Published papers

Background: Historical and updated information provided by time-course data collected during an entire treatment period proves to be more useful than information provided by single-point data. Accurate predictions made using time-course data on multiple biomarkers that indicate a patient’s response to therapy contribute positively to the decision-making process associated with designing effective treatment programs for various diseases. Therefore, the development of prediction methods incorporating time-course data on multiple markers is necessary. Results: We proposed new methods that may be used for prediction and gene selection via time-course gene expression profiles. Our prediction method consolidated multiple probabilities calculated using gene expression profiles collected over a series of time points to predict therapy response. Using two data sets collected from patients with hepatitis C virus (HCV) infection and multiple sclerosis (MS), we performed numerical experiments that predicted response to therapy and evaluated their accuracies. Our methods were more accurate than conventional methods and successfully selected genes, the functions of which were associated with the pathology of HCV infection and MS. Conclusions: The proposed method accurately predicted response to therapy using data at multiple time points. It showed higher accuracies at early time points compared to those of conventional methods. Furthermore, this method successfully selected genes that were directly associated with diseases.

Alzheimer’s disease (AD) has become a problem, owing to its high prevalence in an aging society with no treatment available after onset. However, early diagnosis is essential for preventive intervention to delay disease onset due to its slow progression. The current AD diagnostic methods are typically invasive and expensive, limiting their potential for widespread use. Thus, the development of biomarkers in available biofluids, such as blood, urine, and saliva, which enables low or non-invasive, reasonable, and objective evaluation of AD status, is an urgent task. Here, we reviewed studies that examined biomarker candidates for the early detection of AD. Some of the candidates showed potential biomarkers, but further validation studies are needed. We also reviewed studies for non-invasive biomarkers of AD. Given the complexity of the AD continuum, multiple biomarkers with machine-learning-classification methods have been recently used to enhance diagnostic accuracy and characterize individual AD phenotypes. Artificial intelligence and new body fluid-based biomarkers, in combination with other risk factors, will provide a novel solution that may revolutionize the early diagnosis of AD.

This study aimed to evaluate volatile compounds in exhaled breath as a non-invasive screening method to detect breast neoplasms. Exhaled breath samples were collected from patients with breast cancer (BC; n = 45) and non-breast cancer (NBC; n = 51) controls. Selected ion-flow tube mass spectrometry was used to quantify the volatile compounds. A multiple logistic regression (MLR) model was developed by combining multiple compounds to discriminate between BC and NBC samples. Amongst the 672 quantified peaks, 17 showed significant differences between BC and NBC samples (P < 0.05 corrected by false discovery rate). Pathway analysis revealed a significant difference in glycerophospholipid metabolism. The MLR model showed an area under the receiver operating characteristic curve (AUC) of 0.719 (95% confidence interval: 0.615-0.822, P < 0.0002). Cross-validation under various conditions resulted in a slight fluctuation in the AUC values, indicating the high generalizability of the MLR model. The model showed a higher BC probability for advanced-stage subjects and higher Ki67 (⩾30) for BC subjects. This study suggests the potential of volatile compounds in exhaled breath as a noninvasive screening method for BC.

Food intake biomarkers can be critical tools that can be used to objectively assess dietary exposure for both epidemiological and clinical nutrition studies. While an accurate estimation of food intake is essential to unravel associations between the intake and specific health conditions, random and systematic errors affect self-reported assessments. This study aimed to clarify how habitual food intake influences the circulating plasma metabolome in a free-living Japanese regional population and to identify potential food intake biomarkers. To achieve this aim, we conducted a cross-sectional analysis as part of a large cohort study. From a baseline survey of the Tsuruoka Metabolome Cohort Study, 7, 012 eligible male and female participants aged 40-69 years were chosen for this study. All data on patients' health status and dietary intake were assessed via a food frequency questionnaire, and plasma samples were obtained during an annual physical examination. Ninety-four charged plasma metabolites were measured using capillary electrophoresis mass spectrometry, by a non-targeted approach. Statistical analysis was performed using partial-least-square regression. A total of 21 plasma metabolites were likely to be associated with long-term food intake of nine food groups. In particular, the influential compounds in each food group were hydroxyproline for meat, trimethylamine- N-oxide for fish, choline for eggs, galactarate for dairy, cystine and betaine for soy products, threonate and galactarate for carotenoid-rich vegetables, proline betaine for fruits, quinate and trigonelline for coffee, and pipecolate for alcohol, and these were considered as prominent food intake markers in Japanese eating habits. A set of circulating plasma metabolites was identified as potential food intake biomarkers in the Japanese community-dwelling population. These results will open the way for the application of new reliable dietary assessment tools not by self-reported measurements but through objective quantification of biofluids.

