Schwann cells in neuromuscular in vitro models.

Neuromuscular cell culture models are used to investigate synapse formation and function, as well as mechanisms of de-and regeneration in neuromuscular diseases. Recent developments including 3D culture technique and hiPSC technology have propelled their ability to complement insights from in vivo models. However, most cultures have not considered Schwann cells, the glial part of NMJs. In the following, a brief overview of different types of neuromuscular cocultures is provided alongside examples for studies that included Schwann cells. From these, findings concerning the effects of Schwann cells on those cultures are summarized and future lines of research are proposed.

Mechanistic modeling and simulation of a complex low and high loading elution behavior of a polypeptide in cation exchange chromatography.

The mechanistic modeling of preparative liquid chromatography is still a challenging task. Nonideal thermodynamic conditions may require activity coefficients for the mechanistic description of preparative chromatography. In this work, a chromatographic cation exchange step with a polypeptide having a complex elution behavior in low and high loading situations is modeled. Model calibration in the linear range of the isotherm is done by applying counterion-induced linear gradient elution experiments between pH 3.3 and 4.3. Inverse fitting with column loads up to 25 mg/mLCV is performed for parameter estimation in the nonlinear range. The polypeptide elution peak shows an anti-Langmuirian behavior with fronting under low loading conditions and a switch to a Langmuirian behavior with increasing load. This unusual elution behavior could be described with an extended version of the sigmoidal Self-Association isotherm including two activity coefficients for the polypeptide and counterion in solution. The activity coefficient of the solute polypeptide shows a strong influence on the model parameters and is crucial in the linear and nonlinear range of the isotherm. The modeling procedure results in a unique and robust model parameter set that is sufficient to describe the complex elution behavior and allows modeling over the full isotherm range.

Glyco-engineered HEK 293-F cell lines for the production of therapeutic glycoproteins with human N-glycosylation and improved pharmacokinetics.

N-glycosylated proteins produced in human embryonic kidney 293 (HEK 293) cells often carry terminal N-acetylgalactosamine (GalNAc) and only low levels of sialylation. On therapeutic proteins, such N-glycans often trigger rapid clearance from the patient's bloodstream via efficient binding to asialoglycoprotein receptor (ASGP-R) and mannose receptor (MR). This currently limits the use of HEK 293 cells for therapeutic protein production. To eliminate terminal GalNAc, we knocked-out GalNAc transferases B4GALNT3 and B4GALNT4 by CRISPR/Cas9 in FreeStyle 293-F cells. The resulting cell line produced a coagulation factor VII-albumin fusion protein without GalNAc but with increased sialylation. This glyco-engineered protein bound less efficiently to both the ASGP-R and MR in vitro and it showed improved recovery, terminal half-life and area under the curve in pharmacokinetic rat experiments. By overexpressing sialyltransferases ST6GAL1 and ST3GAL6 in B4GALNT3 and B4GALNT4 knock-out cells, we further increased factor VII-albumin sialylation; for ST6GAL1 even to the level of human plasma-derived factor VII. Simultaneous knock-out of B4GALNT3 and B4GALNT4 and overexpression of ST6GAL1 further lowered factor VII-albumin binding to ASGP-R and MR. This novel glyco-engineered cell line is well-suited for the production of factor VII-albumin and presumably other therapeutic proteins with fully human N-glycosylation and superior pharmacokinetic properties.

Mechanistic modeling of ligand density variations on anion exchange chromatography.

Ion exchange chromatography is a powerful and ubiquitous unit operation in the purification of therapeutic proteins. However, the performance of an ion-exchange process depends on a complex interrelationship between several parameters, such as protein properties, mobile phase conditions, and chromatographic resin characteristics. Consequently, batch variations of ion exchange resins play a significant role in the robustness of these downstream processing steps. Ligand density is known to be one of the main lot-to-lot variations, affecting protein adsorption and separation performance. The use of a model-based approach can be an effective tool for comprehending the impact of parameter variations (e.g., ligand density) and their influence on the process. The objective of this work was to apply mechanistic modeling to gain a deeper understanding of the influence of ligand density variations in anion exchange chromatography. To achieve this, 13 prototype resins having the same support as the strong anion exchange resin Fractogel® EMD TMAE (M), but differing in ligand density, were analyzed. Linear salt gradient elution experiments were performed to observe the elution behavior of a monoclonal antibody and bovine serum albumin. A proposed isotherm model for ion exchange chromatography, describing the dependence of ligand density variations on protein retention, was successfully applied.

