Unbound Fractions of PFAS in Human and Rodent Tissues: Rat Liver a Suitable Proxy for Evaluating Emerging PFAS?

Adverse health effects associated with exposures to perfluoroalkyl and polyfluoroalkyl substances (PFAS) are a concern for public health and are driven by their elimination half-lives and accumulation in specific tissues. However, data on PFAS binding in human tissues are limited. Accumulation of PFAS in human tissues has been linked to interactions with specific proteins and lipids in target organs. Additional data on PFAS binding and unbound fractions (funbound) in whole human tissues are urgently needed. Here, we address this gap by using rapid equilibrium dialysis to measure the binding and funbound of 16 PFAS with 3 to 13 perfluorinated carbon atoms (ηpfc = 3-13) and several functional headgroups in human liver, lung, kidney, heart, and brain tissue. We compare results to mouse (C57BL/6 and CD-1) and rat tissues. Results show that funbound decreases with increasing fluorinated carbon chain length and hydrophobicity. Among human tissues, PFAS binding was generally greatest in brain > liver ≈ kidneys ≈ heart > lungs. A correlation analysis among human and rodent tissues identified rat liver as a suitable surrogate for predicting funbound for PFAS in human tissues (R2 ≥ 0.98). The funbound data resulting from this work and the rat liver prediction method offer input parameters and tools for toxicokinetic models for legacy and emerging PFAS.

An Online Nanoinformatics Platform Empowering Computational Modeling of Nanomaterials by Nanostructure Annotations and Machine Learning Toolkits.

Modern nanotechnology has generated numerous datasets from in vitro and in vivo studies on nanomaterials, with some available on nanoinformatics portals. However, these existing databases lack the digital data and tools suitable for machine learning studies. Here, we report a nanoinformatics platform that accurately annotates nanostructures into machine-readable data files and provides modeling toolkits. This platform, accessible to the public at https://vinas-toolbox.com/, has annotated nanostructures of 14 material types. The associated nanodescriptor data and assay test results are appropriate for modeling purposes. The modeling toolkits enable data standardization, data visualization, and machine learning model development to predict properties and bioactivities of new nanomaterials. Moreover, a library of virtual nanostructures with their predicted properties and bioactivities is available, directing the synthesis of new nanomaterials. This platform provides a data-driven computational modeling platform for the nanoscience community, significantly aiding in the development of safe and effective nanomaterials.

RNA Coating Promotes Peri-Implant Osseointegration.

In addition to transmitting and carrying genetic information, RNA plays an important abiotic role in the world of nanomaterials. RNA is a natural polyanionic biomacromolecule, and its ability to promote osteogenesis by binding with other inorganic materials as an osteogenic induction agent was discovered only recently. However, whether it can promote osseointegration on implants has not been reported. Here, we investigated the effect of the RNA-containing coating materials on peri-implant osseointegration. Total RNA extracted from rat muscle tissue was used as an osteogenic induction agent, and hyaluronic acid (HA) was used to maintain its negative charge. In simulated body fluids (SBF), in vitro studies demonstrated that the resulting material encouraged calcium salt deposition. Cytological experiments showed that the RNA-containing coating induced greater cell adhesion and osteogenic differentiation in comparison to the control. The results of animal experiments showed that the RNA-containing coating had osteoinductive and bone conduction activities, which are beneficial for bone formation and osseointegration. Therefore, the RNA-containing coatings are useful for the surface modification of titanium implants to promote osseointegration.

Hybrid non-animal modeling: A mechanistic approach to predict chemical hepatotoxicity.

Developing mechanistic non-animal testing methods based on the adverse outcome pathway (AOP) framework must incorporate molecular and cellular key events associated with target toxicity. Using data from an in vitro assay and chemical structures, we aimed to create a hybrid model to predict hepatotoxicants. We first curated a reference dataset of 869 compounds for hepatotoxicity modeling. Then, we profiled them against PubChem for existing in vitro toxicity data. Of the 2560 resulting assays, we selected the mitochondrial membrane potential (MMP) assay, a high-throughput screening (HTS) tool that can test chemical disruptors for mitochondrial function. Machine learning was applied to develop quantitative structure-activity relationship (QSAR) models with 2536 compounds tested in the MMP assay for screening new compounds. The MMP assay results, including QSAR model outputs, yielded hepatotoxicity predictions for reference set compounds with a Correct Classification Ratio (CCR) of 0.59. The predictivity improved by including 37 structural alerts (CCR = 0.8). We validated our model by testing 37 reference set compounds in human HepG2 hepatoma cells, and reliably predicting them for hepatotoxicity (CCR = 0.79). This study introduces a novel AOP modeling strategy that combines public HTS data, computational modeling, and experimental testing to predict chemical hepatotoxicity.