The tertiary lymphoid structure (TLS) plays a central role in cancer immune response, and its gene expression pattern, called the TLS signature, has shown prognostic value in breast cancer. The formation of TLS and tumor-associated high endothelial venules (TA-HEVs), responsible for lymphocytic infiltration within the TLS, is associated with the expression of cancer hallmark genes (CHGs) related to immunity and inflammation. In this study, we performed co-expression network analysis of immune- and inflammation-related CHGs to identify predictive genes for breast cancer. In total, 382 immune- and inflammation-related CHGs with high expression variance were extracted from the GSE86166 microarray dataset of patients with breast cancer. CHGs were classified into five modules by applying weighted gene co-expression network analysis. The survival analysis results for each module showed that one module comprising 45 genes was statistically significant for relapse-free and overall survival. Four network properties identified key genes in this module with high prognostic prediction abilities: CD34, CXCL12, F2RL2, JAM2, PROS1, RAPGEF3, and SELP. The prognostic accuracy of the seven genes in breast cancer was synergistic and exceeded that of other predictors in both small and large public datasets. Enrichment analysis predicted that these genes had functions related to leukocyte infiltration of TA-HEVs. There was a positive correlation between key gene expression and the TLS signature, suggesting that gene expression levels are associated with TLS density. Co-expression network analysis of inflammation- and immune-related CHGs allowed us to identify genes that share a standard function in cancer immunity and have a high prognostic predictive value. This analytical approach may contribute to the identification of prognostic genes in TLS.

Vascular dementia, caused by cerebrovascular disease, is associated with cognitive impairment and reduced hippocampal metabolite levels. Specifically, cognitive impairment can be induced by decreased hippocampal brain-derived neurotrophic factor (BDNF) expression. The development of low or non-invasive biomarkers to characterize these diseases is an urgent task. Disturbance of metabolic pathways has been frequently observed in cognitive impairment, and salivary molecules also showed the potentials to reflect cognitive impairment. Therefore, we evaluated salivary metabolic profiles associated with altered hippocampal BDNF expression levels in a cerebral ischemia mouse model using metabolomic analyses. The effect of tacrine (a cholinesterase inhibitor) administration was also examined. The arteries of ICR mice were occluded with aneurysm clips to generate the cerebral ischemia model. Learning and memory performance was assessed using the elevated plus maze (EPM) test. Hippocampal and blood BDNF levels were quantified using an enzyme-linked immunosorbent assay. Glutamate decarboxylase 1 (GAD1) mRNA expression, is associated with cognitive impairment, was quantified by a real-time polymerase chain reaction. The EPM test revealed impaired spatial working memory in the cerebral ischemia mouse model; tacrine administration ameliorated this memory impairment. Cerebral ischemia suppressed GAD1 expression by decreasing hippocampal BDNF expression. In total, seven salivary metabolites, such as trimethylamine N-oxide and putrescine, were changed by cognitive impairment and tacrine administration. Our data suggest that salivary metabolite patterns were associated with cognitive function.

Liver fibrosis is a result of homeostasis breakdown caused by repetitive injury. The accumulation of collagens disrupts liver structure and function, which causes serious consequences such as cirrhosis. Various mathematical simulation models have been developed to understand these complex processes. We employed the agent-based modelling (ABM) approach and implemented inflammatory processes in central venous regions. Collagens were individually modelled and visualised depending on their origin: myofibroblast and portal fibroblast. Our simulation showed that the administration of toxic compounds induced accumulation of myofibroblast-derived collagens in central venous regions and portal fibroblast-derived collagens in portal areas. Subsequently, these collagens were bridged between central-central areas and spread all over areas. We confirmed the consistent dynamic behaviour of collagen formulation in our simulation and from histological sections obtained via in vivo experiments. Sensitivity analyses identified dead hepatocytes caused by inflammation and the ratio of residential liver cells functioned as a cornerstone for the initiation and progression of liver fibrosis. The validated mathematical model demonstrated here shows virtual experiments that are complementary to biological experiments, which contribute to understanding a new mechanism of liver fibrosis.