An alternative pathway for sweet sensation: possible mechanisms and physiological relevance.

Sweet substances are detected by taste-bud cells upon binding to the sweet-taste receptor, a T1R2/T1R3 heterodimeric G protein-coupled receptor. In addition, experiments with mouse models lacking the sweet-taste receptor or its downstream signaling components led to the proposal of a parallel "alternative pathway" that may serve as metabolic sensor and energy regulator. Indeed, these mice showed residual nerve responses and behavioral attraction to sugars and oligosaccharides but not to artificial sweeteners. In analogy to pancreatic ß cells, such alternative mechanism, to sense glucose in sweet-sensitive taste cells, might involve glucose transporters and KATP channels. Their activation may induce depolarization-dependent Ca2+ signals and release of GLP-1, which binds to its receptors on intragemmal nerve fibers. Via unknown neuronal and/or endocrine mechanisms, this pathway may contribute to both, behavioral attraction and/or induction of cephalic-phase insulin release upon oral sweet stimulation. Here, we critically review the evidence for a parallel sweet-sensitive pathway, involved signaling mechanisms, neural processing, interactions with endocrine hormonal mechanisms, and its sensitivity to different stimuli. Finally, we propose its physiological role in detecting the energy content of food and preparing for digestion.

Sensing Senses: Optical Biosensors to Study Gustation.

The five basic taste modalities, sweet, bitter, umami, salty and sour induce changes of Ca2+ levels, pH and/or membrane potential in taste cells of the tongue and/or in neurons that convey and decode gustatory signals to the brain. Optical biosensors, which can be either synthetic dyes or genetically encoded proteins whose fluorescence spectra depend on levels of Ca2+, pH or membrane potential, have been used in primary cells/tissues or in recombinant systems to study taste-related intra- and intercellular signaling mechanisms or to discover new ligands. Taste-evoked responses were measured by microscopy achieving high spatial and temporal resolution, while plate readers were employed for higher throughput screening. Here, these approaches making use of fluorescent optical biosensors to investigate specific taste-related questions or to screen new agonists/antagonists for the different taste modalities were reviewed systematically. Furthermore, in the context of recent developments in genetically encoded sensors, 3D cultures and imaging technologies, we propose new feasible approaches for studying taste physiology and for compound screening.

Compression Bioreactor-Based Mechanical Loading Induces Mobilization of Human Bone Marrow-Derived Mesenchymal Stromal Cells into Collagen Scaffolds In Vitro.

Articular cartilage (AC) is an avascular tissue composed of scattered chondrocytes embedded in a dense extracellular matrix, in which nourishment takes place via the synovial fluid at the surface. AC has a limited intrinsic healing capacity, and thus mainly surgical techniques have been used to relieve pain and improve function. Approaches to promote regeneration remain challenging. The microfracture (MF) approach targets the bone marrow (BM) as a source of factors and progenitor cells to heal chondral defects in situ by opening small holes in the subchondral bone. However, the original function of AC is not obtained yet. We hypothesize that mechanical stimulation can mobilize mesenchymal stromal cells (MSCs) from BM reservoirs upon MF of the subchondral bone. Thus, the aim of this study was to compare the counts of mobilized human BM-MSCs (hBM-MSCs) in alginate-laminin (alginate-Ln) or collagen-I (col-I) scaffolds upon intermittent mechanical loading. The mechanical set up within an established bioreactor consisted of 10% strain, 0.3 Hz, breaks of 10 s every 180 cycles for 24 h. Contrary to previous findings using porcine MSCs, no significant cell count was found for hBM-MSCs into alginate-Ln scaffolds upon mechanical stimulation (8 ± 5 viable cells/mm3 for loaded and 4 ± 2 viable cells/mm3 for unloaded alginate-Ln scaffolds). However, intermittent mechanical stimulation induced the mobilization of hBM-MSCs into col-I scaffolds 10-fold compared to the unloaded col-I controls (245 ± 42 viable cells/mm3 vs. 22 ± 6 viable cells/mm3, respectively; p-value < 0.0001). Cells that mobilized into the scaffolds by mechanical loading did not show morphological changes. This study confirmed that hBM-MSCs can be mobilized in vitro from a reservoir toward col-I but not alginate-Ln scaffolds upon intermittent mechanical loading, against gravity.

