.88; p<0.0001, respectively) and adenocarcinona histology (OR=6.52; 95% CI: 2.03-20.92; p=0.002, and OR=1.99; 95% CI: 0.91-4.34; p=0.083, respectively) even when adjusted for age, gender, and smoking status. The risk for regional metastases development was associated with poor tumor differentiation (OR=2.91; 95% CI: 1.21-7.02; p=0.017). CONCLUSION: EGFR mutations were not associated with the frequency of regional or distant metastases in SR patients with NSCLC.

Expression of proteins involved in epigenetic regulation in human cutaneous melanoma and peritumoral skin.

Epigenetic processes play a critical role in melanoma development. However, little is known about proteins responsible for epigenetic transformations in melanoma cells. The processes in the peritumoral skin within the excision margin are almost unstudied. We studied the changes in expression of 112 proteins involved in epigenetic regulation of gene expression in the human cutaneous melanoma and its peritumoral zone using "The Proteomic Antibody Microarrays" (GRAA2, Sigma-Aldrich). Dimethylated histone H3 at lysines 4 and 9 as well as proteins involved in the regulation of transcription (histone deacetylases HDAC-1 and HDAC-11, DNA methyl-binding protein Kaiso), cell cycle control (protein kinases Aurora-В and PKR, chromosome protein CENP-E , and phosphorylated and acetylated histone H3), DNA repair (phosphorylated histone H2AX), and nuclear protein import (importins α3 and α5/7) were over-expressed in the melanoma tissue as compared with normal skin. At the same time, HDAC-10 and proliferating cell nuclear antigen PCNA were downregulated. In the peritumoral skin, at the excision margin (1-2 cm from the melanoma edge), we observed similar changes in expression of these proteins and, additionally, over-expression of arginine methyltransferases PRMT5 and NAD-dependent histone deacetylase SIR2. Histone methyltransferase G9a and metastasis-associated protein 2 were downregulated. Therefore, epigenetic regulation that requires histone modifications and expression of some regulatory proteins is of importance for melanoma development and propagation. The observed changes in the peritumoral skin may indicate the epigenetic pre-tuning in this zone possibly involved in malignant transformation. These results can be potentially useful for melanoma diagnostics and targeted therapy.

Efficient approach for simulating distorted materials.

The operation principles of nanoscale devices are based upon both electronic and mechanical properties of materials. Because these properties can be coupled, they need to be investigated simultaneously. At this moment, however, the electronic structure calculations with custom-made long-range mechanical distortions are impossible, or expensive at best. Here we present a unified formalism to solve exactly the electronic structures of nanomaterials with versatile distortions. We illustrate the formalism by investigating twisted armchair graphene nanoribbons with the least possible number of atoms. Apart from enabling versatile material distortions, the formalism is capable of reducing computational costs orders of magnitude in various areas of science and engineering.