This study examined the relationship between plasma amino acid (AA) concentrations, including branched-chain AAs, and brown adipose tissue density (BAT-d). One hundred and seventy-three subjects (69 men, 104 women) aged 22–68 years were recruited during the winter season. AAs were comprehensively quantified using liquid chromatography-time-of-flight-mass spectrometry. The total hemoglobin concentration in the supraclavicular region ([total-Hb]sup), an indicator of BAT-d, was assessed using near-infrared time-resolved spectroscopy. Anthropometric parameters, including age, percentage of body fat, and visceral fat, were evaluated. Factors associated with higher (≥74 µM) or lower (<74 µM) [total-Hb]sup were investigated by multiple logistic regression models that included AA concentrations alone (model 1) or AA concentrations and anthropometric parameters (model 2) as independent variables. When adjusted for the false discovery rate, [total-Hb]sup was positively correlated with glycine and asparagine levels in men and with the serine level in both men and women and was negatively correlated with the branched-chain AA concentration in men. Models 1 and 2 correlated with higher or lower BAT-d for men (r = 0.73, p = 0.015) and women (r = 0.58, p = 0.079) and for men (r = 0.82, p = 0.0070) and women (r = 0.70, p = 0.020), respectively. A combination of anthropometric parameters and plasma AA concentrations could be a reliable biomarker for higher and lower BAT-d.

This is the first report of conducting proof-of-concept study for amperometric acetyltransferase-based l-carnitine sensor by employing the thiol/disulfide exchange reaction. The carnitine acetyltransferase (CrAT) catalyzes the reaction between acetyl-CoA and l-carnitine to produce CoA which is difficult to detect directly by electrochemical methods owing to steric hindrance and electrostatic effect of CoA. The thiol/disulfide exchange reaction between CoA and cystamine was mediated in the enzymatic reaction to produce electrochemically detectable low molecular weight of cationic cysteamine. The sensor exhibited high sensitivity and selectivity for l-carnitine in the concentration range 0.28–50 µM with a limit of detection of 0.28 µM. This is a promising strategy for l-carnitine sensing in point-of-care testing applications. Graphical abstract: [Figure not available: see fulltext.]

Background: Transthoracic impedance values have not been widely used to measure extravascular pulmonary water content due to accuracy and complexity concerns. Our aim was to develop a foundational model for a novel system aiming to non-invasively estimate the intrathoracic condition of heart failure patients. Methods: We employed multi-frequency bioelectrical impedance analysis to simultaneously measure multiple frequencies, collecting electrical, physical, and hematological data from 63 hospitalized heart failure patients and 82 healthy volunteers. Measurements were taken upon admission and after treatment, and longitudinal analysis was conducted. Results: Using a light gradient boosting machine, and a decision tree-based machine learning method, we developed an intrathoracic estimation model based on electrical measurements and clinical findings. Out of the 286 features collected, the model utilized 16 features. Notably, the developed model demonstrated high accuracy in discriminating patients with pleural effusion, achieving an area under the receiver characteristic curves (AUC) of 0.905 (95% CI: 0.870–0.940, p < 0.0001) in the cross-validation test. The accuracy significantly outperformed the conventional frequency-based method with an AUC of 0.740 (95% CI: 0.688–0.792, and p < 0.0001). Conclusions: Our findings indicate the potential of machine learning and transthoracic impedance measurements for estimating pleural effusion. By incorporating noninvasive and easily obtainable clinical and laboratory findings, this approach offers an effective means of assessing intrathoracic conditions.