A novel spheroid-based co-culture model mimics loss of keratinocyte differentiation, melanoma cell invasion, and drug-induced selection of ABCB5-expressing cells.

BACKGROUND: Different 3D-cell culture approaches with varying degrees of complexity have been developed to serve as melanoma models for drug testing or mechanistic studies. While these 3D-culture initiatives are already often superior to classical 2D approaches, they are either composed of only melanoma cells or they are so complex that the behavior of individual cell types is hard to understand, and often they are difficult to establish and expensive. METHODS: This study used low-attachment based generation of spheroids composed of up to three cell types. Characterization of cells and spheroids involved cryosectioning, immunofluorescence, FACS, and quantitative analyses. Statistical evaluation used one-way ANOVA with post-hoc Tukey test or Student's t-test. RESULTS: The tri-culture model allowed to track cellular behavior in a cell-type specific manner and recapitulated different characteristics of early melanoma stages. Cells arranged into a collagen-IV rich fibroblast core, a ring of keratinocytes, and groups of highly proliferating melanoma cells on the outside. Regularly, some melanoma cells were also found to invade the fibroblast core. In the absence of melanoma cells, the keratinocyte ring stratified into central basal-like and peripheral, more differentiated cells. Conversely, keratinocyte differentiation was clearly reduced upon addition of melanoma cells. Treatment with the cytostatic drug, docetaxel, restored keratinocyte differentiation and induced apoptosis of external melanoma cells. Remaining intact external melanoma cells showed a significantly increased amount of ABCB5-immunoreactivity. CONCLUSIONS: In the present work, a novel, simple spheroid-based melanoma tri-culture model composed of fibroblasts, keratinocytes, and melanoma cells was described. This model mimicked features observed in early melanoma stages, including loss of keratinocyte differentiation, melanoma cell invasion, and drug-induced increase of ABCB5 expression in external melanoma cells.

Postnatal Development and Distribution of Sympathetic Innervation in Mouse Skeletal Muscle.

Vertebrate neuromuscular junctions (NMJs) have been conceived as tripartite synapses composed of motor neuron, Schwann cell, and muscle fiber. Recent work has shown the presence of sympathetic neurons in the immediate vicinity of NMJs and experimental and clinical findings suggest that this plays an eminent role in adult NMJ biology. The present study examined the postnatal development and distribution of sympathetic innervation in different muscles using immunofluorescence, confocal microscopy, and Western blot. This demonstrates the proximity of sympathetic neurons in diaphragm, extensor digitorum longus, tibialis anterior, soleus, and levator auris longus muscles. In extensor digitorum longus muscle, sympathetic innervation of NMJs was quantified from perinatal to adult stage and found to increase up to two months of age. In diaphragm muscle, an extensive network of sympathetic neurons was prominent along the characteristic central synapse band. In summary, these data demonstrate that an elaborate sympathetic innervation is present in several mouse skeletal muscles and that this is often next to NMJs. Although the presence of sympathetic neurons at the perisynaptic region of NMJs increased during postnatal development, many synapses were already close to sympathetic neurons at birth. Potential implications of these findings for treatment of neuromuscular diseases are discussed.

N-octanoyl dopamine treatment exerts renoprotective properties in acute kidney injury but not in renal allograft recipients.

BACKGROUND: N-octanoyl dopamine (NOD) treatment improves renal function when applied to brain dead donors and in the setting of warm ischaemia-induced acute kidney injury (AKI). Because it also activates transient receptor potential vanilloid type 1 (TRPV1) channels, we first assessed if NOD conveys its renoprotective properties in warm ischaemia-induced AKI via TRPV1 and secondly, if renal transplant recipients also benefit from NOD treatment. METHODS: We induced warm renal ischaemia in Lewis, wild-type (WT) and TRPV1(-/-) Sprague-Dawley (sd) rats by clamping the left renal artery for 45 min. Transplantations were performed in allogeneic and syngeneic donor-recipient combinations (Fisher to Lewis and Lewis to Lewis) with a cold ischaemia time of 20 h. Treatment was instituted directly after restoration of organ perfusion. Renal function, histology and perfusion were assessed by serum creatinine, microscopy and magnetic resonance imaging (MRI) using arterial spin labelling (ASL). RESULTS: NOD treatment significantly improved renal function in Lewis rats after warm ischaemia-induced AKI. It was, however, not effective after prolonged cold ischaemia. The renoprotective properties of NOD were only observed in Lewis or WT, but not in TRPV1(-/-) sd rats. Renal inflammation was significantly abrogated by NOD. MRI-ASL showed a significantly lower cortical perfusion in ischaemic when compared with non-ischaemic kidneys. No overall differences were observed in renal perfusion between NOD- and NaCl-treated rats. CONCLUSIONS: NOD treatment reduces renal injury in warm ischaemia, but is not effective in renal transplant in our experimental animal models. The salutary effect of NOD appears to be TPRV1-dependent, not involving large changes in renal perfusion.