Bacteroid cerium oxide particles promote macrophage polarization to achieve early vascularization and subsequent osseointegration around implants.

The establishment of an osseointegration is crucial for the long-term stability and functionality of implant materials, and early angiogenesis is the key to successful osseointegration. However, the bioinertness of titanium implants affects osseointegration, limiting their clinical application. In this study, inspired by the rapid polarization of macrophages following the phagocytosis of bacteria, we developed bacteroid cerium oxide particles; these particles were composed of CeO2 and had a size similar to that of Bacillus (0.5 µ m). These particles were constructed on the implant surfaces using a hydrothermal method. In vitro experiments demonstrated that the particles effectively decreased the reactive oxygen species (ROS) levels in macrophages (RAW264.7). Furthermore, these particles exerted effects on M1 macrophage polarization, enhanced nitric oxide (NO) secretion to promote vascular regeneration, and facilitated rapid macrophage transition to the M2 phenotype. Subsequently, the particles facilitated human umbilical vein endothelial cell (HUVEC) migration. In vivo studies showed that these particles rapidly stimulated innate immune responses in animal models, leading to enhanced angiogenesis around the implant and improved osseointegration. In summary, the presence of bacteroid cerium oxide particles on the implant surface regulated and accelerated macrophage polarization, thereby enhancing angiogenesis during the immune response and improving peri-implant osseointegration.

Advancing Computational Toxicology by Interpretable Machine Learning.

Chemical toxicity evaluations for drugs, consumer products, and environmental chemicals have a critical impact on human health. Traditional animal models to evaluate chemical toxicity are expensive, time-consuming, and often fail to detect toxicants in humans. Computational toxicology is a promising alternative approach that utilizes machine learning (ML) and deep learning (DL) techniques to predict the toxicity potentials of chemicals. Although the applications of ML- and DL-based computational models in chemical toxicity predictions are attractive, many toxicity models are "black boxes" in nature and difficult to interpret by toxicologists, which hampers the chemical risk assessments using these models. The recent progress of interpretable ML (IML) in the computer science field meets this urgent need to unveil the underlying toxicity mechanisms and elucidate the domain knowledge of toxicity models. In this review, we focused on the applications of IML in computational toxicology, including toxicity feature data, model interpretation methods, use of knowledge base frameworks in IML development, and recent applications. The challenges and future directions of IML modeling in toxicology are also discussed. We hope this review can encourage efforts in developing interpretable models with new IML algorithms that can assist new chemical assessments by illustrating toxicity mechanisms in humans.

Integrating structure annotation and machine learning approaches to develop graphene toxicity models.

Modern nanotechnology provides efficient and cost-effective nanomaterials (NMs). The increasing usage of NMs arises great concerns regarding nanotoxicity in humans. Traditional animal testing of nanotoxicity is expensive and time-consuming. Modeling studies using machine learning (ML) approaches are promising alternatives to direct evaluation of nanotoxicity based on nanostructure features. However, NMs, including two-dimensional nanomaterials (2DNMs) such as graphenes, have complex structures making them difficult to annotate and quantify the nanostructures for modeling purposes. To address this issue, we constructed a virtual graphenes library using nanostructure annotation techniques. The irregular graphene structures were generated by modifying virtual nanosheets. The nanostructures were digitalized from the annotated graphenes. Based on the annotated nanostructures, geometrical nanodescriptors were computed using Delaunay tessellation approach for ML modeling. The partial least square regression (PLSR) models for the graphenes were built and validated using a leave-one-out cross-validation (LOOCV) procedure. The resulted models showed good predictivity in four toxicity-related endpoints with the coefficient of determination (R2) ranging from 0.558 to 0.822. This study provides a novel nanostructure annotation strategy that can be applied to generate high-quality nanodescriptors for ML model developments, which can be widely applied to nanoinformatics studies of graphenes and other NMs.

Biomimetic Mineralized Fiber Bundle-Inspired Scaffolding Surface on Polyetheretherketone Implants Promotes Osseointegration.

