Machine Learning-Based Interpretable Modeling for Subjective Emotional Dynamics Sensing Using Facial EMG.

Understanding the association between subjective emotional experiences and physiological signals is of practical and theoretical significance. Previous psychophysiological studies have shown a linear relationship between dynamic emotional valence experiences and facial electromyography (EMG) activities. However, whether and how subjective emotional valence dynamics relate to facial EMG changes nonlinearly remains unknown. To investigate this issue, we re-analyzed the data of two previous studies that measured dynamic valence ratings and facial EMG of the corrugator supercilii and zygomatic major muscles from 50 participants who viewed emotional film clips. We employed multilinear regression analyses and two nonlinear machine learning (ML) models: random forest and long short-term memory. In cross-validation, these ML models outperformed linear regression in terms of the mean squared error and correlation coefficient. Interpretation of the random forest model using the SHapley Additive exPlanation tool revealed nonlinear and interactive associations between several EMG features and subjective valence dynamics. These findings suggest that nonlinear ML models can better fit the relationship between subjective emotional valence dynamics and facial EMG than conventional linear models and highlight a nonlinear and complex relationship. The findings encourage emotion sensing using facial EMG and offer insight into the subjective-physiological association.

Role of TLR4 signaling on Porphyromonas gingivalis LPS-induced cardiac dysfunction in mice.

Oral infections, particularly periodontitis, are a well-established risk factor for cardiovascular diseases, although the molecular mechanisms involved remain elusive. The aims of the present study were to investigate the effects of lipopolysaccharide derived from Porphyromonas gingivalis (PG-LPS) on cardiac function in mice, and to elucidate the underlying mechanisms. Mice (C57BL/6) were injected with PG-LPS (0.8 mg/kg/day) with or without an inhibitor of Toll-like receptor 4 (TLR4) signaling (TAK-242, 0.8 mg/kg/day) for 4 weeks. Left ventricular ejection function was significantly decreased at 1 week (from 67 ± 0.5 to 58 ± 1.2%) and remained low at 4 weeks (57 ± 1.0%). The number of apoptotic myocytes was increased (approximately 7.4-fold), the area of fibrosis was increased (approximately 3.3-fold) and the number of 8-hydroxydeoxyguanosine-positive myocytes, a sensitive indicator of oxidative DNA damage, was increased (approximately 7.6-fold) at 4 weeks in the heart of PG-LPS treated mice. However, levels of various serum pro-inflammatory cytokines in PG-LPS-treated mice were similar to those in control mice. The impairment of cardiac function in PG-LPS-treated mice appears to involve activation of TLR4-NADPH oxidase (NOX) 4 signaling, leading to abundant production of reactive oxygen species and Ca2+ leakage from sarcoplastic reticulumn induced by calmodulin kinase II (CaMKII)-mediated phosphorylation of phospholamban (at Thr-17) and ryanodine receptor 2 (at Ser-2448). Pharmacological inhibition of TLR4 with TAK-242 attenuated the changes in cardiac function in PG-LPS-treated mice. Our results indicate that TLR4-NOX4 signaling may be a new therapeutic target for treatment of cardiovascular diseases in patients with periodontitis.

Role of ß-adrenergic signaling in masseter muscle.

In skeletal muscle, the major isoform of ß-adrenergic receptor (ß-AR) is ß2-AR and the minor isoform is ß1-AR, which is opposite to the situation in cardiac muscle. Despite extensive studies in cardiac muscle, the physiological roles of the ß-AR subtypes in skeletal muscle are not fully understood. Therefore, in this work, we compared the effects of chronic ß1- or ß2-AR activation with a specific ß1-AR agonist, dobutamine (DOB), or a specific ß2-AR agonist, clenbuterol (CB), on masseter and cardiac muscles in mice. In cardiac muscle, chronic ß1-AR stimulation induced cardiac hypertrophy, fibrosis and myocyte apoptosis, whereas chronic ß2-AR stimulation induced cardiac hypertrophy without histological abnormalities. In masseter muscle, however, chronic ß1-AR stimulation did not induce muscle hypertrophy, but did induce fibrosis and apoptosis concomitantly with increased levels of p44/42 MAPK (ERK1/2) (Thr-202/Tyr-204), calmodulin kinase II (Thr-286) and mammalian target of rapamycin (mTOR) (Ser-2481) phosphorylation. On the other hand, chronic ß2-AR stimulation in masseter muscle induced muscle hypertrophy without histological abnormalities, as in the case of cardiac muscle, concomitantly with phosphorylation of Akt (Ser-473) and mTOR (Ser-2448) and increased expression of microtubule-associated protein light chain 3-II, an autophagosome marker. These results suggest that the ß1-AR pathway is deleterious and the ß2-AR is protective in masseter muscle. These data should be helpful in developing pharmacological approaches for the treatment of skeletal muscle wasting and weakness.

Effects of chronic Porphyromonas gingivalis lipopolysaccharide infusion on skeletal muscles in mice.