Quantification of Hepcidin-related Iron Accumulation in the Rat Liver.

Hepcidin was originally detected as a liver peptide with antimicrobial activity and it functions as a central regulator in the systemic iron metabolism. Consequently suppression of hepcidin leads to iron accumulation in the liver. AbbVie developed a monoclonal antibody ([mAb]; repulsive guidance molecule [RGMa/c] mAb) that downregulates hepcidin expression by influencing the RGMc/bone morphogenetic protein (BMP)/neogenin receptor complex and causes iron deposition in the liver. In a dose range finding study with RGMa/c mAb, rats were treated with different dose levels for a total of 4 weekly doses. The results of this morphometric analysis in the liver showed that iron accumulation is not homogenous between liver lobes and the left lateral lobe was the most responsive lobe in the rat. Quantitative hepcidin messenger RNA analysis showed that the left lateral lobe was the most responsive lobe showing hepcidin downregulation with increasing antibody dose. In addition, the morphometric analysis had higher sensitivity than the chemical iron extraction and quantification using a colorimetric assay. In conclusion, the Prussian blue stain in combination with semi-quantitative and quantitative morphometric analysis is the most reliable method to demonstrate iron accumulation in the liver compared to direct measurement of iron in unfixed tissue using a colorimetric assay.

Rapsyn mediates subsynaptic anchoring of PKA type I and stabilisation of acetylcholine receptor in vivo.

The stabilisation of acetylcholine receptors (AChRs) at the neuromuscular junction depends on muscle activity and the cooperative action of myosin Va and protein kinase A (PKA) type I. To execute its function, PKA has to be present in a subsynaptic microdomain where it is enriched by anchoring proteins. Here, we show that the AChR-associated protein, rapsyn, interacts with PKA type I in C2C12 and T-REx293 cells as well as in live mouse muscle beneath the neuromuscular junction. Molecular modelling, immunoprecipitation and bimolecular fluorescence complementation approaches identify an α-helical stretch of rapsyn to be crucial for binding to the dimerisation and docking domain of PKA type I. When expressed in live mouse muscle, a peptide encompassing the rapsyn α-helical sequence efficiently delocalises PKA type I from the neuromuscular junction. The same peptide, as well as a rapsyn construct lacking the α-helical domain, induces severe alteration of acetylcholine receptor turnover as well as fragmentation of synapses. This shows that rapsyn anchors PKA type I in close proximity to the postsynaptic membrane and suggests that this function is essential for synapse maintenance.

Modeling of salt and pH gradient elution in ion-exchange chromatography.

The separation of proteins by internally and externally generated pH gradients in chromatofocusing on ion-exchange columns is a well-established analytical method with a large number of applications. In this work, a stoichiometric displacement model was used to describe the retention behavior of lysozyme on SP Sepharose FF and a monoclonal antibody on Fractogel SO3 (S) in linear salt and pH gradient elution. The pH dependence of the binding charge B in the linear gradient elution model is introduced using a protein net charge model, while the pH dependence of the equilibrium constant is based on a thermodynamic approach. The model parameter and pH dependences are calculated from linear salt gradient elutions at different pH values as well as from linear pH gradient elutions at different fixed salt concentrations. The application of the model for the well-characterized protein lysozyme resulted in almost identical model parameters based on either linear salt or pH gradient elution data. For the antibody, only the approach based on linear pH gradients is feasible because of the limited pH range useful for salt gradient elution. The application of the model for the separation of an acid variant of the antibody from the major monomeric form is discussed.