The stress shielding effect caused by traditional metal implants is circumvented by using polyetheretherketone (PEEK), due to its excellent mechanical properties; however, the biologically inert nature of PEEK limits its application. Endowing PEEK with biological activity to promote osseointegration would increase its applicability for bone replacement implants. A biomimetic study is performed, inspired by mineralized collagen fiber bundles that contact bone marrow mesenchymal stem cells (BMMSCs) on the native trabecular bone surface. The PEEK surface (P) is first sulfonated with sulfuric acid to form a porous network structure (sP). The surface is then encapsulated with amorphous hydroxyapatite (HA) by magnetron sputtering to form a biomimetic scaffold that resembles mineralized collagen fiber bundles (sPHA). Amorphous HA simulates the composition of osteogenic regions in vivo and exhibits strong biological activity. In vitro results show that more favorable cell adhesion and osteogenic differentiation can be attained with the novelsurface of sPHA than with SP. The results of in vivo experiments show that sPHA exhibits osteoinductive and osteoconductive activity and facilitates bone formation and osseointegration. Therefore, the surface modification strategy can significantly improve the biological activity of PEEK, facilitate effective osseointegration, and inspire further bionic modification of other inert polymers similar to PEEK.

TiO2nanotubes induce early mitochondrial fission in BMMSCs and promote osseointegration.

Nanotopography can promote osseointegration, but how bone marrow mesenchymal stem cells (BMMSCs) respond to this physical stimulus is unclear. Here, we found that early exposure of BMMSCs to nanotopography (6 h) caused mitochondrial fission rather than fusion, which was necessary for osseointegration. We analyzed the changes in mitochondrial morphology and function of BMMSCs located on the surfaces of NT100 (100 nm nanotubes) and ST (smooth) by super-resolution microscopy and other techniques. Then, we found that both ST and NT100 caused a significant increase in mitochondrial fission early on, but NT100 caused mitochondrial fission much earlier than those on ST. In addition, the mitochondrial functional statuses were good at the 6 h time point, this is at odds with the conventional wisdom that fusion is good. This fission phenomenon adequately protected mitochondrial membrane potential (MMP) and respiration and reduced reactive oxygen species. Interestingly, the MMP and oxygen consumption rate of BMMSCs were reduced when mitochondrial fission was inhibited by Mdivi-1(Inhibition of dynamin-related protein 1 fission) in the early stage. In addition, the effect on osseointegration was significantly worse, and this effect did not improve with time. Taken together, the findings indicate that early mitochondrial fission plays an important role in nanotopography-mediated promotion of osseointegration, which is of great significance to the surface structure design of biomaterials.

Mechanism-driven modeling of chemical hepatotoxicity using structural alerts and an in vitro screening assay.

Traditional experimental approaches to evaluate hepatotoxicity are expensive and time-consuming. As an advanced framework of risk assessment, adverse outcome pathways (AOPs) describe the sequence of molecular and cellular events underlying chemical toxicities. We aimed to develop an AOP that can be used to predict hepatotoxicity by leveraging computational modeling and in vitro assays. We curated 869 compounds with known hepatotoxicity classifications as a modeling set and extracted assay data from PubChem. The antioxidant response element (ARE) assay, which quantifies transcriptional responses to oxidative stress, showed a high correlation to hepatotoxicity (PPV=0.82). Next, we developed quantitative structure-activity relationship (QSAR) models to predict ARE activation for compounds lacking testing results. Potential toxicity alerts were identified and used to construct a mechanistic hepatotoxicity model. For experimental validation, 16 compounds in the modeling set and 12 new compounds were selected and tested using an in-house ARE-luciferase assay in HepG2-C8 cells. The mechanistic model showed good hepatotoxicity predictivity (accuracy = 0.82) for these compounds. Potential false positive hepatotoxicity predictions by only using ARE results can be corrected by incorporating structural alerts and vice versa. This mechanistic model illustrates a potential toxicity pathway for hepatotoxicity, and this strategy can be expanded to develop predictive models for other complex toxicities.

Replacement per- and polyfluoroalkyl substances (PFAS) are potent modulators of lipogenic and drug metabolizing gene expression signatures in primary human hepatocytes.