Background: Saliva is often used as a biomarker for the diagnosis of some oral and systematic diseases, owing to the non-invasive attribute of the fluid. In this study, we aimed to identify salivary biomarkers for distinguishing lung cancer (LC) from benign lung lesion (BLL). Materials and Methods: Unstimulated saliva samples were collected from 41 patients with LC and 21 with BLL. Salivary metabolites were comprehensively analyzed using capillary electrophoresis mass spectrometry. To differentiate between patients with LCs and BLLs, the discriminatory ability of each biomarker was assessed. Furthermore, a multiple logistic regression (MLR) model was developed for evaluating discriminatory ability of each salivary metabolite. Results: The profiles of 10 salivary metabolites were remarkably different between the LC and BLL samples. Among them, the concentration of salivary tryptophan was significantly lower in the samples from patients with LC than in those from patients with BLL, and the area under the curve (AUC) for discriminating patients with LC from those with BLL was 0.663 (95% confidence interval [CI] = 0.516–0.810, p = 0.036). Furthermore, from the MLR model developed using these metabolites, diethanolamine, cytosine, lysine, and tyrosine, were selected using the back-selection regression method. The MLR model based on these four metabolites had a high discriminatory ability for patients with LC and those with BLL (AUC = 0.729, 95% CI = 0.598–0.861, p = 0.003). Conclusion: The four salivary metabolites can serve as potential non-invasive biomarkers for distinguishing LC from BLL.

PURPOSE. To identify tissue metabolomic profiles in biopsy specimens with IgG4-related ophthalmic disease (IgG4-ROD) and mucosa-associated lymphoid tissue (MALT) lymphoma and investigate their potential implication in the disease pathogenesis and biomarkers. METHODS. We conducted a comprehensive analysis of the metabolomes and lipidomes of biopsy-proven IgG4-ROD (n = 22) and orbital MALT lymphoma (n = 21) specimens and matched adjacent microscopically normal adipose tissues using liquid chromatography time-of-flight mass spectrometry. The altered metabolomic profiles were visualized by heat map and principal component analysis. Metabolic pathway analysis was performed by Metabo Analyst 4.0 using differentially expressed metabolites. The diagnostic performance of the metabolic markers was evaluated using receiver operating characteristic curves. Machine learning algorithms were implemented by random forest using the R environment. Finally, an independent set of 18 IgG4-ROD and 17 orbital MALT lymphoma specimens were used to validate the identified biomarkers. RESULTS. The principal component analysis showed a significant difference of both IgG4-ROD and orbital MALT lymphoma for biopsy specimens and controls. Interestingly, lesions in IgG4-ROD were uniquely enriched in arachidonic metabolism, whereas those in orbital MALT lymphoma were enriched in tricarboxylic acid cycle metabolism. We identified spermine as the best discriminator between IgG4-ROD and orbital MALT lymphoma, and the area under the receiver operating characteristic curve of the spermine to discriminate between the two diseases was 0.89 (95% confidence interval, 0.803–0.984). A random forest model incorporating a panel of five metabolites showed a high area under the receiver operating characteristic curve value of 0.983 (95% confidence interval, 0.981–0.984). The results of validation revealed that four tissue metabolites: N1,N12diacetylspermine, spermine, malate, and glycolate, had statistically significant differences between IgG4-ROD and orbital MALT lymphoma with receiver operating characteristic values from 0.708 to 0.863. CONCLUSIONS. These data revealed the characteristic differences in metabolomic profiles between IgG4-ROD and orbital MALT lymphoma, which may be useful for developing new diagnostic biomarkers and elucidating the pathogenic mechanisms of these common orbital lymphoproliferative disorders.

Arginine-rich dipeptide repeat proteins (R-DPRs), poly(PR) and poly(GR), translated from the hexanucleotide repeat expansion in the amyotrophic lateral sclerosis (ALS)-causative C9ORF72 gene, contribute significantly to pathogenesis of ALS. Although both R-DPRs share many similarities, there are critical differences in their subcellular localization, phase separation, and toxicity mechanisms. We analyzed localization, protein-protein interactions, and phase separation of R-DPR variants and found that sufficient segregation of arginine charges is necessary for nucleolar distribution. Proline not only efficiently separated the charges, but also allowed for weak, but highly multivalent binding. In contrast, because of its high flexibility, glycine cannot fully separate the charges, and poly(GR) behaves similarly to the contiguous arginines, being trapped in the cytoplasm. We conclude that the amino acid that spaces the arginine charges determines the strength and multivalency of the binding, leading to differences in localization and toxicity mechanisms.