Quantitative Analysis of Whole-Mount Fluorescence-Stained Tumor Spheroids in Phenotypic Drug Screens.

Three-dimensional cell cultures, such as spheroids or organoids, serve as important models for drug screening purposes. Optical tissue clearing (OTC) enhances the visualization of fluorescence stainings and enables in toto microscopy of 3D cell culture models. Furthermore, subsequent automated image analysis tools convert qualitative confocal image sets into quantitative data. In this chapter, we describe a detailed protocol for preparation of HT29 cancer spheroids, 3D in toto immunostaining, glycerol-based OTC, whole-mount imaging, and semi-automated downstream image processing and segmentation for nuclear image analysis using open-source software.

Molecular structure of human GM-CSF in complex with a disease-associated anti-human GM-CSF autoantibody and its potential biological implications.

Polyclonal autoantibodies against human GM-CSF (granulocyte/macrophage colony-stimulating factor) are a hallmark of PAP (pulmonary alveolar proteinosis) and several other reported autoimmune diseases. MB007 is a high-affinity anti-(human GM-CSF) autoantibody isolated from a patient suffering from PAP which shows only modest neutralization of GM-CSF bioactivity. We describe the first crystal structure of a cytokine-directed human IgG1λ autoantibody-binding fragment (Fab) at 1.9 Å (1 Å=0.1 nm) resolution. Its CDR3-H substantially differs from all VH7 germline IgG1 structures reported previously. We derive a reliable model of the antigen-autoantibody complex by using NMR chemical shift perturbation data in combination with computational methods. Superposition of the modelled complex structure with the human GM-CSF-GM-CSF ternary receptor complex reveals only little overlap between receptor and Fab when bound to GM-CSF. Our model provides a structural basis for understanding the mode of action of the MB007 autoantibody.

Two peak elution behavior of a monoclonal antibody in cation exchange chromatography as a screening tool for excipients.

Aggregation of proteins is a critical quality attribute and a major concern during the purification of therapeutic proteins, like monoclonal antibodies. In-solution experiments applying different stress scenarios, e.g., mechanical, or physical stresses, can determine the overall conformational stability of the protein to enhance drug product shelf-life. Several groups have reported surface-induced unfolding and aggregation of monoclonal antibodies and their derivatives during cation exchange chromatography, which results in a two-peak elution behavior of the protein and its species. We have investigated universal influencing factors, like temperature and hold time, on this phenomenon. The formation of the second peak is a kinetic process, which is strongly influenced by temperature during the hold time. However, our main focus was the application of excipients and their influence on the two-peak elution behavior. We compared the on-column screening results with results obtained through a "traditional" in-solution screening using nanoDSF. Mostly, stabilizing excipients, like Sucrose, show their stabilizing abilities in both systems, but some discrepancies, e.g., using Arginine, between the two orthogonal techniques show the potential of the on-column screening system to lead to unexpected results, which would not necessarily be visible in in-solution experiments.

On-line identification of P-glycoprotein substrates by monitoring of extracellular acidification and respiration rates in living cells.

The influence of P-glycoprotein (ABCB1) in drug resistance as well as drug absorption and disposition is an important factor to be considered during the development of new drugs. Thus, the early identification and exclusion of compounds showing a high affinity towards P-glycoprotein can help to select drug candidates. The aim of our study was to implement a label-free assay for the identification of P-glycoprotein substrates in living cells. For this approach, a multiparametric, chip-based sensor system was used to determine extracellular acidification, cell respiration and adhesion upon stimulation with P-glycoprotein substrates. Using L-MDR1 cells, a human P-glycoprotein overexpressing cell line, the influence of P-glycoprotein activity was determined for seven different compounds, demonstrating the applicability of the system for P-glycoprotein substrate identification. Effects were concentration dependent, as shown for the P-glycoprotein substrate verapamil, and were associated with cellular acidification and respiration. P-glycoprotein ATPase activation by verapamil could be described by a Michaelis-Menten type kinetic profile showing saturation at high substrate concentrations. The Michaelis-Menten constants K(M) were determined to be 0.92µM (calculated based on extracellular acidification) and 4.9µM (calculated based on cellular respiration). Control experiments using 100nM of the P-glycoprotein inhibitor elacridar indicated that the observed effects were related to P-glycoprotein ATPase activity. In contrast, wild-type LLC-PK1 cells not expressing P-glycoprotein were not responsive towards stimulation with different P-glycoprotein substrates. Summarizing these findings, the used microsensor system is a generic system suitable for the identification of P-glycoprotein substrates. In contrast to biochemical P-glycoprotein assays, activation of the drug efflux pump can be monitored on-line in living cells to identify P-glycoprotein substrates and to study the molecular mechanisms of adenosintriphosphate-dependent active transport.