Per- and polyfluoroalkyl substances (PFAS) are a class of environmental toxicants, and some, such as perfluorooctanesulfonic acid (PFOS) and perfluorooctanoic acid (PFOA), have been associated with hepatic steatosis in rodents and monkeys. It was hypothesized that perfluorosulfonic acids (C4, 6, 8), perfluorocarboxylic acids (C4-14), perfluoro(2-methyl-3-oxahexanoic) acid (HFPO-DA), 1H, 1H, 2H, 2H-perfluorooctanesulfonic acid (6:2 FTS) along with 3 PFOS precursors could induce expression of lipid metabolism genes and lipid deposition in human hepatocytes. Five-donor pooled cryopreserved human hepatocytes were cultured and treated with 0.1% DMSO vehicle or various PFAS (0.25 to 25 µM) in media. After a 48-h treatment, mRNA transcripts related to lipid transport, metabolism, and synthesis were measured using a Quantigene Plex assay. After 72-h treatments, hepatocytes were stained with Nile Red dye to quantify intracellular lipids. Overall, PFAS were transcriptionally active at 25 µM. In this model, lipid accumulation was not observed with C8-C12 treatments. Shorter chain PFAS (C4-C5), 6:2 FTS, and PFOS precursor, metFOSA, induced significant liver lipid accumulation, and gene activation at lower concentrations than legacy PFAS. In summary short chain PFAS and other alternative PFAS were more potent gene inducers, and potential health effects of replacement PFAS should be critically evaluated in humans.

Strontium-loaded titanium implant with rough surface modulates osseointegration by changing sfrp4 in canonical and noncanonical Wnt signaling pathways.

A rough morphology and strontium (Sr) can activate the Wnt pathway to regulate bone mesenchymal stem cells (rBMSCs) osteogenic differentiation, but the mechanism remains unclear. We constructed smooth Ti (ST) surfaces, rough Ti (RT) surfaces subjected to hydrofluoric acid etching, strontium-loaded smooth Ti (ST-Sr) surfaces subjected to magnetron sputtering, and rough strontium-loaded Ti (RT-Sr) surfaces. We systematically studied thein vitroosteogenic differentiation of rBMSCs on these four surfaces by alkaline phosphatase measurement, Alizarin Red staining and polymerase chain reaction (PCR). We also investigated whether crosstalk of the canonical and noncanonical Wnt signaling pathways regulated by sfrp4, which is an inhibitor of canonical and noncanonical Wnt, is the underlying mechanism via PCR on rBMSCs in different stages of osteogenic differentiation. We confirmed the effect of sfrp4 through anin vivosfrp4-siRNA test. Thein vitroosteogenic differentiation of rBMSCs decreased in the order RT-Sr, RT, ST-Sr, and ST. Regarding the mechanism, rough morphology and Sr both enhanced the canonical Wnt pathway to promote osseointegration. Additionally, rough morphology can inhibit sfrp4 to activate the noncanonical Wnt pathway, and then, the activated noncanonical Wnt pathway can suppress the canonical Wnt pathway at the early stage of osteogenic differentiation. Sr continuously enhanced sfrp4 to inhibit the canonical Wnt pathway instead of activating the noncanonical Wnt pathway. Interestingly, the effect of rough morphology on sfrp4 changed from inhibition to enhancement, and the enhancing effect of Sr on sfrp4 was gradually attenuated. The results of thein vivosfrp4-siRNA test showed that osseointegration decreased in the order RT-Sr, RT-Sr-siRNA, and ST. Our results suggest that the lack of sfrp4 could suppress osseointegration, indicating that sfrp4 acts as a crucial regulatory molecule for the canonical and noncanonical Wnt pathways during the response of rBMSCs to rough morphology and Sr.

Construction of a Virtual Opioid Bioprofile: A Data-Driven QSAR Modeling Study to Identify New Analgesic Opioids.

Compared to traditional experimental approaches, computational modeling is a promising strategy to efficiently prioritize new candidates with low cost. In this study, we developed a novel data mining and computational modeling workflow proven to be applicable by screening new analgesic opioids. To this end, a large opioid data set was used as the probe to automatically obtain bioassay data from the PubChem portal. There were 114 PubChem bioassays selected to build quantitative structure-activity relationship (QSAR) models based on the testing results across the probe compounds. The compounds tested in each bioassay were used to develop 12 models using the combination of three machine learning approaches and four types of chemical descriptors. The model performance was evaluated by the coefficient of determination (R 2) obtained from 5-fold cross-validation. In total, 49 models developed for 14 bioassays were selected based on the criteria and were identified to be mainly associated with binding affinities to different opioid receptors. The models for these 14 bioassays were further used to fill data gaps in the probe opioids data set and to predict general drug compounds in the DrugBank data set. This study provides a universal modeling strategy that can take advantage of large public data sets for computer-aided drug design (CADD).

Na/K-ATPase Y260 Phosphorylation-mediated Src Regulation in Control of Aerobic Glycolysis and Tumor Growth.