Periodontitis, which is caused by various oral organisms, predominantly affects adults, and is one of the main causes of tooth loss, as well as leading to progression of numerous systemic diseases. However, its relationship to sarcopenia (aging-associated degenerative loss of skeletal muscle mass and function) remains unclear. The aim of this study was to investigate the effects of Porphyromonas gingivalis lipopolysaccharide (PG-LPS) on skeletal muscle in mice, and to establish the underlying mechanisms. Mice (C57BL/6) were injected with PG-LPS (0.8 mg/kg/day) for 4 weeks. This treatment significantly decreased the weight of fast-twitch skeletal muscles (masseter and tibialis anterior muscles), but not that of slow-twitch skeletal muscle (soleus muscle). The area of fibrosis was significantly increased in masseter muscle, but remained unchanged in the other two muscles. The number of apoptotic myocytes was significantly increased (approximately eightfold) in masseter muscle. These data suggest that persistent subclinical exposure to PG-LPS might reduce the size of fast-twitch skeletal muscle, but not slow-twitch skeletal muscle. Masseter muscle appears to be especially susceptible to the adverse effects of PG-LPS, because muscle remodeling (muscle fibrosis and myocyte apoptosis) was induced solely in masseter muscle. Thus, periodontitis might be one of the major causes of oral sarcopenia.

Relationship between bite size per mouthful and dental arch size in healthy subjects.

Although multiple factors influence food bite size, the relationship between food bite size per mouthful and mandible or tongue size remains poorly understood. Here, we examined the correlations between food bite size and the lower dental arch size (an indicator of tongue size) in human subjects with good oral and general health, using fish sausage and bread as test foods. Notably, bite size of both foods was significantly positively correlated with the lower dental arch size, whereas masticatory performance (measured in terms of glucose extraction from a gummy jelly) showed no dependence on bite size. Further, bite size was significantly positively correlated with the body mass index. Our findings suggest that larger bite size is associated with larger tongue size, which might be a contributory factor to obesity.

Role of G protein-regulated inducer of neurite outgrowth 3 (GRIN3) in ß-arrestin 2-Akt signaling and dopaminergic behaviors.

The G protein-regulated inducer of neurite growth (GRIN) family has three isoforms (GRIN1-3), which bind to the Gαi/o subfamily of G protein that mediate signal processing via G protein-coupled receptors (GPCRs). Here, we show that GRIN3 is involved in regulation of dopamine-dependent behaviors and is essential for activation of the dopamine receptors (DAR)-ß-arrestin signaling cascade. Analysis of functional regions of GRIN3 showed that a di-cysteine motif (Cys751/752) is required for plasma membrane localization. GRIN3 was co-immunoprecipitated with GPCR kinases 2/6 and ß-arrestins 1/2. Among GRINs, only GRIN3, which is highly expressed in striatum, strongly interacted with ß-arrestin 2. We also generated GRIN3-knockout mice (GRIN3KO). GRIN3KO exhibited reduced locomotor activity and increased anxiety-like behavior in the elevated maze test, as well as a reduced locomoter response to dopamine stimulation. We also examined the phosphorylation of Akt at threonine 308 (phospho308-Akt), which is dephosphorylated via a ß-arrestin 2-mediated pathway. Dephosphorylation of phospho308-Akt via the D2R-ß-arrestin 2 signaling pathway was completely abolished in striatum of GRIN3KO. Our results suggest that GRIN3 has a role in recruitment and assembly of proteins involved in ß-arrestin-dependent, G protein-independent signaling.

Trace Material Capture by Controlled Liquid Droplets on a Superhydrophobic/Hydrophilic Surface.

A liquid droplet in contact with a superhydrophobic surface can be used to collect dissolved trace materials after evaporating the solvent. This process effect enhances detection limits, but a liquid droplet easily rolls off a superhydrophobic surface. Keeping it at a specific collecting spot area is challenging. Here the means for controlling and capturing a liquid droplet on a superhydrophobic surface is demonstrated. To induce a liquid droplet to a collecting spot, its rolling direction was controlled by two superhydrophobic fabric guides. The liquid droplet was then captured by hydrophilic polymer and hydrophilic nanoparticles at the measuring spot. After removing the solvent, the trace compounds were evaluated with a colorimetric analysis visible to the naked eye.

Role of phosphodiesterase 4 expression in the Epac1 signaling-dependent skeletal muscle hypertrophic action of clenbuterol.

Clenbuterol (CB), a selective ß2-adrenergic receptor (AR) agonist, induces muscle hypertrophy and counteracts muscle atrophy. However, it is paradoxically less effective in slow-twitch muscle than in fast-twitch muscle, though slow-twitch muscle has a greater density of ß-AR We recently demonstrated that Epac1 (exchange protein activated by cyclic AMP [cAMP]1) plays a pivotal role in ß2-AR-mediated masseter muscle hypertrophy through activation of the Akt and calmodulin kinase II (CaMKII)/histone deacetylase 4 (HDAC4) signaling pathways. Here, we investigated the role of Epac1 in the differential hypertrophic effect of CB using tibialis anterior muscle (TA; typical fast-twitch muscle) and soleus muscle (SOL; typical slow-twitch muscle) of wild-type (WT) and Epac1-null mice (Epac1KO). The TA mass to tibial length (TL) ratio was similar in WT and Epac1KO at baseline and was significantly increased after CB infusion in WT, but not in Epac1KO The SOL mass to TL ratio was also similar in WT and Epac1KO at baseline, but CB-induced hypertrophy was suppressed in both mice. In order to understand the mechanism involved, we measured the protein expression levels of ß-AR signaling-related molecules, and found that phosphodiesterase 4 (PDE4) expression was 12-fold greater in SOL than in TA These results are consistent with the idea that increased PDE4-mediated cAMP hydrolysis occurs in SOL compared to TA, resulting in a reduced cAMP concentration that is insufficient to activate Epac1 and its downstream Akt and CaMKII/HDAC4 hypertrophic signaling pathways in SOL of WT This scenario can account for the differential effects of CB on fast- and slow-twitch muscles.