Application of the Steric Mass Action formalism for modeling under high loading conditions: Part 1. Investigation of the influence of pH on the steric shielding factor.

In ion exchange chromatography, the Steric Mass Action (SMA) formalism is frequently used to simulate sorption processes at low and high column load conditions. To apply the SMA model for describing protein elution over wide ranges of pH, it is necessary to use pH-dependent model parameters. In the past, some publications have already described the pH-dependence of the characteristic protein charge and the equilibrium constant, while the influence of pH on the steric shielding factor has been mostly neglected. In this work, the pH-dependences of all relevant model parameters, including the shielding factor, were investigated, described, and implemented into the SMA model. Therefore, the elution behavior of a bispecific monoclonal antibody on the strong cation exchange resin POROS™ XS was modeled over broad ranges of pH, salt concentrations, and protein concentrations. Linear gradient elution experiments were performed to generate an extensive data set by using increasing column loadings from 0.5 up to 75.0 mgbsAb/mLresin. By using an inverse peak fitting method, shielding factors were estimated at various pH values ranging from 4.5 to 8.9. The results showed that an increasing buffer pH resulted in strongly increasing shielding factors. A semi-empirical correlation describing the shielding factor as a function of pH was established and implemented into the SMA formalism. This approach led to precise prediction of protein elution behavior using a single-component simulation. This was demonstrated by accurate simulation of linear salt, pH and dual gradient elution experiments conducted under high loading conditions.

Development and In Vitro Differentiation of Schwann Cells.

Schwann cells are glial cells of the peripheral nervous system. They exist in several subtypes and perform a variety of functions in nerves. Their derivation and culture in vitro are interesting for applications ranging from disease modeling to tissue engineering. Since primary human Schwann cells are challenging to obtain in large quantities, in vitro differentiation from other cell types presents an alternative. Here, we first review the current knowledge on the developmental signaling mechanisms that determine neural crest and Schwann cell differentiation in vivo. Next, an overview of studies on the in vitro differentiation of Schwann cells from multipotent stem cell sources is provided. The molecules frequently used in those protocols and their involvement in the relevant signaling pathways are put into context and discussed. Focusing on hiPSC- and hESC-based studies, different protocols are described and compared, regarding cell sources, differentiation methods, characterization of cells, and protocol efficiency. A brief insight into developments regarding the culture and differentiation of Schwann cells in 3D is given. In summary, this contribution provides an overview of the current resources and methods for the differentiation of Schwann cells, it supports the comparison and refinement of protocols and aids the choice of suitable methods for specific applications.

3D Melanoma Cocultures as Improved Models for Nanoparticle-Mediated Delivery of RNA to Tumors.

Cancer therapy is an emergent application for mRNA therapeutics. While in tumor immunotherapy, mRNA encoding for tumor-associated antigens is delivered to antigen-presenting cells in spleen and lymph nodes, other therapeutic options benefit from immediate delivery of mRNA nanomedicines directly to the tumor. However, tumor targeting of mRNA therapeutics is still a challenge, since, in addition to delivery of the cargo to the tumor, specifics of the targeted cell type as well as its interplay with the tumor microenvironment are crucial for successful intervention. This study investigated lipoplex nanoparticle-mediated mRNA delivery to spheroid cell culture models of melanoma. Insights into cell-type specific targeting, non-cell-autonomous effects, and penetration capacity in tumor and stroma cells of the mRNA lipoplex nanoparticles were obtained. It was shown that both coculture of different cell types as well as three-dimensional cell growth characteristics can modulate distribution and transfection efficiency of mRNA lipoplex formulations. The results demonstrate that three-dimensional coculture spheroids can provide a valuable surplus of information in comparison to adherent cells. Thus, they may represent in vitro models with enhanced predictivity for the in vivo activity of cancer nanotherapeutics.