Search for: All records

The stein variational gradient descent (SVGD) algorithm is a deterministic particle method for sampling. However, a mean-field analysis reveals that the gradient flow corresponding to the SVGD algorithm (i.e., the Stein Variational Gradient Flow) only provides a constant-order approximation to the Wasserstein gradient flow corresponding to the KL-divergence minimization. In this work, we propose the Regularized Stein Variational Gradient Flow, which interpolates between the Stein Variational Gradient Flow and the Wasserstein gradient flow. We establish various theoretical properties of the Regularized Stein Variational Gradient Flow (and its time-discretization) including convergence to equilibrium, existence and uniqueness of weak solutions, and stability of the solutions. We provide preliminary numerical evidence of the improved performance offered by the regularization.

Extracellular vesicles (EVs) have been identified as promising biomarkers for the noninvasive diagnosis of various diseases. However, challenges in separating EVs from soluble proteins have resulted in variable EV recovery rates and low purities. Here, we report a high-yield ( > 90%) and rapid ( < 10 min) EV isolation method calledFLocculation viaOrbitalAcousticTrapping (FLOAT). The FLOAT approach utilizes an acoustofluidic droplet centrifuge to rotate and controllably heat liquid droplets. By adding a thermoresponsive polymer flocculant, nanoparticles as small as 20 nm can be rapidly and selectively concentrated at the center of the droplet. We demonstrate the ability of FLOAT to separate urinary EVs from the highly abundant Tamm-Horsfall protein, addressing a significant obstacle in the development of EV-based liquid biopsies. Due to its high-yield nature, FLOAT reduces biofluid starting volume requirements by a factor of 100 (from 20 mL to 200 µL), demonstrating its promising potential in point-of-care diagnostics.

While mesenchymal stem cells (MSCs) have gained enormous attention due to their unique properties of self-renewal, colony formation, and differentiation potential, the MSC secretome has become attractive due to its roles in immunomodulation, anti-inflammatory activity, angiogenesis, and anti-apoptosis. However, the precise stimulation and efficient production of the MSC secretome for therapeutic applications are challenging problems to solve. Here, we report on Acoustofluidic Interfaces for the Mechanobiological Secretome of MSCs: AIMS. We create an acoustofluidic mechanobiological environment to form reproducible three-dimensional MSC aggregates, which produce the MSC secretome with high efficiency. We confirm the increased MSC secretome is due to improved cell-cell interactions using AIMS: the key mediator N-cadherin was up-regulated while functional blocking of N-cadherin resulted in no enhancement of the secretome. After being primed by IFN-γ, the secretome profile of the MSC aggregates contains more anti-inflammatory cytokines and can be used to inhibit the pro-inflammatory response of M1 phenotype macrophages, suppress T cell activation, and support B cell functions. As such, the MSC secretome can be modified for personalized secretome-based therapies. AIMS acts as a powerful tool for improving the MSC secretome and precisely tuning the secretory profile to develop new treatments in translational medicine.

Supramolecular self‐assembly in biological systems holds promise to convert and amplify disease‐specific signals to physical or mechanical signals that can direct cell fate. However, it remains challenging to design physiologically stable self‐assembling systems that demonstrate tunable and predictable behavior. Here, the use of zwitterionic tetrapeptide modalities to direct nanoparticle assembly under physiological conditions is reported. The self‐assembly of gold nanoparticles can be activated by enzymatic unveiling of surface‐bound zwitterionic tetrapeptides through matrix metalloprotease‐9 (MMP‐9), which is overexpressed by cancer cells. This robust nanoparticle assembly is achieved by multivalent, self‐complementary interactions of the zwitterionic tetrapeptides. In cancer cells that overexpress MMP‐9, the nanoparticle assembly process occurs near the cell membrane and causes size‐induced selection of cellular uptake mechanism, resulting in diminished cell growth. The enzyme responsiveness, and therefore, indirectly, the uptake route of the system can be programmed by customizing the peptide sequence: a simple inversion of the two amino acids at the cleavage site completely inactivates the enzyme responsiveness, self‐assembly, and consequently changes the endocytic pathway. This robust self‐complementary, zwitterionic peptide design demonstrates the use of enzyme‐activated electrostatic side‐chain patterns as powerful and customizable peptide modalities to program nanoparticle self‐assembly and alter cellular response in biological context.