Liver cancer is one of the most common malignancies and the second leading cause of cancer-related deaths worldwide, particularly in developing countries, where it poses a significant financial burden. Early detection and timely treatment remain challenging due to the complex mechanisms underlying the initiation and progression of liver cancer. This study aims to uncover key genomic features, analyze their functional roles, and propose potential therapeutic drugs identified through molecular docking, utilizing single-cell RNA sequencing (scRNA-seq) data from liver cancer studies. We applied two advanced hybrid methods known for their robust identification of differentially expressed genes (DEGs) regardless of sample size, along with four top-performing individual methods. These approaches were used to analyze four scRNA-seq datasets, leading to the identification of essential DEGs. Through a protein−protein-interaction (PPI) network, we identified 25 hub-of-hub genes (hHubGs) and 20 additional hHubGs from two naturally occurring gene clusters, ultimately validating a total of 36 hHubGs. Functional, pathway, and survival analyses revealed that these hHubGs are strongly linked to liver cancer. Based on molecular docking and binding-affinity scores with 36 receptor proteins, we proposed 10 potential therapeutic drugs, which we selected from a pool of 300 cancer meta-drugs. The choice of these drugs was further validated using 14 top-ranked published receptor proteins from a set of 42. The proposed candidates include Adozelesin, Tivozanib, NVP-BHG712, Nilotinib, Entrectinib, Irinotecan, Ponatinib, and YM201636. This study provides critical insights into the genomic landscape of liver cancer and identifies promising therapeutic candidates, serving as a valuable resource for advancing liver cancer research and treatment strategies.

Lipids are a major component of extracellular vesicles; however, their significance in tumorigenesis and progression has not been well elucidated. As we previously found that lipid profiles drastically changed in breast tumors upon progression, we hypothesized that lipid profiles of plasma-derived extracellular vesicles could be utilized as breast cancer biomarkers. Here, we adopted modified sucrose cushion ultracentrifugation to isolate plasma-derived extracellular vesicles from breast cancer (n = 105), benign (n = 11), and healthy individuals (n = 43) in two independent cohorts (n = 126 and n = 33) and conducted targeted lipidomic analysis. We established a breast cancer diagnostic model comprising three lipids that showed favorable performance with the area under the receiver operating characteristic curve of 0.759, 0.743, and 0.804 in the training, internal validation, and external test sets, respectively. Moreover, we identified several lipids that could effectively discriminate breast cancer progression and subtypes: phosphatidylethanolamines and phosphatidylserines were relatively higher in Stage III, whereas phosphatidylcholines and sphingomyelins were higher in Stage IV; phosphatidylcholines and ceramides were correspondingly concentrated in HER2-positive patients, while lysophosphatidylcholines and polyunsaturated triglycerides were concentrated in the triple-negative breast cancer subtype. Lipid profiling of plasma-derived extracellular vesicles is a non-invasive and promising approach for diagnosing, staging, and subtyping breast cancer.

Availability of hepatic tissue for the investigation of metabolic processes is severely limited. While primary hepatocytes or animal models are widely used in pharmacological applications, a change in methodology towards more sustainable and ethical assays is highly desirable. Stem cell derived hepatic cells are generally regarded as a viable alternative for the above model systems, if current limitations in functionality and maturation can be overcome. By combining microfluidic organ-on-a-chip technology with individually differentiated, multicellular hepatic tissue fractions, we aim to improve overall functionality of hepatocyte-like cells, as well as evaluate cellular composition and interactions with non-parenchymal cell populations towards the formation of mature liver tissue. Utilizing a multi-omic approach, we show the improved maturation profiles of hepatocyte-like cells maintained in a dynamic microenvironment compared to standard tissue culture setups without continuous perfusion. In order to evaluate the resulting tissue, we employ single cell sequencing to distinguish formed subpopulations and spatial localization. While cellular input was strictly defined based on established differentiation protocols of parenchyma, endothelial and stellate cell fractions, resulting hepatic tissue was shown to comprise a complex mixture of epithelial and non-parenchymal fractions with specific local enrichment of phenotypes along the microchannel. Following this approach, we show the importance of passive, paracrine developmental processes in tissue formation. Using such complex tissue models is a crucial first step to develop stem cell-derived in vitro systems that can compare functionally with currently used pharmacological and toxicological applications.