We report here the identification of α1 Na/K-ATPase as a major regulator of the proto-oncogene Src kinase and the role of this regulation in control of Warburg effect and tumor growth. Specifically, we discovered Y260 in α1 Na/K-ATPase as a Src-specific phosphorylation and binding site and that Y260 phosphorylation is required for Src-mediated signal transduction in response to a number of stimuli including EGF. As such, it enables a dynamic control of aerobic glycolysis. However, such regulation appears to be lost or attenuated in human cancers as the expression of Na/K-ATPase α1 was significantly decreased in prostate, breast and kidney cancers, and further reduced in corresponding metastatic lesions in patient samples. Consistently, knockdown of α1 Na/K-ATPase led to a further increase in lactate production and the growth of tumor xenograft. These findings suggest that α1 Na/K-ATPase works as a tumor suppressor and that a loss of Na/K-ATPase-mediated Src regulation may lead to Warburg phenotype in cancer.

A humanized anti-DLL4 antibody promotes dysfunctional angiogenesis and inhibits breast tumor growth.

Blockage of Delta-like 4 (DLL4)-directed Notch signaling induces excessive tip cell formation and endothelial proliferation resulting in dysfunctional angiogenesis in tumors. MMGZ01, as a murine anti-human DLL4 monoclonal antibody, specifically binds to human DLL4 and blocks Notch pathway. Here, the structure of MMGZ01 variable fragment (Fv) was established and framework region (FR) residues which supported complementarily determining region (CDR) loop conformation were identified. Important residues interactions were also identified through docking MMGZ01 Fv with antigen epitope in DLL4. To humanize the murine antibody, we modified MMGZ01 Fv through CDR grafting and the reconstructed antibody (H3L2) maintained similar structure and binding affinity to parental MMGZ01 after back mutation of 12 canonical murine residues in the FRs. Meanwhile, H3L2 promoted human umbilical vein endothelial cell (HUVEC) proliferation through inhibiting DLL4-directed Notch pathway. Moreover, in MDA-MB-231-bearing nude mice, H3L2 induced dysfunctional angiogenesis and tumor cell apoptosis and showed superior anti-tumor activity. In conclusion, H3L2 is an ideal humanized antibody that inhibits tumor growth through targeting DLL4-Notch pathway and has attracting potentials for clinical applications.

MMGZ01, an anti-DLL4 monoclonal antibody, promotes nonfunctional vessels and inhibits breast tumor growth.

Increasing evidence suggests that DLL4 (Delta-like 4)-Notch signaling plays a critical role in cell fate determination and differentiation in tissues. Blocking DLL4-Notch signaling results in inhibition of tumor growth, which is associated with increased nonfunctional vessels and poor perfusion in the tumor. We successfully generated a human DLL4 monoclonal antibody MMGZ01 that binds specifically to DLL4 to disrupt the interaction between DLL4 and Notch1. MMGZ01 showed high affinity to DLL4 to inhibit the DLL4-mediated human umbilical vein endothelial cell (HUVEC) phenotype. Furthermore, MMGZ01 stimulated HUVEC vessel sprouting and tubule formation in vitro. In addition, MMGZ01 had a pronounced effect in promoting immature vessels and reduced breast cancer cell growth in vivo. Finally, MMGZ01 treatment inhibited the proliferation of breast cancer cells, induced tumor cell apoptosis, suppressed mammosphere formation, decreased CD44(+)/CD24(-) cell population, and reduced epithelial mesenchymal transition (EMT). These findings suggest that antagonism of the DLL4-Notch signaling pathway might provide a potential therapeutic approach for breast cancer treatment.

A bispecific protein rG7S-MICA recruits natural killer cells and enhances NKG2D-mediated immunosurveillance against hepatocellular carcinoma.

MHC class I-related chain A (MICA) is a principal immunoligand of the natural killer (NK) cell receptor NK group 2, member D (NKG2D) and plays a key role in NK cell-mediated immune recognition. Shedding of MICA from tumor cells leads to immunosuppression. To reconstitute the immunosurveilance function of NK cells, we constructed a fusion protein rG7S-MICA and explored its potential anti-tumor activity against hepatocellular carcinoma (HCC). rG7S-MICA consists of human MICA and a single-chain antibody fragment (scFv) targeting the tumor-associated antigen cluster of differentiation 24 (CD24). In vitro, rG7S-MICA engaged both NK cells and CD24(+) human HCC cells, and triggered NK cell-mediated cytolysis. Furthermore, in CD24(+) HCC-bearing nude mice, rG7S-MICA specifically targeted to the tumor tissue, where it effectively recruited NK cells and induced the release of cytokines, and showed superior anti-tumor activity. In conclusion, rG7S-MICA provides a new approach for HCC-targeting immunotherapy and has attracting potentials for clinical applications.